From ce86bac08f11d0e43ece2024aeed038bffdc3c97 Mon Sep 17 00:00:00 2001 From: Ron Green <11993626+georgettica@users.noreply.github.com> Date: Tue, 5 Jan 2021 11:27:24 +0200 Subject: [PATCH 01/62] fix(advanced-scheduling): space fixes as the CKA requires taking these code snippets and using them quickly, spaces can be an issue add two spaces to begin of line to make pod spec copying even faster (if this is the case) --- ...03-00-Advanced-Scheduling-In-Kubernetes.md | 103 +++++------------- 1 file changed, 30 insertions(+), 73 deletions(-) diff --git a/content/en/blog/_posts/2017-03-00-Advanced-Scheduling-In-Kubernetes.md b/content/en/blog/_posts/2017-03-00-Advanced-Scheduling-In-Kubernetes.md index 5d7f0383c5..722b1e59b0 100644 --- a/content/en/blog/_posts/2017-03-00-Advanced-Scheduling-In-Kubernetes.md +++ b/content/en/blog/_posts/2017-03-00-Advanced-Scheduling-In-Kubernetes.md @@ -20,21 +20,14 @@ For example, if we want to require scheduling on a node that is in the us-centra ``` -affinity: - - nodeAffinity: - - requiredDuringSchedulingIgnoredDuringExecution: - - nodeSelectorTerms: - - - matchExpressions: - - - key: "failure-domain.beta.kubernetes.io/zone" - - operator: In - - values: ["us-central1-a"] + affinity: + nodeAffinity: + requiredDuringSchedulingIgnoredDuringExecution: + nodeSelectorTerms: + - matchExpressions: + - key: "failure-domain.beta.kubernetes.io/zone" + operator: In + values: ["us-central1-a"] ``` @@ -44,21 +37,14 @@ Preferred rules mean that if nodes match the rules, they will be chosen first, a ``` -affinity: - - nodeAffinity: - - preferredDuringSchedulingIgnoredDuringExecution: - - nodeSelectorTerms: - - - matchExpressions: - - - key: "failure-domain.beta.kubernetes.io/zone" - - operator: In - - values: ["us-central1-a"] + affinity: + nodeAffinity: + preferredDuringSchedulingIgnoredDuringExecution: + nodeSelectorTerms: + - matchExpressions: + - key: "failure-domain.beta.kubernetes.io/zone" + operator: In + values: ["us-central1-a"] ``` @@ -67,21 +53,14 @@ Node anti-affinity can be achieved by using negative operators. So for instance ``` -affinity: - - nodeAffinity: - - requiredDuringSchedulingIgnoredDuringExecution: - - nodeSelectorTerms: - - - matchExpressions: - - - key: "failure-domain.beta.kubernetes.io/zone" - - operator: NotIn - - values: ["us-central1-a"] + affinity: + nodeAffinity: + requiredDuringSchedulingIgnoredDuringExecution: + nodeSelectorTerms: + - matchExpressions: + - key: "failure-domain.beta.kubernetes.io/zone" + operator: NotIn + values: ["us-central1-a"] ``` @@ -99,7 +78,7 @@ The kubectl command allows you to set taints on nodes, for example: ``` kubectl taint nodes node1 key=value:NoSchedule - ``` +``` creates a taint that marks the node as unschedulable by any pods that do not have a toleration for taint with key key, value value, and effect NoSchedule. (The other taint effects are PreferNoSchedule, which is the preferred version of NoSchedule, and NoExecute, which means any pods that are running on the node when the taint is applied will be evicted unless they tolerate the taint.) The toleration you would add to a PodSpec to have the corresponding pod tolerate this taint would look like this @@ -107,15 +86,11 @@ creates a taint that marks the node as unschedulable by any pods that do not hav ``` -tolerations: - -- key: "key" - - operator: "Equal" - - value: "value" - - effect: "NoSchedule" + tolerations: + - key: "key" + operator: "Equal" + value: "value" + effect: "NoSchedule" ``` @@ -138,21 +113,13 @@ Let’s look at an example. Say you have front-ends in service S1, and they comm ``` affinity: - podAffinity: - requiredDuringSchedulingIgnoredDuringExecution: - - labelSelector: - matchExpressions: - - key: service - operator: In - values: [“S1”] - topologyKey: failure-domain.beta.kubernetes.io/zone ``` @@ -172,25 +139,15 @@ Here we have a Pod where we specify the schedulerName field: ``` apiVersion: v1 - kind: Pod - metadata: - name: nginx - labels: - app: nginx - spec: - schedulerName: my-scheduler - containers: - - name: nginx - image: nginx:1.10 ``` From c6f6fb9a3ee274418e457b88640096089e557cea Mon Sep 17 00:00:00 2001 From: CKchen0726 Date: Thu, 7 Jan 2021 18:54:07 +0800 Subject: [PATCH 02/62] Revise the page to accurately document the well-known labels, annotations and taints. --- .../labels-annotations-taints.md | 140 ++++++++++++++++++ 1 file changed, 140 insertions(+) diff --git a/content/en/docs/reference/kubernetes-api/labels-annotations-taints.md b/content/en/docs/reference/kubernetes-api/labels-annotations-taints.md index b0ef5d5a65..10af60d381 100644 --- a/content/en/docs/reference/kubernetes-api/labels-annotations-taints.md +++ b/content/en/docs/reference/kubernetes-api/labels-annotations-taints.md @@ -114,3 +114,143 @@ The scheduler (through the _VolumeZonePredicate_ predicate) also will ensure tha If `PersistentVolumeLabel` does not support automatic labeling of your PersistentVolumes, you should consider adding the labels manually (or adding support for `PersistentVolumeLabel`). With `PersistentVolumeLabel`, the scheduler prevents Pods from mounting volumes in a different zone. If your infrastructure doesn't have this constraint, you don't need to add the zone labels to the volumes at all. +## node.kubernetes.io/windows-build {#nodekubernetesiowindows-build} + +Example: `node.kubernetes.io/windows-build=10.0.17763` + +Used on: Node + +When the kubelet is running on Microsoft Windows, it automatically labels its node to record the version of Windows Server in use. + +The label's value is in the format "MajorVersion.MinorVersion.BuildNumber". + +## service.kubernetes.io/headless {#servicekubernetesioheadless} + +Example: `service.kubernetes.io/headless=""` + +Used on: Service + +The control plane adds this label to an Endpoints object when the owning Service is headless. + +## kubernetes.io/service-name {#kubernetesioservice-name} + +Example: `kubernetes.io/service-name="nginx"` + +Used on: Service + +Kubernetes uses this label to differentiate multiple Services. Used currently for `ELB`(Elastic Load Balancer) only. + +## endpointslice.kubernetes.io/managed-by {#endpointslicekubernetesiomanaged-by} + +Example: `endpointslice.kubernetes.io/managed-by="controller"` + +Used on: EndpointSlices + +The label is used to indicate the controller or entity that manages an EndpointSlice. This label aims to enable different EndpointSlice objects to be managed by different controllers or entities within the same cluster. + +## endpointslice.kubernetes.io/skip-mirror {#endpointslicekubernetesioskip-mirror} + +Example: `endpointslice.kubernetes.io/skip-mirror="true"` + +Used on: Endpoints + +The label can be set to `"true"` on an Endpoints resource to indicate that the EndpointSliceMirroring controller should not mirror this resource with EndpointSlices. + +## service.kubernetes.io/service-proxy-name {#servicekubernetesioservice-proxy-name} + +Example: `service.kubernetes.io/service-proxy-name="foo-bar"` + +Used on: Service + +The kube-proxy has this label for custom proxy, which delegates service control to custom proxy. + +## experimental.windows.kubernetes.io/isolation-type + +Example: `experimental.windows.kubernetes.io/isolation-type: "hyperv"` + +Used on: Pod + +The annotation is used to run Windows containers with Hyper-V isolation. To use Hyper-V isolation feature and create a Hyper-V isolated container, the kubelet should be started with feature gates HyperVContainer=true and the Pod should include the annotation experimental.windows.kubernetes.io/isolation-type=hyperv. + +{{< note >}} +You can only set this annotation on Pods that have a single container. +{{< /note >}} + +## ingressclass.kubernetes.io/is-default-class + +Example: `ingressclass.kubernetes.io/is-default-class: "true"` + +Used on: IngressClass + +When a single IngressClass resource has this annotation set to `"true"`, new Ingress resource without a class specified will be assigned this default class. + +## kubernetes.io/ingress.class (deprecated) + +{{< note >}} Starting in v1.18, this annotation is deprecated in favor of `spec.ingressClassName`. {{< /note >}} + +## alpha.kubernetes.io/provided-node-ip + +Example: `alpha.kubernetes.io/provided-node-ip: "10.0.0.1"` + +Used on: Node + +The kubelet can set this annotation on a Node to denote its configured IPv4 address. + +When kubelet is started with the "external" cloud provider, it sets this annotation on the Node to denote an IP address set from the command line flag (`--node-ip`). This IP is verified with the cloud provider as valid by the cloud-controller-manager. + +**The taints listed below are always used on Nodes** + +## node.kubernetes.io/not-ready + +Example: `node.kubernetes.io/not-ready:NoExecute` + +The node controller detects whether a node is ready by monitoring its health and adds or removes this taint accordingly. + +## node.kubernetes.io/unreachable + +Example: `node.kubernetes.io/unreachable:NoExecute` + +The node controller adds the taint to a node corresponding to the [NodeCondition](/docs/concepts/architecture/nodes/#condition) `Ready` being `Unknown`. + +## node.kubernetes.io/unschedulable + +Example: `node.kubernetes.io/unschedulable:NoSchedule` + +The taint will be added to a node when initializing the node to avoid race condition. + +## node.kubernetes.io/memory-pressure + +Example: `node.kubernetes.io/memory-pressure:NoSchedule` + +The kubelet detects memory pressure based on `memory.available` and `allocatableMemory.available` observed on a Node. The observed values are then compared to the corresponding thresholds that can be set on the kubelet to determine if the Node condition and taint should be added/removed. + +## node.kubernetes.io/disk-pressure + +Example: `node.kubernetes.io/disk-pressure:NoSchedule` + +The kubelet detects disk pressure based on `imagefs.available`, `imagefs.inodesFree`, `nodefs.available` and `nodefs.inodesFree`(Linux only) observed on a Node. The observed values are then compared to the corresponding thresholds that can be set on the kubelet to determine if the Node condition and taint should be added/removed. + +## node.kubernetes.io/network-unavailable + +Example: `node.kubernetes.io/network-unavailable:NoSchedule` + +This is initially set by the kubelet when the cloud provider used indicates a requirement for additional network configuration. Only when the route on the cloud is configured properly will the taint be removed by the cloud provider. + +## node.kubernetes.io/pid-pressure + +Example: `node.kubernetes.io/pid-pressure:NoSchedule` + +The kubelet checks D-value of the size of `/proc/sys/kernel/pid_max` and the PIDs consumed by Kubernetes on a node to get the number of available PIDs that referred to as the `pid.available` metric. The metric is then compared to the corresponding threshold that can be set on the kubelet to determine if the node condition and taint should be added/removed. + +## node.cloudprovider.kubernetes.io/uninitialized + +Example: `node.cloudprovider.kubernetes.io/uninitialized:NoSchedule` + +Sets this taint on a node to mark it as unusable, when kubelet is started with the "external" cloud provider, until a controller from the cloud-controller-manager initializes this node, and then removes the taint. + +## node.cloudprovider.kubernetes.io/shutdown + +Example: `node.cloudprovider.kubernetes.io/shutdown:NoSchedule` + +If a Node is in a cloud provider specified shutdown state, the Node gets tainted accordingly with `node.cloudprovider.kubernetes.io/shutdown` and the taint effect of `NoSchedule`. + From bda3cea5a6b3ea222be305f79d091ac77dd1a633 Mon Sep 17 00:00:00 2001 From: Aldo Culquicondor <1299064+alculquicondor@users.noreply.github.com> Date: Thu, 14 Jan 2021 14:47:31 -0500 Subject: [PATCH 03/62] Remove indication that a failed scheduling attempt removes the Pod /sig scheduling --- content/en/docs/concepts/workloads/pods/pod-lifecycle.md | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/content/en/docs/concepts/workloads/pods/pod-lifecycle.md b/content/en/docs/concepts/workloads/pods/pod-lifecycle.md index df83f7c5f3..832785923a 100644 --- a/content/en/docs/concepts/workloads/pods/pod-lifecycle.md +++ b/content/en/docs/concepts/workloads/pods/pod-lifecycle.md @@ -38,8 +38,7 @@ If a {{< glossary_tooltip term_id="node" >}} dies, the Pods scheduled to that no are [scheduled for deletion](#pod-garbage-collection) after a timeout period. Pods do not, by themselves, self-heal. If a Pod is scheduled to a -{{< glossary_tooltip text="node" term_id="node" >}} that then fails, -or if the scheduling operation itself fails, the Pod is deleted; likewise, a Pod won't +{{< glossary_tooltip text="node" term_id="node" >}} that then fails, the Pod is deleted; likewise, a Pod won't survive an eviction due to a lack of resources or Node maintenance. Kubernetes uses a higher-level abstraction, called a {{< glossary_tooltip term_id="controller" text="controller" >}}, that handles the work of From daf0cc27a332b27cb91339145f1ea6fb2a3d620a Mon Sep 17 00:00:00 2001 From: Qiming Teng Date: Sat, 30 Jan 2021 12:58:54 +0800 Subject: [PATCH 04/62] [zh] Resync HPA task --- .../horizontal-pod-autoscale.md | 121 ++++++++++++++++++ 1 file changed, 121 insertions(+) diff --git a/content/zh/docs/tasks/run-application/horizontal-pod-autoscale.md b/content/zh/docs/tasks/run-application/horizontal-pod-autoscale.md index fa96247305..4d4c8416c2 100644 --- a/content/zh/docs/tasks/run-application/horizontal-pod-autoscale.md +++ b/content/zh/docs/tasks/run-application/horizontal-pod-autoscale.md @@ -434,6 +434,127 @@ usual. 如果延迟（冷却）时间设置的太短，那么副本数量有可能跟以前一样出现抖动。 {{< /note >}} + +## 对资源指标的支持 {#support-for-resource-metrics} + +HPA 的任何目标资源都可以基于其中的 Pods 的资源用量来实现扩缩。 +在定义 Pod 规约时，类似 `cpu` 和 `memory` 这类资源请求必须被设定。 +这些设定值被用来确定资源利用量并被 HPA 控制器用来对目标资源完成扩缩操作。 +要使用基于资源利用率的扩缩，可以像下面这样指定一个指标源： + +```yaml +type: Resource +resource: + name: cpu + target: + type: Utilization + averageUtilization: 60 +``` + + +基于这一指标设定，HPA 控制器会维持扩缩目标中的 Pods 的平均资源利用率在 60%。 +利用率是 Pod 的当前资源用量与其请求值之间的比值。关于如何计算利用率以及如何计算平均值 +的细节可参考[算法](#algorithm-details)小节。 + +{{< note >}} + +由于所有的容器的资源用量都会被累加起来，Pod 的总体资源用量值可能不会精确体现 +各个容器的资源用量。这一现象也会导致一些问题，例如某个容器运行时的资源用量非常 +高，但因为 Pod 层面的资源用量总值让人在可接受的约束范围内，HPA 不会执行扩大 +目标对象规模的操作。 +{{< /note >}} + + +### 容器资源指标 {#container-resource-metrics} + +{{< feature-state for_k8s_version="v1.20" state="alpha" >}} + + +`HorizontalPodAutoscaler` 也支持容器指标源，这时 HPA 可以跟踪记录一组 Pods 中各个容器的 +资源用量，进而触发扩缩目标对象的操作。 +容器资源指标的支持使得你可以为特定 Pod 中最重要的容器配置规模缩放阈值。 +例如，如果你有一个 Web 应用和一个执行日志操作的边车容器，你可以基于 Web 应用的 +资源用量来执行扩缩，忽略边车容器的存在及其资源用量。 + + +如果你更改缩放目标对象，令其使用新的、包含一组不同的容器的 Pod 规约，你就需要 +修改 HPA 的规约才能基于新添加的容器来执行规模扩缩操作。 +如果指标源中指定的容器不存在或者仅存在于部分 Pods 中，那么这些 Pods 会被忽略， +HPA 会重新计算资源用量值。参阅[算法](#algorithm-details)小节进一步了解计算细节。 +要使用容器资源用量来完成自动扩缩，可以像下面这样定义指标源： + +```yaml +type: ContainerResource +containerResource: + name: cpu + container: application + target: + type: Utilization + averageUtilization: 60 +``` + + +在上面的例子中，HPA 控制器会对目标对象执行扩缩操作以确保所有 Pods 中 +`application` 容器的平均 CPU 用量为 60%。 + +{{< note >}} + +如果你要更改 HorizontalPodAutoscaler 所跟踪记录的容器的名称，你可以按一定顺序 +来执行这一更改，确保在应用更改的过程中用来判定扩缩行为的容器可用。 +在更新定义容器的资源（如 Deployment）之前，你需要更新相关的 HPA，使之能够同时 +跟踪记录新的和老的容器名称。这样，HPA 就能够在整个更新过程中继续计算并提供扩缩操作建议。 + + +一旦你已经将容器名称变更这一操作应用到整个负载对象至上，就可以从 HPA +的规约中去掉老的容器名称，完成清理操作。 +{{< /note >}} + 在温度计的例子中，如果房间很冷，那么某个控制器可能还会启动一个防冻加热器。 diff --git a/content/zh/docs/concepts/architecture/nodes.md b/content/zh/docs/concepts/architecture/nodes.md index 676f6f64b9..7c6000268f 100644 --- a/content/zh/docs/concepts/architecture/nodes.md +++ b/content/zh/docs/concepts/architecture/nodes.md @@ -17,9 +17,10 @@ weight: 10 Kubernetes 通过将容器放入在节点（Node）上运行的 Pod 中来执行你的工作负载。 -节点可以是一个虚拟机或者物理机器，取决于所在的集群配置。每个节点都包含用于运行 -{{< glossary_tooltip text="Pod" term_id="pod" >}} 所需要的服务，这些服务由 -{{< glossary_tooltip text="控制面" term_id="control-plane" >}}负责管理。 +节点可以是一个虚拟机或者物理机器，取决于所在的集群配置。每个节点由 +{{< glossary_tooltip text="控制面" term_id="control-plane" >}} 负责管理， +并包含运行 {{< glossary_tooltip text="Pods" term_id="pod" >}} 所需的服务。 通常集群中会有若干个节点；而在一个学习用或者资源受限的环境中，你的集群中也可能 只有一个节点。 From 90d92bac064e3d58d318ff3ad546224a73a86540 Mon Sep 17 00:00:00 2001 From: Karen Bradshaw Date: Mon, 8 Feb 2021 12:52:08 -0500 Subject: [PATCH 06/62] clean up capture stmts --- .../es/docs/concepts/configuration/secret.md | 75 +++++++++++-------- 1 file changed, 44 insertions(+), 31 deletions(-) diff --git a/content/es/docs/concepts/configuration/secret.md b/content/es/docs/concepts/configuration/secret.md index f1e68ee0f5..7120f0476b 100644 --- a/content/es/docs/concepts/configuration/secret.md +++ b/content/es/docs/concepts/configuration/secret.md @@ -10,16 +10,13 @@ feature: weight: 50 --- - -{{% capture overview %}} + Los objetos de tipo {{< glossary_tooltip text="Secret" term_id="secret" >}} en Kubernetes te permiten almacenar y administrar información confidencial, como contraseñas, tokens OAuth y llaves ssh. Poniendo esta información en un Secret es más seguro y más flexible que ponerlo en la definición de un {{< glossary_tooltip term_id="pod" >}} o en un {{< glossary_tooltip text="container image" term_id="image" >}}. Ver [Secrets design document](https://git.k8s.io/community/contributors/design-proposals/auth/secrets.md) para más información. -{{% /capture %}} - -{{% capture body %}} + ## Introducción a Secrets @@ -58,9 +55,11 @@ empaqueta esos archivos en un Secret y crea el objeto en el Apiserver. ```shell kubectl create secret generic db-user-pass --from-file=./username.txt --from-file=./password.txt ``` -``` + +```none Secret "db-user-pass" created ``` + {{< note >}} Si la contraseña que está utilizando tiene caracteres especiales como por ejemplo `$`, `\`, `*`, o `!`, es posible que sean interpretados por tu intérprete de comandos y es necesario escapar cada carácter utilizando `\` o introduciéndolos entre comillas simples `'`. Por ejemplo, si tú password actual es `S!B\*d$zDsb`, deberías ejecutar el comando de esta manera: @@ -76,14 +75,17 @@ Puedes comprobar que el Secret se haya creado, así: ```shell kubectl get secrets ``` -``` + +```none NAME TYPE DATA AGE db-user-pass Opaque 2 51s ``` + ```shell kubectl describe secrets/db-user-pass ``` -``` + +```none Name: db-user-pass Namespace: default Labels: @@ -137,7 +139,8 @@ Ahora escribe un Secret usando [`kubectl apply`](/docs/reference/generated/kubec ```shell kubectl apply -f ./secret.yaml ``` -``` + +```none secret "mysecret" created ``` @@ -242,6 +245,7 @@ desde 1.14. Con esta nueva característica, puedes tambien crear un Secret a partir de un generador y luego aplicarlo para crear el objeto en el Apiserver. Los generadores deben ser especificados en un `kustomization.yaml` dentro de un directorio. Por ejemplo, para generar un Secret a partir de los archivos `./username.txt` y `./password.txt` + ```shell # Crear un fichero llamado kustomization.yaml con SecretGenerator cat <./kustomization.yaml @@ -281,9 +285,10 @@ username.txt: 5 bytes Por ejemplo, para generar un Secret a partir de literales `username=admin` y `password=secret`, puedes especificar el generador del Secret en `kustomization.yaml` como: + ```shell # Crea un fichero kustomization.yaml con SecretGenerator -$ cat <./kustomization.yaml +cat <./kustomization.yaml secretGenerator: - name: db-user-pass literals: @@ -291,11 +296,14 @@ secretGenerator: - password=secret EOF ``` + Aplica el directorio kustomization para crear el objeto Secret. + ```shell -$ kubectl apply -k . +kubectl apply -k . secret/db-user-pass-dddghtt9b5 created ``` + {{< note >}} El nombre generado del Secret tiene un sufijo agregado al hashing de los contenidos. Esto asegura que se genera un nuevo Secret cada vez que el contenido es modificado. {{< /note >}} @@ -307,7 +315,8 @@ Los Secrets se pueden recuperar a través del comando `kubectl get secret` . Por ```shell kubectl get secret mysecret -o yaml ``` -``` + +```none apiVersion: v1 kind: Secret metadata: @@ -328,7 +337,8 @@ Decodifica el campo de contraseña: ```shell echo 'MWYyZDFlMmU2N2Rm' | base64 --decode ``` -``` + +```none 1f2d1e2e67df ``` @@ -480,7 +490,8 @@ Este es el resultado de comandos ejecutados dentro del contenedor del ejemplo an ```shell ls /etc/foo/ ``` -``` + +```none username password ``` @@ -488,15 +499,16 @@ password ```shell cat /etc/foo/username ``` -``` + +```none admin ``` - ```shell cat /etc/foo/password ``` -``` + +```none 1f2d1e2e67df ``` @@ -562,13 +574,16 @@ Este es el resultado de comandos ejecutados dentro del contenedor del ejemplo an ```shell echo $SECRET_USERNAME ``` -``` + +```none admin ``` + ```shell echo $SECRET_PASSWORD ``` -``` + +```none 1f2d1e2e67df ``` @@ -641,7 +656,7 @@ Cree un fichero kustomization.yaml con SecretGenerator conteniendo algunas llave kubectl create secret generic ssh-key-secret --from-file=ssh-privatekey=/path/to/.ssh/id_rsa --from-file=ssh-publickey=/path/to/.ssh/id_rsa.pub ``` -``` +```none secret "ssh-key-secret" created ``` @@ -649,7 +664,6 @@ secret "ssh-key-secret" created Piense detenidamente antes de enviar tus propias llaves ssh: otros usuarios del cluster pueden tener acceso al Secret. Utilice una cuenta de servicio a la que desee que estén accesibles todos los usuarios con los que comparte el cluster de Kubernetes, y pueda revocarlas si se ven comprometidas. {{< /caution >}} - Ahora podemos crear un pod que haga referencia al Secret con la llave ssh key y lo consuma en un volumen: ```yaml @@ -691,16 +705,19 @@ Crear un fichero kustomization.yaml con SecretGenerator ```shell kubectl create secret generic prod-db-secret --from-literal=username=produser --from-literal=password=Y4nys7f11 ``` -``` + +```none secret "prod-db-secret" created ``` ```shell kubectl create secret generic test-db-secret --from-literal=username=testuser --from-literal=password=iluvtests ``` -``` + +```none secret "test-db-secret" created ``` + {{< note >}} Caracteres especiales como `$`, `\*`, y `!` requieren ser escapados. Si el password que estas usando tiene caracteres especiales, necesitas escaparlos usando el caracter `\\` . Por ejemplo, si tu password actual es `S!B\*d$zDsb`, deberías ejecutar el comando de esta forma: @@ -715,7 +732,7 @@ No necesitas escapar caracteres especiales en contraseñas de los archivos (`--f Ahora haz los pods: ```shell -$ cat < pod.yaml +cat < pod.yaml apiVersion: v1 kind: List items: @@ -759,8 +776,9 @@ EOF ``` Añade los pods a el mismo fichero kustomization.yaml + ```shell -$ cat <> kustomization.yaml +cat <> kustomization.yaml resources: - pod.yaml EOF @@ -833,7 +851,6 @@ spec: mountPath: "/etc/secret-volume" ``` - El `secret-volume` contendrá un solo archivo, llamado `.secret-file`, y el `dotfile-test-container` tendrá este fichero presente en el path `/etc/secret-volume/.secret-file`. @@ -874,7 +891,6 @@ para que los clientes puedan `watch` recursos individuales, y probablemente esta ## Propiedades de seguridad - ### Protecciones Debido a que los objetos `Secret` se pueden crear independientemente de los `Pods` que los usan, hay menos riesgo de que el Secret expuesto durante el flujo de trabajo de la creación, visualización, y edición de pods. El sistema también puede tomar precausiones con los objetos`Secret`, tal como eviar escribirlos en el disco siempre que sea posible. @@ -906,7 +922,4 @@ para datos secretos, para que los Secrets no se almacenen en claro en {{< glossa - Un usuario que puede crear un pod que usa un Secret también puede ver el valor del Secret. Incluso si una política del apiserver no permite que ese usuario lea el objeto Secret, el usuario puede ejecutar el pod que expone el Secret. - Actualmente, cualquier persona con root en cualquier nodo puede leer _cualquier_ secret del apiserver, haciéndose pasar por el kubelet. Es una característica planificada enviar Secrets a los nodos que realmente lo requieran, para restringir el impacto de una explosión de root en un single node. - -{{% capture whatsnext %}} - -{{% /capture %}} +## {{% heading "whatsnext" %}} From dc1de9e3ae48706f5a44d4d795dd841fb32ae0df Mon Sep 17 00:00:00 2001 From: Zhang Yong Date: Tue, 9 Feb 2021 11:21:29 +0800 Subject: [PATCH 07/62] Dual-stack docs correction --- .../zh/docs/concepts/services-networking/dual-stack.md | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/content/zh/docs/concepts/services-networking/dual-stack.md b/content/zh/docs/concepts/services-networking/dual-stack.md index 4c44d75016..ef811874b5 100644 --- a/content/zh/docs/concepts/services-networking/dual-stack.md +++ b/content/zh/docs/concepts/services-networking/dual-stack.md @@ -125,14 +125,14 @@ IPv6 CIDR 的一个例子：`fdXY:IJKL:MNOP:15::/64`（这里演示的是格式 如果你的集群启用了 IPv4/IPv6 双协议栈网络，则可以使用 IPv4 或 IPv6 地址来创建 {{< glossary_tooltip text="Service" term_id="service" >}}。 -服务的地址族默认为第一个服务集群 IP 范围的地址族（通过 kube-controller-manager 的 `--service-cluster-ip-range` 参数配置） +服务的地址族默认为第一个服务集群 IP 范围的地址族（通过 kube-apiserver 的 `--service-cluster-ip-range` 参数配置） 当你定义服务时，可以选择将其配置为双栈。若要指定所需的行为，你可以设置 `.spec.ipFamilyPolicy` 字段为以下值之一： 2. 在集群上启用双栈时，带有选择算符的现有 [无头服务](/zh/docs/concepts/services-networking/service/#headless-services) 由控制面设置 `.spec.ipFamilyPolicy` 为 `SingleStack` - 并设置 `.spec.ipFamilies` 为第一个服务群集 IP 范围的地址族（通过配置 kube-controller-manager 的 + 并设置 `.spec.ipFamilies` 为第一个服务群集 IP 范围的地址族（通过配置 kube-apiserver 的 `--service-cluster-ip-range` 参数），即使 `.spec.ClusterIP` 的设置值为 `None` 也如此。 {{< codenew file="service/networking/dual-stack-default-svc.yaml" >}} From 2471ced338cb86e698c72b6ede842b070f984a45 Mon Sep 17 00:00:00 2001 From: Yang Jiao Date: Tue, 9 Feb 2021 15:36:57 +0800 Subject: [PATCH 08/62] [zh] sync and update statefulset.md update based on English version of statefulset.md 1. added a block inside #stable-network-id 2. sync tables and blocks 3. added a sentence in #parallel-pod-management --- .../workloads/controllers/statefulset.md | 53 ++++++++++++++++++- 1 file changed, 51 insertions(+), 2 deletions(-) diff --git a/content/zh/docs/concepts/workloads/controllers/statefulset.md b/content/zh/docs/concepts/workloads/controllers/statefulset.md index cdc0383316..7a44e249e8 100644 --- a/content/zh/docs/concepts/workloads/controllers/statefulset.md +++ b/content/zh/docs/concepts/workloads/controllers/statefulset.md @@ -1,7 +1,7 @@ --- title: StatefulSets content_type: concept -weight: 40 +weight: 30 --- +上述例子中： + * 名为 `nginx` 的 Headless Service 用来控制网络域名。 * 名为 `web` 的 StatefulSet 有一个 Spec，它表明将在独立的 3 个 Pod 副本中启动 nginx 容器。 * `volumeClaimTemplates` 将通过 PersistentVolumes 驱动提供的 [PersistentVolumes](/zh/docs/concepts/storage/persistent-volumes/) 来提供稳定的存储。 +StatefulSet 的命名需要遵循 [DNS 子域名](zh/docs/concepts/overview/working-with-objects/names#dns-subdomain-names)。 + @@ -217,9 +227,46 @@ StatefulSet 可以使用 [无头服务](/zh/docs/concepts/services-networking/se 一旦每个 Pod 创建成功，就会得到一个匹配的 DNS 子域，格式为： `$(pod 名称).$(所属服务的 DNS 域名)`，其中所属服务由 StatefulSet 的 `serviceName` 域来设定。 + +取决于集群域内部 DNS 的配置，有可能无法查询一个刚刚启动的 Pod 上的 DNS 命名。这个情况会在其他集群域内部的客户端在主机 Pod 创建完成前 +发出查询请求。负缓存 (通常来说在 DNS 内) 意味着之前失败的查询结果在特定秒数内被记录和被重用，甚至直至 Pod 正常运行。 + +如果需要及时查询创建之后的 Pod ，有以下选项： + +- 直接查询 Kubernetes API（比如，利用 watch 机制）而不是依赖于 DNS 查询 +- 减少 Kubernetes DNS 提供商的 缓存时效（通常这意味着修改 CoreDNS 的 config map，目前 config map 会缓存30秒信息） + +正如 [限制](#limitations) 中提到的，创建 [无头服务](zh/docs/concepts/services-networking/service/#headless-services) +需要对其 Pod 的网络标识负责。 + 下面给出一些选择集群域、服务名、StatefulSet 名、及其怎样影响 StatefulSet 的 Pod 上的 DNS 名称的示例： @@ -350,12 +397,14 @@ described [above](#deployment-and-scaling-guarantees). `Parallel` pod management tells the StatefulSet controller to launch or terminate all Pods in parallel, and to not wait for Pods to become Running and Ready or completely terminated prior to launching or terminating another -Pod. +Pod. This option only affects the behavior for scaling operations. Updates are not affected. + --> #### 并行 Pod 管理 {#parallel-pod-management} `Parallel` Pod 管理让 StatefulSet 控制器并行的启动或终止所有的 Pod， 启动或者终止其他 Pod 前，无需等待 Pod 进入 Running 和 ready 或者完全停止状态。 +这个选项只会影响伸缩操作的行为，更新则不会被影响。 在 Ingress 中引用此 Secret 将会告诉 Ingress 控制器使用 TLS 加密从客户端到负载均衡器的通道。 -你需要确保创建的 TLS Secret 创建自包含 `sslexample.foo.com` 的公用名称（CN）的证书。 +你需要确保创建的 TLS Secret 创建自包含 `https-example.foo.com` 的公用名称（CN）的证书。 这里的公共名称也被称为全限定域名（FQDN）。 {{< note >}} From 31b2260133df68b71c0ba98a1638ff4fc4243abd Mon Sep 17 00:00:00 2001 From: Qiming Teng Date: Thu, 11 Feb 2021 15:14:47 +0800 Subject: [PATCH 10/62] [zh] Fix some bad links in tasks section --- .../connecting-frontend-backend.md | 2 +- .../administer-cluster/safely-drain-node.md | 22 ++++++++----------- 2 files changed, 10 insertions(+), 14 deletions(-) diff --git a/content/zh/docs/tasks/access-application-cluster/connecting-frontend-backend.md b/content/zh/docs/tasks/access-application-cluster/connecting-frontend-backend.md index a485d9ba3a..42ebe8d250 100644 --- a/content/zh/docs/tasks/access-application-cluster/connecting-frontend-backend.md +++ b/content/zh/docs/tasks/access-application-cluster/connecting-frontend-backend.md @@ -345,5 +345,5 @@ kubectl delete deployment frontend backend --> * 进一步了解 [Service](/zh/docs/concepts/services-networking/service/) * 进一步了解 [ConfigMap](/zh/docs/tasks/configure-pod-container/configure-pod-configmap/) -* 进一步了解 [Service 和 Pods 的 DNS](/docs/concepts/services-networking/dns-pod-service/) +* 进一步了解 [Service 和 Pods 的 DNS](/zh/docs/concepts/services-networking/dns-pod-service/) diff --git a/content/zh/docs/tasks/administer-cluster/safely-drain-node.md b/content/zh/docs/tasks/administer-cluster/safely-drain-node.md index 06fdf6a07d..063288cb93 100644 --- a/content/zh/docs/tasks/administer-cluster/safely-drain-node.md +++ b/content/zh/docs/tasks/administer-cluster/safely-drain-node.md @@ -4,7 +4,6 @@ content_type: task min-kubernetes-server-version: 1.5 --- @@ -264,7 +262,14 @@ eviction API will never return anything other than 429 or 500. For example: this can happen if ReplicaSet is creating Pods for your application but the replacement Pods do not become `Ready`. You can also see similar symptoms if the last Pod evicted has a very long termination grace period. +--> +## 驱逐阻塞 +在某些情况下，应用程序可能会到达一个中断状态，除了 429 或 500 之外，它将永远不会返回任何内容。 +例如 ReplicaSet 创建的替换 Pod 没有变成就绪状态，或者被驱逐的最后一个 +Pod 有很长的终止宽限期，就会发生这种情况。 + + -## 驱逐阻塞 - -在某些情况下，应用程序可能会到达一个中断状态，除了 429 或 500 之外，它将永远不会返回任何内容。 -例如 ReplicaSet 创建的替换 Pod 没有变成就绪状态，或者被驱逐的最后一个 -Pod 有很长的终止宽限期，就会发生这种情况。 - 在这种情况下，有两种可能的解决方案： - 中止或暂停自动操作。调查应用程序卡住的原因，并重新启动自动化。 -- 经过适当的长时间等待后， 从集群中删除 Pod 而不是使用驱逐 API。 +- 经过适当的长时间等待后，从集群中删除 Pod 而不是使用驱逐 API。 Kubernetes 并没有具体说明在这种情况下应该采取什么行为， 这应该由应用程序所有者和集群所有者紧密沟通，并达成对行动一致意见。 ## {{% heading "whatsnext" %}} - +--> * 执行[配置 PDB](/zh/docs/tasks/run-application/configure-pdb/)中的各个步骤， 保护你的应用 -* 进一步了解[节点维护](/zh/docs/tasks/administer-cluster/cluster-management/#maintenance-on-a-node)。 From 52920a9a853dc4e45124ec09c032a90a0961b4f1 Mon Sep 17 00:00:00 2001 From: Peri Thompson <11655452+perithompson@users.noreply.github.com> Date: Thu, 11 Feb 2021 10:32:24 +0000 Subject: [PATCH 11/62] Added limitation with windows network services --- .../windows/intro-windows-in-kubernetes.md | 1 + 1 file changed, 1 insertion(+) diff --git a/content/en/docs/setup/production-environment/windows/intro-windows-in-kubernetes.md b/content/en/docs/setup/production-environment/windows/intro-windows-in-kubernetes.md index 014df012c7..57f867d0db 100644 --- a/content/en/docs/setup/production-environment/windows/intro-windows-in-kubernetes.md +++ b/content/en/docs/setup/production-environment/windows/intro-windows-in-kubernetes.md @@ -308,6 +308,7 @@ The following networking functionality is not supported on Windows nodes * Host networking mode is not available for Windows pods * Local NodePort access from the node itself fails (works for other nodes or external clients) * Accessing service VIPs from nodes will be available with a future release of Windows Server +* A single service can only support up to 64 backend pods / unique destination IPs * Overlay networking support in kube-proxy is an alpha release. In addition, it requires [KB4482887](https://support.microsoft.com/en-us/help/4482887/windows-10-update-kb4482887) to be installed on Windows Server 2019 * Local Traffic Policy and DSR mode * Windows containers connected to l2bridge, l2tunnel, or overlay networks do not support communicating over the IPv6 stack. There is outstanding Windows platform work required to enable these network drivers to consume IPv6 addresses and subsequent Kubernetes work in kubelet, kube-proxy, and CNI plugins. From 6d1f6e594369fb7b7ed55b643f89f0d26d083afa Mon Sep 17 00:00:00 2001 From: luzg Date: Thu, 11 Feb 2021 17:23:02 +0800 Subject: [PATCH 12/62] [zh] translate blog Dockershim Deprecation FAQ fix according to tengqm --- .../blog/_posts/2020-12-02-dockershim-faq.md | 316 ++++++++++++++++++ 1 file changed, 316 insertions(+) create mode 100644 content/zh/blog/_posts/2020-12-02-dockershim-faq.md diff --git a/content/zh/blog/_posts/2020-12-02-dockershim-faq.md b/content/zh/blog/_posts/2020-12-02-dockershim-faq.md new file mode 100644 index 0000000000..b169eea610 --- /dev/null +++ b/content/zh/blog/_posts/2020-12-02-dockershim-faq.md @@ -0,0 +1,316 @@ +--- +layout: blog +title: "弃用 Dockershim 的常见问题" +date: 2020-12-02 +slug: dockershim-faq +aliases: [ '/dockershim' ] +--- + + + +本文回顾了自 Kubernetes v1.20 版宣布弃用 Dockershim 以来所引发的一些常见问题。 +关于 Kubernetes kubelets 从容器运行时的角度弃用 Docker 的细节以及这些细节背后的含义，请参考博文 +[别慌: Kubernetes 和 Docker](/blog/2020/12/02/dont-panic-kubernetes-and-docker/) + + +### 为什么弃用 dockershim {#why-is-dockershim-being-deprecated} + + +维护 dockershim 已经成为 Kubernetes 维护者肩头一个沉重的负担。 +创建 CRI 标准就是为了减轻这个负担，同时也可以增加不同容器运行时之间平滑的互操作性。 +但反观 Docker 却至今也没有实现 CRI，所以麻烦就来了。 + + +Dockershim 向来都是一个临时解决方案（因此得名：shim）。 +你可以进一步阅读 +[移除 Kubernetes 增强方案 Dockershim](https://github.com/kubernetes/enhancements/tree/master/keps/sig-node/1985-remove-dockershim) +以了解相关的社区讨论和计划。 + + +此外，与 dockershim 不兼容的一些特性，例如：控制组（cgoups）v2 和用户名字空间（user namespace），已经在新的 CRI 运行时中被实现。 +移除对 dockershim 的支持将加速这些领域的发展。 + + +### 在 Kubernetes 1.20 版本中，我还可以用 Docker 吗？ {#can-I-still-use-docker-in-kubernetes-1.20} + + +当然可以，在 1.20 版本中仅有的改变就是：如果使用 Docker 运行时，启动 +[kubelet](/zh/docs/reference/command-line-tools-reference/kubelet/) +的过程中将打印一条警告日志。 + + +### 什么时候移除 dockershim {#when-will-dockershim-be-removed} + + +考虑到此改变带来的影响，我们使用了一个加长的废弃时间表。 +在 Kubernetes 1.22 版之前，它不会被彻底移除；换句话说，dockershim 被移除的最早版本会是 2021 年底发布 1.23 版。 +我们将与供应商以及其他生态团队紧密合作，确保顺利过渡，并将依据事态的发展评估后续事项。 + + +### 我现有的 Docker 镜像还能正常工作吗？ {#will-my-existing-docker-image-still-work} + + +当然可以，`docker build` 创建的镜像适用于任何 CRI 实现。 +所有你的现有镜像将和往常一样工作。 + + +### 私有镜像呢？{#what-about-private-images} + + +当然可以。所有 CRI 运行时均支持 Kubernetes 中相同的拉取（pull）Secret 配置， +不管是通过 PodSpec 还是通过 ServiceAccount 均可。 + + +### Docker 和容器是一回事吗？ {#are-docker-and-containers-the-same-thing} + + +虽然 Linux 的容器技术已经存在了很久， +但 Docker 普及了 Linux 容器这种技术模式，并在开发底层技术方面发挥了重要作用。 +容器的生态相比于单纯的 Docker，已经进化到了一个更宽广的领域。 +像 OCI 和 CRI 这类标准帮助许多工具在我们的生态中成长和繁荣， +其中一些工具替代了 Docker 的某些部分，另一些增强了现有功能。 + + +### 现在是否有在生产系统中使用其他运行时的例子？ {#are-there-example-of-folks-using-other-runtimes-in-production-today} + + +Kubernetes 所有项目在所有版本中出产的工件（Kubernetes 二进制文件）都经过了验证。 + + +此外，[kind](https://kind.sigs.k8s.io/) 项目使用 containerd 已经有年头了， +并且在这个场景中，稳定性还明显得到提升。 +Kind 和 containerd 每天都会做多次协调，以验证对 Kubernetes 代码库的所有更改。 +其他相关项目也遵循同样的模式，从而展示了其他容器运行时的稳定性和可用性。 +例如，OpenShift 4.x 从 2019 年 6 月以来，就一直在生产环境中使用 [CRI-O](https://cri-o.io/) 运行时。 + + +至于其他示例和参考资料，你可以查看 containerd 和 CRI-O 的使用者列表， +这两个容器运行时是云原生基金会（[CNCF]）下的项目。 + +- [containerd](https://github.com/containerd/containerd/blob/master/ADOPTERS.md) +- [CRI-O](https://github.com/cri-o/cri-o/blob/master/ADOPTERS.md) + + +### 人们总在谈论 OCI，那是什么？ {#people-keep-referenceing-oci-what-is-that} + + +OCI 代表[开放容器标准](https://opencontainers.org/about/overview/)， +它标准化了容器工具和底层实现（technologies）之间的大量接口。 +他们维护了打包容器镜像（OCI image-spec）和运行容器（OCI runtime-spec）的标准规范。 +他们还以 [runc](https://github.com/opencontainers/runc) +的形式维护了一个 runtime-spec 的真实实现， +这也是 [containerd](https://containerd.io/) 和 [CRI-O](https://cri-o.io/) 依赖的默认运行时。 +CRI 建立在这些底层规范之上，为管理容器提供端到端的标准。 + + +### 我应该用哪个 CRI 实现？ {#which-cri-implementation-should-I-use} + + +这是一个复杂的问题，依赖于许多因素。 +在 Docker 工作良好的情况下，迁移到 containerd 是一个相对容易的转换，并将获得更好的性能和更少的开销。 +然而，我们建议你先探索 [CNCF 全景图](https://landscape.cncf.io/category=container-runtime&format=card-mode&grouping=category) +提供的所有选项，以做出更适合你的环境的选择。 + + +### 当切换 CRI 底层实现时，我应该注意什么？ {#what-should-I-look-out-for-when-changing-CRI-implementation} + + +Docker 和大多数 CRI（包括 containerd）的底层容器化代码是相同的，但其周边部分却存在一些不同。 +迁移时一些常见的关注点是： + + + +- 日志配置 +- 运行时的资源限制 +- 直接访问 docker 命令或通过控制套接字调用 Docker 的节点供应脚本 +- 需要访问 docker 命令或控制套接字的 kubectl 插件 +- 需要直接访问 Docker 的 Kubernetes 工具（例如：kube-imagepuller） +- 像 `registry-mirrors` 和不安全的注册表这类功能的配置 +- 需要 Docker 保持可用、且运行在 Kubernetes 之外的，其他支持脚本或守护进程（例如：监视或安全代理） +- GPU 或特殊硬件，以及它们如何与你的运行时和 Kubernetes 集成 + + +如果你只是用了 Kubernetes 资源请求/限制或基于文件的日志收集 DaemonSet，它们将继续稳定工作， +但是如果你用了自定义了 dockerd 配置，则可能需要为新容器运行时做一些适配工作。 + + +另外还有一个需要关注的点，那就是当创建镜像时，系统维护或嵌入容器方面的任务将无法工作。 +对于前者，可以用 [`crictl`](https://github.com/kubernetes-sigs/cri-tools) 工具作为临时替代方案 +(参见 [从 docker 命令映射到 crictl](https://kubernetes.io/zh/docs/tasks/debug-application-cluster/crictl/#mapping-from-docker-cli-to-crictl))； +对于后者，可以用新的容器创建选项，比如 +[img](https://github.com/genuinetools/img)、 +[buildah](https://github.com/containers/buildah)、 +[kaniko](https://github.com/GoogleContainerTools/kaniko)、或 +[buildkit-cli-for-kubectl](https://github.com/vmware-tanzu/buildkit-cli-for-kubectl +)， +他们均不需要访问 Docker。 + + +对于 containerd，你可以从它们的 +[文档](https://github.com/containerd/cri/blob/master/docs/registry.md) +开始，看看在迁移过程中有哪些配置选项可用。 + + +对于如何协同 Kubernetes 使用 containerd 和 CRI-O 的说明，参见 Kubernetes 文档中这部分： +[容器运行时](/zh/docs/setup/production-environment/container-runtimes)。 + + +### 我还有问题怎么办？{#what-if-I-have-more-question} + + +如果你使用了一个有供应商支持的 Kubernetes 发行版，你可以咨询供应商他们产品的升级计划。 +对于最终用户的问题，请把问题发到我们的最终用户社区的论坛：https://discuss.kubernetes.io/。 + + +你也可以看看这篇优秀的博文： +[等等，Docker 刚刚被 Kubernetes 废掉了？](https://dev.to/inductor/wait-docker-is-deprecated-in-kubernetes-now-what-do-i-do-e4m) +一个对此变化更深入的技术讨论。 + + +### 我可以加入吗？{#can-I-have-a-hug} + + +只要你愿意，随时随地欢迎加入！ + From 807eeda15f3e749afd0e463751f051f8ac0eecb2 Mon Sep 17 00:00:00 2001 From: Yang Jiao Date: Fri, 12 Feb 2021 15:40:54 +0800 Subject: [PATCH 13/62] Apply suggestions from code review Co-authored-by: Qiming Teng --- .../workloads/controllers/statefulset.md | 17 +++++++++-------- 1 file changed, 9 insertions(+), 8 deletions(-) diff --git a/content/zh/docs/concepts/workloads/controllers/statefulset.md b/content/zh/docs/concepts/workloads/controllers/statefulset.md index 7a44e249e8..4cd6606a38 100644 --- a/content/zh/docs/concepts/workloads/controllers/statefulset.md +++ b/content/zh/docs/concepts/workloads/controllers/statefulset.md @@ -163,7 +163,7 @@ The name of a StatefulSet object must be a valid * `volumeClaimTemplates` 将通过 PersistentVolumes 驱动提供的 [PersistentVolumes](/zh/docs/concepts/storage/persistent-volumes/) 来提供稳定的存储。 -StatefulSet 的命名需要遵循 [DNS 子域名](zh/docs/concepts/overview/working-with-objects/names#dns-subdomain-names)。 +StatefulSet 的命名需要遵循[DNS 子域名](zh/docs/concepts/overview/working-with-objects/names#dns-subdomain-names)规范。 -取决于集群域内部 DNS 的配置，有可能无法查询一个刚刚启动的 Pod 上的 DNS 命名。这个情况会在其他集群域内部的客户端在主机 Pod 创建完成前 -发出查询请求。负缓存 (通常来说在 DNS 内) 意味着之前失败的查询结果在特定秒数内被记录和被重用，甚至直至 Pod 正常运行。 +取决于集群域内部 DNS 的配置，有可能无法查询一个刚刚启动的 Pod 的 DNS 命名。 +当集群内其他客户端在 Pod 创建完成前发出 Pod 主机名查询时，就会发生这种情况。 +负缓存 (在 DNS 中较为常见) 意味着之前失败的查询结果会被记录和重用至少若干秒钟， +即使 Pod 已经正常运行了也是如此。 -如果需要及时查询创建之后的 Pod ，有以下选项： +如果需要在 Pod 被创建之后及时发现它们，有以下选项： - 直接查询 Kubernetes API（比如，利用 watch 机制）而不是依赖于 DNS 查询 -- 减少 Kubernetes DNS 提供商的 缓存时效（通常这意味着修改 CoreDNS 的 config map，目前 config map 会缓存30秒信息） +- 缩短 Kubernetes DNS 驱动的缓存时长（通常这意味着修改 CoreDNS 的 ConfigMap，目前缓存时长为 30 秒） -正如 [限制](#limitations) 中提到的，创建 [无头服务](zh/docs/concepts/services-networking/service/#headless-services) -需要对其 Pod 的网络标识负责。 +正如[限制](#limitations)中所述，你需要负责创建[无头服务](/zh/docs/concepts/services-networking/service/#headless-services) +以便为 Pod 提供网络标识。 + +As the Kubernetes API evolves, APIs are periodically reorganized or upgraded. +When APIs evolve, the old API is deprecated and eventually removed. +This page contains information you need to know when migrating from +deprecated API versions to newer and more stable API versions. + + + +## Removed APIs by release + + +### v1.25 + +The **v1.25** release will stop serving the following deprecated API versions: + +#### Event + +The **events.k8s.io/v1beta1** API version of Event will no longer be served in v1.25. + +* Migrate manifests and API clients to use the **events.k8s.io/v1** API version, available since v1.19. +* All existing persisted objects are accessible via the new API +* Notable changes + +#### RuntimeClass + +RuntimeClass in the **node.k8s.io/v1beta1** API version will no longer be served in v1.25. + +* Migrate manifests and API clients to use the **node.k8s.io/v1** API version, available since v1.20. +* All existing persisted objects are accessible via the new API +* No notable changes + +### v1.22 + +The **v1.22** release will stop serving the following deprecated API versions: + +#### MutatingWebhookConfiguration and ValidatingWebhookConfiguration + +The **admissionregistration.k8s.io/v1beta1** API version of MutatingWebhookConfiguration and ValidatingWebhookConfiguration will no longer be served in v1.22. + +* Migrate manifests and API clients to use the **admissionregistration.k8s.io/v1** API version, available since v1.16. +* All existing persisted objects are accessible via the new API +* Notable changes: + * `webhooks[*].failurePolicy` default changed from `Ignore` to `Fail` for v1 + * `webhooks[*].matchPolicy` default changed from `Exact` to `Equivalent` for v1 + * `webhooks[*].timeoutSeconds` default changed from `30s` to `10s` for v1 + * `webhooks[*].sideEffects` default value is removed, and the field made required, and only `None` and `NoneOnDryRun` are permitted for v1 + * `webhooks[*].admissionReviewVersions` default value is removed and the field made required for v1 (supported versions for AdmissionReview are `v1` and `v1beta1`) + * `webhooks[*].name` must be unique in the list for objects created via `admissionregistration.k8s.io/v1` + +#### CustomResourceDefinitions + +The **apiextensions.k8s.io/v1beta1** API version of CustomResourceDefinition will no longer be served in v1.22. + +* Migrate manifests and API clients to use the **apiextensions.k8s.io/v1beta1** API version, available since v1.16. +* All existing persisted objects are accessible via the new API +* Notable changes: + * `spec.scope` is no longer defaulted to `Namespaced` and must be explicitly specified + * `spec.version` is removed in v1; use `spec.versions` instead + * `spec.validation` is removed in v1; use `spec.versions[*].schema` instead + * `spec.subresources` is removed in v1; use `spec.versions[*].subresources` instead + * `spec.additionalPrinterColumns` is removed in v1; use `spec.versions[*].additionalPrinterColumns` instead + * `spec.conversion.webhookClientConfig` is moved to `spec.conversion.webhook.clientConfig` in v1 + * `spec.conversion.conversionReviewVersions` is moved to `spec.conversion.webhook.conversionReviewVersions` in v1 + * `spec.versions[*].schema.openAPIV3Schema` is now required when creating v1 CustomResourceDefinitions, and must be a [structural schema](/docs/tasks/extend-kubernetes/custom-resources/custom-resource-definitions/#specifying-a-structural-schema) + * `spec.preserveUnknownFields: true` is disallowed when creating v1 CustomResourceDefinitions; it must be specified within schema definitions as `x-kubernetes-preserve-unknown-fields: true` + * In `additionalPrinterColumns` items, the `JSONPath` field was renamed to `jsonPath` in v1 (fixes [#66531](https://github.com/kubernetes/kubernetes/issues/66531)) + +#### APIService + +The **apiregistration.k8s.io/v1beta1** API version of APIService will no longer be served in v1.22. + +* Migrate manifests and API clients to use the **apiregistration.k8s.io/v1** API version, available since v1.10. +* All existing persisted objects are accessible via the new API +* No notable changes + +#### TokenReview + +The **authentication.k8s.io/v1beta1** API version of TokenReview will no longer be served in v1.22. + +* Migrate manifests and API clients to use the **authentication.k8s.io/v1** API version, available since v1.6. +* No notable changes + +#### SubjectAccessReview + +The **authorization.k8s.io/v1beta1** API version of LocalSubjectAccessReview, SelfSubjectAccessReview, and SubjectAccessReview will no longer be served in v1.22. + +* Migrate manifests and API clients to use the **authorization.k8s.io/v1** API version, available since v1.6. +* Notable changes: + * `spec.group` was renamed to `spec.groups` in v1 (fixes [#32709](https://github.com/kubernetes/kubernetes/issues/32709)) + +#### CertificateSigningRequest + +The **certificates.k8s.io/v1beta1** API version of CertificateSigningRequest will no longer be served in v1.22. + +* Migrate manifests and API clients to use the **certificates.k8s.io/v1** API version, available since v1.19. +* All existing persisted objects are accessible via the new API +* Notable changes in `certificates.k8s.io/v1`: + * For API clients requesting certificates: + * `spec.signerName` is now required (see [known Kubernetes signers](https://kubernetes.io/docs/reference/access-authn-authz/certificate-signing-requests/#kubernetes-signers)), and requests for `kubernetes.io/legacy-unknown` are not allowed to be created via the `certificates.k8s.io/v1` API + * `spec.usages` is now required, may not contain duplicate values, and must only contain known usages + * For API clients approving or signing certificates: + * `status.conditions` may not contain duplicate types + * `status.conditions[*].status` is now required + * `status.certificate` must be PEM-encoded, and contain only `CERTIFICATE` blocks + +#### Lease + +The **coordination.k8s.io/v1beta1** API version of Lease will no longer be served in v1.22. + +* Migrate manifests and API clients to use the **coordination.k8s.io/v1** API version, available since v1.14. +* All existing persisted objects are accessible via the new API +* No notable changes + +#### Ingress + +The **extensions/v1beta1** and **networking.k8s.io/v1beta1** API versions of Ingress will no longer be served in v1.22. + +* Migrate manifests and API clients to use the **networking.k8s.io/v1** API version, available since v1.19. +* All existing persisted objects are accessible via the new API +* Notable changes: + * `spec.backend` is renamed to `spec.defaultBackend` + * The backend `serviceName` field is renamed to `service.name` + * Numeric backend `servicePort` fields are renamed to `service.port.number` + * String backend `servicePort` fields are renamed to `service.port.name` + * `pathType` is now required for each specified path. Options are `Prefix`, `Exact`, and `ImplementationSpecific`. To match the undefined `v1beta1` behavior, use `ImplementationSpecific`. + +#### IngressClass + +The **networking.k8s.io/v1beta1** API version of IngressClass will no longer be served in v1.22. + +* Migrate manifests and API clients to use the **networking.k8s.io/v1** API version, available since v1.19. +* All existing persisted objects are accessible via the new API +* No notable changes + +#### RBAC + +The **rbac.authorization.k8s.io/v1beta1** API version of ClusterRole, ClusterRoleBinding, Role, and RoleBinding will no longer be served in v1.22. + +* Migrate manifests and API clients to use the **networking.k8s.io/v1** API version, available since v1.8. +* All existing persisted objects are accessible via the new API +* No notable changes + +#### PriorityClass + +The **scheduling.k8s.io/v1beta1** API version of PriorityClass will no longer be served in v1.22. + +* Migrate manifests and API clients to use the **scheduling.k8s.io/v1** API version, available since v1.14. +* All existing persisted objects are accessible via the new API +* No notable changes + +#### Storage + +The **storage.k8s.io/v1beta1** API version of CSIDriver, CSINode, StorageClass, and VolumeAttachment will no longer be served in v1.22. + +* Migrate manifests and API clients to use the **storage.k8s.io/v1** API version + * CSIDriver is available in **storage.k8s.io/v1** since v1.19. + * CSINode is available in **storage.k8s.io/v1** since v1.17 + * StorageClass is available in **storage.k8s.io/v1** since v1.6 + * VolumeAttachment is available in **storage.k8s.io/v1** v1.13 +* All existing persisted objects are accessible via the new API +* No notable changes + +### v1.16 + +The **v1.16** release stopped serving the following deprecated API versions: + +#### NetworkPolicy + +The **extensions/v1beta1** API version of NetworkPolicy is no longer served as of v1.16. + +* Migrate manifests and API clients to use the **networking.k8s.io/v1** API version, available since v1.8. +* All existing persisted objects are accessible via the new API + +#### DaemonSet + +The **extensions/v1beta1** and **apps/v1beta2** API versions of DaemonSet are no longer served as of v1.16. + +* Migrate manifests and API clients to use the **apps/v1** API version, available since v1.9. +* All existing persisted objects are accessible via the new API +* Notable changes: + * `spec.templateGeneration` is removed + * `spec.selector` is now required and immutable after creation; use the existing template labels as the selector for seamless upgrades + * `spec.updateStrategy.type` now defaults to `RollingUpdate` (the default in `extensions/v1beta1` was `OnDelete`) + +#### Deployment + +The **extensions/v1beta1**, **apps/v1beta1**, and **apps/v1beta2** API versions of Deployment are no longer served as of v1.16. + +* Migrate manifests and API clients to use the **apps/v1** API version, available since v1.9. +* All existing persisted objects are accessible via the new API +* Notable changes: + * `spec.rollbackTo` is removed + * `spec.selector` is now required and immutable after creation; use the existing template labels as the selector for seamless upgrades + * `spec.progressDeadlineSeconds` now defaults to `600` seconds (the default in `extensions/v1beta1` was no deadline) + * `spec.revisionHistoryLimit` now defaults to `10` (the default in `apps/v1beta1` was `2`, the default in `extensions/v1beta1` was to retain all) + * `maxSurge` and `maxUnavailable` now default to `25%` (the default in `extensions/v1beta1` was `1`) + +#### StatefulSet + +The **apps/v1beta1** and **apps/v1beta2** API versions of StatefulSet are no longer served as of v1.16. + +* Migrate manifests and API clients to use the **apps/v1** API version, available since v1.9. +* All existing persisted objects are accessible via the new API +* Notable changes: + * `spec.selector` is now required and immutable after creation; use the existing template labels as the selector for seamless upgrades + * `spec.updateStrategy.type` now defaults to `RollingUpdate` (the default in `apps/v1beta1` was `OnDelete`) + +#### ReplicaSet + +The **extensions/v1beta1**, **apps/v1beta1**, and **apps/v1beta2** API versions of ReplicaSet are no longer served as of v1.16. + +* Migrate manifests and API clients to use the **apps/v1** API version, available since v1.9. +* All existing persisted objects are accessible via the new API +* Notable changes: + * `spec.selector` is now required and immutable after creation; use the existing template labels as the selector for seamless upgrades + +## What To Do + +Kubernetes 1.22 will be released later in 2021, so be sure to audit +your configuration and integrations now! + +### Test with deprecated APIs disabled + +You can test your clusters by starting an API server with specific API versions disabled +to simulate upcoming removals. Add the following flag to the API server startup arguments: + +`--runtime-config=$group/$version=false` + +For example: + +`--runtime-config=admissionregistration.k8s.io/v1beta1=false,apiextensions.k8s.io/v1beta1,...` + +### Locate use of deprecated APIs + +Use [client warnings, metrics, and audit information available in 1.19+](https://kubernetes.io/blog/2020/09/03/warnings/#deprecation-warnings) +to locate use of deprecated APIs. + +### Migrate to non-deprecated APIs + +* Update custom integrations and controllers to call the non-deprecated APIs +* Change YAML files to reference the non-deprecated APIs + + You can use the `kubectl-convert` command (`kubectl convert` prior to v1.20) + to automatically convert an existing object: + + `kubectl-convert -f --output-version /`. + + For example, to convert an older Deployment to `apps/v1`, you can run: + + `kubectl-convert -f ./my-deployment.yaml --output-version apps/v1` + + Note that this may use non-ideal default values. To learn more about a specific + resource, check the Kubernetes [api reference](https://kubernetes.io/docs/reference/#api-reference). From c44f397eb85650c8891078002513f6197a44abe6 Mon Sep 17 00:00:00 2001 From: Arnaud Lemaire Date: Sat, 13 Feb 2021 23:19:44 +0100 Subject: [PATCH 17/62] Update control-plane-flags.md --- .../production-environment/tools/kubeadm/control-plane-flags.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/content/en/docs/setup/production-environment/tools/kubeadm/control-plane-flags.md b/content/en/docs/setup/production-environment/tools/kubeadm/control-plane-flags.md index 1bcdad0092..1dd44e9b0b 100644 --- a/content/en/docs/setup/production-environment/tools/kubeadm/control-plane-flags.md +++ b/content/en/docs/setup/production-environment/tools/kubeadm/control-plane-flags.md @@ -78,7 +78,7 @@ kind: ClusterConfiguration kubernetesVersion: v1.16.0 scheduler: extraArgs: - address: 0.0.0.0 + bind-address: 0.0.0.0 config: /home/johndoe/schedconfig.yaml kubeconfig: /home/johndoe/kubeconfig.yaml ``` From 0477c9abbcf597a9b7b4d1ccfdb9739d00723793 Mon Sep 17 00:00:00 2001 From: Minchao Date: Sun, 14 Feb 2021 09:10:41 +0800 Subject: [PATCH 18/62] [zh] Sync taint and toleration example --- .../docs/concepts/scheduling-eviction/taint-and-toleration.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/content/zh/docs/concepts/scheduling-eviction/taint-and-toleration.md b/content/zh/docs/concepts/scheduling-eviction/taint-and-toleration.md index 430f0d60b6..e20c15890f 100644 --- a/content/zh/docs/concepts/scheduling-eviction/taint-and-toleration.md +++ b/content/zh/docs/concepts/scheduling-eviction/taint-and-toleration.md @@ -67,7 +67,7 @@ kubectl taint nodes node1 key1=value1:NoSchedule- 若要移除上述命令所添加的污点，你可以执行： ```shell -kubectl taint nodes node1 key:NoSchedule- +kubectl taint nodes node1 key1=value1:NoSchedule- ``` -## GitHub 发布版本 - -Kompose 通过 GitHub 发布版本，发布周期为三星期。 +Kompose 通过 GitHub 发布，发布周期为三星期。 你可以在 [GitHub 发布页面](https://github.com/kubernetes/kompose/releases) 上看到所有当前版本。 ```shell # Linux -curl -L https://github.com/kubernetes/kompose/releases/download/v1.16.0/kompose-linux-amd64 -o kompose +curl -L https://github.com/kubernetes/kompose/releases/download/v1.22.0/kompose-linux-amd64 -o kompose # macOS -curl -L https://github.com/kubernetes/kompose/releases/download/v1.16.0/kompose-darwin-amd64 -o kompose +curl -L https://github.com/kubernetes/kompose/releases/download/v1.22.0/kompose-darwin-amd64 -o kompose # Windows -curl -L https://github.com/kubernetes/kompose/releases/download/v1.16.0/kompose-windows-amd64.exe -o kompose.exe +curl -L https://github.com/kubernetes/kompose/releases/download/v1.22.0/kompose-windows-amd64.exe -o kompose.exe chmod +x kompose sudo mv ./kompose /usr/local/bin/kompose @@ -68,9 +67,10 @@ sudo mv ./kompose /usr/local/bin/kompose -或者，你可以下载 [tarball](https://github.com/kubernetes/kompose/releases)。 +或者，你可以下载 [tar 包](https://github.com/kubernetes/kompose/releases)。 -## Go +{{% /tab %}} +{{% tab name="基于源代码构建" %}} @@ -135,129 +141,139 @@ you need is an existing `docker-compose.yml` file. 再需几步，我们就把你从 Docker Compose 带到 Kubernetes。 你只需要一个现有的 `docker-compose.yml` 文件。 -1. - 进入 `docker-compose.yml` 文件所在的目录。如果没有，请使用下面这个进行测试。 +1. + 进入 `docker-compose.yml` 文件所在的目录。如果没有，请使用下面这个进行测试。 - ```yaml - version: "2" + ```yaml + version: "2" - services: + services: - redis-master: - image: k8s.gcr.io/redis:e2e - ports: - - "6379" + redis-master: + image: k8s.gcr.io/redis:e2e + ports: + - "6379" - redis-slave: - image: gcr.io/google_samples/gb-redisslave:v3 - ports: - - "6379" - environment: - - GET_HOSTS_FROM=dns + redis-slave: + image: gcr.io/google_samples/gb-redisslave:v3 + ports: + - "6379" + environment: + - GET_HOSTS_FROM=dns - frontend: - image: gcr.io/google-samples/gb-frontend:v4 - ports: - - "80:80" - environment: - - GET_HOSTS_FROM=dns - labels: - kompose.service.type: LoadBalancer - ``` + frontend: + image: gcr.io/google-samples/gb-frontend:v4 + ports: + - "80:80" + environment: + - GET_HOSTS_FROM=dns + labels: + kompose.service.type: LoadBalancer + ``` -2. - 运行 `kompose up` 命令直接部署到 Kubernetes，或者跳到下一步，生成 `kubectl` 使用的文件。 + +2. 要将 `docker-compose.yml` 转换为 `kubectl` 可用的文件，请运行 `kompose convert` + 命令进行转换，然后运行 `kubectl create -f ` 进行创建。 - ```bash - $ kompose up - We are going to create Kubernetes Deployments, Services and PersistentVolumeClaims for your Dockerized application. - If you need different kind of resources, use the 'kompose convert' and 'kubectl create -f' commands instead. + ```shell + kompose convert + ``` - INFO Successfully created Service: redis - INFO Successfully created Service: web - INFO Successfully created Deployment: redis - INFO Successfully created Deployment: web + ```none + INFO Kubernetes file "frontend-service.yaml" created + INFO Kubernetes file "frontend-service.yaml" created + INFO Kubernetes file "frontend-service.yaml" created + INFO Kubernetes file "redis-master-service.yaml" created + INFO Kubernetes file "redis-master-service.yaml" created + INFO Kubernetes file "redis-master-service.yaml" created + INFO Kubernetes file "redis-slave-service.yaml" created + INFO Kubernetes file "redis-slave-service.yaml" created + INFO Kubernetes file "redis-slave-service.yaml" created + INFO Kubernetes file "frontend-deployment.yaml" created + INFO Kubernetes file "frontend-deployment.yaml" created + INFO Kubernetes file "frontend-deployment.yaml" created + INFO Kubernetes file "redis-master-deployment.yaml" created + INFO Kubernetes file "redis-master-deployment.yaml" created + INFO Kubernetes file "redis-master-deployment.yaml" created + INFO Kubernetes file "redis-slave-deployment.yaml" created + INFO Kubernetes file "redis-slave-deployment.yaml" created + INFO Kubernetes file "redis-slave-deployment.yaml" created + ``` - Your application has been deployed to Kubernetes. You can run 'kubectl get deployment,svc,pods,pvc' for details. - ``` + ```bash + kubectl apply -f frontend-service.yaml,redis-master-service.yaml,redis-slave-service.yaml,frontend-deployment.yaml, + ``` -3. - 要将 `docker-compose.yml` 转换为 `kubectl` 可用的文件，请运行 `kompose convert` 命令进行转换， - 然后运行 `kubectl create -f ` 进行创建。 + + 输出类似于： - ```shell - kompose convert - ``` + ```none + service/frontend created + service/redis-master created + service/redis-slave created + deployment.apps/frontend created + deployment.apps/redis-master created + deployment.apps/redis-slave created + ``` - ``` - INFO Kubernetes file "frontend-service.yaml" created - INFO Kubernetes file "redis-master-service.yaml" created - INFO Kubernetes file "redis-slave-service.yaml" created - INFO Kubernetes file "frontend-deployment.yaml" created - INFO Kubernetes file "redis-master-deployment.yaml" created - INFO Kubernetes file "redis-slave-deployment.yaml" created - ``` + + 你部署的应用在 Kubernetes 中运行起来了。 - ```shell - kubectl create -f frontend-service.yaml,redis-master-service.yaml,redis-slave-service.yaml,frontend-deployment.yaml,redis-master-deployment.yaml,redis-slave-deployment.yaml - ``` + +3. 访问你的应用 - ``` - service/frontend created - service/redis-master created - service/redis-slave created - deployment.apps/frontend created - deployment.apps/redis-master created - deployment.apps/redis-slave created - ``` + - - 你部署的应用在 Kubernetes 中运行起来了。 + 如果你在开发过程中使用 `minikube`，请执行： -4. - 访问你的应用 + ```shell + minikube service frontend + ``` - + + 否则，我们要查看一下你的服务使用了什么 IP！ - 如果你在开发过程中使用 `minikube`，请执行： + ```shell + kubectl describe svc frontend + ``` - ```shell - minikube service frontend - ``` + ```none + Name: frontend + Namespace: default + Labels: service=frontend + Selector: service=frontend + Type: LoadBalancer + IP: 10.0.0.183 + LoadBalancer Ingress: 192.0.2.89 + Port: 80 80/TCP + NodePort: 80 31144/TCP + Endpoints: 172.17.0.4:80 + Session Affinity: None + No events. + ``` - - 否则，我们要查看一下你的服务使用了什么 IP！ + + 如果你使用的是云提供商，你的 IP 将在 `LoadBalancer Ingress` 字段给出。 - ```shell - kubectl describe svc frontend - ``` - - ``` - Name: frontend - Namespace: default - Labels: service=frontend - Selector: service=frontend - Type: LoadBalancer - IP: 10.0.0.183 - LoadBalancer Ingress: 192.0.2.89 - Port: 80 80/TCP - NodePort: 80 31144/TCP - Endpoints: 172.17.0.4:80 - Session Affinity: None - No events. - ``` - - - 如果你使用的是云提供商，你的 IP 将在 `LoadBalancer Ingress` 字段给出。 - - ```shell - curl http://192.0.2.89 - ``` + ```shell + curl http://192.0.2.89 + ``` @@ -284,29 +300,37 @@ you need is an existing `docker-compose.yml` file. - [`kompose down`](#kompose-down) - 文档 - - [构建和推送 Docker 镜像](#构建和推送-docker-镜像) + - [构建和推送 Docker 镜像](#build-and-push-docker-images) - [其他转换方式](#其他转换方式) - - [标签](#标签) - - [重启](#重启) - - [Docker Compose 版本](#docker-compose-版本) + - [标签](#labels) + - [重启](#restart) + - [Docker Compose 版本](#docker-compose-versions) Kompose 支持两种驱动：OpenShift 和 Kubernetes。 -你可以通过全局选项 `--provider` 选择驱动方式。如果没有指定，会将 Kubernetes 作为默认驱动。 +你可以通过全局选项 `--provider` 选择驱动。如果没有指定， +会将 Kubernetes 作为默认驱动。 ## `kompose convert` + Kompose 支持将 V1、V2 和 V3 版本的 Docker Compose 文件转换为 Kubernetes 和 OpenShift 资源对象。 -### Kubernetes + +### Kubernetes `kompose convert` 示例 ```shell kompose --file docker-voting.yml convert ``` -``` + +```none WARN Unsupported key networks - ignoring WARN Unsupported key build - ignoring INFO Kubernetes file "worker-svc.yaml" created @@ -325,7 +349,7 @@ INFO Kubernetes file "db-deployment.yaml" created ls ``` -``` +```none db-deployment.yaml docker-compose.yml docker-gitlab.yml redis-deployment.yaml result-deployment.yaml vote-deployment.yaml worker-deployment.yaml db-svc.yaml docker-voting.yml redis-svc.yaml result-svc.yaml vote-svc.yaml worker-svc.yaml ``` @@ -338,7 +362,8 @@ You can also provide multiple docker-compose files at the same time: ```shell kompose -f docker-compose.yml -f docker-guestbook.yml convert ``` -``` + +```none INFO Kubernetes file "frontend-service.yaml" created INFO Kubernetes file "mlbparks-service.yaml" created INFO Kubernetes file "mongodb-service.yaml" created @@ -368,7 +393,10 @@ When multiple docker-compose files are provided the configuration is merged. Any --> 当提供多个 docker-compose 文件时，配置将会合并。任何通用的配置都将被后续文件覆盖。 -### OpenShift + +### OpenShift `kompose convert` 示例 ```shell kompose --provider openshift --file docker-voting.yml convert @@ -403,7 +431,7 @@ kompose 还支持为服务中的构建指令创建 buildconfig。 kompose --provider openshift --file buildconfig/docker-compose.yml convert ``` -``` +```none WARN [foo] Service cannot be created because of missing port. INFO OpenShift Buildconfig using git@github.com:rtnpro/kompose.git::master as source. INFO OpenShift file "foo-deploymentconfig.yaml" created @@ -424,15 +452,19 @@ imagestream 工件，以解决 Openshift 的这个问题：https://github.com/op -Kompose 支持通过 `kompose up` 直接将你的"复合的（composed）" 应用程序部署到 Kubernetes 或 OpenShift。 +Kompose 支持通过 `kompose up` 直接将你的"复合的（composed）" 应用程序 +部署到 Kubernetes 或 OpenShift。 -### Kubernetes + +### Kubernetes `kompose up` 示例 ```shell kompose --file ./examples/docker-guestbook.yml up ``` -``` +```none We are going to create Kubernetes deployments and services for your Dockerized application. If you need different kind of resources, use the 'kompose convert' and 'kubectl create -f' commands instead. @@ -468,26 +500,27 @@ pod/redis-master-1432129712-63jn8 1/1 Running 0 4m pod/redis-slave-2504961300-nve7b 1/1 Running 0 4m ``` + +{{< note >}} - -**注意**： - - 你必须有一个运行正常的 Kubernetes 集群，该集群具有预先配置的 kubectl 上下文。 - 此操作仅生成 Deployment 和 Service 对象并将其部署到 Kubernetes。 如果需要部署其他不同类型的资源，请使用 `kompose convert` 和 `kubectl create -f` 命令。 +{{< /note >}} - -### OpenShift + +### OpenShift `kompose up` 示例 ```shell kompose --file ./examples/docker-guestbook.yml --provider openshift up ``` -``` +```none We are going to create OpenShift DeploymentConfigs and Services for your Dockerized application. If you need different kind of resources, use the 'kompose convert' and 'oc create -f' commands instead. @@ -508,7 +541,7 @@ Your application has been deployed to OpenShift. You can run 'oc get dc,svc,is' oc get dc,svc,is ``` -``` +```none NAME REVISION DESIRED CURRENT TRIGGERED BY dc/frontend 0 1 0 config,image(frontend:v4) dc/redis-master 0 1 0 config,image(redis-master:e2e) @@ -523,20 +556,18 @@ is/redis-master 172.30.12.200:5000/fff/redis-master is/redis-slave 172.30.12.200:5000/fff/redis-slave v1 ``` +{{< note >}} -**注意**： - -- 你必须有一个运行正常的 OpenShift 集群，该集群具有预先配置的 `oc` 上下文 (`oc login`)。 +你必须有一个运行正常的 OpenShift 集群，该集群具有预先配置的 `oc` 上下文 (`oc login`)。 +{{< /note >}} ## `kompose down` - 你一旦将"复合(composed)" 应用部署到 Kubernetes，`kompose down` 命令将能帮你通过删除 Deployment 和 Service 对象来删除应用。 如果需要删除其他资源，请使用 'kubectl' 命令。 @@ -554,26 +585,27 @@ INFO Successfully deleted service: frontend INFO Successfully deleted deployment: frontend ``` +{{< note >}} +- 你必须有一个运行正常的 Kubernetes 集群，该集群具有预先配置的 kubectl 上下文。 +{{< /note >}} + +## 构建和推送 Docker 镜像 {#build-and-push-docker-images} -**注意**： - -- 你必须有一个运行正常的 Kubernetes 集群，该集群具有预先配置的 kubectl 上下文。 - -## 构建和推送 Docker 镜像 - -Kompose 支持构建和推送 Docker 镜像。如果 Docker Compose 文件中使用了 `build` 关键字，你的镜像将会： +Kompose 支持构建和推送 Docker 镜像。如果 Docker Compose 文件中使用了 `build` +关键字，你的镜像将会： - 使用文档中指定的 `image` 键自动构建 Docker 镜像 - 使用本地凭据推送到正确的 Docker 仓库 @@ -598,7 +630,7 @@ Using `kompose up` with a `build` key: kompose up ``` -``` +```none INFO Build key detected. Attempting to build and push image 'docker.io/foo/bar' INFO Building image 'docker.io/foo/bar' from directory 'build' INFO Image 'docker.io/foo/bar' from directory 'build' built successfully @@ -621,10 +653,10 @@ In order to disable the functionality, or choose to use BuildConfig generation ( 可以通过传递 `--build (local|build-config|none)` 参数来实现。 ```shell -# Disable building/pushing Docker images +# 禁止构造和推送 Docker 镜像 kompose up --build none -# Generate Build Config artifacts for OpenShift +# 为 OpenShift 生成 Build Config 工件 kompose up --provider openshift --build build-config ``` @@ -633,7 +665,7 @@ kompose up --provider openshift --build build-config The default `kompose` transformation will generate Kubernetes [Deployments](/docs/concepts/workloads/controllers/deployment/) and [Services](/docs/concepts/services-networking/service/), in yaml format. You have alternative option to generate json with `-j`. Also, you can alternatively generate [Replication Controllers](/docs/concepts/workloads/controllers/replicationcontroller/) objects, [Daemon Sets](/docs/concepts/workloads/controllers/daemonset/), or [Helm](https://github.com/helm/helm) charts. --> -## 其他转换方式 +## 其他转换方式 {#alternative-conversions} 默认的 `kompose` 转换会生成 yaml 格式的 Kubernetes [Deployment](/zh/docs/concepts/workloads/controllers/deployment/) 和 @@ -646,7 +678,8 @@ The default `kompose` transformation will generate Kubernetes [Deployments](/doc ```shell kompose convert -j ``` -``` + +```none INFO Kubernetes file "redis-svc.json" created INFO Kubernetes file "web-svc.json" created INFO Kubernetes file "redis-deployment.json" created @@ -661,7 +694,8 @@ The `*-deployment.json` files contain the Deployment objects. ```shell kompose convert --replication-controller ``` -``` + +```none INFO Kubernetes file "redis-svc.yaml" created INFO Kubernetes file "web-svc.yaml" created INFO Kubernetes file "redis-replicationcontroller.yaml" created @@ -671,7 +705,6 @@ INFO Kubernetes file "web-replicationcontroller.yaml" created - `*-replicationcontroller.yaml` 文件包含 Replication Controller 对象。 如果你想指定副本数（默认为 1），可以使用 `--replicas` 参数： `kompose convert --replication-controller --replicas 3` @@ -680,7 +713,7 @@ The `*-replicationcontroller.yaml` files contain the Replication Controller obje kompose convert --daemon-set ``` -``` +```none INFO Kubernetes file "redis-svc.yaml" created INFO Kubernetes file "web-svc.yaml" created INFO Kubernetes file "redis-daemonset.yaml" created @@ -688,17 +721,19 @@ INFO Kubernetes file "web-daemonset.yaml" created ``` -`*-daemonset.yaml` 文件包含 Daemon Set 对象。 +`*-daemonset.yaml` 文件包含 DaemonSet 对象。 -如果你想生成 [Helm](https://github.com/kubernetes/helm) 可用的 Chart，只需简单的执行下面的命令： +如果你想生成 [Helm](https://github.com/kubernetes/helm) 可用的 Chart， +只需简单的执行下面的命令： ```shell kompose convert -c ``` -``` + +```none INFO Kubernetes file "web-svc.yaml" created INFO Kubernetes file "redis-svc.yaml" created INFO Kubernetes file "web-deployment.yaml" created @@ -734,9 +769,10 @@ The chart structure is aimed at providing a skeleton for building your Helm char For example: --> -## 标签 +## 标签 {#labels} -`kompose` 支持 `docker-compose.yml` 文件中用于 Kompose 的标签，以便在转换时明确定义 Service 的行为。 +`kompose` 支持 `docker-compose.yml` 文件中用于 Kompose 的标签，以便 +在转换时明确定义 Service 的行为。 - `kompose.service.type` 定义要创建的 Service 类型。例如： @@ -761,11 +797,13 @@ For example: For example: --> - `kompose.service.expose` 定义是否允许从集群外部访问 Service。 - 如果该值被设置为 "true"，提供程序将自动设置端点，对于任何其他值，该值将被设置为主机名。 + 如果该值被设置为 "true"，提供程序将自动设置端点， + 对于任何其他值，该值将被设置为主机名。 如果在 Service 中定义了多个端口，则选择第一个端口作为公开端口。 - - 对于 Kubernetes 驱动程序，创建了一个 Ingress 资源，并且假定已经配置了相应的 Ingress 控制器。 - - 对于 OpenShift 驱动程序, 创建一个 route。 + - 如果使用 Kubernetes 驱动，会有一个 Ingress 资源被创建，并且假定 + 已经配置了相应的 Ingress 控制器。 + - 如果使用 OpenShift 驱动, 则会有一个 route 被创建。 例如： @@ -793,19 +831,18 @@ The currently supported options are: | kompose.service.type | nodeport / clusterip / loadbalancer | | kompose.service.expose| true / hostname | --> - 当前支持的选项有: -| 键 | 值 | -|----------------------|-------------------------------------| -| kompose.service.type | nodeport / clusterip / loadbalancer | -| kompose.service.expose| true / hostname | +| 键 | 值 | +|------------------------|-------------------------------------| +| kompose.service.type | nodeport / clusterip / loadbalancer | +| kompose.service.expose | true / hostname | +{{< note >}} -{{< note >}} -`kompose.service.type` 标签应该只用`ports`来定义，否则 `kompose` 会失败。 +`kompose.service.type` 标签应该只用 `ports` 来定义，否则 `kompose` 会失败。 {{< /note >}} -## 重启 +## 重启 {#restart} -如果你想创建没有控制器的普通 Pod，可以使用 docker-compose 的 `restart` 结构来定义它。 -请参考下表了解 `restart` 的不同参数。 +如果你想创建没有控制器的普通 Pod，可以使用 docker-compose 的 `restart` +结构来指定这一行为。请参考下表了解 `restart` 的不同参数。 -| `docker-compose` `restart` | 创建的对象 | Pod `restartPolicy` | +| `docker-compose` `restart` | 创建的对象 | Pod `restartPolicy` | |----------------------------|-------------------|---------------------| -| `""` | 控制器对象 | `Always` | -| `always` | 控制器对象 | `Always` | +| `""` | 控制器对象 | `Always` | +| `always` | 控制器对象 | `Always` | | `on-failure` | Pod | `OnFailure` | | `no` | Pod | `Never` | @@ -843,9 +880,9 @@ The controller object could be `deployment` or `replicationcontroller`, etc. {{< /note >}} -例如，`pival` Service 将在这里变成 Pod。这个容器的计算值为 `pi`。 +例如，`pival` Service 将在这里变成 Pod。这个容器计算 `pi` 的取值。 ```yaml version: '2' @@ -858,23 +895,22 @@ services: ``` - ### 关于 Deployment Config 的提醒 -如果 Docker Compose 文件中为服务声明了卷，Deployment (Kubernetes) 或 DeploymentConfig (OpenShift) -的策略会从 "RollingUpdate" (默认) 变为 "Recreate"。 +如果 Docker Compose 文件中为服务声明了卷，Deployment (Kubernetes) 或 +DeploymentConfig (OpenShift) 策略会从 "RollingUpdate" (默认) 变为 "Recreate"。 这样做的目的是为了避免服务的多个实例同时访问卷。 -如果 Docker Compose 文件中的服务名包含 `_` (例如 `web_service`)， -那么将会被替换为 `-`，服务也相应的会重命名(例如 `web-service`)。 +如果 Docker Compose 文件中的服务名包含 `_`（例如 `web_service`）， +那么将会被替换为 `-`，服务也相应的会重命名（例如 `web-service`）。 Kompose 这样做的原因是 "Kubernetes" 不允许对象名称中包含 `_`。 请注意，更改服务名称可能会破坏一些 `docker-compose` 文件。 @@ -883,14 +919,15 @@ Kompose 这样做的原因是 "Kubernetes" 不允许对象名称中包含 `_`。 ## Docker Compose Versions Kompose supports Docker Compose versions: 1, 2 and 3. We have limited support on versions 2.1 and 3.2 due to their experimental nature. + A full list on compatibility between all three versions is listed in our [conversion document](https://github.com/kubernetes/kompose/blob/master/docs/conversion.md) including a list of all incompatible Docker Compose keys. --> -## Docker Compose 版本 +## Docker Compose 版本 {#docker-compose-versions} -Kompose 支持的 Docker Compose 版本包括：1、2 和 3。有限支持 2.1 和 3.2 版本，因为它们还在实验阶段。 +Kompose 支持的 Docker Compose 版本包括：1、2 和 3。 +对 2.1 和 3.2 版本的支持还有限，因为它们还在实验阶段。 所有三个版本的兼容性列表请查看我们的 [转换文档](https://github.com/kubernetes/kompose/blob/master/docs/conversion.md)， 文档中列出了所有不兼容的 Docker Compose 关键字。 - From 0ac2e87d8758c07b4bb1fe61cf69076a639f1f35 Mon Sep 17 00:00:00 2001 From: Xianglin Gao Date: Sun, 14 Feb 2021 16:35:35 +0800 Subject: [PATCH 20/62] Update flow-control.md tiny for how to enable flowcontrol.apiserver.k8s.io/v1alpha1 --- content/en/docs/concepts/cluster-administration/flow-control.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/content/en/docs/concepts/cluster-administration/flow-control.md b/content/en/docs/concepts/cluster-administration/flow-control.md index 6cb4c386d3..e8fd3e4061 100644 --- a/content/en/docs/concepts/cluster-administration/flow-control.md +++ b/content/en/docs/concepts/cluster-administration/flow-control.md @@ -59,7 +59,7 @@ kube-apiserver \ ``` Alternatively, you can enable the v1alpha1 version of the API group -with `--runtime-config=flowcontrol.apiserver.k8s.io/v1beta1=true`. +with `--runtime-config=flowcontrol.apiserver.k8s.io/v1alpha1=true`. The command-line flag `--enable-priority-and-fairness=false` will disable the API Priority and Fairness feature, even if other flags have enabled it. From 9ee0e9c07719243c227368c5d07fbffeb93f1eb2 Mon Sep 17 00:00:00 2001 From: Qiming Teng Date: Sun, 14 Feb 2021 20:42:47 +0800 Subject: [PATCH 21/62] [zh] Sync changes that remove logging solutions --- .../cluster-administration/logging.md | 290 ++++---- .../events-stackdriver.md | 156 ----- .../logging-elasticsearch-kibana.md | 197 ------ .../stateful-application/zookeeper.md | 652 ++++++++++-------- 4 files changed, 489 insertions(+), 806 deletions(-) delete mode 100644 content/zh/docs/tasks/debug-application-cluster/events-stackdriver.md delete mode 100644 content/zh/docs/tasks/debug-application-cluster/logging-elasticsearch-kibana.md diff --git a/content/zh/docs/concepts/cluster-administration/logging.md b/content/zh/docs/concepts/cluster-administration/logging.md index 2a820144be..44689dcac3 100644 --- a/content/zh/docs/concepts/cluster-administration/logging.md +++ b/content/zh/docs/concepts/cluster-administration/logging.md @@ -14,44 +14,43 @@ weight: 60 应用日志可以让你了解应用内部的运行状况。日志对调试问题和监控集群活动非常有用。 -大部分现代化应用都有某种日志记录机制；同样地，大多数容器引擎也被设计成支持某种日志记录机制。 -针对容器化应用，最简单且受欢迎的日志记录方式就是写入标准输出和标准错误流。 +大部分现代化应用都有某种日志记录机制。同样地，容器引擎也被设计成支持日志记录。 +针对容器化应用，最简单且最广泛采用的日志记录方式就是写入标准输出和标准错误流。 -但是，由容器引擎或运行时提供的原生功能通常不足以满足完整的日志记录方案。 -例如，如果发生容器崩溃、Pod 被逐出或节点宕机等情况，你仍然想访问到应用日志。 -因此，日志应该具有独立的存储和生命周期，与节点、Pod 或容器的生命周期相独立。 -这个概念叫 _集群级的日志_ 。集群级日志方案需要一个独立的后台来存储、分析和查询日志。 -Kubernetes 没有为日志数据提供原生存储方案，但是你可以集成许多现有的日志解决方案到 Kubernetes 集群中。 +但是，由容器引擎或运行时提供的原生功能通常不足以构成完整的日志记录方案。 +例如，如果发生容器崩溃、Pod 被逐出或节点宕机等情况，你可能想访问应用日志。 +在集群中，日志应该具有独立的存储和生命周期，与节点、Pod 或容器的生命周期相独立。 +这个概念叫 _集群级的日志_ 。 -集群级日志架构假定在集群内部或者外部有一个日志后台。 -如果你对集群级日志不感兴趣，你仍会发现关于如何在节点上存储和处理日志的描述对你是有用的。 +集群级日志架构需要一个独立的后端用来存储、分析和查询日志。 +Kubernetes 并不为日志数据提供原生的存储解决方案。 +相反，有很多现成的日志方案可以集成到 Kubernetes 中. +下面各节描述如何在节点上处理和存储日志。 ## Kubernetes 中的基本日志记录 -本节，你会看到一个kubernetes 中生成基本日志的例子，该例子中数据被写入到标准输出。 -这里的示例为包含一个容器的 Pod 规约，该容器每秒钟向标准输出写入数据。 +这里的示例使用包含一个容器的 Pod 规约，每秒钟向标准输出写入数据。 {{< codenew file="debug/counter-pod.yaml" >}} @@ -76,7 +75,7 @@ pod/counter created -使用 `kubectl logs` 命令获取日志: +像下面这样，使用 `kubectl logs` 命令获取日志: ```shell kubectl logs counter @@ -95,10 +94,10 @@ The output is: ``` -一旦发生容器崩溃，你可以使用命令 `kubectl logs` 和参数 `--previous` 检索之前的容器日志。 -如果 pod 中有多个容器，你应该向该命令附加一个容器名以访问对应容器的日志。 +你可以使用命令 `kubectl logs --previous` 检索之前容器实例的日志。 +如果 Pod 中有多个容器，你应该为该命令附加容器名以访问对应容器的日志。 详见 [`kubectl logs` 文档](/docs/reference/generated/kubectl/kubectl-commands#logs)。 容器化应用写入 `stdout` 和 `stderr` 的任何数据，都会被容器引擎捕获并被重定向到某个位置。 例如，Docker 容器引擎将这两个输出流重定向到某个 -[日志驱动](https://docs.docker.com/engine/admin/logging/overview) ， +[日志驱动（Logging Driver）](https://docs.docker.com/engine/admin/logging/overview) ， 该日志驱动在 Kubernetes 中配置为以 JSON 格式写入文件。 -节点级日志记录中，需要重点考虑实现日志的轮转，以此来保证日志不会消耗节点上所有的可用空间。 -Kubernetes 当前并不负责轮转日志，而是通过部署工具建立一个解决问题的方案。 -例如，在 Kubernetes 集群中，用 `kube-up.sh` 部署一个每小时运行的工具 -[`logrotate`](https://linux.die.net/man/8/logrotate)。 -你也可以设置容器 runtime 来自动地轮转应用日志，比如使用 Docker 的 `log-opt` 选项。 -在 `kube-up.sh` 脚本中，使用后一种方式来处理 GCP 上的 COS 镜像，而使用前一种方式来处理其他环境。 -这两种方式，默认日志超过 10MB 大小时都会触发日志轮转。 +节点级日志记录中，需要重点考虑实现日志的轮转，以此来保证日志不会消耗节点上全部可用空间。 +Kubernetes 并不负责轮转日志，而是通过部署工具建立一个解决问题的方案。 +例如，在用 `kube-up.sh` 部署的 Kubernetes 集群中，存在一个 +[`logrotate`](https://linux.die.net/man/8/logrotate)，每小时运行一次。 +你也可以设置容器运行时来自动地轮转应用日志。 例如，你可以找到关于 `kube-up.sh` 为 GCP 环境的 COS 镜像设置日志的详细信息， -相应的脚本在 -[这里](https://github.com/kubernetes/kubernetes/blob/{{< param "githubbranch" >}}/cluster/gce/gci/configure-helper.sh)。 +脚本为 +[`configure-helper` 脚本](https://github.com/kubernetes/kubernetes/blob/{{< param "githubbranch" >}}/cluster/gce/gci/configure-helper.sh)。 当运行 [`kubectl logs`](/docs/reference/generated/kubectl/kubectl-commands#logs) 时， 节点上的 kubelet 处理该请求并直接读取日志文件，同时在响应中返回日志文件内容。 {{< note >}} -当前，如果有其他系统机制执行日志轮转，那么 `kubectl logs` 仅可查询到最新的日志内容。 -比如，一个 10MB 大小的文件，通过`logrotate` 执行轮转后生成两个文件，一个 10MB 大小， -一个为空，所以 `kubectl logs` 将返回空。 +如果有外部系统执行日志轮转，那么 `kubectl logs` 仅可查询到最新的日志内容。 +比如，对于一个 10MB 大小的文件，通过 `logrotate` 执行轮转后生成两个文件， +一个 10MB 大小，一个为空，`kubectl logs` 返回最新的日志文件，而该日志文件 +在这个例子中为空。 {{< /note >}} * 在容器中运行的 kube-scheduler 和 kube-proxy。 -* 不在容器中运行的 kubelet 和容器运行时（例如 Docker）。 +* 不在容器中运行的 kubelet 和容器运行时。 -在使用 systemd 机制的服务器上，kubelet 和容器 runtime 写入日志到 journald。 -如果没有 systemd，他们写入日志到 `/var/log` 目录的 `.log` 文件。 -容器中的系统组件通常将日志写到 `/var/log` 目录，绕过了默认的日志机制。他们使用 -[klog](https://github.com/kubernetes/klog) 日志库。 -你可以在[日志开发文档](https://github.com/kubernetes/community/blob/master/contributors/devel/sig-instrumentation/logging.md)找到这些组件的日志告警级别协议。 +在使用 systemd 机制的服务器上，kubelet 和容器容器运行时将日志写入到 journald 中。 +如果没有 systemd，它们将日志写入到 `/var/log` 目录下的 `.log` 文件中。 +容器中的系统组件通常将日志写到 `/var/log` 目录，绕过了默认的日志机制。 +他们使用 [klog](https://github.com/kubernetes/klog) 日志库。 +你可以在[日志开发文档](https://github.com/kubernetes/community/blob/master/contributors/devel/sig-instrumentation/logging.md) +找到这些组件的日志告警级别约定。 和容器日志类似，`/var/log` 目录中的系统组件日志也应该被轮转。 -通过脚本 `kube-up.sh` 启动的 Kubernetes 集群中，日志被工具 `logrotate` 执行每日轮转， -或者日志大小超过 100MB 时触发轮转。 +通过脚本 `kube-up.sh` 启动的 Kubernetes 集群中，日志被工具 `logrotate` +执行每日轮转，或者日志大小超过 100MB 时触发轮转。 ## 集群级日志架构 -虽然Kubernetes没有为集群级日志记录提供原生的解决方案，但你可以考虑几种常见的方法。以下是一些选项： +虽然Kubernetes没有为集群级日志记录提供原生的解决方案，但你可以考虑几种常见的方法。 +以下是一些选项： * 使用在每个节点上运行的节点级日志记录代理。 -* 在应用程序的 pod 中，包含专门记录日志的 sidecar 容器。 +* 在应用程序的 Pod 中，包含专门记录日志的边车（Sidecar）容器。 * 将日志直接从应用程序中推送到日志记录后端。 -由于日志记录代理必须在每个节点上运行，它可以用 DaemonSet 副本，Pod 或 本机进程来实现。 -然而，后两种方法被弃用并且非常不别推荐。 +由于日志记录代理必须在每个节点上运行，通常可以用 `DaemonSet` 的形式运行该代理。 +节点级日志在每个节点上仅创建一个代理，不需要对节点上的应用做修改。 -对于 Kubernetes 集群来说，使用节点级的日志代理是最常用和被推荐的方式， -因为在每个节点上仅创建一个代理，并且不需要对节点上的应用做修改。 -但是，节点级的日志 _仅适用于应用程序的标准输出和标准错误输出_。 +容器向标准输出和标准错误输出写出数据，但在格式上并不统一。 +节点级代理 +收集这些日志并将其进行转发以完成汇总。 -Kubernetes 并不指定日志代理，但是有两个可选的日志代理与 Kubernetes 发行版一起发布。 -[Stackdriver 日志](/zh/docs/tasks/debug-application-cluster/logging-stackdriver/) -适用于 Google Cloud Platform，和 -[Elasticsearch](/zh/docs/tasks/debug-application-cluster/logging-elasticsearch-kibana/)。 -你可以在专门的文档中找到更多的信息和说明。 -两者都使用 [fluentd](https://www.fluentd.org/) 与自定义配置作为节点上的代理。 - - -### 使用 sidecar 容器和日志代理 +### 使用 sidecar 容器运行日志代理 {#sidecar-container-with-logging-agent} -你可以通过以下方式之一使用 sidecar 容器： +你可以通过以下方式之一使用边车（Sidecar）容器： -* sidecar 容器将应用程序日志传送到自己的标准输出。 -* sidecar 容器运行一个日志代理，配置该日志代理以便从应用容器收集日志。 +* 边车容器将应用程序日志传送到自己的标准输出。 +* 边车容器运行一个日志代理，配置该日志代理以便从应用容器收集日志。 #### 传输数据流的 sidecar 容器 -![数据流容器的 Sidecar 容器](/images/docs/user-guide/logging/logging-with-streaming-sidecar.png) +![带数据流容器的边车容器](/images/docs/user-guide/logging/logging-with-streaming-sidecar.png) -利用 sidecar 容器向自己的 `stdout` 和 `stderr` 传输流的方式， +利用边车容器向自己的 `stdout` 和 `stderr` 传输流的方式， 你就可以利用每个节点上的 kubelet 和日志代理来处理日志。 -sidecar 容器从文件、套接字或 journald 读取日志。 -每个 sidecar 容器打印其自己的 `stdout` 和 `stderr` 流。 +边车容器从文件、套接字或 journald 读取日志。 +每个边车容器向自己的 `stdout` 和 `stderr` 流中输出日志。 -考虑接下来的例子。pod 的容器向两个文件写不同格式的日志，下面是这个 pod 的配置文件: +例如，某 Pod 中运行一个容器，该容器向两个文件写不同格式的日志。 +下面是这个 pod 的配置文件: {{< codenew file="admin/logging/two-files-counter-pod.yaml" >}} -在同一个日志流中有两种不同格式的日志条目，这有点混乱，即使你试图重定向它们到容器的 `stdout` 流。 -取而代之的是，你可以引入两个 sidecar 容器。 -每一个 sidecar 容器可以从共享卷跟踪特定的日志文件，并重定向文件内容到各自的 `stdout` 流。 +不建议在同一个日志流中写入不同格式的日志条目，即使你成功地将其重定向到容器的 +`stdout` 流。相反，你可以创建两个边车容器。每个边车容器可以从共享卷 +跟踪特定的日志文件，并将文件内容重定向到各自的 `stdout` 流。 -这是运行两个 sidecar 容器的 Pod 文件。 +下面是运行两个边车容器的 Pod 的配置文件： {{< codenew file="admin/logging/two-files-counter-pod-streaming-sidecar.yaml" >}} @@ -358,12 +350,18 @@ Here's a configuration file for a pod that has two sidecar containers: Now when you run this pod, you can access each log stream separately by running the following commands: --> -现在当你运行这个 Pod 时，你可以分别地访问每一个日志流，运行如下命令： +现在当你运行这个 Pod 时，你可以运行如下命令分别访问每个日志流： ```shell kubectl logs counter count-log-1 ``` -``` + + +输出为： + +```console 0: Mon Jan 1 00:00:00 UTC 2001 1: Mon Jan 1 00:00:01 UTC 2001 2: Mon Jan 1 00:00:02 UTC 2001 @@ -373,7 +371,13 @@ kubectl logs counter count-log-1 ```shell kubectl logs counter count-log-2 ``` -``` + + +输出为： + +```console Mon Jan 1 00:00:00 UTC 2001 INFO 0 Mon Jan 1 00:00:01 UTC 2001 INFO 1 Mon Jan 1 00:00:02 UTC 2001 INFO 2 @@ -385,7 +389,8 @@ The node-level agent installed in your cluster picks up those log streams automatically without any further configuration. If you like, you can configure the agent to parse log lines depending on the source container. --> -集群中安装的节点级代理会自动获取这些日志流，而无需进一步配置。如果你愿意，你可以配置代理程序来解析源容器的日志行。 +集群中安装的节点级代理会自动获取这些日志流，而无需进一步配置。 +如果你愿意，你也可以配置代理程序来解析源容器的日志行。 -注意，尽管 CPU 和内存使用率都很低（以多个 cpu millicores 指标排序或者按内存的兆字节排序）， +注意，尽管 CPU 和内存使用率都很低（以多个 CPU 毫核指标排序或者按内存的兆字节排序）， 向文件写日志然后输出到 `stdout` 流仍然会成倍地增加磁盘使用率。 -如果你的应用向单一文件写日志，通常最好设置 `/dev/stdout` 作为目标路径，而不是使用流式的 sidecar 容器方式。 +如果你的应用向单一文件写日志，通常最好设置 `/dev/stdout` 作为目标路径， +而不是使用流式的边车容器方式。 -应用本身如果不具备轮转日志文件的功能，可以通过 sidecar 容器实现。 -该方式的一个例子是运行一个定期轮转日志的容器。 -然而，还是推荐直接使用 `stdout` 和 `stderr`，将日志的轮转和保留策略交给 kubelet。 +应用本身如果不具备轮转日志文件的功能，可以通过边车容器实现。 +该方式的一个例子是运行一个小的、定期轮转日志的容器。 +然而，还是推荐直接使用 `stdout` 和 `stderr`，将日志的轮转和保留策略 +交给 kubelet。 -### 具有日志代理功能的 sidecar 容器 +### 具有日志代理功能的边车容器 -![日志记录代理功能的 sidecar 容器](/images/docs/user-guide/logging/logging-with-sidecar-agent.png) +![含日志代理的边车容器](/images/docs/user-guide/logging/logging-with-sidecar-agent.png) -如果节点级日志记录代理程序对于你的场景来说不够灵活，你可以创建一个带有单独日志记录代理程序的 -sidecar 容器，将代理程序专门配置为与你的应用程序一起运行。 +如果节点级日志记录代理程序对于你的场景来说不够灵活，你可以创建一个 +带有单独日志记录代理的边车容器，将代理程序专门配置为与你的应用程序一起运行。 - -{{< note >}} -在 sidecar 容器中使用日志代理会导致严重的资源损耗。 +在边车容器中使用日志代理会带来严重的资源损耗。 此外，你不能使用 `kubectl logs` 命令访问日志，因为日志并没有被 kubelet 管理。 {{< /note >}} -例如，你可以使用 [Stackdriver](/zh/docs/tasks/debug-application-cluster/logging-stackdriver/)， -它使用 fluentd 作为日志记录代理。 -以下是两个可用于实现此方法的配置文件。 -第一个文件包含配置 fluentd 的 +下面是两个配置文件，可以用来实现一个带日志代理的边车容器。 +第一个文件包含用来配置 fluentd 的 [ConfigMap](/zh/docs/tasks/configure-pod-container/configure-pod-configmap/)。 {{< codenew file="admin/logging/fluentd-sidecar-config.yaml" >}} +{{< note >}} -{{< note >}} -配置 fluentd 超出了本文的范围。要进一步了解如何配置 fluentd， -请参考 [fluentd 官方文档](https://docs.fluentd.org/). +要进一步了解如何配置 fluentd，请参考 [fluentd 官方文档](https://docs.fluentd.org/). {{< /note >}} -第二个文件描述了运行 fluentd sidecar 容器的 Pod 。flutend 通过 Pod 的挂载卷获取它的配置数据。 +第二个文件描述了运行 fluentd 边车容器的 Pod 。 +flutend 通过 Pod 的挂载卷获取它的配置数据。 {{< codenew file="admin/logging/two-files-counter-pod-agent-sidecar.yaml" >}} -一段时间后，你可以在 Stackdriver 界面看到日志消息。 - - -记住，这只是一个例子，事实上你可以用任何一个日志代理替换 fluentd ，并从应用容器中读取任何资源。 +在示例配置中，你可以将 fluentd 替换为任何日志代理，从应用容器内 +的任何来源读取数据。 - ### 从应用中直接暴露日志目录 ![直接从应用程序暴露日志](/images/docs/user-guide/logging/logging-from-application.png) -通过暴露或推送每个应用的日志，你可以实现集群级日志记录； -然而，这种日志记录机制的实现已超出 Kubernetes 的范围。 - +从各个应用中直接暴露和推送日志数据的集群日志机制 +已超出 Kubernetes 的范围。 diff --git a/content/zh/docs/tasks/debug-application-cluster/events-stackdriver.md b/content/zh/docs/tasks/debug-application-cluster/events-stackdriver.md deleted file mode 100644 index 04a20a7174..0000000000 --- a/content/zh/docs/tasks/debug-application-cluster/events-stackdriver.md +++ /dev/null @@ -1,156 +0,0 @@ ---- -content_type: concept -title: StackDriver 中的事件 ---- - - - - - - - -Kubernetes 事件是一种对象，它为用户提供了洞察集群内发生的事情的能力， -例如调度程序做出了什么决定，或者为什么某些 Pod 被逐出节点。 -你可以在[应用程序自检和调试](/zh/docs/tasks/debug-application-cluster/debug-application-introspection/) -中阅读有关使用事件调试应用程序的更多信息。 - - -因为事件是 API 对象，所以它们存储在主控节点上的 API 服务器中。 -为了避免主节点磁盘空间被填满，将强制执行保留策略：事件在最后一次发生的一小时后将会被删除。 -为了提供更长的历史记录和聚合能力，应该安装第三方解决方案来捕获事件。 - - -本文描述了一个将 Kubernetes 事件导出为 Stackdriver Logging 的解决方案，在这里可以对它们进行处理和分析。 - - -{{< note >}} -不能保证集群中发生的所有事件都将导出到 Stackdriver。 -事件不能导出的一种可能情况是事件导出器没有运行（例如，在重新启动或升级期间）。 -在大多数情况下，可以将事件用于设置 -[metrics](https://cloud.google.com/logging/docs/view/logs_based_metrics) 和 -[alerts](https://cloud.google.com/logging/docs/view/logs_based_metrics#creating_an_alerting_policy) -等目的，但你应该注意其潜在的不准确性。 -{{< /note >}} - - - - -## 部署 {#deployment} - -### Google Kubernetes Engine - - - -在 Google Kubernetes Engine 中，如果启用了云日志，那么事件导出器默认部署在主节点运行版本为 1.7 及更高版本的集群中。 -为了防止干扰你的工作负载，事件导出器没有设置资源，并且处于尽力而为的 QoS 类型中，这意味着它将在资源匮乏的情况下第一个被杀死。 -如果要导出事件，请确保有足够的资源给事件导出器 Pod 使用。 -这可能会因为工作负载的不同而有所不同，但平均而言，需要大约 100MB 的内存和 100m 的 CPU。 - - -### 部署到现有集群 - -使用下面的命令将事件导出器部署到你的集群： - -```shell -kubectl create -f https://k8s.io/examples/debug/event-exporter.yaml -``` - - - -由于事件导出器访问 Kubernetes API，因此它需要权限才能访问。 -以下的部署配置为使用 RBAC 授权。 -它设置服务帐户和集群角色绑定，以允许事件导出器读取事件。 -为了确保事件导出器 Pod 不会从节点中退出，你可以另外设置资源请求。 -如前所述，100MB 内存和 100m CPU 应该就足够了。 - -{{< codenew file="debug/event-exporter.yaml" >}} - - -## 用户指南 {#user-guide} - -事件在 Stackdriver Logging 中被导出到 `GKE Cluster` 资源。 -你可以通过从可用资源的下拉菜单中选择适当的选项来找到它们： - - -Stackdriver 日志接口中事件的位置 - - -你可以使用 Stackdriver Logging 的 -[过滤机制](https://cloud.google.com/logging/docs/view/advanced_filters) -基于事件对象字段进行过滤。 -例如，下面的查询将显示调度程序中有关 Deployment `nginx-deployment` 中的 Pod 的事件： - -``` -resource.type="gke_cluster" -jsonPayload.kind="Event" -jsonPayload.source.component="default-scheduler" -jsonPayload.involvedObject.name:"nginx-deployment" -``` - -{{< figure src="/images/docs/stackdriver-event-exporter-filter.png" alt="在 Stackdriver 接口中过滤的事件" width="500" >}} - - diff --git a/content/zh/docs/tasks/debug-application-cluster/logging-elasticsearch-kibana.md b/content/zh/docs/tasks/debug-application-cluster/logging-elasticsearch-kibana.md deleted file mode 100644 index 2dbabad039..0000000000 --- a/content/zh/docs/tasks/debug-application-cluster/logging-elasticsearch-kibana.md +++ /dev/null @@ -1,197 +0,0 @@ ---- -content_type: concept -title: 使用 ElasticSearch 和 Kibana 进行日志管理 ---- - - - - - - -在 Google Compute Engine (GCE) 平台上，默认的日志管理支持目标是 -[Stackdriver Logging](https://cloud.google.com/logging/)， -在[使用 Stackdriver Logging 管理日志](/zh/docs/tasks/debug-application-cluster/logging-stackdriver/) -中详细描述了这一点。 - - -本文介绍了如何设置一个集群，将日志导入 -[Elasticsearch](https://www.elastic.co/products/elasticsearch)，并使用 -[Kibana](https://www.elastic.co/products/kibana) 查看日志，作为在 GCE 上 -运行应用时使用 Stackdriver Logging 管理日志的替代方案。 - - -{{< note >}} -你不能在 Google Kubernetes Engine 平台运行的 Kubernetes 集群上自动部署 -Elasticsearch 和 Kibana。你必须手动部署它们。 -{{< /note >}} - - - - -要使用 Elasticsearch 和 Kibana 处理集群日志，你应该在使用 kube-up.sh -脚本创建集群时设置下面所示的环境变量： - -```shell -KUBE_LOGGING_DESTINATION=elasticsearch -``` - - -你还应该确保设置了 `KUBE_ENABLE_NODE_LOGGING=true` （这是 GCE 平台的默认设置）。 - - -现在，当你创建集群时，将有一条消息将指示每个节点上运行的 fluentd 日志收集守护进程 -以 ElasticSearch 为日志输出目标： - -```shell -cluster/kube-up.sh -``` - -``` -... -Project: kubernetes-satnam -Zone: us-central1-b -... calling kube-up -Project: kubernetes-satnam -Zone: us-central1-b -+++ Staging server tars to Google Storage: gs://kubernetes-staging-e6d0e81793/devel -+++ kubernetes-server-linux-amd64.tar.gz uploaded (sha1 = 6987c098277871b6d69623141276924ab687f89d) -+++ kubernetes-salt.tar.gz uploaded (sha1 = bdfc83ed6b60fa9e3bff9004b542cfc643464cd0) -Looking for already existing resources -Starting master and configuring firewalls -Created [https://www.googleapis.com/compute/v1/projects/kubernetes-satnam/zones/us-central1-b/disks/kubernetes-master-pd]. -NAME ZONE SIZE_GB TYPE STATUS -kubernetes-master-pd us-central1-b 20 pd-ssd READY -Created [https://www.googleapis.com/compute/v1/projects/kubernetes-satnam/regions/us-central1/addresses/kubernetes-master-ip]. -+++ Logging using Fluentd to elasticsearch -``` - - -每个节点的 Fluentd Pod、Elasticsearch Pod 和 Kibana Pod 都应该在集群启动后不久运行在 -kube-system 名字空间中。 - -```shell -kubectl get pods --namespace=kube-system -``` - -``` -NAME READY STATUS RESTARTS AGE -elasticsearch-logging-v1-78nog 1/1 Running 0 2h -elasticsearch-logging-v1-nj2nb 1/1 Running 0 2h -fluentd-elasticsearch-kubernetes-node-5oq0 1/1 Running 0 2h -fluentd-elasticsearch-kubernetes-node-6896 1/1 Running 0 2h -fluentd-elasticsearch-kubernetes-node-l1ds 1/1 Running 0 2h -fluentd-elasticsearch-kubernetes-node-lz9j 1/1 Running 0 2h -kibana-logging-v1-bhpo8 1/1 Running 0 2h -kube-dns-v3-7r1l9 3/3 Running 0 2h -monitoring-heapster-v4-yl332 1/1 Running 1 2h -monitoring-influx-grafana-v1-o79xf 2/2 Running 0 2h -``` - - -`fluentd-elasticsearch` Pod 从每个节点收集日志并将其发送到 `elasticsearch-logging` Pod， -该 Pod 是名为 `elasticsearch-logging` 的 -[服务](/zh/docs/concepts/services-networking/service/)的一部分。 -这些 ElasticSearch pod 存储日志，并通过 REST API 将其公开。 -`kibana-logging` pod 提供了一个用于读取 ElasticSearch 中存储的日志的 Web UI， -它是名为 `kibana-logging` 的服务的一部分。 - - - -Elasticsearch 和 Kibana 服务都位于 `kube-system` 名字空间中，并且没有通过 -可公开访问的 IP 地址直接暴露。要访问它们，请参照 -[访问集群中运行的服务](/zh/docs/tasks/access-application-cluster/access-cluster/#accessing-services-running-on-the-cluster) -的说明进行操作。 - - -如果你想在浏览器中访问 `elasticsearch-logging` 服务，你将看到类似下面的状态页面： - -![Elasticsearch Status](/images/docs/es-browser.png) - - -现在你可以直接在浏览器中输入 Elasticsearch 查询，如果你愿意的话。 -请参考 [Elasticsearch 的文档](https://www.elastic.co/guide/en/elasticsearch/reference/current/search-uri-request.html) -以了解这样做的更多细节。 - - - -或者，你可以使用 Kibana 查看集群的日志（再次使用 -[访问集群中运行的服务的说明](/zh/docs/tasks/access-application-cluster/access-cluster/#accessing-services-running-on-the-cluster)）。 -第一次访问 Kibana URL 时，将显示一个页面，要求你配置所接收日志的视图。 -选择时间序列值的选项，然后选择 `@timestamp`。 -在下面的页面中选择 `Discover` 选项卡，然后你应该能够看到所摄取的日志。 -你可以将刷新间隔设置为 5 秒，以便定期刷新日志。 - - - -以下是从 Kibana 查看器中摄取日志的典型视图： - -![Kibana logs](/images/docs/kibana-logs.png) - -## {{% heading "whatsnext" %}} - - -Kibana 为浏览你的日志提供了各种强大的选项！有关如何深入研究它的一些想法， -请查看 [Kibana 的文档](https://www.elastic.co/guide/en/kibana/current/discover.html)。 - diff --git a/content/zh/docs/tutorials/stateful-application/zookeeper.md b/content/zh/docs/tutorials/stateful-application/zookeeper.md index 500259f5fa..3f5baf3c27 100644 --- a/content/zh/docs/tutorials/stateful-application/zookeeper.md +++ b/content/zh/docs/tutorials/stateful-application/zookeeper.md @@ -1,5 +1,10 @@ --- -approvers: +title: 运行 ZooKeeper，一个分布式协调系统 +content_type: tutorial +weight: 40 +--- + @@ -20,8 +25,11 @@ Kubernetes using [StatefulSets](/docs/concepts/workloads/controllers/statefulset [PodDisruptionBudgets](/docs/concepts/workloads/pods/disruptions/#pod-disruption-budget), and [PodAntiAffinity](/docs/concepts/scheduling-eviction/assign-pod-node/#affinity-and-anti-affinity). --> - -本教程展示了在 Kubernetes 上使用 [StatefulSets](/zh/docs/concepts/workloads/controllers/statefulset/)，[PodDisruptionBudgets](/zh/docs/concepts/workloads/pods/disruptions/#pod-disruption-budget) 和 [PodAntiAffinity](/zh/docs/concepts/scheduling-eviction/assign-pod-node/#亲和与反亲和) 特性运行 [Apache Zookeeper](https://zookeeper.apache.org)。 +本教程展示了在 Kubernetes 上使用 +[StatefulSet](/zh/docs/concepts/workloads/controllers/statefulset/)， +[PodDisruptionBudget](/zh/docs/concepts/workloads/pods/disruptions/#pod-disruption-budget) 和 +[PodAntiAffinity](/zh/docs/concepts/scheduling-eviction/assign-pod-node/#亲和与反亲和) +特性运行 [Apache Zookeeper](https://zookeeper.apache.org)。 ## {{% heading "prerequisites" %}} @@ -29,44 +37,45 @@ and [PodAntiAffinity](/docs/concepts/scheduling-eviction/assign-pod-node/#affini Before starting this tutorial, you should be familiar with the following Kubernetes concepts. --> - 在开始本教程前，你应该熟悉以下 Kubernetes 概念。 -- [Pods](/zh/docs/concepts/workloads/pods/) -- [Cluster DNS](/zh/docs/concepts/services-networking/dns-pod-service/) -- [Headless Services](/zh/docs/concepts/services-networking/service/#headless-services) -- [PersistentVolumes](/zh/docs/concepts/storage/persistent-volumes/) -- [PersistentVolume Provisioning](https://github.com/kubernetes/examples/tree/{{< param "githubbranch" >}}/staging/persistent-volume-provisioning/) -- [StatefulSets](/zh/docs/concepts/workloads/controllers/statefulset/) -- [PodDisruptionBudgets](/zh/docs/concepts/workloads/pods/disruptions/#pod-disruption-budget) -- [PodAntiAffinity](/zh/docs/concepts/scheduling-eviction/assign-pod-node/#亲和与反亲和) -- [kubectl CLI](/zh/docs/reference/kubectl/kubectl/) +- [Pods](/zh/docs/concepts/workloads/pods/) +- [集群 DNS](/zh/docs/concepts/services-networking/dns-pod-service/) +- [无头服务（Headless Service）](/zh/docs/concepts/services-networking/service/#headless-services) +- [PersistentVolumes](/zh/docs/concepts/storage/persistent-volumes/) +- [PersistentVolume 制备](https://github.com/kubernetes/examples/tree/{{< param "githubbranch" >}}/staging/persistent-volume-provisioning/) +- [StatefulSet](/zh/docs/concepts/workloads/controllers/statefulset/) +- [PodDisruptionBudget](/zh/docs/concepts/workloads/pods/disruptions/#pod-disruption-budget) +- [PodAntiAffinity](/zh/docs/concepts/scheduling-eviction/assign-pod-node/#亲和与反亲和) +- [kubectl CLI](/zh/docs/reference/kubectl/kubectl/) +你需要一个至少包含四个节点的集群，每个节点至少 2 CPUs 和 4 GiB 内存。 +在本教程中你将会隔离（Cordon）和腾空（Drain ）集群的节点。 +**这意味着集群节点上所有的 Pods 将会被终止并移除。这些节点也会暂时变为不可调度**。 +在本教程中你应该使用一个独占的集群，或者保证你造成的干扰不会影响其它租户。 + - -你需要一个至少包含四个节点的集群，每个节点至少 2 CPUs 和 4 GiB 内存。在本教程中你将会 cordon 和 drain 集群的节点。**这意味着集群节点上所有的 Pods 将会被终止并移除**。**这些节点也会暂时变为不可调度**。在本教程中你应该使用一个独占的集群，或者保证你造成的干扰不会影响其它租户。 - -本教程假设你的集群配置为动态的提供 PersistentVolumes。如果你的集群没有配置成这样，在开始本教程前，你需要手动准备三个 20 GiB 的卷。 - +本教程假设你的集群配置为动态的提供 PersistentVolumes。 +如果你的集群没有配置成这样，在开始本教程前，你需要手动准备三个 20 GiB 的卷。 ## {{% heading "objectives" %}} - 在学习本教程后，你将熟悉下列内容。 * 如何使用 StatefulSet 部署一个 ZooKeeper ensemble。 @@ -74,11 +83,10 @@ After this tutorial, you will know the following. * 如何在 ensemble 中 分布 ZooKeeper 服务器的部署。 * 如何在计划维护中使用 PodDisruptionBudgets 确保服务可用性。 - +### ZooKeeper {#zookeeper-basics} +[Apache ZooKeeper](https://zookeeper.apache.org/doc/current/) +是一个分布式的开源协调服务，用于分布式系统。 +ZooKeeper 允许你读取、写入数据和发现数据更新。 +数据按层次结构组织在文件系统中，并复制到 ensemble（一个 ZooKeeper 服务器的集合） +中所有的 ZooKeeper 服务器。对数据的所有操作都是原子的和顺序一致的。 +ZooKeeper 通过 +[Zab](https://pdfs.semanticscholar.org/b02c/6b00bd5dbdbd951fddb00b906c82fa80f0b3.pdf) +一致性协议在 ensemble 的所有服务器之间复制一个状态机来确保这个特性。 + + +Ensemble 使用 Zab 协议选举一个领导者，在选举出领导者前不能写入数据。 +一旦选举出了领导者，ensemble 使用 Zab 保证所有写入被复制到一个 quorum， +然后这些写入操作才会被确认并对客户端可用。 +如果没有遵照加权 quorums，一个 quorum 表示包含当前领导者的 ensemble 的多数成员。 +例如，如果 ensemble 有 3 个服务器，一个包含领导者的成员和另一个服务器就组成了一个 +quorum。 +如果 ensemble 不能达成一个 quorum，数据将不能被写入。 + - -### ZooKeeper 基础 - -[Apache ZooKeeper](https://zookeeper.apache.org/doc/current/) 是一个分布式的开源协调服务，用于分布式系统。ZooKeeper 允许你读取、写入数据和发现数据更新。数据按层次结构组织在文件系统中，并复制到 ensemble（一个 ZooKeeper 服务器的集合） 中所有的 ZooKeeper 服务器。对数据的所有操作都是原子的和顺序一致的。ZooKeeper 通过 [Zab](https://pdfs.semanticscholar.org/b02c/6b00bd5dbdbd951fddb00b906c82fa80f0b3.pdf) 一致性协议在 ensemble 的所有服务器之间复制一个状态机来确保这个特性。 - -ensemble 使用 Zab 协议选举一个 leader，在选举出 leader 前不能写入数据。一旦选举出了 leader，ensemble 使用 Zab 保证所有写入被复制到一个 quorum，然后这些写入操作才会被确认并对客户端可用。如果没有遵照加权 quorums，一个 quorum 表示包含当前 leader 的 ensemble 的多数成员。例如，如果 ensemble 有3个服务器，一个包含 leader 的成员和另一个服务器就组成了一个 quorum。如果 ensemble 不能达成一个 quorum，数据将不能被写入。 - -ZooKeeper 在内存中保存它们的整个状态机，但是每个改变都被写入一个在存储介质上的持久 WAL（Write Ahead Log）。当一个服务器故障时，它能够通过回放 WAL 恢复之前的状态。为了防止 WAL 无限制的增长，ZooKeeper 服务器会定期的将内存状态快照保存到存储介质。这些快照能够直接加载到内存中，所有在这个快照之前的 WAL 条目都可以被安全的丢弃。 +ZooKeeper 在内存中保存它们的整个状态机，但是每个改变都被写入一个在存储介质上的 +持久 WAL（Write Ahead Log）。 +当一个服务器出现故障时，它能够通过回放 WAL 恢复之前的状态。 +为了防止 WAL 无限制的增长，ZooKeeper 服务器会定期的将内存状态快照保存到存储介质。 +这些快照能够直接加载到内存中，所有在这个快照之前的 WAL 条目都可以被安全的丢弃。 - ## 创建一个 ZooKeeper Ensemble 下面的清单包含一个 -[Headless Service](/zh/docs/concepts/services-networking/service/#headless-services)， +[无头服务](/zh/docs/concepts/services-networking/service/#headless-services)， 一个 [Service](/zh/docs/concepts/services-networking/service/)， 一个 [PodDisruptionBudget](/zh/docs/concepts/workloads/pods/disruptions/#specifying-a-poddisruptionbudget)， 和一个 [StatefulSet](/zh/docs/concepts/workloads/controllers/statefulset/)。 @@ -127,8 +152,8 @@ Open a terminal, and use the [`kubectl apply`](/docs/reference/generated/kubectl/kubectl-commands/#apply) command to create the manifest. --> - -打开一个命令行终端，使用 [`kubectl apply`](/docs/reference/generated/kubectl/kubectl-commands/#apply) +打开一个命令行终端，使用命令 +[`kubectl apply`](/docs/reference/generated/kubectl/kubectl-commands/#apply) 创建这个清单。 ```shell @@ -139,8 +164,8 @@ kubectl apply -f https://k8s.io/examples/application/zookeeper/zookeeper.yaml This creates the `zk-hs` Headless Service, the `zk-cs` Service, the `zk-pdb` PodDisruptionBudget, and the `zk` StatefulSet. --> - -这个操作创建了 `zk-hs` Headless Service、`zk-cs` Service、`zk-pdb` PodDisruptionBudget 和 `zk` StatefulSet。 +这个操作创建了 `zk-hs` 无头服务、`zk-cs` 服务、`zk-pdb` PodDisruptionBudget +和 `zk` StatefulSet。 ``` service/zk-hs created @@ -153,8 +178,9 @@ statefulset.apps/zk created Use [`kubectl get`](/docs/reference/generated/kubectl/kubectl-commands/#get) to watch the StatefulSet controller create the StatefulSet's Pods. --> - -使用 [`kubectl get`](/docs/reference/generated/kubectl/kubectl-commands/#get) 查看 StatefulSet 控制器创建的 Pods。 +使用命令 +[`kubectl get`](/docs/reference/generated/kubectl/kubectl-commands/#get) +查看 StatefulSet 控制器创建的 Pods。 ```shell kubectl get pods -w -l app=zk @@ -163,7 +189,6 @@ kubectl get pods -w -l app=zk - 一旦 `zk-2` Pod 变成 Running 和 Ready 状态，使用 `CRTL-C` 结束 kubectl。 ``` @@ -189,8 +214,8 @@ zk-2 1/1 Running 0 40s The StatefulSet controller creates three Pods, and each Pod has a container with a [ZooKeeper](https://www-us.apache.org/dist/zookeeper/stable/) server. --> - -StatefulSet 控制器创建了3个 Pods，每个 Pod 包含一个 [ZooKeeper](https://www-us.apache.org/dist/zookeeper/stable/) 服务器。 +StatefulSet 控制器创建 3 个 Pods，每个 Pod 包含一个 +[ZooKeeper](https://www-us.apache.org/dist/zookeeper/stable/) 服务器。 +### 促成 Leader 选举 {#facilitating-leader-election} -### 促成 Leader 选举 +由于在匿名网络中没有用于选举 leader 的终止算法，Zab 要求显式的进行成员关系配置， +以执行 leader 选举。Ensemble 中的每个服务器都需要具有一个独一无二的标识符， +所有的服务器均需要知道标识符的全集，并且每个标识符都需要和一个网络地址相关联。 -由于在匿名网络中没有用于选举 leader 的终止算法，Zab 要求显式的进行成员关系配置，以执行 leader 选举。Ensemble 中的每个服务器都需要具有一个独一无二的标识符，所有的服务器均需要知道标识符的全集，并且每个标识符都需要和一个网络地址相关联。 - -使用 [`kubectl exec`](/docs/reference/generated/kubectl/kubectl-commands/#exec) 获取 `zk` StatefulSet 中 Pods 的主机名。 +使用命令 +[`kubectl exec`](/docs/reference/generated/kubectl/kubectl-commands/#exec) +获取 `zk` StatefulSet 中 Pods 的主机名。 ```shell for i in 0 1 2; do kubectl exec zk-$i -- hostname; done @@ -215,8 +243,10 @@ for i in 0 1 2; do kubectl exec zk-$i -- hostname; done The StatefulSet controller provides each Pod with a unique hostname based on its ordinal index. The hostnames take the form of `-`. Because the `replicas` field of the `zk` StatefulSet is set to `3`, the Set's controller creates three Pods with their hostnames set to `zk-0`, `zk-1`, and `zk-2`. --> - -StatefulSet 控制器基于每个 Pod 的序号索引为它们各自提供一个唯一的主机名。主机名采用 `-` 的形式。由于 `zk` StatefulSet 的 `replicas` 字段设置为3，这个 Set 的控制器将创建3个 Pods，主机名为：`zk-0`、`zk-1` 和 `zk-2`。 +StatefulSet 控制器基于每个 Pod 的序号索引为它们各自提供一个唯一的主机名。 +主机名采用 `-<序数索引>` 的形式。 +由于 `zk` StatefulSet 的 `replicas` 字段设置为 3，这个集合的控制器将创建 +3 个 Pods，主机名为：`zk-0`、`zk-1` 和 `zk-2`。 ``` zk-0 @@ -229,8 +259,8 @@ The servers in a ZooKeeper ensemble use natural numbers as unique identifiers, a To examine the contents of the `myid` file for each server use the following command. --> - -ZooKeeper ensemble 中的服务器使用自然数作为唯一标识符，每个服务器的标识符都保存在服务器的数据目录中一个名为 `myid` 的文件里。 +ZooKeeper ensemble 中的服务器使用自然数作为唯一标识符， +每个服务器的标识符都保存在服务器的数据目录中一个名为 `myid` 的文件里。 检查每个服务器的 `myid` 文件的内容。 @@ -241,7 +271,6 @@ for i in 0 1 2; do echo "myid zk-$i";kubectl exec zk-$i -- cat /var/lib/zookeepe - 由于标识符为自然数并且序号索引是非负整数，你可以在序号上加 1 来生成一个标识符。 ``` @@ -256,8 +285,7 @@ myid zk-2 - -获取 `zk` StatefulSet 中每个 Pod 的 FQDN (Fully Qualified Domain Name，正式域名)。 +获取 `zk` StatefulSet 中每个 Pod 的全限定域名（Fully Qualified Domain Name，FQDN）。 ```shell for i in 0 1 2; do kubectl exec zk-$i -- hostname -f; done @@ -267,8 +295,7 @@ for i in 0 1 2; do kubectl exec zk-$i -- hostname -f; done The `zk-hs` Service creates a domain for all of the Pods, `zk-hs.default.svc.cluster.local`. --> - -`zk-hs` Service 为所有 Pods 创建了一个 domain：`zk-hs.default.svc.cluster.local`。 +`zk-hs` Service 为所有 Pods 创建了一个域：`zk-hs.default.svc.cluster.local`。 ``` zk-0.zk-hs.default.svc.cluster.local @@ -281,10 +308,13 @@ The A records in [Kubernetes DNS](/docs/concepts/services-networking/dns-pod-ser ZooKeeper stores its application configuration in a file named `zoo.cfg`. Use `kubectl exec` to view the contents of the `zoo.cfg` file in the `zk-0` Pod. --> +[Kubernetes DNS](/zh/docs/concepts/services-networking/dns-pod-service/) +中的 A 记录将 FQDNs 解析成为 Pods 的 IP 地址。 +如果 Pods 被调度，这个 A 记录将会使用 Pods 的新 IP 地址完成更新， +但 A 记录的名称不会改变。 -[Kubernetes DNS](/zh/docs/concepts/services-networking/dns-pod-service/) 中的 A 记录将 FQDNs 解析成为 Pods 的 IP 地址。如果 Pods 被调度，这个 A 记录将会使用 Pods 的新 IP 地址更新，但 A 记录的名称不会改变。 - -ZooKeeper 在一个名为 `zoo.cfg` 的文件中保存它的应用配置。使用 `kubectl exec` 在 `zk-0` Pod 中查看 `zoo.cfg` 文件的内容。 +ZooKeeper 在一个名为 `zoo.cfg` 的文件中保存它的应用配置。 +使用 `kubectl exec` 在 `zk-0` Pod 中查看 `zoo.cfg` 文件的内容。 ```shell kubectl exec zk-0 -- cat /opt/zookeeper/conf/zoo.cfg @@ -296,8 +326,9 @@ the file, the `1`, `2`, and `3` correspond to the identifiers in the ZooKeeper servers' `myid` files. They are set to the FQDNs for the Pods in the `zk` StatefulSet. --> - -文件底部为 `server.1`、`server.2` 和 `server.3`，其中的 `1`、`2`和`3`分别对应 ZooKeeper 服务器的 `myid` 文件中的标识符。它们被设置为 `zk` StatefulSet 中的 Pods 的 FQDNs。 +文件底部为 `server.1`、`server.2` 和 `server.3`，其中的 `1`、`2` 和 `3` +分别对应 ZooKeeper 服务器的 `myid` 文件中的标识符。 +它们被设置为 `zk` StatefulSet 中的 Pods 的 FQDNs。 ``` clientPort=2181 @@ -317,14 +348,17 @@ server.3=zk-2.zk-hs.default.svc.cluster.local:2888:3888 ``` +### 达成共识 {#achieving-consensus} -### 达成一致 - - 一致性协议要求每个参与者的标识符唯一。在 Zab 协议里任何两个参与者都不应该声明相同的唯一标识符。对于让系统中的进程协商哪些进程已经提交了哪些数据而言，这是必须的。如果有两个 Pods 使用相同的序号启动，这两个 ZooKeeper 服务器会将自己识别为相同的服务器。 + 一致性协议要求每个参与者的标识符唯一。 +在 Zab 协议里任何两个参与者都不应该声明相同的唯一标识符。 +对于让系统中的进程协商哪些进程已经提交了哪些数据而言，这是必须的。 +如果有两个 Pods 使用相同的序号启动，这两个 ZooKeeper 服务器 +会将自己识别为相同的服务器。 ```shell kubectl get pods -w -l app=zk @@ -355,8 +389,10 @@ the FQDNs of the ZooKeeper servers will resolve to a single endpoint, and that endpoint will be the unique ZooKeeper server claiming the identity configured in its `myid` file. --> - -每个 Pod 的 A 记录仅在 Pod 变成 Ready状态时被录入。因此，ZooKeeper 服务器的 FQDNs 只会解析到一个 endpoint，而那个 endpoint 将会是一个唯一的 ZooKeeper 服务器，这个服务器声明了配置在它的 `myid` 文件中的标识符。 +每个 Pod 的 A 记录仅在 Pod 变成 Ready状态时被录入。 +因此，ZooKeeper 服务器的 FQDNs 只会解析到一个端点，而那个端点将会是 +一个唯一的 ZooKeeper 服务器，这个服务器声明了配置在它的 `myid` +文件中的标识符。 ``` zk-0.zk-hs.default.svc.cluster.local @@ -369,7 +405,8 @@ This ensures that the `servers` properties in the ZooKeepers' `zoo.cfg` files represents a correctly configured ensemble. --> -这保证了 ZooKeepers 的 `zoo.cfg` 文件中的 `servers` 属性代表了一个正确配置的 ensemble。 +这保证了 ZooKeepers 的 `zoo.cfg` 文件中的 `servers` 属性代表了 +一个正确配置的 ensemble。 ``` server.1=zk-0.zk-hs.default.svc.cluster.local:2888:3888 @@ -380,8 +417,10 @@ server.3=zk-2.zk-hs.default.svc.cluster.local:2888:3888 - -当服务器使用 Zab 协议尝试提交一个值的时候，它们会达成一致并成功提交这个值（如果 leader 选举成功并且至少有两个 Pods 处于 Running 和 Ready状态），或者将会失败（如果没有满足上述条件中的任意一条）。当一个服务器承认另一个服务器的代写时不会有状态产生。 +当服务器使用 Zab 协议尝试提交一个值的时候，它们会达成一致并成功提交这个值 +（如果领导者选举成功并且至少有两个 Pods 处于 Running 和 Ready状态）， +或者将会失败（如果没有满足上述条件中的任意一条）。 +当一个服务器承认另一个服务器的代写时不会有状态产生。 - ### Ensemble 健康检查 -最基本的健康检查是向一个 ZooKeeper 服务器写入一些数据，然后从另一个服务器读取这些数据。 +最基本的健康检查是向一个 ZooKeeper 服务器写入一些数据，然后从 +另一个服务器读取这些数据。 使用 `zkCli.sh` 脚本在 `zk-0` Pod 上写入 `world` 到路径 `/hello`。 @@ -411,8 +450,7 @@ Created /hello - -从 `zk-1` Pod 获取数据。 +使用下面的命令从 `zk-1` Pod 获取数据。 ```shell kubectl exec zk-1 zkCli.sh get /hello @@ -422,8 +460,7 @@ kubectl exec zk-1 zkCli.sh get /hello The data that you created on `zk-0` is available on all the servers in the ensemble. --> - -你在 `zk-0` 创建的数据在 ensemble 中所有的服务器上都是可用的。 +你在 `zk-0` 上创建的数据在 ensemble 中所有的服务器上都是可用的。 ``` WATCHER:: @@ -455,12 +492,15 @@ state machine. Use the [`kubectl delete`](/docs/reference/generated/kubectl/kubectl-commands/#delete) command to delete the `zk` StatefulSet. --> +### 提供持久存储 -### 准备持久存储 +如同在 [ZooKeeper](#zookeeper-basics) 一节所提到的，ZooKeeper 提交 +所有的条目到一个持久 WAL，并周期性的将内存快照写入存储介质。 +对于使用一致性协议实现一个复制状态机的应用来说，使用 WALs 提供持久化 +是一种常用的技术，对于普通的存储应用也是如此。 -如同在 [ZooKeeper 基础](#zookeeper-基础) 一节所提到的，ZooKeeper 提交所有的条目到一个持久 WAL，并周期性的将内存快照写入存储介质。对于使用一致性协议实现一个复制状态机的应用来说，使用 WALs 提供持久化是一种常用的技术，对于普通的存储应用也是如此。 - -使用 [`kubectl delete`](/docs/reference/generated/kubectl/kubectl-commands/#delete) 删除 `zk` StatefulSet。 +使用 [`kubectl delete`](/docs/reference/generated/kubectl/kubectl-commands/#delete) +删除 `zk` StatefulSet。 ```shell kubectl delete statefulset zk @@ -473,7 +513,6 @@ statefulset.apps "zk" deleted - 观察 StatefulSet 中的 Pods 变为终止状态。 ```shell @@ -483,7 +522,6 @@ kubectl get pods -w -l app=zk - 当 `zk-0` 完全终止时，使用 `CRTL-C` 结束 kubectl。 ``` @@ -504,8 +542,7 @@ zk-0 0/1 Terminating 0 11m - -重新应用 `zookeeper.yaml` 中的代码清单。 +重新应用 `zookeeper.yaml` 中的清单。 ```shell kubectl apply -f https://k8s.io/examples/application/zookeeper/zookeeper.yaml @@ -516,7 +553,6 @@ This creates the `zk` StatefulSet object, but the other API objects in the manif Watch the StatefulSet controller recreate the StatefulSet's Pods. --> - `zk` StatefulSet 将会被创建。由于清单中的其他 API 对象已经存在，所以它们不会被修改。 观察 StatefulSet 控制器重建 StatefulSet 的 Pods。 @@ -528,7 +564,6 @@ kubectl get pods -w -l app=zk - 一旦 `zk-2` Pod 处于 Running 和 Ready 状态，使用 `CRTL-C` 停止 kubectl命令。 ``` @@ -554,7 +589,6 @@ zk-2 1/1 Running 0 40s Use the command below to get the value you entered during the [sanity test](#sanity-testing-the-ensemble), from the `zk-2` Pod. --> - 从 `zk-2` Pod 中获取你在[健康检查](#Ensemble-健康检查)中输入的值。 ```shell @@ -564,8 +598,8 @@ kubectl exec zk-2 zkCli.sh get /hello - -尽管 `zk` StatefulSet 中所有的 Pods 都已经被终止并重建过，ensemble 仍然使用原来的数值提供服务。 +尽管 `zk` StatefulSet 中所有的 Pods 都已经被终止并重建过，ensemble +仍然使用原来的数值提供服务。 ``` WATCHER:: @@ -588,8 +622,8 @@ numChildren = 0 - -`zk` StatefulSet 的 `spec` 中的 `volumeClaimTemplates` 字段标识了将要为每个 Pod 准备的 PersistentVolume。 +`zk` StatefulSet 的 `spec` 中的 `volumeClaimTemplates` 字段标识了 +将要为每个 Pod 准备的 PersistentVolume。 ```yaml volumeClaimTemplates: @@ -610,10 +644,9 @@ the `StatefulSet`. Use the following command to get the `StatefulSet`'s `PersistentVolumeClaims`. --> +`StatefulSet` 控制器为 `StatefulSet` 中的每个 Pod 生成一个 `PersistentVolumeClaim`。 -StatefulSet 控制器为 StatefulSet 中的每个 Pod 生成一个 PersistentVolumeClaim。 - -获取 StatefulSet 的 PersistentVolumeClaims。 +获取 `StatefulSet` 的 `PersistentVolumeClaim`。 ```shell kubectl get pvc -l app=zk @@ -622,8 +655,7 @@ kubectl get pvc -l app=zk - -当 StatefulSet 重新创建它的 Pods时，Pods 的 PersistentVolumes 会被重新挂载。 +当 `StatefulSet` 重新创建它的 Pods 时，Pods 的 PersistentVolumes 会被重新挂载。 ``` NAME STATUS VOLUME CAPACITY ACCESSMODES AGE @@ -635,8 +667,8 @@ datadir-zk-2 Bound pvc-bee0817e-bcb1-11e6-994f-42010a800002 20Gi R - -StatefulSet 的容器 `template` 中的 `volumeMounts` 一节使得 PersistentVolumes 被挂载到 ZooKeeper 服务器的数据目录。 +StatefulSet 的容器 `template` 中的 `volumeMounts` 一节使得 +PersistentVolumes 被挂载到 ZooKeeper 服务器的数据目录。 ```shell volumeMounts: @@ -650,11 +682,13 @@ same `PersistentVolume` mounted to the ZooKeeper server's data directory. Even when the Pods are rescheduled, all the writes made to the ZooKeeper servers' WALs, and all their snapshots, remain durable. --> - -当 `zk` StatefulSet 中的一个 Pod 被（重新）调度时，它总是拥有相同的 PersistentVolume，挂载到 ZooKeeper 服务器的数据目录。即使在 Pods 被重新调度时，所有对 ZooKeeper 服务器的 WALs 的写入和它们的全部快照都仍然是持久的。 +当 `zk` StatefulSet 中的一个 Pod 被（重新）调度时，它总是拥有相同的 PersistentVolume， +挂载到 ZooKeeper 服务器的数据目录。 +即使在 Pods 被重新调度时，所有对 ZooKeeper 服务器的 WALs 的写入和它们的 +全部快照都仍然是持久的。 - ## 确保一致性配置 -如同在 [促成 leader 选举](#促成-Leader-选举) 和 [达成一致](#达成一致) 小节中提到的，ZooKeeper ensemble 中的服务器需要一致性的配置来选举一个 leader 并形成一个 quorum。它们还需要 Zab 协议的一致性配置来保证这个协议在网络中正确的工作。在这次的样例中，我们通过直接将配置写入代码清单中来达到该目的。 +如同在[促成领导者选举](#facilitating-leader-election) 和[达成一致](#achieving-consensus) +小节中提到的，ZooKeeper ensemble 中的服务器需要一致性的配置来选举一个领导者并形成一个 +quorum。它们还需要 Zab 协议的一致性配置来保证这个协议在网络中正确的工作。 +在这次的示例中，我们通过直接将配置写入代码清单中来达到该目的。 获取 `zk` StatefulSet。 @@ -677,8 +713,8 @@ Get the `zk` StatefulSet. kubectl get sts zk -o yaml ``` ``` -… -command: + ... + command: - sh - -c - "start-zookeeper \ @@ -699,14 +735,14 @@ command: --max_session_timeout=40000 \ --min_session_timeout=4000 \ --log_level=INFO" -… +... ``` - -用于启动 ZooKeeper 服务器的命令将这些配置作为命令行参数传给了 ensemble。你也可以通过环境变量来传入这些配置。 +用于启动 ZooKeeper 服务器的命令将这些配置作为命令行参数传给了 ensemble。 +你也可以通过环境变量来传入这些配置。 +### 配置日志 {#configuring-logging} -### 配置日志 - -`zkGenConfig.sh` 脚本产生的一个文件控制了 ZooKeeper 的日志行为。ZooKeeper 使用了 [Log4j](http://logging.apache.org/log4j/2.x/) 并默认使用基于文件大小和时间的滚动文件追加器作为日志配置。 +`zkGenConfig.sh` 脚本产生的一个文件控制了 ZooKeeper 的日志行为。 +ZooKeeper 使用了 [Log4j](http://logging.apache.org/log4j/2.x/) 并默认使用 +基于文件大小和时间的滚动文件追加器作为日志配置。 从 `zk` StatefulSet 的一个 Pod 中获取日志配置。 @@ -732,7 +769,6 @@ kubectl exec zk-0 cat /usr/etc/zookeeper/log4j.properties The logging configuration below will cause the ZooKeeper process to write all of its logs to the standard output file stream. --> - 下面的日志配置会使 ZooKeeper 进程将其所有的日志写入标志输出文件流中。 ``` @@ -753,11 +789,13 @@ standard out and standard error do not exhaust local storage media. Use [`kubectl logs`](/docs/reference/generated/kubectl/kubectl-commands/#logs) to retrieve the last 20 log lines from one of the Pods. --> +这是在容器里安全记录日志的最简单的方法。 +由于应用的日志被写入标准输出，Kubernetes 将会为你处理日志轮转。 +Kubernetes 还实现了一个智能保存策略，保证写入标准输出和标准错误流 +的应用日志不会耗尽本地存储媒介。 -这是在容器里安全记录日志的最简单的方法。由于应用的日志被写入标准输出，Kubernetes 将会为你处理日志轮转。Kubernetes 还实现了一个智能保存策略，保证写入标准输出和标准错误流的应用日志不会耗尽本地存储媒介。 - - -使用 [`kubectl logs`](/docs/reference/generated/kubectl/kubectl-commands/#logs) 从一个 Pod 中取回最后几行日志。 +使用命令 [`kubectl logs`](/docs/reference/generated/kubectl/kubectl-commands/#logs) +从一个 Pod 中取回最后 20 行日志。 ```shell kubectl logs zk-0 --tail 20 @@ -766,7 +804,6 @@ kubectl logs zk-0 --tail 20 - 使用 `kubectl logs` 或者从 Kubernetes Dashboard 可以查看写入到标准输出和标准错误流中的应用日志。 ``` @@ -793,18 +830,17 @@ You can view application logs written to standard out or standard error using `k ``` - -Kubernetes 支持与 [Stackdriver](/docs/tasks/debug-application-cluster/logging-stackdriver/) 和 [Elasticsearch and Kibana](/docs/tasks/debug-application-cluster/logging-elasticsearch-kibana/) 的整合以获得复杂但更为强大的日志功能。 -对于集群级别的日志输出与整合，可以考虑部署一个 [sidecar](https://kubernetes.io/blog/2015/06/the-distributed-system-toolkit-patterns) 容器。 +Kubernetes 支持与多种日志方案集成。你可以选择一个最适合你的集群和应用 +的日志解决方案。对于集群级别的日志输出与整合，可以考虑部署一个 +[边车容器](/zh/docs/concepts/cluster-administration/logging#sidecar-container-with-logging-agent) +来轮转和提供日志数据。 - ### 配置非特权用户 -在容器中允许应用以特权用户运行这条最佳实践是值得商讨的。如果你的组织要求应用以非特权用户运行，你可以使用 [SecurityContext](/zh/docs/tasks/configure-pod-container/security-context/) 控制运行容器入口点的用户。 +在容器中允许应用以特权用户运行这条最佳实践是值得商讨的。 +如果你的组织要求应用以非特权用户运行，你可以使用 +[SecurityContext](/zh/docs/tasks/configure-pod-container/security-context/) +控制运行容器入口点所使用的用户。 `zk` StatefulSet 的 Pod 的 `template` 包含了一个 `SecurityContext`。 @@ -833,8 +871,7 @@ corresponds to the zookeeper group. Get the ZooKeeper process information from the `zk-0` Pod. --> - -在 Pods 的容器内部，UID 1000 对应用户 zookeeper，GID 1000对应用户组 zookeeper。 +在 Pods 的容器内部，UID 1000 对应用户 zookeeper，GID 1000 对应用户组 zookeeper。 从 `zk-0` Pod 获取 ZooKeeper 进程信息。 @@ -846,8 +883,8 @@ kubectl exec zk-0 -- ps -elf As the `runAsUser` field of the `securityContext` object is set to 1000, instead of running as root, the ZooKeeper process runs as the zookeeper user. --> - -由于 `securityContext` 对象的 `runAsUser` 字段被设置为1000而不是 root，ZooKeeper 进程将以 zookeeper 用户运行。 +由于 `securityContext` 对象的 `runAsUser` 字段被设置为 1000 而不是 root， +ZooKeeper 进程将以 zookeeper 用户运行。 ``` F S UID PID PPID C PRI NI ADDR SZ WCHAN STIME TTY TIME CMD @@ -860,8 +897,8 @@ By default, when the Pod's PersistentVolumes is mounted to the ZooKeeper server' Use the command below to get the file permissions of the ZooKeeper data directory on the `zk-0` Pod. --> - -默认情况下，当 Pod 的 PersistentVolume 被挂载到 ZooKeeper 服务器的数据目录时，它只能被 root 用户访问。这个配置将阻止 ZooKeeper 进程写入它的 WAL 及保存快照。 +默认情况下，当 Pod 的 PersistentVolume 被挂载到 ZooKeeper 服务器的数据目录时， +它只能被 root 用户访问。这个配置将阻止 ZooKeeper 进程写入它的 WAL 及保存快照。 在 `zk-0` Pod 上获取 ZooKeeper 数据目录的文件权限。 @@ -872,8 +909,9 @@ kubectl exec -ti zk-0 -- ls -ld /var/lib/zookeeper/data - -由于 `securityContext` 对象的 `fsGroup` 字段设置为1000，Pods 的 PersistentVolumes 的所有权属于 zookeeper 用户组，因而 ZooKeeper 进程能够成功的读写数据。 +由于 `securityContext` 对象的 `fsGroup` 字段设置为 1000，Pods 的 +PersistentVolumes 的所有权属于 zookeeper 用户组，因而 ZooKeeper +进程能够成功地读写数据。 ``` drwxr-sr-x 3 zookeeper zookeeper 4096 Dec 5 20:45 /var/lib/zookeeper/data @@ -890,19 +928,19 @@ common pattern. When deploying an application in Kubernetes, rather than using an external utility as a supervisory process, you should use Kubernetes as the watchdog for your application. --> - ## 管理 ZooKeeper 进程 -[ZooKeeper documentation](https://zookeeper.apache.org/doc/current/zookeeperAdmin.html#sc_supervision) 文档指出“你将需要一个监管程序用于管理每个 ZooKeeper 服务进程（JVM）”。在分布式系统中，使用一个看门狗（监管程序）来重启故障进程是一种常用的模式。 +[ZooKeeper 文档](https://zookeeper.apache.org/doc/current/zookeeperAdmin.html#sc_supervision) +指出“你将需要一个监管程序用于管理每个 ZooKeeper 服务进程（JVM）”。 +在分布式系统中，使用一个看门狗（监管程序）来重启故障进程是一种常用的模式。 - ### 更新 Ensemble `zk` `StatefulSet` 的更新策略被设置为了 `RollingUpdate`。 @@ -912,6 +950,7 @@ You can use `kubectl patch` to update the number of `cpus` allocated to the serv ```shell kubectl patch sts zk --type='json' -p='[{"op": "replace", "path": "/spec/template/spec/containers/0/resources/requests/cpu", "value":"0.3"}]' ``` + ``` statefulset.apps/zk patched ``` @@ -919,12 +958,12 @@ statefulset.apps/zk patched - 使用 `kubectl rollout status` 观测更新状态。 ```shell kubectl rollout status sts/zk ``` + ``` waiting for statefulset rolling update to complete 0 pods at revision zk-5db4499664... Waiting for 1 pods to be ready... @@ -943,8 +982,8 @@ This terminates the Pods, one at a time, in reverse ordinal order, and recreates Use the `kubectl rollout history` command to view a history or previous configurations. --> - -这项操作会逆序地依次终止每一个 Pod，并用新的配置重新创建。这样做确保了在滚动更新的过程中 quorum 依旧保持工作。 +这项操作会逆序地依次终止每一个 Pod，并用新的配置重新创建。 +这样做确保了在滚动更新的过程中 quorum 依旧保持工作。 使用 `kubectl rollout history` 命令查看历史或先前的配置。 @@ -962,7 +1001,6 @@ REVISION - 使用 `kubectl rollout undo` 命令撤销这次的改动。 ```shell @@ -974,7 +1012,7 @@ statefulset.apps/zk rolled back ``` - ### 处理进程故障 -[Restart Policies](/zh/docs/concepts/workloads/pods/pod-lifecycle/#restart-policy) 控制 Kubernetes 如何处理一个 Pod 中容器入口点的进程故障。对于 StatefulSet 中的 Pods 来说，Always 是唯一合适的 RestartPolicy，这也是默认值。你应该**绝不**覆盖 stateful 应用的默认策略。 +[重启策略](/zh/docs/concepts/workloads/pods/pod-lifecycle/#restart-policy) +控制 Kubernetes 如何处理一个 Pod 中容器入口点的进程故障。 +对于 StatefulSet 中的 Pods 来说，Always 是唯一合适的 RestartPolicy，也是默认值。 +你应该**绝不**覆盖有状态应用的默认策略。 检查 `zk-0` Pod 中运行的 ZooKeeper 服务器的进程树。 @@ -999,8 +1039,8 @@ kubectl exec zk-0 -- ps -ef The command used as the container's entry point has PID 1, and the ZooKeeper process, a child of the entry point, has PID 27. --> - -作为容器入口点的命令的 PID 为 1，Zookeeper 进程是入口点的子进程，PID 为27。 +作为容器入口点的命令的 PID 为 1，Zookeeper 进程是入口点的子进程， +PID 为 27。 ``` UID PID PPID C STIME TTY TIME CMD @@ -1011,8 +1051,7 @@ zookeep+ 27 1 0 15:03 ? 00:00:03 /usr/lib/jvm/java-8-openjdk-amd6 - -在一个终端观察 `zk` StatefulSet 中的 Pods。 +在一个终端观察 `zk` `StatefulSet` 中的 Pods。 ```shell kubectl get pod -w -l app=zk @@ -1021,7 +1060,6 @@ kubectl get pod -w -l app=zk - 在另一个终端杀掉 Pod `zk-0` 中的 ZooKeeper 进程。 ```shell @@ -1031,8 +1069,8 @@ In another terminal, terminate the ZooKeeper process in Pod `zk-0` with the foll - -ZooKeeper 进程的终结导致了它父进程的终止。由于容器的 RestartPolicy 是 Always，父进程被重启。 +ZooKeeper 进程的终结导致了它父进程的终止。由于容器的 `RestartPolicy` +是 Always，父进程被重启。 ``` NAME READY STATUS RESTARTS AGE @@ -1051,11 +1089,12 @@ that implements the application's business logic, the script must terminate with child process. This ensures that Kubernetes will restart the application's container when the process implementing the application's business logic fails. --> - -如果你的应用使用一个脚本（例如 zkServer.sh）来启动一个实现了应用业务逻辑的进程，这个脚本必须和子进程一起结束。这保证了当实现应用业务逻辑的进程故障时，Kubernetes 会重启这个应用的容器。 +如果你的应用使用一个脚本（例如 `zkServer.sh`）来启动一个实现了应用业务逻辑的进程， +这个脚本必须和子进程一起结束。这保证了当实现应用业务逻辑的进程故障时， +Kubernetes 会重启这个应用的容器。 - ### 存活性测试 -你的应用配置为自动重启故障进程，但这对于保持一个分布式系统的健康来说是不够的。许多场景下，一个系统进程可以是活动状态但不响应请求，或者是不健康状态。你应该使用 liveness probes 来通知 Kubernetes 你的应用进程处于不健康状态，需要被重启。 +你的应用配置为自动重启故障进程，但这对于保持一个分布式系统的健康来说是不够的。 +许多场景下，一个系统进程可以是活动状态但不响应请求，或者是不健康状态。 +你应该使用存活性探针来通知 Kubernetes 你的应用进程处于不健康状态，需要被重启。 `zk` StatefulSet 的 Pod 的 `template` 一节指定了一个存活探针。 ```yaml livenessProbe: - exec: - command: - - sh - - -c - - "zookeeper-ready 2181" - initialDelaySeconds: 15 - timeoutSeconds: 5 + exec: + command: + - sh + - -c + - "zookeeper-ready 2181" + initialDelaySeconds: 15 + timeoutSeconds: 5 ``` - -这个探针调用一个简单的 bash 脚本，使用 ZooKeeper 的四字缩写 `ruok` 来测试服务器的健康状态。 +这个探针调用一个简单的 Bash 脚本，使用 ZooKeeper 的四字缩写 `ruok` +来测试服务器的健康状态。 ``` OK=$(echo ruok | nc 127.0.0.1 $1) @@ -1102,8 +1142,7 @@ fi - -在一个终端窗口观察 `zk` StatefulSet 中的 Pods。 +在一个终端窗口中使用下面的命令观察 `zk` StatefulSet 中的 Pods。 ```shell kubectl get pod -w -l app=zk @@ -1112,7 +1151,6 @@ kubectl get pod -w -l app=zk - 在另一个窗口中，从 Pod `zk-0` 的文件系统中删除 `zookeeper-ready` 脚本。 ```shell @@ -1124,8 +1162,8 @@ When the liveness probe for the ZooKeeper process fails, Kubernetes will automatically restart the process for you, ensuring that unhealthy processes in the ensemble are restarted. --> - -当 ZooKeeper 进程的存活探针探测失败时，Kubernetes 将会为你自动重启这个进程，从而保证 ensemble 中不健康状态的进程都被重启。 +当 ZooKeeper 进程的存活探针探测失败时，Kubernetes 将会为你自动重启这个进程， +从而保证 ensemble 中不健康状态的进程都被重启。 ```shell kubectl get pod -w -l app=zk @@ -1143,28 +1181,32 @@ zk-0 1/1 Running 1 1h ``` +### 就绪性测试 +就绪不同于存活。如果一个进程是存活的，它是可调度和健康的。 +如果一个进程是就绪的，它应该能够处理输入。存活是就绪的必要非充分条件。 +在许多场景下，特别是初始化和终止过程中，一个进程可以是存活但没有就绪的。 + + +如果你指定了一个就绪探针，Kubernetes 将保证在就绪检查通过之前， +你的应用不会接收到网络流量。 -### 就绪性测试 - -就绪不同于存活。如果一个进程是存活的，它是可调度和健康的。如果一个进程是就绪的，它应该能够处理输入。存活是就绪的必要非充分条件。在许多场景下，特别是初始化和终止过程中，一个进程可以是存活但没有就绪的。 - -如果你指定了一个就绪探针，Kubernetes将保证在就绪检查通过之前，你的应用不会接收到网络流量。 - -对于一个 ZooKeeper 服务器来说，存活即就绪。因此 `zookeeper.yaml` 清单中的就绪探针和存活探针完全相同。 +对于一个 ZooKeeper 服务器来说，存活即就绪。 +因此 `zookeeper.yaml` 清单中的就绪探针和存活探针完全相同。 ```yaml readinessProbe: @@ -1182,11 +1224,11 @@ Even though the liveness and readiness probes are identical, it is important to specify both. This ensures that only healthy servers in the ZooKeeper ensemble receive network traffic. --> - -虽然存活探针和就绪探针是相同的，但同时指定它们两者仍然重要。这保证了 ZooKeeper ensemble 中只有健康的服务器能接收网络流量。 +虽然存活探针和就绪探针是相同的，但同时指定它们两者仍然重要。 +这保证了 ZooKeeper ensemble 中只有健康的服务器能接收网络流量。 +## 容忍节点故障 +ZooKeeper 需要一个 quorum 来提交数据变动。对于一个拥有 3 个服务器的 ensemble 来说， +必须有两个服务器是健康的，写入才能成功。 +在基于 quorum 的系统里，成员被部署在多个故障域中以保证可用性。 +为了防止由于某台机器断连引起服务中断，最佳实践是防止应用的多个实例在相同的机器上共存。 + + +默认情况下，Kubernetes 可以把 StatefulSet 的 Pods 部署在相同节点上。 +对于你创建的 3 个服务器的 ensemble 来说，如果有两个服务器并存于 +相同的节点上并且该节点发生故障时，ZooKeeper 服务将中断， +直至至少一个 Pods 被重新调度。 + - -## 容忍节点故障 - -ZooKeeper 需要一个 quorum 来提交数据变动。对于一个拥有 3 个服务器的 ensemble来说，必须有两个服务器是健康的，写入才能成功。在基于 quorum 的系统里，成员被部署在故障域之间以保证可用性。为了防止由于某台机器断连引起服务中断，最佳实践是防止应用的多个示例在相同的机器上共存。 - -默认情况下，Kubernetes 可以把 StatefulSet 的 Pods 部署在相同节点上。对于你创建的 3 个服务器的 ensemble 来说，如果有两个服务器并存于相同的节点上并且该节点发生故障时，ZooKeeper 服务将中断，直至至少一个 Pods 被重新调度。 - -你应该总是提供额外的容量以允许关键系统进程在节点故障时能够被重新调度。如果你这样做了，服务故障就只会持续到 Kubernetes 调度器重新调度某个 ZooKeeper 服务器为止。但是，如果希望你的服务在容忍节点故障时无停服时间，你应该设置 `podAntiAffinity`。 +你应该总是提供多余的容量以允许关键系统进程在节点故障时能够被重新调度。 +如果你这样做了，服务故障就只会持续到 Kubernetes 调度器重新调度某个 +ZooKeeper 服务器为止。 +但是，如果希望你的服务在容忍节点故障时无停服时间，你应该设置 `podAntiAffinity`。 获取 `zk` Stateful Set 中的 Pods 的节点。 @@ -1225,8 +1277,7 @@ for i in 0 1 2; do kubectl get pod zk-$i --template {{.spec.nodeName}}; echo ""; - -`zk` StatefulSe 中所有的 Pods 都被部署在不同的节点。 +`zk` `StatefulSet` 中所有的 Pods 都被部署在不同的节点。 ``` kubernetes-node-cxpk @@ -1238,19 +1289,19 @@ kubernetes-node-2g2d This is because the Pods in the `zk` `StatefulSet` have a `PodAntiAffinity` specified. --> -这是因为 `zk` StatefulSet 中的 Pods 指定了 `PodAntiAffinity`。 +这是因为 `zk` `StatefulSet` 中的 Pods 指定了 `PodAntiAffinity`。 ```yaml - affinity: - podAntiAffinity: - requiredDuringSchedulingIgnoredDuringExecution: - - labelSelector: - matchExpressions: - - key: "app" - operator: In - values: - - zk - topologyKey: "kubernetes.io/hostname" +affinity: + podAntiAffinity: + requiredDuringSchedulingIgnoredDuringExecution: + - labelSelector: + matchExpressions: + - key: "app" + operator: In + values: + - zk + topologyKey: "kubernetes.io/hostname" ``` - -`requiredDuringSchedulingIgnoredDuringExecution` 告诉 Kubernetes 调度器，在以 `topologyKey` 指定的域中，绝对不要把带有键为 `app`，值为 `zk` 的标签的两个 Pods 调度到相同的节点。`topologyKey` -`kubernetes.io/hostname` 表示这个域是一个单独的节点。使用不同的 rules、labels 和 selectors，你能够通过这种技术把你的 ensemble 分布在不同的物理、网络和电力故障域之间。 +`requiredDuringSchedulingIgnoredDuringExecution` 告诉 Kubernetes 调度器， +在以 `topologyKey` 指定的域中，绝对不要把带有键为 `app`、值为 `zk` 的标签 +的两个 Pods 调度到相同的节点。`topologyKey` `kubernetes.io/hostname` 表示 +这个域是一个单独的节点。 +使用不同的规则、标签和选择算符，你能够通过这种技术把你的 ensemble 分布 +在不同的物理、网络和电力故障域之间。 +## 节点维护期间保持应用可用 -## 存活管理 +**在本节中你将会隔离（Cordon）和腾空（Drain）节点。 +如果你是在一个共享的集群里使用本教程，请保证不会影响到其他租户。** -**在本节中你将会 cordon 和 drain 节点。如果你是在一个共享的集群里使用本教程，请保证不会影响到其他租户** +上一小节展示了如何在节点之间分散 Pods 以在计划外的节点故障时保证服务存活。 +但是你也需要为计划内维护引起的临时节点故障做准备。 -上一小节展示了如何在节点之间分散 Pods 以在计划外的节点故障时保证服务存活。但是你也需要为计划内维护引起的临时节点故障做准备。 - -获取你集群中的节点。 +使用此命令获取你的集群中的节点。 ```shell kubectl get nodes @@ -1295,7 +1350,8 @@ Use [`kubectl cordon`](/docs/reference/generated/kubectl/kubectl-commands/#cordo cordon all but four of the nodes in your cluster. --> -使用 [`kubectl cordon`](/docs/reference/generated/kubectl/kubectl-commands/#cordon) cordon 你的集群中除4个节点以外的所有节点。 +使用 [`kubectl cordon`](/docs/reference/generated/kubectl/kubectl-commands/#cordon) +隔离你的集群中除 4 个节点以外的所有节点。 ```shell kubectl cordon @@ -1304,8 +1360,7 @@ kubectl cordon - -获取 `zk-pdb` `PodDisruptionBudget`。 +使用下面的命令获取 `zk-pdb` `PodDisruptionBudget`。 ```shell kubectl get pdb zk-pdb @@ -1315,8 +1370,8 @@ kubectl get pdb zk-pdb The `max-unavailable` field indicates to Kubernetes that at most one Pod from `zk` `StatefulSet` can be unavailable at any time. --> - -`max-unavailable` 字段指示 Kubernetes 在任何时候，`zk` `StatefulSet` 至多有一个 Pod 是不可用的。 +`max-unavailable` 字段指示 Kubernetes 在任何时候，`zk` `StatefulSet` +至多有一个 Pod 是不可用的。 ``` NAME MIN-AVAILABLE MAX-UNAVAILABLE ALLOWED-DISRUPTIONS AGE @@ -1326,8 +1381,7 @@ zk-pdb N/A 1 1 - -在一个终端观察 `zk` `StatefulSet` 中的 Pods。 +在一个终端中，使用下面的命令观察 `zk` `StatefulSet` 中的 Pods。 ```shell kubectl get pods -w -l app=zk @@ -1337,7 +1391,7 @@ kubectl get pods -w -l app=zk In another terminal, use this command to get the nodes that the Pods are currently scheduled on. --> -在另一个终端获取 Pods 当前调度的节点。 +在另一个终端中，使用下面的命令获取 Pods 当前调度的节点。 ```shell for i in 0 1 2; do kubectl get pod zk-$i --template {{.spec.nodeName}}; echo ""; done @@ -1354,7 +1408,8 @@ Use [`kubectl drain`](/docs/reference/generated/kubectl/kubectl-commands/#drain) drain the node on which the `zk-0` Pod is scheduled. --> -使用 [`kubectl drain`](/docs/reference/generated/kubectl/kubectl-commands/#drain) 来 cordon 和 drain `zk-0` Pod 调度的节点。 +使用 [`kubectl drain`](/docs/reference/generated/kubectl/kubectl-commands/#drain) +来隔离和腾空 `zk-0` Pod 调度所在的节点。 ```shell kubectl drain $(kubectl get pod zk-0 --template {{.spec.nodeName}}) --ignore-daemonsets --force --delete-local-data @@ -1372,8 +1427,7 @@ node "kubernetes-node-pb41" drained As there are four nodes in your cluster, `kubectl drain`, succeeds and the `zk-0` is rescheduled to another node. --> - -由于你的集群中有4个节点, `kubectl drain` 执行成功，`zk-0 被调度到其它节点。 +由于你的集群中有 4 个节点, `kubectl drain` 执行成功，`zk-0` 被调度到其它节点。 ``` NAME READY STATUS RESTARTS AGE @@ -1396,8 +1450,7 @@ zk-0 1/1 Running 0 1m Keep watching the `StatefulSet`'s Pods in the first terminal and drain the node on which `zk-1` is scheduled. --> - -在第一个终端持续观察 StatefulSet 的 Pods并 drain `zk-1` 调度的节点。 +在第一个终端中持续观察 StatefulSet 的 Pods 并腾空 `zk-1` 调度所在的节点。 ```shell kubectl drain $(kubectl get pod zk-1 --template {{.spec.nodeName}}) --ignore-daemonsets --force --delete-local-data "kubernetes-node-ixsl" cordoned @@ -1413,42 +1466,42 @@ node "kubernetes-node-ixsl" drained The `zk-1` Pod cannot be scheduled because the `zk` `StatefulSet` contains a `PodAntiAffinity` rule preventing co-location of the Pods, and as only two nodes are schedulable, the Pod will remain in a Pending state. --> - -`zk-1` Pod 不能被调度。由于 `zk` StatefulSet 包含了一个防止 Pods 共存的 PodAntiAffinity 规则，而且只有两个节点可用于调度，这个 Pod 将保持在 Pending 状态。 +`zk-1` Pod 不能被调度，这是因为 `zk` `StatefulSet` 包含了一个防止 Pods +共存的 PodAntiAffinity 规则，而且只有两个节点可用于调度， +这个 Pod 将保持在 Pending 状态。 ```shell kubectl get pods -w -l app=zk ``` ``` -NAME READY STATUS RESTARTS AGE -zk-0 1/1 Running 2 1h -zk-1 1/1 Running 0 1h -zk-2 1/1 Running 0 1h -NAME READY STATUS RESTARTS AGE -zk-0 1/1 Terminating 2 2h -zk-0 0/1 Terminating 2 2h -zk-0 0/1 Terminating 2 2h -zk-0 0/1 Terminating 2 2h -zk-0 0/1 Pending 0 0s -zk-0 0/1 Pending 0 0s -zk-0 0/1 ContainerCreating 0 0s -zk-0 0/1 Running 0 51s -zk-0 1/1 Running 0 1m -zk-1 1/1 Terminating 0 2h -zk-1 0/1 Terminating 0 2h -zk-1 0/1 Terminating 0 2h -zk-1 0/1 Terminating 0 2h -zk-1 0/1 Pending 0 0s -zk-1 0/1 Pending 0 0s +NAME READY STATUS RESTARTS AGE +zk-0 1/1 Running 2 1h +zk-1 1/1 Running 0 1h +zk-2 1/1 Running 0 1h +NAME READY STATUS RESTARTS AGE +zk-0 1/1 Terminating 2 2h +zk-0 0/1 Terminating 2 2h +zk-0 0/1 Terminating 2 2h +zk-0 0/1 Terminating 2 2h +zk-0 0/1 Pending 0 0s +zk-0 0/1 Pending 0 0s +zk-0 0/1 ContainerCreating 0 0s +zk-0 0/1 Running 0 51s +zk-0 1/1 Running 0 1m +zk-1 1/1 Terminating 0 2h +zk-1 0/1 Terminating 0 2h +zk-1 0/1 Terminating 0 2h +zk-1 0/1 Terminating 0 2h +zk-1 0/1 Pending 0 0s +zk-1 0/1 Pending 0 0s ``` - -继续观察 stateful set 的 Pods 并 drain `zk-2` 调度的节点。 +继续观察 StatefulSet 中的 Pods 并腾空 `zk-2` 调度所在的节点。 ```shell kubectl drain $(kubectl get pod zk-2 --template {{.spec.nodeName}}) --ignore-daemonsets --force --delete-local-data @@ -1469,11 +1522,10 @@ You cannot drain the third node because evicting `zk-2` would violate `zk-budget Use `zkCli.sh` to retrieve the value you entered during the sanity test from `zk-0`. --> - 使用 `CRTL-C` 终止 kubectl。 -你不能 drain 第三个节点，因为删除 `zk-2` 将和 `zk-budget` 冲突。然而这个节点仍然保持 cordoned。 - +你不能腾空第三个节点，因为驱逐 `zk-2` 将和 `zk-budget` 冲突。 +然而这个节点仍然处于隔离状态（Cordoned）。 使用 `zkCli.sh` 从 `zk-0` 取回你的健康检查中输入的数值。 @@ -1484,7 +1536,6 @@ kubectl exec zk-0 zkCli.sh get /hello - 由于遵守了 PodDisruptionBudget，服务仍然可用。 ``` @@ -1506,12 +1557,13 @@ numChildren = 0 - -使用 [`kubectl uncordon`](/docs/reference/generated/kubectl/kubectl-commands/#uncordon) 来取消对第一个节点的隔离。 +使用 [`kubectl uncordon`](/docs/reference/generated/kubectl/kubectl-commands/#uncordon) +来取消对第一个节点的隔离。 ```shell kubectl uncordon kubernetes-node-pb41 ``` + ``` node "kubernetes-node-pb41" uncordoned ``` @@ -1519,44 +1571,43 @@ node "kubernetes-node-pb41" uncordoned - `zk-1` 被重新调度到了这个节点。等待 `zk-1` 变为 Running 和 Ready 状态。 ```shell kubectl get pods -w -l app=zk ``` + ``` -NAME READY STATUS RESTARTS AGE -zk-0 1/1 Running 2 1h -zk-1 1/1 Running 0 1h -zk-2 1/1 Running 0 1h -NAME READY STATUS RESTARTS AGE -zk-0 1/1 Terminating 2 2h -zk-0 0/1 Terminating 2 2h -zk-0 0/1 Terminating 2 2h -zk-0 0/1 Terminating 2 2h -zk-0 0/1 Pending 0 0s -zk-0 0/1 Pending 0 0s -zk-0 0/1 ContainerCreating 0 0s -zk-0 0/1 Running 0 51s -zk-0 1/1 Running 0 1m -zk-1 1/1 Terminating 0 2h -zk-1 0/1 Terminating 0 2h -zk-1 0/1 Terminating 0 2h -zk-1 0/1 Terminating 0 2h -zk-1 0/1 Pending 0 0s -zk-1 0/1 Pending 0 0s -zk-1 0/1 Pending 0 12m -zk-1 0/1 ContainerCreating 0 12m -zk-1 0/1 Running 0 13m -zk-1 1/1 Running 0 13m +NAME READY STATUS RESTARTS AGE +zk-0 1/1 Running 2 1h +zk-1 1/1 Running 0 1h +zk-2 1/1 Running 0 1h +NAME READY STATUS RESTARTS AGE +zk-0 1/1 Terminating 2 2h +zk-0 0/1 Terminating 2 2h +zk-0 0/1 Terminating 2 2h +zk-0 0/1 Terminating 2 2h +zk-0 0/1 Pending 0 0s +zk-0 0/1 Pending 0 0s +zk-0 0/1 ContainerCreating 0 0s +zk-0 0/1 Running 0 51s +zk-0 1/1 Running 0 1m +zk-1 1/1 Terminating 0 2h +zk-1 0/1 Terminating 0 2h +zk-1 0/1 Terminating 0 2h +zk-1 0/1 Terminating 0 2h +zk-1 0/1 Pending 0 0s +zk-1 0/1 Pending 0 0s +zk-1 0/1 Pending 0 12m +zk-1 0/1 ContainerCreating 0 12m +zk-1 0/1 Running 0 13m +zk-1 1/1 Running 0 13m ``` - -尝试 drain `zk-2` 调度的节点。 +尝试腾空 `zk-2` 调度所在的节点。 ```shell kubectl drain $(kubectl get pod zk-2 --template {{.spec.nodeName}}) --ignore-daemonsets --force --delete-local-data @@ -1565,7 +1616,6 @@ kubectl drain $(kubectl get pod zk-2 --template {{.spec.nodeName}}) --ignore-dae - 输出： ``` @@ -1581,10 +1631,9 @@ This time `kubectl drain` succeeds. Uncordon the second node to allow `zk-2` to be rescheduled. --> - 这次 `kubectl drain` 执行成功。 -Uncordon 第二个节点以允许 `zk-2` 被重新调度。 +取消第二个节点的隔离，以允许 `zk-2` 被重新调度。 ```shell kubectl uncordon kubernetes-node-ixsl @@ -1600,19 +1649,20 @@ If drain is used to cordon nodes and evict pods prior to taking the node offline services that express a disruption budget will have that budget respected. You should always allocate additional capacity for critical services so that their Pods can be immediately rescheduled. --> - -你可以同时使用 `kubectl drain` 和 `PodDisruptionBudgets` 来保证你的服务在维护过程中仍然可用。如果使用了 drain 来隔离节点并在节点离线之前排出了 pods，那么表达了 disruption budget 的服务将会遵守该 budget。你应该总是为关键服务分配额外容量，这样它们的 Pods 就能够迅速的重新调度。 +你可以同时使用 `kubectl drain` 和 `PodDisruptionBudgets` 来保证你的服务 +在维护过程中仍然可用。如果使用了腾空操作来隔离节点并在节点离线之前驱逐了 pods， +那么设置了干扰预算的服务将会遵守该预算。 +你应该总是为关键服务分配额外容量，这样它们的 Pods 就能够迅速的重新调度。 ## {{% heading "cleanup" %}} - - * 使用 `kubectl uncordon` 解除你集群中所有节点的隔离。 -* 你需要删除在本教程中使用的 PersistentVolumes 的持久存储媒介。请遵循必须的步骤，基于你的环境、存储配置和准备方法，保证回收所有的存储。 +* 你需要删除在本教程中使用的 PersistentVolumes 的持久存储媒介。 + 请遵循必须的步骤，基于你的环境、存储配置和制备方法，保证回收所有的存储。 + From ec4df0f2cf8016093c5c7d7e9a1f55960694c1cd Mon Sep 17 00:00:00 2001 From: Jordan Liggitt Date: Sun, 14 Feb 2021 15:27:04 -0500 Subject: [PATCH 22/62] Address review comments --- .../en/docs/contribute/style/style-guide.md | 4 ++ .../reference/using-api/deprecation-guide.md | 69 ++++++++++--------- 2 files changed, 41 insertions(+), 32 deletions(-) diff --git a/content/en/docs/contribute/style/style-guide.md b/content/en/docs/contribute/style/style-guide.md index 0047799fb0..2f0c39168c 100644 --- a/content/en/docs/contribute/style/style-guide.md +++ b/content/en/docs/contribute/style/style-guide.md @@ -576,6 +576,10 @@ Avoid making promises or giving hints about the future. If you need to talk abou an alpha feature, put the text under a heading that identifies it as alpha information. +An exception to this rule is documentation about announced deprecations +targeting removal in future versions. One example of documentation like this +is the [Deprecated API migration guide](/docs/reference/using-api/deprecation-guide/). + ### Avoid statements that will soon be out of date Avoid words like "currently" and "new." A feature that is new today might not be diff --git a/content/en/docs/reference/using-api/deprecation-guide.md b/content/en/docs/reference/using-api/deprecation-guide.md index 10ef2a642b..e885646f2f 100755 --- a/content/en/docs/reference/using-api/deprecation-guide.md +++ b/content/en/docs/reference/using-api/deprecation-guide.md @@ -6,7 +6,7 @@ reviewers: - smarterclayton title: "Deprecated API Migration Guide" weight: 45 -content_type: api_reference +content_type: reference --- @@ -25,15 +25,23 @@ deprecated API versions to newer and more stable API versions. The **v1.25** release will stop serving the following deprecated API versions: -#### Event +#### Event {#event-v125} The **events.k8s.io/v1beta1** API version of Event will no longer be served in v1.25. * Migrate manifests and API clients to use the **events.k8s.io/v1** API version, available since v1.19. * All existing persisted objects are accessible via the new API -* Notable changes +* Notable changes in **events.k8s.io/v1**: + * `type` is limited to `Normal` and `Warning` + * `involvedObject` is renamed to `regarding` + * `action`, `reason`, `reportingComponent`, and `reportingInstance` are required when creating new **events.k8s.io/v1** Events + * use `eventTime` instead of the deprecated `firstTimestamp` field (which is renamed to `deprecatedFirstTimestamp` and not permitted in new **events.k8s.io/v1** Events) + * use `series.lastObservedTime` instead of the deprecated `lastTimestamp` field (which is renamed to `deprecatedLastTimestamp` and not permitted in new **events.k8s.io/v1** Events) + * use `series.count` instead of the deprecated `count` field (which is renamed to `deprecatedCount` and not permitted in new **events.k8s.io/v1** Events) + * use `reportingComponent` instead of the deprecated `source.component` field (which is renamed to `deprecatedSource.component` and not permitted in new **events.k8s.io/v1** Events) + * use `reportingInstance` instead of the deprecated `source.host` field (which is renamed to `deprecatedSource.host` and not permitted in new **events.k8s.io/v1** Events) -#### RuntimeClass +#### RuntimeClass {#runtimeclass-v125} RuntimeClass in the **node.k8s.io/v1beta1** API version will no longer be served in v1.25. @@ -45,12 +53,12 @@ RuntimeClass in the **node.k8s.io/v1beta1** API version will no longer be served The **v1.22** release will stop serving the following deprecated API versions: -#### MutatingWebhookConfiguration and ValidatingWebhookConfiguration +#### Webhook resources {#webhook-resources-v122} The **admissionregistration.k8s.io/v1beta1** API version of MutatingWebhookConfiguration and ValidatingWebhookConfiguration will no longer be served in v1.22. * Migrate manifests and API clients to use the **admissionregistration.k8s.io/v1** API version, available since v1.16. -* All existing persisted objects are accessible via the new API +* All existing persisted objects are accessible via the new APIs * Notable changes: * `webhooks[*].failurePolicy` default changed from `Ignore` to `Fail` for v1 * `webhooks[*].matchPolicy` default changed from `Exact` to `Equivalent` for v1 @@ -59,7 +67,7 @@ The **admissionregistration.k8s.io/v1beta1** API version of MutatingWebhookConfi * `webhooks[*].admissionReviewVersions` default value is removed and the field made required for v1 (supported versions for AdmissionReview are `v1` and `v1beta1`) * `webhooks[*].name` must be unique in the list for objects created via `admissionregistration.k8s.io/v1` -#### CustomResourceDefinitions +#### CustomResourceDefinition {#customresourcedefinition-v122} The **apiextensions.k8s.io/v1beta1** API version of CustomResourceDefinition will no longer be served in v1.22. @@ -73,11 +81,11 @@ The **apiextensions.k8s.io/v1beta1** API version of CustomResourceDefinition wil * `spec.additionalPrinterColumns` is removed in v1; use `spec.versions[*].additionalPrinterColumns` instead * `spec.conversion.webhookClientConfig` is moved to `spec.conversion.webhook.clientConfig` in v1 * `spec.conversion.conversionReviewVersions` is moved to `spec.conversion.webhook.conversionReviewVersions` in v1 - * `spec.versions[*].schema.openAPIV3Schema` is now required when creating v1 CustomResourceDefinitions, and must be a [structural schema](/docs/tasks/extend-kubernetes/custom-resources/custom-resource-definitions/#specifying-a-structural-schema) - * `spec.preserveUnknownFields: true` is disallowed when creating v1 CustomResourceDefinitions; it must be specified within schema definitions as `x-kubernetes-preserve-unknown-fields: true` + * `spec.versions[*].schema.openAPIV3Schema` is now required when creating v1 CustomResourceDefinition objects, and must be a [structural schema](/docs/tasks/extend-kubernetes/custom-resources/custom-resource-definitions/#specifying-a-structural-schema) + * `spec.preserveUnknownFields: true` is disallowed when creating v1 CustomResourceDefinition objects; it must be specified within schema definitions as `x-kubernetes-preserve-unknown-fields: true` * In `additionalPrinterColumns` items, the `JSONPath` field was renamed to `jsonPath` in v1 (fixes [#66531](https://github.com/kubernetes/kubernetes/issues/66531)) -#### APIService +#### APIService {#apiservice-v122} The **apiregistration.k8s.io/v1beta1** API version of APIService will no longer be served in v1.22. @@ -85,14 +93,14 @@ The **apiregistration.k8s.io/v1beta1** API version of APIService will no longer * All existing persisted objects are accessible via the new API * No notable changes -#### TokenReview +#### TokenReview {#tokenreview-v122} The **authentication.k8s.io/v1beta1** API version of TokenReview will no longer be served in v1.22. * Migrate manifests and API clients to use the **authentication.k8s.io/v1** API version, available since v1.6. * No notable changes -#### SubjectAccessReview +#### SubjectAccessReview resources {#subjectaccessreview-resources-v122} The **authorization.k8s.io/v1beta1** API version of LocalSubjectAccessReview, SelfSubjectAccessReview, and SubjectAccessReview will no longer be served in v1.22. @@ -100,7 +108,7 @@ The **authorization.k8s.io/v1beta1** API version of LocalSubjectAccessReview, Se * Notable changes: * `spec.group` was renamed to `spec.groups` in v1 (fixes [#32709](https://github.com/kubernetes/kubernetes/issues/32709)) -#### CertificateSigningRequest +#### CertificateSigningRequest {#certificatesigningrequest-v122} The **certificates.k8s.io/v1beta1** API version of CertificateSigningRequest will no longer be served in v1.22. @@ -115,7 +123,7 @@ The **certificates.k8s.io/v1beta1** API version of CertificateSigningRequest wil * `status.conditions[*].status` is now required * `status.certificate` must be PEM-encoded, and contain only `CERTIFICATE` blocks -#### Lease +#### Lease {#lease-v122} The **coordination.k8s.io/v1beta1** API version of Lease will no longer be served in v1.22. @@ -123,7 +131,7 @@ The **coordination.k8s.io/v1beta1** API version of Lease will no longer be serve * All existing persisted objects are accessible via the new API * No notable changes -#### Ingress +#### Ingress {#ingress-v122} The **extensions/v1beta1** and **networking.k8s.io/v1beta1** API versions of Ingress will no longer be served in v1.22. @@ -136,7 +144,7 @@ The **extensions/v1beta1** and **networking.k8s.io/v1beta1** API versions of Ing * String backend `servicePort` fields are renamed to `service.port.name` * `pathType` is now required for each specified path. Options are `Prefix`, `Exact`, and `ImplementationSpecific`. To match the undefined `v1beta1` behavior, use `ImplementationSpecific`. -#### IngressClass +#### IngressClass {#ingressclass-v122} The **networking.k8s.io/v1beta1** API version of IngressClass will no longer be served in v1.22. @@ -144,15 +152,15 @@ The **networking.k8s.io/v1beta1** API version of IngressClass will no longer be * All existing persisted objects are accessible via the new API * No notable changes -#### RBAC +#### RBAC resources {#rbac-resources-v122} The **rbac.authorization.k8s.io/v1beta1** API version of ClusterRole, ClusterRoleBinding, Role, and RoleBinding will no longer be served in v1.22. * Migrate manifests and API clients to use the **networking.k8s.io/v1** API version, available since v1.8. -* All existing persisted objects are accessible via the new API +* All existing persisted objects are accessible via the new APIs * No notable changes -#### PriorityClass +#### PriorityClass {#priorityclass-v122} The **scheduling.k8s.io/v1beta1** API version of PriorityClass will no longer be served in v1.22. @@ -160,7 +168,7 @@ The **scheduling.k8s.io/v1beta1** API version of PriorityClass will no longer be * All existing persisted objects are accessible via the new API * No notable changes -#### Storage +#### Storage resources {#storage-resources-v122} The **storage.k8s.io/v1beta1** API version of CSIDriver, CSINode, StorageClass, and VolumeAttachment will no longer be served in v1.22. @@ -169,21 +177,21 @@ The **storage.k8s.io/v1beta1** API version of CSIDriver, CSINode, StorageClass, * CSINode is available in **storage.k8s.io/v1** since v1.17 * StorageClass is available in **storage.k8s.io/v1** since v1.6 * VolumeAttachment is available in **storage.k8s.io/v1** v1.13 -* All existing persisted objects are accessible via the new API +* All existing persisted objects are accessible via the new APIs * No notable changes ### v1.16 The **v1.16** release stopped serving the following deprecated API versions: -#### NetworkPolicy +#### NetworkPolicy {#networkpolicy-v116} The **extensions/v1beta1** API version of NetworkPolicy is no longer served as of v1.16. * Migrate manifests and API clients to use the **networking.k8s.io/v1** API version, available since v1.8. * All existing persisted objects are accessible via the new API -#### DaemonSet +#### DaemonSet {#daemonset-v116} The **extensions/v1beta1** and **apps/v1beta2** API versions of DaemonSet are no longer served as of v1.16. @@ -194,7 +202,7 @@ The **extensions/v1beta1** and **apps/v1beta2** API versions of DaemonSet are no * `spec.selector` is now required and immutable after creation; use the existing template labels as the selector for seamless upgrades * `spec.updateStrategy.type` now defaults to `RollingUpdate` (the default in `extensions/v1beta1` was `OnDelete`) -#### Deployment +#### Deployment {#deployment-v116} The **extensions/v1beta1**, **apps/v1beta1**, and **apps/v1beta2** API versions of Deployment are no longer served as of v1.16. @@ -207,7 +215,7 @@ The **extensions/v1beta1**, **apps/v1beta1**, and **apps/v1beta2** API versions * `spec.revisionHistoryLimit` now defaults to `10` (the default in `apps/v1beta1` was `2`, the default in `extensions/v1beta1` was to retain all) * `maxSurge` and `maxUnavailable` now default to `25%` (the default in `extensions/v1beta1` was `1`) -#### StatefulSet +#### StatefulSet {#statefulset-v116} The **apps/v1beta1** and **apps/v1beta2** API versions of StatefulSet are no longer served as of v1.16. @@ -217,7 +225,7 @@ The **apps/v1beta1** and **apps/v1beta2** API versions of StatefulSet are no lon * `spec.selector` is now required and immutable after creation; use the existing template labels as the selector for seamless upgrades * `spec.updateStrategy.type` now defaults to `RollingUpdate` (the default in `apps/v1beta1` was `OnDelete`) -#### ReplicaSet +#### ReplicaSet {#replicaset-v116} The **extensions/v1beta1**, **apps/v1beta1**, and **apps/v1beta2** API versions of ReplicaSet are no longer served as of v1.16. @@ -226,17 +234,14 @@ The **extensions/v1beta1**, **apps/v1beta1**, and **apps/v1beta2** API versions * Notable changes: * `spec.selector` is now required and immutable after creation; use the existing template labels as the selector for seamless upgrades -## What To Do - -Kubernetes 1.22 will be released later in 2021, so be sure to audit -your configuration and integrations now! +## What to do ### Test with deprecated APIs disabled You can test your clusters by starting an API server with specific API versions disabled to simulate upcoming removals. Add the following flag to the API server startup arguments: -`--runtime-config=$group/$version=false` +`--runtime-config=/=false` For example: @@ -262,4 +267,4 @@ to locate use of deprecated APIs. `kubectl-convert -f ./my-deployment.yaml --output-version apps/v1` Note that this may use non-ideal default values. To learn more about a specific - resource, check the Kubernetes [api reference](https://kubernetes.io/docs/reference/#api-reference). + resource, check the Kubernetes [API reference](/docs/reference/kubernetes-api/). From 0b4bbec54b8cee7e62f952562a6578806833d593 Mon Sep 17 00:00:00 2001 From: Arhell Date: Mon, 15 Feb 2021 00:42:08 +0200 Subject: [PATCH 23/62] update cheatsheet.md --- content/ru/docs/reference/kubectl/cheatsheet.md | 3 +++ 1 file changed, 3 insertions(+) diff --git a/content/ru/docs/reference/kubectl/cheatsheet.md b/content/ru/docs/reference/kubectl/cheatsheet.md index d2be7e9c0c..02a8a9bc4a 100644 --- a/content/ru/docs/reference/kubectl/cheatsheet.md +++ b/content/ru/docs/reference/kubectl/cheatsheet.md @@ -186,6 +186,9 @@ kubectl get pods --show-labels JSONPATH='{range .items[*]}{@.metadata.name}:{range @.status.conditions[*]}{@.type}={@.status};{end}{end}' \ && kubectl get nodes -o jsonpath="$JSONPATH" | grep "Ready=True" +# Вывод декодированных секретов без внешних инструментов +kubectl get secret my-secret -o go-template='{{range $k,$v := .data}}{{"### "}}{{$k}}{{"\n"}}{{$v|base64decode}}{{"\n\n"}}{{end}}' + # Вывести все секреты, используемые сейчас в поде. kubectl get pods -o json | jq '.items[].spec.containers[].env[]?.valueFrom.secretKeyRef.name' | grep -v null | sort | uniq From c6557c608e8b8eb0a06932a7b7b3612cef5b432f Mon Sep 17 00:00:00 2001 From: timyinshi Date: Mon, 15 Feb 2021 09:39:46 +0800 Subject: [PATCH 24/62] modify the chinese word of organizations Signed-off-by: timyinshi --- content/zh/docs/concepts/overview/what-is-kubernetes.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/content/zh/docs/concepts/overview/what-is-kubernetes.md b/content/zh/docs/concepts/overview/what-is-kubernetes.md index 3fa58f5da0..b976570edb 100644 --- a/content/zh/docs/concepts/overview/what-is-kubernetes.md +++ b/content/zh/docs/concepts/overview/what-is-kubernetes.md @@ -61,11 +61,11 @@ Early on, organizations ran applications on physical servers. There was no way t --> **传统部署时代：** -早期，组织在物理服务器上运行应用程序。无法为物理服务器中的应用程序定义资源边界，这会导致资源分配问题。 +早期，各个组织机构在物理服务器上运行应用程序。无法为物理服务器中的应用程序定义资源边界，这会导致资源分配问题。 例如，如果在物理服务器上运行多个应用程序，则可能会出现一个应用程序占用大部分资源的情况， 结果可能导致其他应用程序的性能下降。 一种解决方案是在不同的物理服务器上运行每个应用程序，但是由于资源利用不足而无法扩展， -并且组织维护许多物理服务器的成本很高。 +并且维护许多物理服务器的成本很高。 ## Prerequisites * Run etcd as a cluster of odd members. -* etcd is a leader-based distributed system. Ensure that the leader periodically send heartbeats on time to all followers to keep the cluster stable. +* etcd is a leader-based distributed system. Ensure that the leader + periodically send heartbeats on time to all followers to keep the cluster + stable. * Ensure that no resource starvation occurs. - Performance and stability of the cluster is sensitive to network and disk IO. Any resource starvation can lead to heartbeat timeout, causing instability of the cluster. An unstable etcd indicates that no leader is elected. Under such circumstances, a cluster cannot make any changes to its current state, which implies no new pods can be scheduled. + Performance and stability of the cluster is sensitive to network and disk + IO. Any resource starvation can lead to heartbeat timeout, causing instability + of the cluster. An unstable etcd indicates that no leader is elected. Under + such circumstances, a cluster cannot make any changes to its current state, + which implies no new pods can be scheduled. -* Keeping stable etcd clusters is critical to the stability of Kubernetes clusters. Therefore, run etcd clusters on dedicated machines or isolated environments for [guaranteed resource requirements](https://github.com/coreos/etcd/blob/master/Documentation/op-guide/hardware.md#hardware-recommendations). +* Keeping stable etcd clusters is critical to the stability of Kubernetes + clusters. Therefore, run etcd clusters on dedicated machines or isolated + environments for [guaranteed resource requirements](https://github.com/coreos/etcd/blob/master/Documentation/op-guide/hardware.md#hardware-recommendations). * The minimum recommended version of etcd to run in production is `3.2.10+`. ## Resource requirements -Operating etcd with limited resources is suitable only for testing purposes. For deploying in production, advanced hardware configuration is required. Before deploying etcd in production, see [resource requirement reference documentation](https://github.com/coreos/etcd/blob/master/Documentation/op-guide/hardware.md#example-hardware-configurations). +Operating etcd with limited resources is suitable only for testing purposes. +For deploying in production, advanced hardware configuration is required. +Before deploying etcd in production, see +[resource requirement reference documentation](https://github.com/coreos/etcd/blob/master/Documentation/op-guide/hardware.md#example-hardware-configurations). ## Starting etcd clusters @@ -50,33 +55,43 @@ Use a single-node etcd cluster only for testing purpose. 1. Run the following: - ```sh - ./etcd --listen-client-urls=http://$PRIVATE_IP:2379 --advertise-client-urls=http://$PRIVATE_IP:2379 - ``` + ```sh + etcd --listen-client-urls=http://$PRIVATE_IP:2379 \ + --advertise-client-urls=http://$PRIVATE_IP:2379 + ``` -2. Start Kubernetes API server with the flag `--etcd-servers=$PRIVATE_IP:2379`. +2. Start Kubernetes API server with the flag + `--etcd-servers=$PRIVATE_IP:2379`. - Replace `PRIVATE_IP` with your etcd client IP. + Make sure `PRIVATE_IP` is set to your etcd client IP. ### Multi-node etcd cluster -For durability and high availability, run etcd as a multi-node cluster in production and back it up periodically. A five-member cluster is recommended in production. For more information, see [FAQ Documentation](https://github.com/coreos/etcd/blob/master/Documentation/faq.md#what-is-failure-tolerance). +For durability and high availability, run etcd as a multi-node cluster in +production and back it up periodically. A five-member cluster is recommended +in production. For more information, see [FAQ +Documentation](https://github.com/coreos/etcd/blob/master/Documentation/faq.md#what-is-failure-tolerance). -Configure an etcd cluster either by static member information or by dynamic discovery. For more information on clustering, see [etcd Clustering Documentation](https://github.com/coreos/etcd/blob/master/Documentation/op-guide/clustering.md). +Configure an etcd cluster either by static member information or by dynamic +discovery. For more information on clustering, see [etcd Clustering +Documentation](https://github.com/coreos/etcd/blob/master/Documentation/op-guide/clustering.md). -For an example, consider a five-member etcd cluster running with the following client URLs: `http://$IP1:2379`, `http://$IP2:2379`, `http://$IP3:2379`, `http://$IP4:2379`, and `http://$IP5:2379`. To start a Kubernetes API server: +For an example, consider a five-member etcd cluster running with the following +client URLs: `http://$IP1:2379`, `http://$IP2:2379`, `http://$IP3:2379`, +`http://$IP4:2379`, and `http://$IP5:2379`. To start a Kubernetes API server: 1. Run the following: - ```sh - ./etcd --listen-client-urls=http://$IP1:2379, http://$IP2:2379, http://$IP3:2379, http://$IP4:2379, http://$IP5:2379 --advertise-client-urls=http://$IP1:2379, http://$IP2:2379, http://$IP3:2379, http://$IP4:2379, http://$IP5:2379 - ``` + ```shell + etcd --listen-client-urls=http://$IP1:2379,http://$IP2:2379,http://$IP3:2379,http://$IP4:2379,http://$IP5:2379 --advertise-client-urls=http://$IP1:2379,http://$IP2:2379,http://$IP3:2379,http://$IP4:2379,http://$IP5:2379 + ``` -2. Start Kubernetes API servers with the flag `--etcd-servers=$IP1:2379, $IP2:2379, $IP3:2379, $IP4:2379, $IP5:2379`. +2. Start Kubernetes API servers with the flag + `--etcd-servers=$IP1:2379,$IP2:2379,$IP3:2379,$IP4:2379,$IP5:2379`. - Replace `IP` with your client IP addresses. + Replace `IP` with your client IP addresses. -### Multi-node etcd cluster with load balancer +### Multi-node etcd cluster with load balancer To run a load balancing etcd cluster: @@ -87,92 +102,159 @@ To run a load balancing etcd cluster: ## Securing etcd clusters -Access to etcd is equivalent to root permission in the cluster so ideally only the API server should have access to it. Considering the sensitivity of the data, it is recommended to grant permission to only those nodes that require access to etcd clusters. +Access to etcd is equivalent to root permission in the cluster so ideally only +the API server should have access to it. Considering the sensitivity of the +data, it is recommended to grant permission to only those nodes that require +access to etcd clusters. -To secure etcd, either set up firewall rules or use the security features provided by etcd. etcd security features depend on x509 Public Key Infrastructure (PKI). To begin, establish secure communication channels by generating a key and certificate pair. For example, use key pairs `peer.key` and `peer.cert` for securing communication between etcd members, and `client.key` and `client.cert` for securing communication between etcd and its clients. See the [example scripts](https://github.com/coreos/etcd/tree/master/hack/tls-setup) provided by the etcd project to generate key pairs and CA files for client authentication. +To secure etcd, either set up firewall rules or use the security features +provided by etcd. etcd security features depend on x509 Public Key +Infrastructure (PKI). To begin, establish secure communication channels by +generating a key and certificate pair. For example, use key pairs `peer.key` +and `peer.cert` for securing communication between etcd members, and +`client.key` and `client.cert` for securing communication between etcd and its +clients. See the [example scripts](https://github.com/coreos/etcd/tree/master/hack/tls-setup) +provided by the etcd project to generate key pairs and CA files for client +authentication. ### Securing communication -To configure etcd with secure peer communication, specify flags `--peer-key-file=peer.key` and `--peer-cert-file=peer.cert`, and use https as URL schema. +To configure etcd with secure peer communication, specify flags +`--peer-key-file=peer.key` and `--peer-cert-file=peer.cert`, and use HTTPS as +URL schema. -Similarly, to configure etcd with secure client communication, specify flags `--key-file=k8sclient.key` and `--cert-file=k8sclient.cert`, and use https as URL schema. +Similarly, to configure etcd with secure client communication, specify flags +`--key-file=k8sclient.key` and `--cert-file=k8sclient.cert`, and use HTTPS as +URL schema. ### Limiting access of etcd clusters -After configuring secure communication, restrict the access of etcd cluster to only the Kubernetes API server. Use TLS authentication to do so. +After configuring secure communication, restrict the access of etcd cluster to +only the Kubernetes API server. Use TLS authentication to do so. -For example, consider key pairs `k8sclient.key` and `k8sclient.cert` that are trusted by the CA `etcd.ca`. When etcd is configured with `--client-cert-auth` along with TLS, it verifies the certificates from clients by using system CAs or the CA passed in by `--trusted-ca-file` flag. Specifying flags `--client-cert-auth=true` and `--trusted-ca-file=etcd.ca` will restrict the access to clients with the certificate `k8sclient.cert`. +For example, consider key pairs `k8sclient.key` and `k8sclient.cert` that are +trusted by the CA `etcd.ca`. When etcd is configured with `--client-cert-auth` +along with TLS, it verifies the certificates from clients by using system CAs +or the CA passed in by `--trusted-ca-file` flag. Specifying flags +`--client-cert-auth=true` and `--trusted-ca-file=etcd.ca` will restrict the +access to clients with the certificate `k8sclient.cert`. -Once etcd is configured correctly, only clients with valid certificates can access it. To give Kubernetes API server the access, configure it with the flags `--etcd-certfile=k8sclient.cert`,`--etcd-keyfile=k8sclient.key` and `--etcd-cafile=ca.cert`. +Once etcd is configured correctly, only clients with valid certificates can +access it. To give Kubernetes API server the access, configure it with the +flags `--etcd-certfile=k8sclient.cert`,`--etcd-keyfile=k8sclient.key` and +`--etcd-cafile=ca.cert`. {{< note >}} -etcd authentication is not currently supported by Kubernetes. For more information, see the related issue [Support Basic Auth for Etcd v2](https://github.com/kubernetes/kubernetes/issues/23398). +etcd authentication is not currently supported by Kubernetes. For more +information, see the related issue +[Support Basic Auth for Etcd v2](https://github.com/kubernetes/kubernetes/issues/23398). {{< /note >}} ## Replacing a failed etcd member -etcd cluster achieves high availability by tolerating minor member failures. However, to improve the overall health of the cluster, replace failed members immediately. When multiple members fail, replace them one by one. Replacing a failed member involves two steps: removing the failed member and adding a new member. +etcd cluster achieves high availability by tolerating minor member failures. +However, to improve the overall health of the cluster, replace failed members +immediately. When multiple members fail, replace them one by one. Replacing a +failed member involves two steps: removing the failed member and adding a new +member. -Though etcd keeps unique member IDs internally, it is recommended to use a unique name for each member to avoid human errors. For example, consider a three-member etcd cluster. Let the URLs be, member1=http://10.0.0.1, member2=http://10.0.0.2, and member3=http://10.0.0.3. When member1 fails, replace it with member4=http://10.0.0.4. +Though etcd keeps unique member IDs internally, it is recommended to use a +unique name for each member to avoid human errors. For example, consider a +three-member etcd cluster. Let the URLs be, member1=http://10.0.0.1, +member2=http://10.0.0.2, and member3=http://10.0.0.3. When member1 fails, +replace it with member4=http://10.0.0.4. 1. Get the member ID of the failed member1: - `etcdctl --endpoints=http://10.0.0.2,http://10.0.0.3 member list` + ```shell + etcdctl --endpoints=http://10.0.0.2,http://10.0.0.3 member list + ``` - The following message is displayed: + The following message is displayed: - 8211f1d0f64f3269, started, member1, http://10.0.0.1:2380, http://10.0.0.1:2379 - 91bc3c398fb3c146, started, member2, http://10.0.0.2:2380, http://10.0.0.2:2379 - fd422379fda50e48, started, member3, http://10.0.0.3:2380, http://10.0.0.3:2379 + ```console + 8211f1d0f64f3269, started, member1, http://10.0.0.1:2380, http://10.0.0.1:2379 + 91bc3c398fb3c146, started, member2, http://10.0.0.2:2380, http://10.0.0.2:2379 + fd422379fda50e48, started, member3, http://10.0.0.3:2380, http://10.0.0.3:2379 + ``` 2. Remove the failed member: - `etcdctl member remove 8211f1d0f64f3269` + ```shell + etcdctl member remove 8211f1d0f64f3269 + ``` - The following message is displayed: + The following message is displayed: - Removed member 8211f1d0f64f3269 from cluster + ```console + Removed member 8211f1d0f64f3269 from cluster + ``` 3. Add the new member: - `./etcdctl member add member4 --peer-urls=http://10.0.0.4:2380` + ```shell + etcdctl member add member4 --peer-urls=http://10.0.0.4:2380 + ``` - The following message is displayed: + The following message is displayed: - Member 2be1eb8f84b7f63e added to cluster ef37ad9dc622a7c4 + ```console + Member 2be1eb8f84b7f63e added to cluster ef37ad9dc622a7c4 + ``` 4. Start the newly added member on a machine with the IP `10.0.0.4`: - export ETCD_NAME="member4" - export ETCD_INITIAL_CLUSTER="member2=http://10.0.0.2:2380,member3=http://10.0.0.3:2380,member4=http://10.0.0.4:2380" - export ETCD_INITIAL_CLUSTER_STATE=existing - etcd [flags] + ```shell + export ETCD_NAME="member4" + export ETCD_INITIAL_CLUSTER="member2=http://10.0.0.2:2380,member3=http://10.0.0.3:2380,member4=http://10.0.0.4:2380" + export ETCD_INITIAL_CLUSTER_STATE=existing + etcd [flags] + ``` 5. Do either of the following: - 1. Update its `--etcd-servers` flag to make Kubernetes aware of the configuration changes, then restart the Kubernetes API server. - 2. Update the load balancer configuration if a load balancer is used in the deployment. + 1. Update its `--etcd-servers` flag to make Kubernetes aware of the + configuration changes, then restart the Kubernetes API server. + 2. Update the load balancer configuration if a load balancer is used in the + deployment. -For more information on cluster reconfiguration, see [etcd Reconfiguration Documentation](https://github.com/coreos/etcd/blob/master/Documentation/op-guide/runtime-configuration.md#remove-a-member). +For more information on cluster reconfiguration, see [etcd Reconfiguration +Documentation](https://github.com/coreos/etcd/blob/master/Documentation/op-guide/runtime-configuration.md#remove-a-member). ## Backing up an etcd cluster -All Kubernetes objects are stored on etcd. Periodically backing up the etcd cluster data is important to recover Kubernetes clusters under disaster scenarios, such as losing all master nodes. The snapshot file contains all the Kubernetes states and critical information. In order to keep the sensitive Kubernetes data safe, encrypt the snapshot files. +All Kubernetes objects are stored on etcd. Periodically backing up the etcd +cluster data is important to recover Kubernetes clusters under disaster +scenarios, such as losing all master nodes. The snapshot file contains all the +Kubernetes states and critical information. In order to keep the sensitive +Kubernetes data safe, encrypt the snapshot files. -Backing up an etcd cluster can be accomplished in two ways: etcd built-in snapshot and volume snapshot. +Backing up an etcd cluster can be accomplished in two ways: etcd built-in +snapshot and volume snapshot. ### Built-in snapshot -etcd supports built-in snapshot. A snapshot may either be taken from a live member with the `etcdctl snapshot save` command or by copying the `member/snap/db` file from an etcd [data directory](https://github.com/coreos/etcd/blob/master/Documentation/op-guide/configuration.md#--data-dir) that is not currently used by an etcd process. Taking the snapshot will normally not affect the performance of the member. +etcd supports built-in snapshot. A snapshot may either be taken from a live +member with the `etcdctl snapshot save` command or by copying the +`member/snap/db` file from an etcd [data +directory](https://github.com/coreos/etcd/blob/master/Documentation/op-guide/configuration.md#--data-dir) +that is not currently used by an etcd process. Taking the snapshot will +normally not affect the performance of the member. -Below is an example for taking a snapshot of the keyspace served by `$ENDPOINT` to the file `snapshotdb`: +Below is an example for taking a snapshot of the keyspace served by +`$ENDPOINT` to the file `snapshotdb`: -```sh +```shell ETCDCTL_API=3 etcdctl --endpoints $ENDPOINT snapshot save snapshotdb -# exit 0 +``` -# verify the snapshot +Verify the snapshot: + +```shell ETCDCTL_API=3 etcdctl --write-out=table snapshot status snapshotdb +``` + +```console +----------+----------+------------+------------+ | HASH | REVISION | TOTAL KEYS | TOTAL SIZE | +----------+----------+------------+------------+ @@ -182,35 +264,67 @@ ETCDCTL_API=3 etcdctl --write-out=table snapshot status snapshotdb ### Volume snapshot -If etcd is running on a storage volume that supports backup, such as Amazon Elastic Block Store, back up etcd data by taking a snapshot of the storage volume. +If etcd is running on a storage volume that supports backup, such as Amazon +Elastic Block Store, back up etcd data by taking a snapshot of the storage +volume. ## Scaling up etcd clusters -Scaling up etcd clusters increases availability by trading off performance. Scaling does not increase cluster performance nor capability. A general rule is not to scale up or down etcd clusters. Do not configure any auto scaling groups for etcd clusters. It is highly recommended to always run a static five-member etcd cluster for production Kubernetes clusters at any officially supported scale. +Scaling up etcd clusters increases availability by trading off performance. +Scaling does not increase cluster performance nor capability. A general rule +is not to scale up or down etcd clusters. Do not configure any auto scaling +groups for etcd clusters. It is highly recommended to always run a static +five-member etcd cluster for production Kubernetes clusters at any officially +supported scale. -A reasonable scaling is to upgrade a three-member cluster to a five-member one, when more reliability is desired. See [etcd Reconfiguration Documentation](https://github.com/coreos/etcd/blob/master/Documentation/op-guide/runtime-configuration.md#remove-a-member) for information on how to add members into an existing cluster. +A reasonable scaling is to upgrade a three-member cluster to a five-member +one, when more reliability is desired. See [etcd Reconfiguration +Documentation](https://github.com/coreos/etcd/blob/master/Documentation/op-guide/runtime-configuration.md#remove-a-member) +for information on how to add members into an existing cluster. ## Restoring an etcd cluster -etcd supports restoring from snapshots that are taken from an etcd process of the [major.minor](http://semver.org/) version. Restoring a version from a different patch version of etcd also is supported. A restore operation is employed to recover the data of a failed cluster. +etcd supports restoring from snapshots that are taken from an etcd process of +the [major.minor](http://semver.org/) version. Restoring a version from a +different patch version of etcd also is supported. A restore operation is +employed to recover the data of a failed cluster. -Before starting the restore operation, a snapshot file must be present. It can either be a snapshot file from a previous backup operation, or from a remaining [data directory](https://github.com/coreos/etcd/blob/master/Documentation/op-guide/configuration.md#--data-dir). For more information and examples on restoring a cluster from a snapshot file, see [etcd disaster recovery documentation](https://github.com/coreos/etcd/blob/master/Documentation/op-guide/recovery.md#restoring-a-cluster). +Before starting the restore operation, a snapshot file must be present. It can +either be a snapshot file from a previous backup operation, or from a +remaining [data +directory](https://github.com/coreos/etcd/blob/master/Documentation/op-guide/configuration.md#--data-dir). +For more information and examples on restoring a cluster from a snapshot file, +see [etcd disaster recovery +documentation](https://github.com/coreos/etcd/blob/master/Documentation/op-guide/recovery.md#restoring-a-cluster). -If the access URLs of the restored cluster is changed from the previous cluster, the Kubernetes API server must be reconfigured accordingly. In this case, restart Kubernetes API server with the flag `--etcd-servers=$NEW_ETCD_CLUSTER` instead of the flag `--etcd-servers=$OLD_ETCD_CLUSTER`. Replace `$NEW_ETCD_CLUSTER` and `$OLD_ETCD_CLUSTER` with the respective IP addresses. If a load balancer is used in front of an etcd cluster, you might need to update the load balancer instead. +If the access URLs of the restored cluster is changed from the previous +cluster, the Kubernetes API server must be reconfigured accordingly. In this +case, restart Kubernetes API server with the flag +`--etcd-servers=$NEW_ETCD_CLUSTER` instead of the flag +`--etcd-servers=$OLD_ETCD_CLUSTER`. Replace `$NEW_ETCD_CLUSTER` and +`$OLD_ETCD_CLUSTER` with the respective IP addresses. If a load balancer is +used in front of an etcd cluster, you might need to update the load balancer +instead. -If the majority of etcd members have permanently failed, the etcd cluster is considered failed. In this scenario, Kubernetes cannot make any changes to its current state. Although the scheduled pods might continue to run, no new pods can be scheduled. In such cases, recover the etcd cluster and potentially reconfigure Kubernetes API server to fix the issue. +If the majority of etcd members have permanently failed, the etcd cluster is +considered failed. In this scenario, Kubernetes cannot make any changes to its +current state. Although the scheduled pods might continue to run, no new pods +can be scheduled. In such cases, recover the etcd cluster and potentially +reconfigure Kubernetes API server to fix the issue. {{< note >}} -If any API servers are running in your cluster, you should not attempt to restore instances of etcd. +If any API servers are running in your cluster, you should not attempt to +restore instances of etcd. Instead, follow these steps to restore etcd: - stop *all* kube-apiserver instances - restore state in all etcd instances - restart all kube-apiserver instances -We also recommend restarting any components (e.g. kube-scheduler, kube-controller-manager, kubelet) to ensure that they don't -rely on some stale data. Note that in practice, the restore takes a bit of time. -During the restoration, critical components will lose leader lock and restart themselves. +We also recommend restarting any components (e.g. kube-scheduler, +kube-controller-manager, kubelet) to ensure that they don't rely on some stale +data. Note that in practice, the restore takes a bit of time. During the +restoration, critical components will lose leader lock and restart themselves. {{< /note >}} ## Upgrading and rolling back etcd clusters @@ -238,18 +352,32 @@ cluster was deployed and configured. We recommend that you consult your cluster provider's documentation to see if there is a predefined solution. If your cluster was created via `kube-up.sh` and is still using etcd2 as its -storage backend, please consult the [Kubernetes v1.12 etcd cluster upgrade docs](https://v1-12.docs.kubernetes.io/docs/tasks/administer-cluster/configure-upgrade-etcd/#upgrading-and-rolling-back-etcd-clusters) +storage backend, please consult the [Kubernetes v1.12 etcd cluster upgrade +docs](https://v1-12.docs.kubernetes.io/docs/tasks/administer-cluster/configure-upgrade-etcd/#upgrading-and-rolling-back-etcd-clusters) ## Known issue: etcd client balancer with secure endpoints -The etcd v3 client, released in etcd v3.3.13 or earlier, has a [critical bug](https://github.com/kubernetes/kubernetes/issues/72102) which affects the kube-apiserver and HA deployments. The etcd client balancer failover does not properly work against secure endpoints. As a result, etcd servers may fail or disconnect briefly from the kube-apiserver. This affects kube-apiserver HA deployments. +The etcd v3 client, released in etcd v3.3.13 or earlier, has a [critical +bug](https://github.com/kubernetes/kubernetes/issues/72102) which affects the +kube-apiserver and HA deployments. The etcd client balancer failover does not +properly work against secure endpoints. As a result, etcd servers may fail or +disconnect briefly from the kube-apiserver. This affects kube-apiserver HA +deployments. -The fix was made in [etcd v3.4](https://github.com/etcd-io/etcd/pull/10911) (and backported to v3.3.14 or later): the new client now creates its own credential bundle to correctly set authority target in dial function. +The fix was made in [etcd v3.4](https://github.com/etcd-io/etcd/pull/10911) +(and backported to v3.3.14 or later): the new client now creates its own +credential bundle to correctly set authority target in dial function. -Because the fix requires gRPC dependency upgrade (to v1.23.0), downstream Kubernetes [did not backport etcd upgrades](https://github.com/kubernetes/kubernetes/issues/72102#issuecomment-526645978). Which means the [etcd fix in kube-apiserver](https://github.com/etcd-io/etcd/pull/10911/commits/db61ee106ca9363ba3f188ecf27d1a8843da33ab) is only available from Kubernetes 1.16. - -To urgently fix this bug for Kubernetes 1.15 or earlier, build a custom kube-apiserver. You can make local changes to [`vendor/google.golang.org/grpc/credentials/credentials.go`](https://github.com/kubernetes/kubernetes/blob/7b85be021cd2943167cd3d6b7020f44735d9d90b/vendor/google.golang.org/grpc/credentials/credentials.go#L135) with [etcd@db61ee106](https://github.com/etcd-io/etcd/pull/10911/commits/db61ee106ca9363ba3f188ecf27d1a8843da33ab). - -See ["kube-apiserver 1.13.x refuses to work when first etcd-server is not available"](https://github.com/kubernetes/kubernetes/issues/72102). +Because the fix requires gRPC dependency upgrade (to v1.23.0), downstream +Kubernetes [did not backport etcd +upgrades](https://github.com/kubernetes/kubernetes/issues/72102#issuecomment-526645978). +Which means the [etcd fix in +kube-apiserver](https://github.com/etcd-io/etcd/pull/10911/commits/db61ee106ca9363ba3f188ecf27d1a8843da33ab) +is only available from Kubernetes 1.16. +To urgently fix this bug for Kubernetes 1.15 or earlier, build a custom +kube-apiserver. You can make local changes to +[`vendor/google.golang.org/grpc/credentials/credentials.go`](https://github.com/kubernetes/kubernetes/blob/7b85be021cd2943167cd3d6b7020f44735d9d90b/vendor/google.golang.org/grpc/credentials/credentials.go#L135) +with +[etcd@db61ee106](https://github.com/etcd-io/etcd/pull/10911/commits/db61ee106ca9363ba3f188ecf27d1a8843da33ab). From 9cac41b0a84f67144b9bf74e0052c197eb6f68b6 Mon Sep 17 00:00:00 2001 From: Qiming Teng Date: Tue, 16 Feb 2021 11:29:50 +0800 Subject: [PATCH 27/62] Fix links to etcd official documentation --- .../configure-upgrade-etcd.md | 96 +++++++++---------- 1 file changed, 47 insertions(+), 49 deletions(-) diff --git a/content/en/docs/tasks/administer-cluster/configure-upgrade-etcd.md b/content/en/docs/tasks/administer-cluster/configure-upgrade-etcd.md index 90ab3b70f9..4be21db388 100644 --- a/content/en/docs/tasks/administer-cluster/configure-upgrade-etcd.md +++ b/content/en/docs/tasks/administer-cluster/configure-upgrade-etcd.md @@ -27,14 +27,14 @@ content_type: task * Ensure that no resource starvation occurs. Performance and stability of the cluster is sensitive to network and disk - IO. Any resource starvation can lead to heartbeat timeout, causing instability + I/O. Any resource starvation can lead to heartbeat timeout, causing instability of the cluster. An unstable etcd indicates that no leader is elected. Under such circumstances, a cluster cannot make any changes to its current state, which implies no new pods can be scheduled. -* Keeping stable etcd clusters is critical to the stability of Kubernetes +* Keeping etcd clusters stable is critical to the stability of Kubernetes clusters. Therefore, run etcd clusters on dedicated machines or isolated - environments for [guaranteed resource requirements](https://github.com/coreos/etcd/blob/master/Documentation/op-guide/hardware.md#hardware-recommendations). + environments for [guaranteed resource requirements](https://etcd.io/docs/current/op-guide/hardware/). * The minimum recommended version of etcd to run in production is `3.2.10+`. @@ -43,7 +43,7 @@ content_type: task Operating etcd with limited resources is suitable only for testing purposes. For deploying in production, advanced hardware configuration is required. Before deploying etcd in production, see -[resource requirement reference documentation](https://github.com/coreos/etcd/blob/master/Documentation/op-guide/hardware.md#example-hardware-configurations). +[resource requirement reference](https://etcd.io/docs/current/op-guide/hardware/#example-hardware-configurations). ## Starting etcd clusters @@ -60,7 +60,7 @@ Use a single-node etcd cluster only for testing purpose. --advertise-client-urls=http://$PRIVATE_IP:2379 ``` -2. Start Kubernetes API server with the flag +2. Start the Kubernetes API server with the flag `--etcd-servers=$PRIVATE_IP:2379`. Make sure `PRIVATE_IP` is set to your etcd client IP. @@ -69,12 +69,12 @@ Use a single-node etcd cluster only for testing purpose. For durability and high availability, run etcd as a multi-node cluster in production and back it up periodically. A five-member cluster is recommended -in production. For more information, see [FAQ -Documentation](https://github.com/coreos/etcd/blob/master/Documentation/faq.md#what-is-failure-tolerance). +in production. For more information, see +[FAQ documentation](https://etcd.io/docs/current/faq/#what-is-failure-tolerance). Configure an etcd cluster either by static member information or by dynamic -discovery. For more information on clustering, see [etcd Clustering -Documentation](https://github.com/coreos/etcd/blob/master/Documentation/op-guide/clustering.md). +discovery. For more information on clustering, see +[etcd clustering documentation](https://etcd.io/docs/current/op-guide/clustering/). For an example, consider a five-member etcd cluster running with the following client URLs: `http://$IP1:2379`, `http://$IP2:2379`, `http://$IP3:2379`, @@ -86,10 +86,10 @@ client URLs: `http://$IP1:2379`, `http://$IP2:2379`, `http://$IP3:2379`, etcd --listen-client-urls=http://$IP1:2379,http://$IP2:2379,http://$IP3:2379,http://$IP4:2379,http://$IP5:2379 --advertise-client-urls=http://$IP1:2379,http://$IP2:2379,http://$IP3:2379,http://$IP4:2379,http://$IP5:2379 ``` -2. Start Kubernetes API servers with the flag +2. Start the Kubernetes API servers with the flag `--etcd-servers=$IP1:2379,$IP2:2379,$IP3:2379,$IP4:2379,$IP5:2379`. - Replace `IP` with your client IP addresses. + Make sure the `IP` variables are set to your client IP addresses. ### Multi-node etcd cluster with load balancer @@ -121,16 +121,16 @@ authentication. To configure etcd with secure peer communication, specify flags `--peer-key-file=peer.key` and `--peer-cert-file=peer.cert`, and use HTTPS as -URL schema. +the URL schema. Similarly, to configure etcd with secure client communication, specify flags `--key-file=k8sclient.key` and `--cert-file=k8sclient.cert`, and use HTTPS as -URL schema. +the URL schema. ### Limiting access of etcd clusters After configuring secure communication, restrict the access of etcd cluster to -only the Kubernetes API server. Use TLS authentication to do so. +only the Kubernetes API servers. Use TLS authentication to do so. For example, consider key pairs `k8sclient.key` and `k8sclient.cert` that are trusted by the CA `etcd.ca`. When etcd is configured with `--client-cert-auth` @@ -140,7 +140,7 @@ or the CA passed in by `--trusted-ca-file` flag. Specifying flags access to clients with the certificate `k8sclient.cert`. Once etcd is configured correctly, only clients with valid certificates can -access it. To give Kubernetes API server the access, configure it with the +access it. To give Kubernetes API servers the access, configure them with the flags `--etcd-certfile=k8sclient.cert`,`--etcd-keyfile=k8sclient.key` and `--etcd-cafile=ca.cert`. @@ -160,11 +160,11 @@ member. Though etcd keeps unique member IDs internally, it is recommended to use a unique name for each member to avoid human errors. For example, consider a -three-member etcd cluster. Let the URLs be, member1=http://10.0.0.1, -member2=http://10.0.0.2, and member3=http://10.0.0.3. When member1 fails, -replace it with member4=http://10.0.0.4. +three-member etcd cluster. Let the URLs be, `member1=http://10.0.0.1`, +`member2=http://10.0.0.2`, and `member3=http://10.0.0.3`. When `member1` fails, +replace it with `member4=http://10.0.0.4`. -1. Get the member ID of the failed member1: +1. Get the member ID of the failed `member1`: ```shell etcdctl --endpoints=http://10.0.0.2,http://10.0.0.3 member list @@ -213,21 +213,22 @@ replace it with member4=http://10.0.0.4. 5. Do either of the following: - 1. Update its `--etcd-servers` flag to make Kubernetes aware of the - configuration changes, then restart the Kubernetes API server. + 1. Update the `--etcd-servers` flag for the Kubernetes API servers to make + Kubernetes aware of the configuration changes, then restart the + Kubernetes API servers. 2. Update the load balancer configuration if a load balancer is used in the deployment. -For more information on cluster reconfiguration, see [etcd Reconfiguration -Documentation](https://github.com/coreos/etcd/blob/master/Documentation/op-guide/runtime-configuration.md#remove-a-member). +For more information on cluster reconfiguration, see +[etcd reconfiguration documentation](https://etcd.io/docs/current/op-guide/runtime-configuration/#remove-a-member). ## Backing up an etcd cluster All Kubernetes objects are stored on etcd. Periodically backing up the etcd cluster data is important to recover Kubernetes clusters under disaster -scenarios, such as losing all master nodes. The snapshot file contains all the -Kubernetes states and critical information. In order to keep the sensitive -Kubernetes data safe, encrypt the snapshot files. +scenarios, such as losing all control plane nodes. The snapshot file contains +all the Kubernetes states and critical information. In order to keep the +sensitive Kubernetes data safe, encrypt the snapshot files. Backing up an etcd cluster can be accomplished in two ways: etcd built-in snapshot and volume snapshot. @@ -236,10 +237,10 @@ snapshot and volume snapshot. etcd supports built-in snapshot. A snapshot may either be taken from a live member with the `etcdctl snapshot save` command or by copying the -`member/snap/db` file from an etcd [data -directory](https://github.com/coreos/etcd/blob/master/Documentation/op-guide/configuration.md#--data-dir) +`member/snap/db` file from an etcd +[data directory](https://etcd.io/docs/current/op-guide/configuration/#--data-dir) that is not currently used by an etcd process. Taking the snapshot will -normally not affect the performance of the member. +not affect the performance of the member. Below is an example for taking a snapshot of the keyspace served by `$ENDPOINT` to the file `snapshotdb`: @@ -278,8 +279,8 @@ five-member etcd cluster for production Kubernetes clusters at any officially supported scale. A reasonable scaling is to upgrade a three-member cluster to a five-member -one, when more reliability is desired. See [etcd Reconfiguration -Documentation](https://github.com/coreos/etcd/blob/master/Documentation/op-guide/runtime-configuration.md#remove-a-member) +one, when more reliability is desired. See +[etcd reconfiguration documentation](https://etcd.io/docs/current/op-guide/runtime-configuration/#remove-a-member) for information on how to add members into an existing cluster. ## Restoring an etcd cluster @@ -290,16 +291,14 @@ different patch version of etcd also is supported. A restore operation is employed to recover the data of a failed cluster. Before starting the restore operation, a snapshot file must be present. It can -either be a snapshot file from a previous backup operation, or from a -remaining [data -directory](https://github.com/coreos/etcd/blob/master/Documentation/op-guide/configuration.md#--data-dir). -For more information and examples on restoring a cluster from a snapshot file, -see [etcd disaster recovery -documentation](https://github.com/coreos/etcd/blob/master/Documentation/op-guide/recovery.md#restoring-a-cluster). +either be a snapshot file from a previous backup operation, or from a remaining +[data directory]( https://etcd.io/docs/current/op-guide/configuration/#--data-dir). +For more information and examples on restoring a cluster from a snapshot file, see +[etcd disaster recovery documentation](https://etcd.io/docs/current/op-guide/recovery/#restoring-a-cluster). If the access URLs of the restored cluster is changed from the previous cluster, the Kubernetes API server must be reconfigured accordingly. In this -case, restart Kubernetes API server with the flag +case, restart Kubernetes API servers with the flag `--etcd-servers=$NEW_ETCD_CLUSTER` instead of the flag `--etcd-servers=$OLD_ETCD_CLUSTER`. Replace `$NEW_ETCD_CLUSTER` and `$OLD_ETCD_CLUSTER` with the respective IP addresses. If a load balancer is @@ -310,20 +309,19 @@ If the majority of etcd members have permanently failed, the etcd cluster is considered failed. In this scenario, Kubernetes cannot make any changes to its current state. Although the scheduled pods might continue to run, no new pods can be scheduled. In such cases, recover the etcd cluster and potentially -reconfigure Kubernetes API server to fix the issue. +reconfigure Kubernetes API servers to fix the issue. {{< note >}} If any API servers are running in your cluster, you should not attempt to -restore instances of etcd. -Instead, follow these steps to restore etcd: +restore instances of etcd. Instead, follow these steps to restore etcd: -- stop *all* kube-apiserver instances +- stop *all* API server instances - restore state in all etcd instances -- restart all kube-apiserver instances +- restart all API server instances -We also recommend restarting any components (e.g. kube-scheduler, -kube-controller-manager, kubelet) to ensure that they don't rely on some stale -data. Note that in practice, the restore takes a bit of time. During the +We also recommend restarting any components (e.g. `kube-scheduler`, +`kube-controller-manager`, `kubelet`) to ensure that they don't rely on some +stale data. Note that in practice, the restore takes a bit of time. During the restoration, critical components will lose leader lock and restart themselves. {{< /note >}} @@ -364,9 +362,9 @@ properly work against secure endpoints. As a result, etcd servers may fail or disconnect briefly from the kube-apiserver. This affects kube-apiserver HA deployments. -The fix was made in [etcd v3.4](https://github.com/etcd-io/etcd/pull/10911) -(and backported to v3.3.14 or later): the new client now creates its own -credential bundle to correctly set authority target in dial function. +The fix was made in etcd v3.4 (and backported to v3.3.14 or later): the new +client now creates its own credential bundle to correctly set authority target +in dial function. Because the fix requires gRPC dependency upgrade (to v1.23.0), downstream Kubernetes [did not backport etcd From 3b916b3366f7b60b0d8e5f3e88e8e64c43b1c56a Mon Sep 17 00:00:00 2001 From: Qiming Teng Date: Tue, 16 Feb 2021 11:31:31 +0800 Subject: [PATCH 28/62] Drop outdated information from the page --- .../configure-upgrade-etcd.md | 54 ------------------- 1 file changed, 54 deletions(-) diff --git a/content/en/docs/tasks/administer-cluster/configure-upgrade-etcd.md b/content/en/docs/tasks/administer-cluster/configure-upgrade-etcd.md index 4be21db388..25cd61a710 100644 --- a/content/en/docs/tasks/administer-cluster/configure-upgrade-etcd.md +++ b/content/en/docs/tasks/administer-cluster/configure-upgrade-etcd.md @@ -325,57 +325,3 @@ stale data. Note that in practice, the restore takes a bit of time. During the restoration, critical components will lose leader lock and restart themselves. {{< /note >}} -## Upgrading and rolling back etcd clusters - -As of Kubernetes v1.13.0, etcd2 is no longer supported as a storage backend for -new or existing Kubernetes clusters. The timeline for Kubernetes support for -etcd2 and etcd3 is as follows: - -- Kubernetes v1.0: etcd2 only -- Kubernetes v1.5.1: etcd3 support added, new clusters still default to etcd2 -- Kubernetes v1.6.0: new clusters created with `kube-up.sh` default to etcd3, - and `kube-apiserver` defaults to etcd3 -- Kubernetes v1.9.0: deprecation of etcd2 storage backend announced -- Kubernetes v1.13.0: etcd2 storage backend removed, `kube-apiserver` will - refuse to start with `--storage-backend=etcd2`, with the - message `etcd2 is no longer a supported storage backend` - -Before upgrading a v1.12.x kube-apiserver using `--storage-backend=etcd2` to -v1.13.x, etcd v2 data must be migrated to the v3 storage backend and -kube-apiserver invocations must be changed to use `--storage-backend=etcd3`. - -The process for migrating from etcd2 to etcd3 is highly dependent on how the -etcd cluster was deployed and configured, as well as how the Kubernetes -cluster was deployed and configured. We recommend that you consult your cluster -provider's documentation to see if there is a predefined solution. - -If your cluster was created via `kube-up.sh` and is still using etcd2 as its -storage backend, please consult the [Kubernetes v1.12 etcd cluster upgrade -docs](https://v1-12.docs.kubernetes.io/docs/tasks/administer-cluster/configure-upgrade-etcd/#upgrading-and-rolling-back-etcd-clusters) - -## Known issue: etcd client balancer with secure endpoints - -The etcd v3 client, released in etcd v3.3.13 or earlier, has a [critical -bug](https://github.com/kubernetes/kubernetes/issues/72102) which affects the -kube-apiserver and HA deployments. The etcd client balancer failover does not -properly work against secure endpoints. As a result, etcd servers may fail or -disconnect briefly from the kube-apiserver. This affects kube-apiserver HA -deployments. - -The fix was made in etcd v3.4 (and backported to v3.3.14 or later): the new -client now creates its own credential bundle to correctly set authority target -in dial function. - -Because the fix requires gRPC dependency upgrade (to v1.23.0), downstream -Kubernetes [did not backport etcd -upgrades](https://github.com/kubernetes/kubernetes/issues/72102#issuecomment-526645978). -Which means the [etcd fix in -kube-apiserver](https://github.com/etcd-io/etcd/pull/10911/commits/db61ee106ca9363ba3f188ecf27d1a8843da33ab) -is only available from Kubernetes 1.16. - -To urgently fix this bug for Kubernetes 1.15 or earlier, build a custom -kube-apiserver. You can make local changes to -[`vendor/google.golang.org/grpc/credentials/credentials.go`](https://github.com/kubernetes/kubernetes/blob/7b85be021cd2943167cd3d6b7020f44735d9d90b/vendor/google.golang.org/grpc/credentials/credentials.go#L135) -with -[etcd@db61ee106](https://github.com/etcd-io/etcd/pull/10911/commits/db61ee106ca9363ba3f188ecf27d1a8843da33ab). - From 18c22d133e573d85700cdfb7b5e74d130a654ea0 Mon Sep 17 00:00:00 2001 From: Qiming Teng Date: Tue, 16 Feb 2021 12:17:57 +0800 Subject: [PATCH 29/62] Fix dockerhub URL --- .../pull-image-private-registry.md | 16 ++++------------ 1 file changed, 4 insertions(+), 12 deletions(-) diff --git a/content/en/docs/tasks/configure-pod-container/pull-image-private-registry.md b/content/en/docs/tasks/configure-pod-container/pull-image-private-registry.md index ce0b5b3656..32b857c156 100644 --- a/content/en/docs/tasks/configure-pod-container/pull-image-private-registry.md +++ b/content/en/docs/tasks/configure-pod-container/pull-image-private-registry.md @@ -9,18 +9,13 @@ weight: 100 This page shows how to create a Pod that uses a Secret to pull an image from a private Docker registry or repository. - - ## {{% heading "prerequisites" %}} - * {{< include "task-tutorial-prereqs.md" >}} {{< version-check >}} * To do this exercise, you need a [Docker ID](https://docs.docker.com/docker-id/) and password. - - ## Log in to Docker @@ -106,7 +101,8 @@ kubectl create secret docker-registry regcred --docker-server=` is your Private Docker Registry FQDN. (https://index.docker.io/v1/ for DockerHub) +* `` is your Private Docker Registry FQDN. + Use `https://index.docker.io/v2/` for DockerHub. * `` is your Docker username. * `` is your Docker password. * `` is your Docker email. @@ -192,7 +188,8 @@ your.private.registry.example.com/janedoe/jdoe-private:v1 ``` To pull the image from the private registry, Kubernetes needs credentials. -The `imagePullSecrets` field in the configuration file specifies that Kubernetes should get the credentials from a Secret named `regcred`. +The `imagePullSecrets` field in the configuration file specifies that +Kubernetes should get the credentials from a Secret named `regcred`. Create a Pod that uses your Secret, and verify that the Pod is running: @@ -201,11 +198,8 @@ kubectl apply -f my-private-reg-pod.yaml kubectl get pod private-reg ``` - - ## {{% heading "whatsnext" %}} - * Learn more about [Secrets](/docs/concepts/configuration/secret/). * Learn more about [using a private registry](/docs/concepts/containers/images/#using-a-private-registry). * Learn more about [adding image pull secrets to a service account](/docs/tasks/configure-pod-container/configure-service-account/#add-imagepullsecrets-to-a-service-account). @@ -213,5 +207,3 @@ kubectl get pod private-reg * See [Secret](/docs/reference/generated/kubernetes-api/{{< param "version" >}}/#secret-v1-core). * See the `imagePullSecrets` field of [PodSpec](/docs/reference/generated/kubernetes-api/{{< param "version" >}}/#podspec-v1-core). - - From 3b502ed114c011ff974085dbccddf6aa47f3ff48 Mon Sep 17 00:00:00 2001 From: timyinshi Date: Tue, 16 Feb 2021 15:40:24 +0800 Subject: [PATCH 30/62] modify the word of and from english to chinese Signed-off-by: timyinshi --- .../zh/docs/concepts/overview/working-with-objects/labels.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/content/zh/docs/concepts/overview/working-with-objects/labels.md b/content/zh/docs/concepts/overview/working-with-objects/labels.md index d5dc5821ae..30e2635695 100644 --- a/content/zh/docs/concepts/overview/working-with-objects/labels.md +++ b/content/zh/docs/concepts/overview/working-with-objects/labels.md @@ -255,7 +255,7 @@ LIST and WATCH operations may specify label selectors to filter the sets of obje --> ### LIST 和 WATCH 过滤 -LIST and WATCH 操作可以使用查询参数指定标签选择算符过滤一组对象。 +LIST 和 WATCH 操作可以使用查询参数指定标签选择算符过滤一组对象。 两种需求都是允许的。（这里显示的是它们出现在 URL 查询字符串中） @@ -36,7 +32,7 @@ weight: 60 클러스터가 {{< glossary_tooltip term_id="rbac" text="RBAC" >}}을 사용하는 경우, 메트릭을 읽으려면 `/metrics` 에 접근을 허용하는 클러스터롤(ClusterRole)을 가지는 사용자, 그룹 또는 서비스어카운트(ServiceAccount)를 통한 권한이 필요하다. 예를 들면, 다음과 같다. -``` +```yaml apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRole metadata: @@ -156,5 +152,4 @@ kube-scheduler는 각 파드에 대해 구성된 리소스 [요청과 제한](/k ## {{% heading "whatsnext" %}} * 메트릭에 대한 [프로메테우스 텍스트 형식](https://github.com/prometheus/docs/blob/master/content/docs/instrumenting/exposition_formats.md#text-based-format)에 대해 읽어본다 -* [안정적인 쿠버네티스 메트릭](https://github.com/kubernetes/kubernetes/blob/master/test/instrumentation/testdata/stable-metrics-list.yaml) 목록을 참고한다 * [쿠버네티스 사용 중단 정책](/docs/reference/using-api/deprecation-policy/#deprecating-a-feature-or-behavior)에 대해 읽어본다 diff --git a/content/ko/docs/concepts/configuration/secret.md b/content/ko/docs/concepts/configuration/secret.md index e5466b8dec..073eb2ff4a 100644 --- a/content/ko/docs/concepts/configuration/secret.md +++ b/content/ko/docs/concepts/configuration/secret.md @@ -22,6 +22,16 @@ weight: 30 명세나 이미지에 포함될 수 있다. 사용자는 시크릿을 만들 수 있고 시스템도 일부 시크릿을 만들 수 있다. +{{< caution >}} +쿠버네티스 시크릿은 기본적으로 암호화되지 않은 base64 인코딩 문자열로 저장된다. +기본적으로 API 액세스 권한이 있는 모든 사용자 또는 쿠버네티스의 기본 데이터 저장소 etcd에 +액세스할 수 있는 모든 사용자가 일반 텍스트로 검색 할 수 있다. +시크릿을 안전하게 사용하려면 (최소한) 다음과 같이 하는 것이 좋다. + +1. 시크릿에 대한 [암호화 활성화](/docs/tasks/administer-cluster/encrypt-data/). +2. 시크릿 읽기 및 쓰기를 제한하는 [RBAC 규칙 활성화 또는 구성](/docs/reference/access-authn-authz/authorization/). 파드를 만들 권한이 있는 모든 사용자는 시크릿을 암묵적으로 얻을 수 있다. +{{< /caution >}} + ## 시크릿 개요 @@ -269,6 +279,13 @@ SSH 인증 시크릿 타입은 사용자 편의만을 위해서 제공된다. API 서버는 요구되는 키가 시크릿 구성에서 제공되고 있는지 검증도 한다. +{{< caution >}} +SSH 개인 키는 자체적으로 SSH 클라이언트와 호스트 서버간에 신뢰할 수있는 통신을 +설정하지 않는다. ConfigMap에 추가된 `known_hosts` 파일과 같은 +"중간자(man in the middle)" 공격을 완화하려면 신뢰를 설정하는 +2차 수단이 필요하다. +{{< /caution >}} + ### TLS 시크릿 쿠버네티스는 보통 TLS를 위해 사용되는 인증서와 관련된 키를 저장하기 위해서 @@ -786,7 +803,6 @@ immutable: true 수동으로 생성된 시크릿(예: GitHub 계정에 접근하기 위한 토큰이 포함된 시크릿)은 시크릿의 서비스 어카운트를 기반한 파드에 자동으로 연결될 수 있다. -해당 프로세스에 대한 자세한 설명은 [파드프리셋(PodPreset)을 사용하여 파드에 정보 주입하기](/docs/tasks/inject-data-application/podpreset/)를 참고한다. ## 상세 내용 @@ -1233,3 +1249,4 @@ API 서버에서 kubelet으로의 통신은 SSL/TLS로 보호된다. - [`kubectl` 을 사용한 시크릿 관리](/docs/tasks/configmap-secret/managing-secret-using-kubectl/)하는 방법 배우기 - [구성 파일을 사용한 시크릿 관리](/docs/tasks/configmap-secret/managing-secret-using-config-file/)하는 방법 배우기 - [kustomize를 사용한 시크릿 관리](/docs/tasks/configmap-secret/managing-secret-using-kustomize/)하는 방법 배우기 + diff --git a/content/ko/docs/concepts/containers/container-lifecycle-hooks.md b/content/ko/docs/concepts/containers/container-lifecycle-hooks.md index 662ac71522..d4bdc17403 100644 --- a/content/ko/docs/concepts/containers/container-lifecycle-hooks.md +++ b/content/ko/docs/concepts/containers/container-lifecycle-hooks.md @@ -54,7 +54,7 @@ Angular와 같이, 컴포넌트 라이프사이클 훅을 가진 많은 프로 컨테이너 라이프사이클 관리 훅이 호출되면, 쿠버네티스 관리 시스템은 훅 동작에 따라 핸들러를 실행하고, -`exec` 와 `tcpSocket` 은 컨테이너에서 실행되고, `httpGet` 은 kubelet 프로세스에 의해 실행된다. +`httpGet` 와 `tcpSocket` 은 kubelet 프로세스에 의해 실행되고, `exec` 은 컨테이너에서 실행된다. 훅 핸들러 호출은 해당 컨테이너를 포함하고 있는 파드의 컨텍스트와 동기적으로 동작한다. 이것은 `PostStart` 훅에 대해서, diff --git a/content/ko/docs/concepts/containers/runtime-class.md b/content/ko/docs/concepts/containers/runtime-class.md index b31cabe888..3d7c89b65c 100644 --- a/content/ko/docs/concepts/containers/runtime-class.md +++ b/content/ko/docs/concepts/containers/runtime-class.md @@ -1,7 +1,4 @@ --- - - - title: 런타임클래스(RuntimeClass) content_type: concept weight: 20 @@ -35,10 +32,6 @@ weight: 20 ## 셋업 -런타임클래스 기능 게이트가 활성화(기본값)된 것을 확인한다. -기능 게이트 활성화에 대한 설명은 [기능 게이트](/ko/docs/reference/command-line-tools-reference/feature-gates/)를 -참고한다. `RuntimeClass` 기능 게이트는 API 서버 _및_ kubelets에서 활성화되어야 한다. - 1. CRI 구현(implementation)을 노드에 설정(런타임에 따라서). 2. 상응하는 런타임클래스 리소스 생성. @@ -144,11 +137,9 @@ https://github.com/containerd/cri/blob/master/docs/config.md {{< feature-state for_k8s_version="v1.16" state="beta" >}} -쿠버네티스 v1.16 부터, 런타임 클래스는 `scheduling` 필드를 통해 이종의 클러스터 -지원을 포함한다. 이 필드를 사용하면, 이 런타임 클래스를 갖는 파드가 이를 지원하는 -노드로 스케줄된다는 것을 보장할 수 있다. 이 스케줄링 기능을 사용하려면, -[런타임 클래스 어드미션(admission) 컨트롤러](/docs/reference/access-authn-authz/admission-controllers/#runtimeclass)를 -활성화(1.16 부터 기본값)해야 한다. +RuntimeClass에 `scheduling` 필드를 지정하면, 이 RuntimeClass로 실행되는 파드가 +이를 지원하는 노드로 예약되도록 제약 조건을 설정할 수 있다. +`scheduling`이 설정되지 않은 경우 이 RuntimeClass는 모든 노드에서 지원되는 것으로 간주된다. 파드가 지정된 런타임클래스를 지원하는 노드에 안착한다는 것을 보장하려면, 해당 노드들은 `runtimeClass.scheduling.nodeSelector` 필드에서 선택되는 공통 레이블을 가져야한다. diff --git a/content/ko/docs/concepts/extend-kubernetes/extend-cluster.md b/content/ko/docs/concepts/extend-kubernetes/extend-cluster.md index d1eecd6fdc..ee9763a769 100644 --- a/content/ko/docs/concepts/extend-kubernetes/extend-cluster.md +++ b/content/ko/docs/concepts/extend-kubernetes/extend-cluster.md @@ -69,7 +69,7 @@ weight: 10 웹훅 모델에서 쿠버네티스는 원격 서비스에 네트워크 요청을 한다. *바이너리 플러그인* 모델에서 쿠버네티스는 바이너리(프로그램)를 실행한다. 바이너리 플러그인은 kubelet(예: -[Flex Volume 플러그인](/ko/docs/concepts/storage/volumes/#flexvolume)과 +[Flex 볼륨 플러그인](/ko/docs/concepts/storage/volumes/#flexvolume)과 [네트워크 플러그인](/ko/docs/concepts/extend-kubernetes/compute-storage-net/network-plugins/))과 kubectl에서 사용한다. @@ -157,7 +157,7 @@ API를 추가해도 기존 API(예: 파드)의 동작에 직접 영향을 미치 ### 스토리지 플러그인 -[Flex Volumes](/ko/docs/concepts/storage/volumes/#flexvolume)을 사용하면 +[Flex 볼륨](/ko/docs/concepts/storage/volumes/#flexvolume)을 사용하면 Kubelet이 바이너리 플러그인을 호출하여 볼륨을 마운트하도록 함으로써 빌트인 지원 없이 볼륨 유형을 마운트 할 수 있다. diff --git a/content/ko/docs/concepts/services-networking/ingress-controllers.md b/content/ko/docs/concepts/services-networking/ingress-controllers.md index 3af939488e..06340143a0 100644 --- a/content/ko/docs/concepts/services-networking/ingress-controllers.md +++ b/content/ko/docs/concepts/services-networking/ingress-controllers.md @@ -9,11 +9,11 @@ weight: 40 인그레스 리소스가 작동하려면, 클러스터는 실행 중인 인그레스 컨트롤러가 반드시 필요하다. -kube-controller-manager 바이너리의 일부로 실행되는 컨트롤러의 다른 타입과 달리 인그레스 컨트롤러는 +`kube-controller-manager` 바이너리의 일부로 실행되는 컨트롤러의 다른 타입과 달리 인그레스 컨트롤러는 클러스터와 함께 자동으로 실행되지 않는다. 클러스터에 가장 적합한 인그레스 컨트롤러 구현을 선택하는데 이 페이지를 사용한다. -프로젝트로써 쿠버네티스는 [AWS](https://github.com/kubernetes-sigs/aws-load-balancer-controller#readme), [GCE](https://git.k8s.io/ingress-gce/README.md#readme)와 +프로젝트로서 쿠버네티스는 [AWS](https://github.com/kubernetes-sigs/aws-load-balancer-controller#readme), [GCE](https://git.k8s.io/ingress-gce/README.md#readme)와 [nginx](https://git.k8s.io/ingress-nginx/README.md#readme) 인그레스 컨트롤러를 지원하고 유지한다. @@ -26,6 +26,7 @@ kube-controller-manager 바이너리의 일부로 실행되는 컨트롤러의 * [AKS 애플리케이션 게이트웨이 인그레스 컨트롤러] (https://azure.github.io/application-gateway-kubernetes-ingress/)는 [Azure 애플리케이션 게이트웨이](https://docs.microsoft.com)를 구성하는 인그레스 컨트롤러다. * [Ambassador](https://www.getambassador.io/) API 게이트웨이는 [Envoy](https://www.envoyproxy.io) 기반 인그레스 컨트롤러다. +* [Apache APISIX 인그레스 컨트롤러](https://github.com/apache/apisix-ingress-controller)는 [Apache APISIX](https://github.com/apache/apisix) 기반의 인그레스 컨트롤러이다. * [Avi 쿠버네티스 오퍼레이터](https://github.com/vmware/load-balancer-and-ingress-services-for-kubernetes)는 [VMware NSX Advanced Load Balancer](https://avinetworks.com/)을 사용하는 L4-L7 로드 밸런싱을 제공한다. * [Citrix 인그레스 컨트롤러](https://github.com/citrix/citrix-k8s-ingress-controller#readme)는 Citrix 애플리케이션 딜리버리 컨트롤러에서 작동한다. @@ -42,7 +43,7 @@ kube-controller-manager 바이너리의 일부로 실행되는 컨트롤러의 기반 인그레스 컨트롤러다. * [쿠버네티스 용 Kong 인그레스 컨트롤러](https://github.com/Kong/kubernetes-ingress-controller#readme)는 [Kong 게이트웨이](https://konghq.com/kong/)를 구동하는 인그레스 컨트롤러다. -* [쿠버네티스 용 NGINX 인그레스 컨트롤러](https://www.nginx.com/products/nginx/kubernetes-ingress-controller)는 [NGINX](https://www.nginx.com/resources/glossary) +* [쿠버네티스 용 NGINX 인그레스 컨트롤러](https://www.nginx.com/products/nginx-ingress-controller/)는 [NGINX](https://www.nginx.com/resources/glossary/nginx/) 웹서버(프록시로 사용)와 함께 작동한다. * [Skipper](https://opensource.zalando.com/skipper/kubernetes/ingress-controller/)는 사용자의 커스텀 프록시를 구축하기 위한 라이브러리로 설계된 쿠버네티스 인그레스와 같은 유스케이스를 포함한 서비스 구성을 위한 HTTP 라우터 및 역방향 프록시다. * [Traefik 쿠버네티스 인그레스 제공자](https://doc.traefik.io/traefik/providers/kubernetes-ingress/)는 diff --git a/content/ko/docs/concepts/services-networking/ingress.md b/content/ko/docs/concepts/services-networking/ingress.md index 5b91356437..a55302f059 100644 --- a/content/ko/docs/concepts/services-networking/ingress.md +++ b/content/ko/docs/concepts/services-networking/ingress.md @@ -376,7 +376,7 @@ graph LR; 트래픽을 일치 시킬 수 있다. 예를 들어, 다음 인그레스는 `first.bar.com`에 요청된 트래픽을 -`service1`로, `second.foo.com`는 `service2`로, 호스트 이름이 정의되지 +`service1`로, `second.bar.com`는 `service2`로, 호스트 이름이 정의되지 않은(즉, 요청 헤더가 표시 되지 않는) IP 주소로의 모든 트래픽은 `service3`로 라우팅 한다. diff --git a/content/ko/docs/concepts/services-networking/service.md b/content/ko/docs/concepts/services-networking/service.md index da9d353d6f..5d1acf5392 100644 --- a/content/ko/docs/concepts/services-networking/service.md +++ b/content/ko/docs/concepts/services-networking/service.md @@ -134,7 +134,7 @@ spec: * 한 서비스에서 다른 {{< glossary_tooltip term_id="namespace" text="네임스페이스">}} 또는 다른 클러스터의 서비스를 지정하려고 한다. * 워크로드를 쿠버네티스로 마이그레이션하고 있다. 해당 방식을 평가하는 동안, - 쿠버네티스에서는 일정 비율의 백엔드만 실행한다. + 쿠버네티스에서는 백엔드의 일부만 실행한다. 이러한 시나리오 중에서 파드 셀렉터 _없이_ 서비스를 정의 할 수 있다. 예를 들면 diff --git a/content/ko/docs/concepts/workloads/controllers/job.md b/content/ko/docs/concepts/workloads/controllers/job.md index 0f04051ff1..64b5d3879d 100644 --- a/content/ko/docs/concepts/workloads/controllers/job.md +++ b/content/ko/docs/concepts/workloads/controllers/job.md @@ -13,7 +13,7 @@ weight: 50 -잡에서 하나 이상의 파드를 생성하고 지정된 수의 파드가 성공적으로 종료되도록 한다. +잡에서 하나 이상의 파드를 생성하고 지정된 수의 파드가 성공적으로 종료될 때까지 계속해서 파드의 실행을 재시도한다. 파드가 성공적으로 완료되면, 성공적으로 완료된 잡을 추적한다. 지정된 수의 성공 완료에 도달하면, 작업(즉, 잡)이 완료된다. 잡을 삭제하면 잡이 생성한 파드가 정리된다. diff --git a/content/ko/docs/concepts/workloads/controllers/ttlafterfinished.md b/content/ko/docs/concepts/workloads/controllers/ttlafterfinished.md index a941266230..5ed869fb57 100644 --- a/content/ko/docs/concepts/workloads/controllers/ttlafterfinished.md +++ b/content/ko/docs/concepts/workloads/controllers/ttlafterfinished.md @@ -76,4 +76,4 @@ TTL 컨트롤러는 쿠버네티스 리소스에 * [자동으로 잡 정리](/ko/docs/concepts/workloads/controllers/job/#완료된-잡을-자동으로-정리) -* [디자인 문서](https://github.com/kubernetes/enhancements/blob/master/keps/sig-apps/0026-ttl-after-finish.md) +* [디자인 문서](https://github.com/kubernetes/enhancements/blob/master/keps/sig-apps/592-ttl-after-finish/README.md) diff --git a/content/ko/docs/reference/_index.md b/content/ko/docs/reference/_index.md index 14fee6ee9c..c294f2efb5 100644 --- a/content/ko/docs/reference/_index.md +++ b/content/ko/docs/reference/_index.md @@ -18,7 +18,8 @@ content_type: concept ## API 레퍼런스 -* [쿠버네티스 API 레퍼런스 {{< param "version" >}}](/docs/reference/generated/kubernetes-api/{{< param "version" >}}/) +* [쿠버네티스 API 레퍼런스](/docs/reference/kubernetes-api/) +* [쿠버네티스 {{< param "version" >}}용 원페이지(One-page) API 레퍼런스](/docs/reference/generated/kubernetes-api/{{< param "version" >}}/) * [쿠버네티스 API 사용](/ko/docs/reference/using-api/) - 쿠버네티스 API에 대한 개요 ## API 클라이언트 라이브러리 diff --git a/content/ko/docs/reference/command-line-tools-reference/feature-gates.md b/content/ko/docs/reference/command-line-tools-reference/feature-gates.md index 6fa2c58a56..715dbf4801 100644 --- a/content/ko/docs/reference/command-line-tools-reference/feature-gates.md +++ b/content/ko/docs/reference/command-line-tools-reference/feature-gates.md @@ -48,13 +48,15 @@ kubelet과 같은 컴포넌트의 기능 게이트를 설정하려면, 기능 | 기능 | 디폴트 | 단계 | 도입 | 종료 | |---------|---------|-------|-------|-------| -| `AnyVolumeDataSource` | `false` | 알파 | 1.18 | | | `APIListChunking` | `false` | 알파 | 1.8 | 1.8 | | `APIListChunking` | `true` | 베타 | 1.9 | | | `APIPriorityAndFairness` | `false` | 알파 | 1.17 | 1.19 | | `APIPriorityAndFairness` | `true` | 베타 | 1.20 | | -| `APIResponseCompression` | `false` | 알파 | 1.7 | | +| `APIResponseCompression` | `false` | 알파 | 1.7 | 1.15 | +| `APIResponseCompression` | `false` | 베타 | 1.16 | | | `APIServerIdentity` | `false` | 알파 | 1.20 | | +| `AllowInsecureBackendProxy` | `true` | 베타 | 1.17 | | +| `AnyVolumeDataSource` | `false` | 알파 | 1.18 | | | `AppArmor` | `true` | 베타 | 1.4 | | | `BalanceAttachedNodeVolumes` | `false` | 알파 | 1.11 | | | `BoundServiceAccountTokenVolume` | `false` | 알파 | 1.13 | | @@ -77,7 +79,8 @@ kubelet과 같은 컴포넌트의 기능 게이트를 설정하려면, 기능 | `CSIMigrationGCE` | `false` | 알파 | 1.14 | 1.16 | | `CSIMigrationGCE` | `false` | 베타 | 1.17 | | | `CSIMigrationGCEComplete` | `false` | 알파 | 1.17 | | -| `CSIMigrationOpenStack` | `false` | 알파 | 1.14 | | +| `CSIMigrationOpenStack` | `false` | 알파 | 1.14 | 1.17 | +| `CSIMigrationOpenStack` | `true` | 베타 | 1.18 | | | `CSIMigrationOpenStackComplete` | `false` | 알파 | 1.17 | | | `CSIMigrationvSphere` | `false` | 베타 | 1.19 | | | `CSIMigrationvSphereComplete` | `false` | 베타 | 1.19 | | @@ -89,26 +92,23 @@ kubelet과 같은 컴포넌트의 기능 게이트를 설정하려면, 기능 | `ConfigurableFSGroupPolicy` | `true` | 베타 | 1.20 | | | `CronJobControllerV2` | `false` | 알파 | 1.20 | | | `CustomCPUCFSQuotaPeriod` | `false` | 알파 | 1.12 | | -| `CustomResourceDefaulting` | `false` | 알파| 1.15 | 1.15 | -| `CustomResourceDefaulting` | `true` | 베타 | 1.16 | | | `DefaultPodTopologySpread` | `false` | 알파 | 1.19 | 1.19 | | `DefaultPodTopologySpread` | `true` | 베타 | 1.20 | | | `DevicePlugins` | `false` | 알파 | 1.8 | 1.9 | | `DevicePlugins` | `true` | 베타 | 1.10 | | | `DisableAcceleratorUsageMetrics` | `false` | 알파 | 1.19 | 1.19 | -| `DisableAcceleratorUsageMetrics` | `true` | 베타 | 1.20 | 1.22 | +| `DisableAcceleratorUsageMetrics` | `true` | 베타 | 1.20 | | | `DownwardAPIHugePages` | `false` | 알파 | 1.20 | | -| `DryRun` | `false` | 알파 | 1.12 | 1.12 | -| `DryRun` | `true` | 베타 | 1.13 | | | `DynamicKubeletConfig` | `false` | 알파 | 1.4 | 1.10 | | `DynamicKubeletConfig` | `true` | 베타 | 1.11 | | +| `EfficientWatchResumption` | `false` | 알파 | 1.20 | | | `EndpointSlice` | `false` | 알파 | 1.16 | 1.16 | | `EndpointSlice` | `false` | 베타 | 1.17 | | | `EndpointSlice` | `true` | 베타 | 1.18 | | | `EndpointSliceNodeName` | `false` | 알파 | 1.20 | | | `EndpointSliceProxying` | `false` | 알파 | 1.18 | 1.18 | | `EndpointSliceProxying` | `true` | 베타 | 1.19 | | -| `EndpointSliceTerminating` | `false` | 알파 | 1.20 | | +| `EndpointSliceTerminatingCondition` | `false` | 알파 | 1.20 | | | `EphemeralContainers` | `false` | 알파 | 1.16 | | | `ExpandCSIVolumes` | `false` | 알파 | 1.14 | 1.15 | | `ExpandCSIVolumes` | `true` | 베타 | 1.16 | | @@ -119,19 +119,22 @@ kubelet과 같은 컴포넌트의 기능 게이트를 설정하려면, 기능 | `ExperimentalHostUserNamespaceDefaulting` | `false` | 베타 | 1.5 | | | `GenericEphemeralVolume` | `false` | 알파 | 1.19 | | | `GracefulNodeShutdown` | `false` | 알파 | 1.20 | | +| `HPAContainerMetrics` | `false` | 알파 | 1.20 | | | `HPAScaleToZero` | `false` | 알파 | 1.16 | | | `HugePageStorageMediumSize` | `false` | 알파 | 1.18 | 1.18 | | `HugePageStorageMediumSize` | `true` | 베타 | 1.19 | | -| `HyperVContainer` | `false` | 알파 | 1.10 | | +| `IPv6DualStack` | `false` | 알파 | 1.15 | | | `ImmutableEphemeralVolumes` | `false` | 알파 | 1.18 | 1.18 | | `ImmutableEphemeralVolumes` | `true` | 베타 | 1.19 | | -| `IPv6DualStack` | `false` | 알파 | 1.16 | | -| `LegacyNodeRoleBehavior` | `true` | 알파 | 1.16 | | +| `KubeletCredentialProviders` | `false` | 알파 | 1.20 | | +| `KubeletPodResources` | `true` | 알파 | 1.13 | 1.14 | +| `KubeletPodResources` | `true` | 베타 | 1.15 | | +| `LegacyNodeRoleBehavior` | `false` | 알파 | 1.16 | 1.18 | +| `LegacyNodeRoleBehavior` | `true` | True | 1.19 | | | `LocalStorageCapacityIsolation` | `false` | 알파 | 1.7 | 1.9 | | `LocalStorageCapacityIsolation` | `true` | 베타 | 1.10 | | | `LocalStorageCapacityIsolationFSQuotaMonitoring` | `false` | 알파 | 1.15 | | | `MixedProtocolLBService` | `false` | 알파 | 1.20 | | -| `MountContainers` | `false` | 알파 | 1.9 | | | `NodeDisruptionExclusion` | `false` | 알파 | 1.16 | 1.18 | | `NodeDisruptionExclusion` | `true` | 베타 | 1.19 | | | `NonPreemptingPriority` | `false` | 알파 | 1.15 | 1.18 | @@ -143,25 +146,27 @@ kubelet과 같은 컴포넌트의 기능 게이트를 설정하려면, 기능 | `ProcMountType` | `false` | 알파 | 1.12 | | | `QOSReserved` | `false` | 알파 | 1.11 | | | `RemainingItemCount` | `false` | 알파 | 1.15 | | +| `RemoveSelfLink` | `false` | 알파 | 1.16 | 1.19 | +| `RemoveSelfLink` | `true` | 베타 | 1.20 | | | `RootCAConfigMap` | `false` | 알파 | 1.13 | 1.19 | | `RootCAConfigMap` | `true` | 베타 | 1.20 | | | `RotateKubeletServerCertificate` | `false` | 알파 | 1.7 | 1.11 | | `RotateKubeletServerCertificate` | `true` | 베타 | 1.12 | | | `RunAsGroup` | `true` | 베타 | 1.14 | | -| `RuntimeClass` | `false` | 알파 | 1.12 | 1.13 | -| `RuntimeClass` | `true` | 베타 | 1.14 | | | `SCTPSupport` | `false` | 알파 | 1.12 | 1.18 | | `SCTPSupport` | `true` | 베타 | 1.19 | | | `ServerSideApply` | `false` | 알파 | 1.14 | 1.15 | | `ServerSideApply` | `true` | 베타 | 1.16 | | -| `ServiceAccountIssuerDiscovery` | `false` | 알파 | 1.18 | | -| `ServiceLBNodePortControl` | `false` | 알파 | 1.20 | 1.20 | +| `ServiceAccountIssuerDiscovery` | `false` | 알파 | 1.18 | 1.19 | +| `ServiceAccountIssuerDiscovery` | `true` | 베타 | 1.20 | | +| `ServiceLBNodePortControl` | `false` | 알파 | 1.20 | | | `ServiceNodeExclusion` | `false` | 알파 | 1.8 | 1.18 | | `ServiceNodeExclusion` | `true` | 베타 | 1.19 | | | `ServiceTopology` | `false` | 알파 | 1.17 | | -| `SizeMemoryBackedVolumes` | `false` | 알파 | 1.20 | | | `SetHostnameAsFQDN` | `false` | 알파 | 1.19 | 1.19 | | `SetHostnameAsFQDN` | `true` | 베타 | 1.20 | | +| `SizeMemoryBackedVolumes` | `false` | 알파 | 1.20 | | +| `StorageVersionAPI` | `false` | 알파 | 1.20 | | | `StorageVersionHash` | `false` | 알파 | 1.14 | 1.14 | | `StorageVersionHash` | `true` | 베타 | 1.15 | | | `Sysctls` | `true` | 베타 | 1.11 | | @@ -170,11 +175,11 @@ kubelet과 같은 컴포넌트의 기능 게이트를 설정하려면, 기능 | `TopologyManager` | `true` | 베타 | 1.18 | | | `ValidateProxyRedirects` | `false` | 알파 | 1.12 | 1.13 | | `ValidateProxyRedirects` | `true` | 베타 | 1.14 | | -| `WindowsEndpointSliceProxying` | `false` | 알파 | 1.19 | | -| `WindowsGMSA` | `false` | 알파 | 1.14 | | -| `WindowsGMSA` | `true` | 베타 | 1.16 | | +| `WarningHeaders` | `true` | 베타 | 1.19 | | | `WinDSR` | `false` | 알파 | 1.14 | | -| `WinOverlay` | `false` | 알파 | 1.14 | | +| `WinOverlay` | `false` | 알파 | 1.14 | 1.19 | +| `WinOverlay` | `true` | 베타 | 1.20 | | +| `WindowsEndpointSliceProxying` | `false` | 알파 | 1.19 | | {{< /table >}} ### GA 또는 사용 중단된 기능을 위한 기능 게이트 @@ -228,6 +233,9 @@ kubelet과 같은 컴포넌트의 기능 게이트를 설정하려면, 기능 | `CustomResourceWebhookConversion` | `false` | 알파 | 1.13 | 1.14 | | `CustomResourceWebhookConversion` | `true` | 베타 | 1.15 | 1.15 | | `CustomResourceWebhookConversion` | `true` | GA | 1.16 | - | +| `DryRun` | `false` | 알파 | 1.12 | 1.12 | +| `DryRun` | `true` | 베타 | 1.13 | 1.18 | +| `DryRun` | `true` | GA | 1.19 | - | | `DynamicAuditing` | `false` | 알파 | 1.13 | 1.18 | | `DynamicAuditing` | - | 사용중단 | 1.19 | - | | `DynamicProvisioningScheduling` | `false` | 알파 | 1.11 | 1.11 | @@ -247,23 +255,28 @@ kubelet과 같은 컴포넌트의 기능 게이트를 설정하려면, 기능 | `HugePages` | `false` | 알파 | 1.8 | 1.9 | | `HugePages` | `true` | 베타| 1.10 | 1.13 | | `HugePages` | `true` | GA | 1.14 | - | +| `HyperVContainer` | `false` | 알파 | 1.10 | 1.19 | +| `HyperVContainer` | `false` | 사용중단 | 1.20 | - | | `Initializers` | `false` | 알파 | 1.7 | 1.13 | | `Initializers` | - | 사용중단 | 1.14 | - | | `KubeletConfigFile` | `false` | 알파 | 1.8 | 1.9 | | `KubeletConfigFile` | - | 사용중단 | 1.10 | - | -| `KubeletCredentialProviders` | `false` | 알파 | 1.20 | 1.20 | | `KubeletPluginsWatcher` | `false` | 알파 | 1.11 | 1.11 | | `KubeletPluginsWatcher` | `true` | 베타 | 1.12 | 1.12 | | `KubeletPluginsWatcher` | `true` | GA | 1.13 | - | | `KubeletPodResources` | `false` | 알파 | 1.13 | 1.14 | | `KubeletPodResources` | `true` | 베타 | 1.15 | | | `KubeletPodResources` | `true` | GA | 1.20 | | +| `MountContainers` | `false` | 알파 | 1.9 | 1.16 | +| `MountContainers` | `false` | 사용중단 | 1.17 | - | | `MountPropagation` | `false` | 알파 | 1.8 | 1.9 | | `MountPropagation` | `true` | 베타 | 1.10 | 1.11 | | `MountPropagation` | `true` | GA | 1.12 | - | | `NodeLease` | `false` | 알파 | 1.12 | 1.13 | | `NodeLease` | `true` | 베타 | 1.14 | 1.16 | | `NodeLease` | `true` | GA | 1.17 | - | +| `PVCProtection` | `false` | 알파 | 1.9 | 1.9 | +| `PVCProtection` | - | 사용중단 | 1.10 | - | | `PersistentLocalVolumes` | `false` | 알파 | 1.7 | 1.9 | | `PersistentLocalVolumes` | `true` | 베타 | 1.10 | 1.13 | | `PersistentLocalVolumes` | `true` | GA | 1.14 | - | @@ -276,8 +289,6 @@ kubelet과 같은 컴포넌트의 기능 게이트를 설정하려면, 기능 | `PodShareProcessNamespace` | `false` | 알파 | 1.10 | 1.11 | | `PodShareProcessNamespace` | `true` | 베타 | 1.12 | 1.16 | | `PodShareProcessNamespace` | `true` | GA | 1.17 | - | -| `PVCProtection` | `false` | 알파 | 1.9 | 1.9 | -| `PVCProtection` | - | 사용중단 | 1.10 | - | | `RequestManagement` | `false` | 알파 | 1.15 | 1.16 | | `ResourceLimitsPriorityFunction` | `false` | 알파 | 1.9 | 1.18 | | `ResourceLimitsPriorityFunction` | - | 사용중단 | 1.19 | - | @@ -398,62 +409,131 @@ kubelet과 같은 컴포넌트의 기능 게이트를 설정하려면, 기능 각 기능 게이트는 특정 기능을 활성화/비활성화하도록 설계되었다. +- `APIListChunking`: API 클라이언트가 API 서버에서 (`LIST` 또는 `GET`) + 리소스를 청크(chunks)로 검색할 수 있도록 한다. +- `APIPriorityAndFairness`: 각 서버의 우선 순위와 공정성을 통해 동시 요청을 + 관리할 수 있다. (`RequestManagement` 에서 이름이 변경됨) +- `APIResponseCompression`: `LIST` 또는 `GET` 요청에 대한 API 응답을 압축한다. +- `APIServerIdentity`: 클러스터의 각 API 서버에 ID를 할당한다. - `Accelerators`: 도커 사용 시 Nvidia GPU 지원 활성화한다. - `AdvancedAuditing`: [고급 감사](/docs/tasks/debug-application-cluster/audit/#advanced-audit) 기능을 활성화한다. -- `AffinityInAnnotations`(*사용 중단됨*): [파드 어피니티 또는 안티-어피니티](/ko/docs/concepts/scheduling-eviction/assign-pod-node/#어피니티-affinity-와-안티-어피니티-anti-affinity) 설정을 활성화한다. +- `AffinityInAnnotations`(*사용 중단됨*): [파드 어피니티 또는 안티-어피니티](/ko/docs/concepts/scheduling-eviction/assign-pod-node/#어피니티-affinity-와-안티-어피니티-anti-affinity) + 설정을 활성화한다. - `AllowExtTrafficLocalEndpoints`: 서비스가 외부 요청을 노드의 로컬 엔드포인트로 라우팅할 수 있도록 한다. +- `AllowInsecureBackendProxy`: 사용자가 파드 로그 요청에서 kubelet의 + TLS 확인을 건너뛸 수 있도록 한다. - `AnyVolumeDataSource`: {{< glossary_tooltip text="PVC" term_id="persistent-volume-claim" >}}의 `DataSource` 로 모든 사용자 정의 리소스 사용을 활성화한다. -- `APIListChunking`: API 클라이언트가 API 서버에서 (`LIST` 또는 `GET`) 리소스를 청크(chunks)로 검색할 수 있도록 한다. -- `APIPriorityAndFairness`: 각 서버의 우선 순위와 공정성을 통해 동시 요청을 관리할 수 있다. (`RequestManagement` 에서 이름이 변경됨) -- `APIResponseCompression`: `LIST` 또는 `GET` 요청에 대한 API 응답을 압축한다. -- `APIServerIdentity`: 클러스터의 각 kube-apiserver에 ID를 할당한다. - `AppArmor`: 도커를 사용할 때 리눅스 노드에서 AppArmor 기반의 필수 접근 제어를 활성화한다. - 자세한 내용은 [AppArmor 튜토리얼](/ko/docs/tutorials/clusters/apparmor/)을 참고한다. + 자세한 내용은 [AppArmor 튜토리얼](/ko/docs/tutorials/clusters/apparmor/)을 참고한다. - `AttachVolumeLimit`: 볼륨 플러그인이 노드에 연결될 수 있는 볼륨 수에 대한 제한을 보고하도록 한다. - 자세한 내용은 [동적 볼륨 제한](/ko/docs/concepts/storage/storage-limits/#동적-볼륨-한도)을 참고한다. + 자세한 내용은 [동적 볼륨 제한](/ko/docs/concepts/storage/storage-limits/#동적-볼륨-한도)을 참고한다. - `BalanceAttachedNodeVolumes`: 스케줄링 시 균형 잡힌 리소스 할당을 위해 고려할 노드의 볼륨 수를 포함한다. 스케줄러가 결정을 내리는 동안 CPU, 메모리 사용률 및 볼륨 수가 더 가까운 노드가 선호된다. - `BlockVolume`: 파드에서 원시 블록 장치의 정의와 사용을 활성화한다. - 자세한 내용은 [원시 블록 볼륨 지원](/ko/docs/concepts/storage/persistent-volumes/#원시-블록-볼륨-지원)을 - 참고한다. + 자세한 내용은 [원시 블록 볼륨 지원](/ko/docs/concepts/storage/persistent-volumes/#원시-블록-볼륨-지원)을 + 참고한다. - `BoundServiceAccountTokenVolume`: ServiceAccountTokenVolumeProjection으로 구성된 프로젝션 볼륨을 사용하도록 서비스어카운트 볼륨을 - 마이그레이션한다. 클러스터 관리자는 `serviceaccount_stale_tokens_total` 메트릭을 사용하여 - 확장 토큰에 의존하는 워크로드를 모니터링 할 수 있다. 이러한 워크로드가 없는 경우 `--service-account-extend-token-expiration=false` 플래그로 - `kube-apiserver`를 시작하여 확장 토큰 기능을 끈다. - 자세한 내용은 [바운드 서비스 계정 토큰](https://github.com/kubernetes/enhancements/blob/master/keps/sig-auth/1205-bound-service-account-tokens/README.md)을 - 확인한다. -- `ConfigurableFSGroupPolicy`: 파드에 볼륨을 마운트할 때 fsGroups에 대한 볼륨 권한 변경 정책을 구성할 수 있다. 자세한 내용은 [파드에 대한 볼륨 권한 및 소유권 변경 정책 구성](/docs/tasks/configure-pod-container/security-context/#configure-volume-permission-and-ownership-change-policy-for-pods)을 참고한다. --`CronJobControllerV2` : {{< glossary_tooltip text="크론잡" term_id="cronjob" >}} 컨트롤러의 대체 구현을 사용한다. 그렇지 않으면 동일한 컨트롤러의 버전 1이 선택된다. 버전 2 컨트롤러는 실험적인 성능 향상을 제공한다. -- `CPUManager`: 컨테이너 수준의 CPU 어피니티 지원을 활성화한다. [CPU 관리 정책](/docs/tasks/administer-cluster/cpu-management-policies/)을 참고한다. + 마이그레이션한다. 클러스터 관리자는 `serviceaccount_stale_tokens_total` 메트릭을 사용하여 + 확장 토큰에 의존하는 워크로드를 모니터링 할 수 있다. 이러한 워크로드가 없는 경우 `--service-account-extend-token-expiration=false` 플래그로 + `kube-apiserver`를 시작하여 확장 토큰 기능을 끈다. + 자세한 내용은 [바운드 서비스 계정 토큰](https://github.com/kubernetes/enhancements/blob/master/keps/sig-auth/1205-bound-service-account-tokens/README.md)을 + 확인한다. +- `CPUManager`: 컨테이너 수준의 CPU 어피니티 지원을 활성화한다. + [CPU 관리 정책](/docs/tasks/administer-cluster/cpu-management-policies/)을 참고한다. - `CRIContainerLogRotation`: cri 컨테이너 런타임에 컨테이너 로그 로테이션을 활성화한다. -- `CSIBlockVolume`: 외부 CSI 볼륨 드라이버가 블록 스토리지를 지원할 수 있게 한다. 자세한 내용은 [`csi` 원시 블록 볼륨 지원](/ko/docs/concepts/storage/volumes/#csi-원시-raw-블록-볼륨-지원) 문서를 참고한다. -- `CSIDriverRegistry`: csi.storage.k8s.io에서 CSIDriver API 오브젝트와 관련된 모든 로직을 활성화한다. +- `CSIBlockVolume`: 외부 CSI 볼륨 드라이버가 블록 스토리지를 지원할 수 있게 한다. + 자세한 내용은 [`csi` 원시 블록 볼륨 지원](/ko/docs/concepts/storage/volumes/#csi-원시-raw-블록-볼륨-지원) + 문서를 참고한다. +- `CSIDriverRegistry`: csi.storage.k8s.io에서 CSIDriver API 오브젝트와 관련된 + 모든 로직을 활성화한다. - `CSIInlineVolume`: 파드에 대한 CSI 인라인 볼륨 지원을 활성화한다. -- `CSIMigration`: shim 및 변환 로직을 통해 볼륨 작업을 인-트리 플러그인에서 사전 설치된 해당 CSI 플러그인으로 라우팅할 수 있다. -- `CSIMigrationAWS`: shim 및 변환 로직을 통해 볼륨 작업을 AWS-EBS 인-트리 플러그인에서 EBS CSI 플러그인으로 라우팅할 수 있다. 노드에 EBS CSI 플러그인이 설치와 구성이 되어 있지 않은 경우 인-트리 EBS 플러그인으로 폴백(falling back)을 지원한다. CSIMigration 기능 플래그가 필요하다. -- `CSIMigrationAWSComplete`: kubelet 및 볼륨 컨트롤러에서 EBS 인-트리 플러그인 등록을 중지하고 shim 및 변환 로직을 사용하여 볼륨 작업을 AWS-EBS 인-트리 플러그인에서 EBS CSI 플러그인으로 라우팅할 수 있다. 클러스터의 모든 노드에 CSIMigration과 CSIMigrationAWS 기능 플래그가 활성화되고 EBS CSI 플러그인이 설치 및 구성이 되어 있어야 한다. -- `CSIMigrationAzureDisk`: shim 및 변환 로직을 통해 볼륨 작업을 Azure-Disk 인-트리 플러그인에서 AzureDisk CSI 플러그인으로 라우팅할 수 있다. 노드에 AzureDisk CSI 플러그인이 설치와 구성이 되어 있지 않은 경우 인-트리 AzureDisk 플러그인으로 폴백을 지원한다. CSIMigration 기능 플래그가 필요하다. -- `CSIMigrationAzureDiskComplete`: kubelet 및 볼륨 컨트롤러에서 Azure-Disk 인-트리 플러그인 등록을 중지하고 shim 및 변환 로직을 사용하여 볼륨 작업을 Azure-Disk 인-트리 플러그인에서 AzureDisk CSI 플러그인으로 라우팅할 수 있다. 클러스터의 모든 노드에 CSIMigration과 CSIMigrationAzureDisk 기능 플래그가 활성화되고 AzureDisk CSI 플러그인이 설치 및 구성이 되어 있어야 한다. -- `CSIMigrationAzureFile`: shim 및 변환 로직을 통해 볼륨 작업을 Azure-File 인-트리 플러그인에서 AzureFile CSI 플러그인으로 라우팅할 수 있다. 노드에 AzureFile CSI 플러그인이 설치 및 구성이 되어 있지 않은 경우 인-트리 AzureFile 플러그인으로 폴백을 지원한다. CSIMigration 기능 플래그가 필요하다. -- `CSIMigrationAzureFileComplete`: kubelet 및 볼륨 컨트롤러에서 Azure 파일 인-트리 플러그인 등록을 중지하고 shim 및 변환 로직을 통해 볼륨 작업을 Azure 파일 인-트리 플러그인에서 AzureFile CSI 플러그인으로 라우팅할 수 있다. 클러스터의 모든 노드에 CSIMigration과 CSIMigrationAzureFile 기능 플래그가 활성화되고 AzureFile CSI 플러그인이 설치 및 구성이 되어 있어야 한다. -- `CSIMigrationGCE`: shim 및 변환 로직을 통해 볼륨 작업을 GCE-PD 인-트리 플러그인에서 PD CSI 플러그인으로 라우팅할 수 있다. 노드에 PD CSI 플러그인이 설치 및 구성이 되어 있지 않은 경우 인-트리 GCE 플러그인으로 폴백을 지원한다. CSIMigration 기능 플래그가 필요하다. -- `CSIMigrationGCEComplete`: kubelet 및 볼륨 컨트롤러에서 GCE-PD 인-트리 플러그인 등록을 중지하고 shim 및 변환 로직을 통해 볼륨 작업을 GCE-PD 인-트리 플러그인에서 PD CSI 플러그인으로 라우팅할 수 있다. CSIMigration과 CSIMigrationGCE 기능 플래그가 필요하다. -- `CSIMigrationOpenStack`: shim 및 변환 로직을 통해 볼륨 작업을 Cinder 인-트리 플러그인에서 Cinder CSI 플러그인으로 라우팅할 수 있다. 노드에 Cinder CSI 플러그인이 설치 및 구성이 되어 있지 않은 경우 인-트리 Cinder 플러그인으로 폴백을 지원한다. CSIMigration 기능 플래그가 필요하다. -- `CSIMigrationOpenStackComplete`: kubelet 및 볼륨 컨트롤러에서 Cinder 인-트리 플러그인 등록을 중지하고 shim 및 변환 로직이 Cinder 인-트리 플러그인에서 Cinder CSI 플러그인으로 볼륨 작업을 라우팅할 수 있도록 한다. 클러스터의 모든 노드에 CSIMigration과 CSIMigrationOpenStack 기능 플래그가 활성화되고 Cinder CSI 플러그인이 설치 및 구성이 되어 있어야 한다. -- `CSIMigrationvSphere`: vSphere 인-트리 플러그인에서 vSphere CSI 플러그인으로 볼륨 작업을 라우팅하는 shim 및 변환 로직을 사용한다. 노드에 vSphere CSI 플러그인이 설치 및 구성이 되어 있지 않은 경우 인-트리 vSphere 플러그인으로 폴백을 지원한다. CSIMigration 기능 플래그가 필요하다. -- `CSIMigrationvSphereComplete`: kubelet 및 볼륨 컨트롤러에서 vSphere 인-트리 플러그인 등록을 중지하고 shim 및 변환 로직을 활성화하여 vSphere 인-트리 플러그인에서 vSphere CSI 플러그인으로 볼륨 작업을 라우팅할 수 있도록 한다. CSIMigration 및 CSIMigrationvSphere 기능 플래그가 활성화되고 vSphere CSI 플러그인이 클러스터의 모든 노드에 설치 및 구성이 되어 있어야 한다. +- `CSIMigration`: shim 및 변환 로직을 통해 볼륨 작업을 인-트리 플러그인에서 + 사전 설치된 해당 CSI 플러그인으로 라우팅할 수 있다. +- `CSIMigrationAWS`: shim 및 변환 로직을 통해 볼륨 작업을 + AWS-EBS 인-트리 플러그인에서 EBS CSI 플러그인으로 라우팅할 수 있다. 노드에 + EBS CSI 플러그인이 설치와 구성이 되어 있지 않은 경우 인-트리 EBS 플러그인으로 + 폴백(falling back)을 지원한다. CSIMigration 기능 플래그가 필요하다. +- `CSIMigrationAWSComplete`: kubelet 및 볼륨 컨트롤러에서 EBS 인-트리 + 플러그인 등록을 중지하고 shim 및 변환 로직을 사용하여 볼륨 작업을 AWS-EBS + 인-트리 플러그인에서 EBS CSI 플러그인으로 라우팅할 수 있다. + 클러스터의 모든 노드에 CSIMigration과 CSIMigrationAWS 기능 플래그가 활성화되고 + EBS CSI 플러그인이 설치 및 구성이 되어 있어야 한다. +- `CSIMigrationAzureDisk`: shim 및 변환 로직을 통해 볼륨 작업을 + Azure-Disk 인-트리 플러그인에서 AzureDisk CSI 플러그인으로 라우팅할 수 있다. + 노드에 AzureDisk CSI 플러그인이 설치와 구성이 되어 있지 않은 경우 인-트리 + AzureDisk 플러그인으로 폴백을 지원한다. CSIMigration 기능 플래그가 + 필요하다. +- `CSIMigrationAzureDiskComplete`: kubelet 및 볼륨 컨트롤러에서 Azure-Disk 인-트리 + 플러그인 등록을 중지하고 shim 및 변환 로직을 사용하여 볼륨 작업을 + Azure-Disk 인-트리 플러그인에서 AzureDisk CSI 플러그인으로 + 라우팅할 수 있다. 클러스터의 모든 노드에 CSIMigration과 CSIMigrationAzureDisk 기능 + 플래그가 활성화되고 AzureDisk CSI 플러그인이 설치 및 구성이 되어 + 있어야 한다. +- `CSIMigrationAzureFile`: shim 및 변환 로직을 통해 볼륨 작업을 + Azure-File 인-트리 플러그인에서 AzureFile CSI 플러그인으로 라우팅할 수 있다. + 노드에 AzureFile CSI 플러그인이 설치 및 구성이 되어 있지 않은 경우 인-트리 + AzureFile 플러그인으로 폴백을 지원한다. CSIMigration 기능 플래그가 + 필요하다. +- `CSIMigrationAzureFileComplete`: kubelet 및 볼륨 컨트롤러에서 Azure 파일 인-트리 + 플러그인 등록을 중지하고 shim 및 변환 로직을 통해 볼륨 작업을 + Azure 파일 인-트리 플러그인에서 AzureFile CSI 플러그인으로 + 라우팅할 수 있다. 클러스터의 모든 노드에 CSIMigration과 CSIMigrationAzureFile 기능 + 플래그가 활성화되고 AzureFile CSI 플러그인이 설치 및 구성이 되어 + 있어야 한다. +- `CSIMigrationGCE`: shim 및 변환 로직을 통해 볼륨 작업을 + GCE-PD 인-트리 플러그인에서 PD CSI 플러그인으로 라우팅할 수 있다. 노드에 + PD CSI 플러그인이 설치 및 구성이 되어 있지 않은 경우 인-트리 GCE 플러그인으로 폴백을 + 지원한다. CSIMigration 기능 플래그가 필요하다. +- `CSIMigrationGCEComplete`: kubelet 및 볼륨 컨트롤러에서 GCE-PD + 인-트리 플러그인 등록을 중지하고 shim 및 변환 로직을 통해 볼륨 작업을 GCE-PD + 인-트리 플러그인에서 PD CSI 플러그인으로 라우팅할 수 있다. + CSIMigration과 CSIMigrationGCE 기능 플래그가 활성화되고 PD CSI + 플러그인이 클러스터의 모든 노드에 설치 및 구성이 되어 있어야 한다. +- `CSIMigrationOpenStack`: shim 및 변환 로직을 통해 볼륨 작업을 + Cinder 인-트리 플러그인에서 Cinder CSI 플러그인으로 라우팅할 수 있다. 노드에 + Cinder CSI 플러그인이 설치 및 구성이 되어 있지 않은 경우 인-트리 + Cinder 플러그인으로 폴백을 지원한다. CSIMigration 기능 플래그가 필요하다. +- `CSIMigrationOpenStackComplete`: kubelet 및 볼륨 컨트롤러에서 + Cinder 인-트리 플러그인 등록을 중지하고 shim 및 변환 로직이 Cinder 인-트리 + 플러그인에서 Cinder CSI 플러그인으로 볼륨 작업을 라우팅할 수 있도록 한다. + 클러스터의 모든 노드에 CSIMigration과 CSIMigrationOpenStack 기능 플래그가 활성화되고 + Cinder CSI 플러그인이 설치 및 구성이 되어 있어야 한다. +- `CSIMigrationvSphere`: vSphere 인-트리 플러그인에서 vSphere CSI 플러그인으로 볼륨 작업을 + 라우팅하는 shim 및 변환 로직을 사용한다. + 노드에 vSphere CSI 플러그인이 설치 및 구성이 되어 있지 않은 경우 + 인-트리 vSphere 플러그인으로 폴백을 지원한다. CSIMigration 기능 플래그가 필요하다. +- `CSIMigrationvSphereComplete`: kubelet 및 볼륨 컨트롤러에서 vSphere 인-트리 + 플러그인 등록을 중지하고 shim 및 변환 로직을 활성화하여 vSphere 인-트리 플러그인에서 + vSphere CSI 플러그인으로 볼륨 작업을 라우팅할 수 있도록 한다. CSIMigration 및 + CSIMigrationvSphere 기능 플래그가 활성화되고 vSphere CSI 플러그인이 + 클러스터의 모든 노드에 설치 및 구성이 되어 있어야 한다. - `CSINodeInfo`: csi.storage.k8s.io에서 CSINodeInfo API 오브젝트와 관련된 모든 로직을 활성화한다. - `CSIPersistentVolume`: [CSI (Container Storage Interface)](https://github.com/kubernetes/community/blob/master/contributors/design-proposals/storage/container-storage-interface.md) 호환 볼륨 플러그인을 통해 프로비저닝된 볼륨을 감지하고 마운트할 수 있다. -- `CSIServiceAccountToken` : 볼륨을 마운트하는 파드의 서비스 계정 토큰을 받을 수 있도록 CSI 드라이버를 활성화한다. [토큰 요청](https://kubernetes-csi.github.io/docs/token-requests.html)을 참조한다. -- `CSIStorageCapacity`: CSI 드라이버가 스토리지 용량 정보를 게시하고 쿠버네티스 스케줄러가 파드를 스케줄할 때 해당 정보를 사용하도록 한다. [스토리지 용량](/docs/concepts/storage/storage-capacity/)을 참고한다. +- `CSIServiceAccountToken` : 볼륨을 마운트하는 파드의 서비스 계정 토큰을 받을 수 있도록 + CSI 드라이버를 활성화한다. + [토큰 요청](https://kubernetes-csi.github.io/docs/token-requests.html)을 참조한다. +- `CSIStorageCapacity`: CSI 드라이버가 스토리지 용량 정보를 게시하고 + 쿠버네티스 스케줄러가 파드를 스케줄할 때 해당 정보를 사용하도록 한다. + [스토리지 용량](/docs/concepts/storage/storage-capacity/)을 참고한다. 자세한 내용은 [`csi` 볼륨 유형](/ko/docs/concepts/storage/volumes/#csi) 문서를 확인한다. -- `CSIVolumeFSGroupPolicy`: CSI드라이버가 `fsGroupPolicy` 필드를 사용하도록 허용한다. 이 필드는 CSI드라이버에서 생성된 볼륨이 마운트될 때 볼륨 소유권과 권한 수정을 지원하는지 여부를 제어한다. -- `CustomCPUCFSQuotaPeriod`: 노드가 CPUCFSQuotaPeriod를 변경하도록 한다. +- `CSIVolumeFSGroupPolicy`: CSI드라이버가 `fsGroupPolicy` 필드를 사용하도록 허용한다. + 이 필드는 CSI드라이버에서 생성된 볼륨이 마운트될 때 볼륨 소유권과 + 권한 수정을 지원하는지 여부를 제어한다. +- `ConfigurableFSGroupPolicy`: 사용자가 파드에 볼륨을 마운트할 때 fsGroups에 대한 + 볼륨 권한 변경 정책을 구성할 수 있다. 자세한 내용은 + [파드의 볼륨 권한 및 소유권 변경 정책 구성](/docs/tasks/configure-pod-container/security-context/#configure-volume-permission-and-ownership-change-policy-for-pods)을 + 참고한다. +- `CronJobControllerV2`: {{< glossary_tooltip text="크론잡(CronJob)" term_id="cronjob" >}} + 컨트롤러의 대체 구현을 사용한다. 그렇지 않으면, + 동일한 컨트롤러의 버전 1이 선택된다. + 버전 2 컨트롤러는 실험적인 성능 향상을 제공한다. +- `CustomCPUCFSQuotaPeriod`: [kubelet config](/docs/tasks/administer-cluster/kubelet-config-file/)에서 + `cpuCFSQuotaPeriod` 를 노드가 변경할 수 있도록 한다. - `CustomPodDNS`: `dnsConfig` 속성을 사용하여 파드의 DNS 설정을 사용자 정의할 수 있다. 자세한 내용은 [파드의 DNS 설정](/ko/docs/concepts/services-networking/dns-pod-service/#pod-dns-config)을 확인한다. @@ -466,147 +546,248 @@ kubelet과 같은 컴포넌트의 기능 게이트를 설정하려면, 기능 - `CustomResourceWebhookConversion`: [커스텀리소스데피니션](/ko/docs/concepts/extend-kubernetes/api-extension/custom-resources/)에서 생성된 리소스에 대해 웹 훅 기반의 변환을 활성화한다. 실행 중인 파드 문제를 해결한다. -- `DisableAcceleratorUsageMetrics`: [kubelet이 수집한 액셀러레이터 지표 비활성화](/ko/docs/concepts/cluster-administration/system-metrics/#액셀러레이터-메트릭-비활성화). -- `DevicePlugins`: 노드에서 [장치 플러그인](/ko/docs/concepts/extend-kubernetes/compute-storage-net/device-plugins/) - 기반 리소스 프로비저닝을 활성화한다. - `DefaultPodTopologySpread`: `PodTopologySpread` 스케줄링 플러그인을 사용하여 [기본 분배](/ko/docs/concepts/workloads/pods/pod-topology-spread-constraints/#내부-기본-제약)를 수행한다. -- `DownwardAPIHugePages`: 다운워드 API에서 hugepages 사용을 활성화한다. +- `DevicePlugins`: 노드에서 [장치 플러그인](/ko/docs/concepts/extend-kubernetes/compute-storage-net/device-plugins/) + 기반 리소스 프로비저닝을 활성화한다. +- `DisableAcceleratorUsageMetrics`: + [kubelet이 수집한 액셀러레이터 지표 비활성화](/ko/docs/concepts/cluster-administration/system-metrics/#액셀러레이터-메트릭-비활성화). +- `DownwardAPIHugePages`: [다운워드 API](/docs/tasks/inject-data-application/downward-api-volume-expose-pod-information)에서 + hugepages 사용을 활성화한다. - `DryRun`: 서버 측의 [dry run](/docs/reference/using-api/api-concepts/#dry-run) 요청을 요청을 활성화하여 커밋하지 않고 유효성 검사, 병합 및 변화를 테스트할 수 있다. - `DynamicAuditing`(*사용 중단됨*): v1.19 이전의 버전에서 동적 감사를 활성화하는 데 사용된다. -- `DynamicKubeletConfig`: kubelet의 동적 구성을 활성화한다. [kubelet 재구성](/docs/tasks/administer-cluster/reconfigure-kubelet/)을 참고한다. -- `DynamicProvisioningScheduling`: 볼륨 스케줄을 인식하고 PV 프로비저닝을 처리하도록 기본 스케줄러를 확장한다. +- `DynamicKubeletConfig`: kubelet의 동적 구성을 활성화한다. + [kubelet 재구성](/docs/tasks/administer-cluster/reconfigure-kubelet/)을 참고한다. +- `DynamicProvisioningScheduling`: 볼륨 토폴로지를 인식하고 PV 프로비저닝을 처리하도록 + 기본 스케줄러를 확장한다. 이 기능은 v1.12의 `VolumeScheduling` 기능으로 대체되었다. -- `DynamicVolumeProvisioning`(*사용 중단됨*): 파드에 퍼시스턴트 볼륨의 [동적 프로비저닝](/ko/docs/concepts/storage/dynamic-provisioning/)을 활성화한다. -- `EnableAggregatedDiscoveryTimeout` (*사용 중단됨*): 수집된 검색 호출에서 5초 시간 초과를 활성화한다. -- `EnableEquivalenceClassCache`: 스케줄러가 파드를 스케줄링할 때 노드의 동등성을 캐시할 수 있게 한다. -- `EphemeralContainers`: 파드를 실행하기 위한 {{< glossary_tooltip text="임시 컨테이너" - term_id="ephemeral-container" >}}를 추가할 수 있다. -- `EvenPodsSpread`: 토폴로지 도메인 간에 파드를 균등하게 스케줄링할 수 있다. [파드 토폴로지 분배 제약 조건](/ko/docs/concepts/workloads/pods/pod-topology-spread-constraints/)을 참고한다. --`ExecProbeTimeout` : kubelet이 exec 프로브 시간 초과를 준수하는지 확인한다. 이 기능 게이트는 기존 워크로드가 쿠버네티스가 exec 프로브 제한 시간을 무시한 현재 수정된 결함에 의존하는 경우 존재한다. [준비성 프로브](/docs/tasks/configure-pod-container/configure-liveness-readiness-startup-probes/#configure-probes)를 참조한다. -- `ExpandInUsePersistentVolumes`: 사용 중인 PVC를 확장할 수 있다. [사용 중인 퍼시스턴트볼륨클레임 크기 조정](/ko/docs/concepts/storage/persistent-volumes/#사용-중인-퍼시스턴트볼륨클레임-크기-조정)을 참고한다. -- `ExpandPersistentVolumes`: 퍼시스턴트 볼륨 확장을 활성화한다. [퍼시스턴트 볼륨 클레임 확장](/ko/docs/concepts/storage/persistent-volumes/#퍼시스턴트-볼륨-클레임-확장)을 참고한다. -- `ExperimentalCriticalPodAnnotation`: 특정 파드에 *critical* 로 어노테이션을 달아서 [스케줄링이 보장되도록](/docs/tasks/administer-cluster/guaranteed-scheduling-critical-addon-pods/) 한다. +- `DynamicVolumeProvisioning`(*사용 중단됨*): 파드에 퍼시스턴트 볼륨의 + [동적 프로비저닝](/ko/docs/concepts/storage/dynamic-provisioning/)을 활성화한다. +- `EfficientWatchResumption`: 스토리지에서 생성된 북마크(진행 + 알림) 이벤트를 사용자에게 전달할 수 있다. 이것은 감시 작업에만 + 적용된다. +- `EnableAggregatedDiscoveryTimeout` (*사용 중단됨*): 수집된 검색 호출에서 5초 + 시간 초과를 활성화한다. +- `EnableEquivalenceClassCache`: 스케줄러가 파드를 스케줄링할 때 노드의 + 동등성을 캐시할 수 있게 한다. +- `EndpointSlice`: 보다 스케일링 가능하고 확장 가능한 네트워크 엔드포인트에 대한 + 엔드포인트슬라이스(EndpointSlices)를 활성화한다. [엔드포인트슬라이스 활성화](/docs/tasks/administer-cluster/enabling-endpointslices/)를 참고한다. +- `EndpointSliceNodeName` : 엔드포인트슬라이스 `nodeName` 필드를 활성화한다. +- `EndpointSliceProxying`: 활성화되면, 리눅스에서 실행되는 + kube-proxy는 엔드포인트 대신 엔드포인트슬라이스를 + 기본 데이터 소스로 사용하여 확장성과 성능을 향상시킨다. + [엔드포인트 슬라이스 활성화](/docs/tasks/administer-cluster/enabling-endpointslices/)를 참고한다. +- `EndpointSliceTerminatingCondition`: 엔드포인트슬라이스 `terminating` 및 `serving` + 조건 필드를 활성화한다. +- `EphemeralContainers`: 파드를 실행하기 위한 + {{< glossary_tooltip text="임시 컨테이너" term_id="ephemeral-container" >}}를 + 추가할 수 있다. +- `EvenPodsSpread`: 토폴로지 도메인 간에 파드를 균등하게 스케줄링할 수 있다. + [파드 토폴로지 분배 제약 조건](/ko/docs/concepts/workloads/pods/pod-topology-spread-constraints/)을 참고한다. +- `ExecProbeTimeout` : kubelet이 exec 프로브 시간 초과를 준수하는지 확인한다. + 이 기능 게이트는 기존 워크로드가 쿠버네티스가 exec 프로브 제한 시간을 무시한 + 현재 수정된 결함에 의존하는 경우 존재한다. + [준비성 프로브](/docs/tasks/configure-pod-container/configure-liveness-readiness-startup-probes/#configure-probes)를 참조한다. +- `ExpandCSIVolumes`: CSI 볼륨 확장을 활성화한다. +- `ExpandInUsePersistentVolumes`: 사용 중인 PVC를 확장할 수 있다. + [사용 중인 퍼시스턴트볼륨클레임 크기 조정](/ko/docs/concepts/storage/persistent-volumes/#사용-중인-퍼시스턴트볼륨클레임-크기-조정)을 참고한다. +- `ExpandPersistentVolumes`: 퍼시스턴트 볼륨 확장을 활성화한다. + [퍼시스턴트 볼륨 클레임 확장](/ko/docs/concepts/storage/persistent-volumes/#퍼시스턴트-볼륨-클레임-확장)을 참고한다. +- `ExperimentalCriticalPodAnnotation`: 특정 파드에 *critical* 로 + 어노테이션을 달아서 [스케줄링이 보장되도록](/docs/tasks/administer-cluster/guaranteed-scheduling-critical-addon-pods/) 한다. 이 기능은 v1.13부터 파드 우선 순위 및 선점으로 인해 사용 중단되었다. - `ExperimentalHostUserNamespaceDefaultingGate`: 사용자 네임스페이스를 호스트로 기본 활성화한다. 이것은 다른 호스트 네임스페이스, 호스트 마운트, 권한이 있는 컨테이너 또는 특정 비-네임스페이스(non-namespaced) 기능(예: `MKNODE`, `SYS_MODULE` 등)을 사용하는 컨테이너를 위한 것이다. 도커 데몬에서 사용자 네임스페이스 재 매핑이 활성화된 경우에만 활성화해야 한다. -- `EndpointSlice`: 보다 스케일링 가능하고 확장 가능한 네트워크 엔드포인트에 대한 - 엔드포인트 슬라이스를 활성화한다. [엔드포인트 슬라이스 활성화](/docs/tasks/administer-cluster/enabling-endpointslices/)를 참고한다. --`EndpointSliceNodeName` : 엔드포인트슬라이스 `nodeName` 필드를 활성화한다. --`EndpointSliceTerminating` : 엔드포인트슬라이스 `terminating` 및 `serving` 조건 필드를 - 활성화한다. -- `EndpointSliceProxying`: 이 기능 게이트가 활성화되면, 리눅스에서 실행되는 - kube-proxy는 엔드포인트 대신 엔드포인트슬라이스를 - 기본 데이터 소스로 사용하여 확장성과 성능을 향상시킨다. - [엔드포인트 슬라이스 활성화](/docs/tasks/administer-cluster/enabling-endpointslices/)를 참고한다. -- `WindowsEndpointSliceProxying`: 이 기능 게이트가 활성화되면, 윈도우에서 실행되는 - kube-proxy는 엔드포인트 대신 엔드포인트슬라이스를 - 기본 데이터 소스로 사용하여 확장성과 성능을 향상시킨다. - [엔드포인트 슬라이스 활성화](/docs/tasks/administer-cluster/enabling-endpointslices/)를 참고한다. - `GCERegionalPersistentDisk`: GCE에서 지역 PD 기능을 활성화한다. -- `GenericEphemeralVolume`: 일반 볼륨의 모든 기능을 지원하는 임시, 인라인 볼륨을 활성화한다(타사 스토리지 공급 업체, 스토리지 용량 추적, 스냅샷으로부터 복원 등에서 제공할 수 있음). [임시 볼륨](/docs/concepts/storage/ephemeral-volumes/)을 참고한다. --`GracefulNodeShutdown` : kubelet에서 정상 종료를 지원한다. 시스템 종료 중에 kubelet은 종료 이벤트를 감지하고 노드에서 실행중인 파드를 정상적으로 종료하려고 시도한다. 자세한 내용은 [Graceful Node Shutdown](/ko/docs/concepts/architecture/nodes/#그레이스풀-graceful-노드-셧다운)을 참조한다. -- `HugePages`: 사전 할당된 [huge page](/ko/docs/tasks/manage-hugepages/scheduling-hugepages/)의 할당 및 사용을 활성화한다. -- `HugePageStorageMediumSize`: 사전 할당된 [huge page](/ko/docs/tasks/manage-hugepages/scheduling-hugepages/)의 여러 크기를 지원한다. -- `HyperVContainer`: 윈도우 컨테이너를 위한 [Hyper-V 격리](https://docs.microsoft.com/ko-kr/virtualization/windowscontainers/manage-containers/hyperv-container) 기능을 활성화한다. -- `HPAScaleToZero`: 사용자 정의 또는 외부 메트릭을 사용할 때 `HorizontalPodAutoscaler` 리소스에 대해 `minReplicas` 를 0으로 설정한다. -- `ImmutableEphemeralVolumes`: 안정성과 성능 향상을 위해 개별 시크릿(Secret)과 컨피그맵(ConfigMap)을 변경할 수 없는(immutable) 것으로 표시할 수 있다. -- `KubeletConfigFile`: 구성 파일을 사용하여 지정된 파일에서 kubelet 구성을 로드할 수 있다. - 자세한 내용은 [구성 파일을 통해 kubelet 파라미터 설정](/docs/tasks/administer-cluster/kubelet-config-file/)을 참고한다. +- `GenericEphemeralVolume`: 일반 볼륨의 모든 기능을 지원하는 임시, 인라인 + 볼륨을 활성화한다(타사 스토리지 공급 업체, 스토리지 용량 추적, 스냅샷으로부터 복원 + 등에서 제공할 수 있음). + [임시 볼륨](/docs/concepts/storage/ephemeral-volumes/)을 참고한다. +- `GracefulNodeShutdown` : kubelet에서 정상 종료를 지원한다. + 시스템 종료 중에 kubelet은 종료 이벤트를 감지하고 노드에서 실행 중인 + 파드를 정상적으로 종료하려고 시도한다. 자세한 내용은 + [Graceful Node Shutdown](/ko/docs/concepts/architecture/nodes/#그레이스풀-graceful-노드-셧다운)을 + 참조한다. +- `HPAContainerMetrics`: `HorizontalPodAutoscaler`를 활성화하여 대상 파드의 + 개별 컨테이너 메트릭을 기반으로 확장한다. +- `HPAScaleToZero`: 사용자 정의 또는 외부 메트릭을 사용할 때 `HorizontalPodAutoscaler` 리소스에 대해 + `minReplicas` 를 0으로 설정한다. +- `HugePages`: 사전 할당된 [huge page](/ko/docs/tasks/manage-hugepages/scheduling-hugepages/)의 + 할당 및 사용을 활성화한다. +- `HugePageStorageMediumSize`: 사전 할당된 [huge page](/ko/docs/tasks/manage-hugepages/scheduling-hugepages/)의 + 여러 크기를 지원한다. +- `HyperVContainer`: 윈도우 컨테이너를 위한 + [Hyper-V 격리](https://docs.microsoft.com/ko-kr/virtualization/windowscontainers/manage-containers/hyperv-container) + 기능을 활성화한다. +- `IPv6DualStack`: IPv6에 대한 [듀얼 스택](/ko/docs/concepts/services-networking/dual-stack/) + 지원을 활성화한다. +- `ImmutableEphemeralVolumes`: 안정성과 성능 향상을 위해 개별 시크릿(Secret)과 컨피그맵(ConfigMap)을 + 변경할 수 없는(immutable) 것으로 표시할 수 있다. +- `KubeletConfigFile`: 구성 파일을 사용하여 지정된 파일에서 + kubelet 구성을 로드할 수 있다. + 자세한 내용은 [구성 파일을 통해 kubelet 파라미터 설정](/docs/tasks/administer-cluster/kubelet-config-file/)을 + 참고한다. - `KubeletCredentialProviders`: 이미지 풀 자격 증명에 대해 kubelet exec 자격 증명 공급자를 활성화한다. - `KubeletPluginsWatcher`: kubelet이 [CSI 볼륨 드라이버](/ko/docs/concepts/storage/volumes/#csi)와 같은 플러그인을 검색할 수 있도록 프로브 기반 플러그인 감시자(watcher) 유틸리티를 사용한다. -- `KubeletPodResources`: kubelet의 파드 리소스 grpc 엔드포인트를 활성화한다. - 자세한 내용은 [장치 모니터링 지원](https://github.com/kubernetes/enhancements/blob/master/keps/sig-node/compute-device-assignment.md)을 참고한다. -- `LegacyNodeRoleBehavior`: 비활성화되면, 서비스 로드 밸런서 및 노드 중단의 레거시 동작은 `NodeDisruptionExclusion` 과 `ServiceNodeExclusion` 에 의해 제공된 기능별 레이블을 대신하여 `node-role.kubernetes.io/master` 레이블을 무시한다. -- `LocalStorageCapacityIsolation`: [로컬 임시 스토리지](/ko/docs/concepts/configuration/manage-resources-containers/)와 [emptyDir 볼륨](/ko/docs/concepts/storage/volumes/#emptydir)의 `sizeLimit` 속성을 사용할 수 있게 한다. -- `LocalStorageCapacityIsolationFSQuotaMonitoring`: [로컬 임시 스토리지](/ko/docs/concepts/configuration/manage-resources-containers/)에 `LocalStorageCapacityIsolation` 이 활성화되고 [emptyDir 볼륨](/ko/docs/concepts/storage/volumes/#emptydir)의 백업 파일시스템이 프로젝트 쿼터를 지원하고 활성화된 경우, 파일시스템 사용보다는 프로젝트 쿼터를 사용하여 [emptyDir 볼륨](/ko/docs/concepts/storage/volumes/#emptydir) 스토리지 사용을 모니터링하여 성능과 정확성을 향상시킨다. -- `MixedProtocolLBService`: 동일한 로드밸런서 유형 서비스 인스턴스에서 다른 프로토콜 사용을 활성화한다. -- `MountContainers`: 호스트의 유틸리티 컨테이너를 볼륨 마운터로 사용할 수 있다. +- `KubeletPodResources`: kubelet의 파드 리소스 GPRC 엔드포인트를 활성화한다. 자세한 내용은 + [장치 모니터링 지원](https://github.com/kubernetes/enhancements/blob/master/keps/sig-node/compute-device-assignment.md)을 + 참고한다. +- `LegacyNodeRoleBehavior`: 비활성화되면, 서비스 로드 밸런서 및 노드 중단의 레거시 동작은 + `NodeDisruptionExclusion` 과 `ServiceNodeExclusion` 에 의해 제공된 기능별 레이블을 대신하여 + `node-role.kubernetes.io/master` 레이블을 무시한다. +- `LocalStorageCapacityIsolation`: [로컬 임시 스토리지](/ko/docs/concepts/configuration/manage-resources-containers/)와 + [emptyDir 볼륨](/ko/docs/concepts/storage/volumes/#emptydir)의 + `sizeLimit` 속성을 사용할 수 있게 한다. +- `LocalStorageCapacityIsolationFSQuotaMonitoring`: [로컬 임시 스토리지](/ko/docs/concepts/configuration/manage-resources-containers/)에 + `LocalStorageCapacityIsolation` 이 활성화되고 + [emptyDir 볼륨](/ko/docs/concepts/storage/volumes/#emptydir)의 + 백업 파일시스템이 프로젝트 쿼터를 지원하고 활성화된 경우, 파일시스템 사용보다는 + 프로젝트 쿼터를 사용하여 [emptyDir 볼륨](/ko/docs/concepts/storage/volumes/#emptydir) + 스토리지 사용을 모니터링하여 성능과 정확성을 + 향상시킨다. +- `MixedProtocolLBService`: 동일한 로드밸런서 유형 서비스 인스턴스에서 다른 프로토콜 + 사용을 활성화한다. +- `MountContainers` (*사용 중단됨*): 호스트의 유틸리티 컨테이너를 볼륨 마운터로 + 사용할 수 있다. - `MountPropagation`: 한 컨테이너에서 다른 컨테이너 또는 파드로 마운트된 볼륨을 공유할 수 있다. 자세한 내용은 [마운트 전파(propagation)](/ko/docs/concepts/storage/volumes/#마운트-전파-propagation)을 참고한다. -- `NodeDisruptionExclusion`: 영역(zone) 장애 시 노드가 제외되지 않도록 노드 레이블 `node.kubernetes.io/exclude-disruption` 사용을 활성화한다. +- `NodeDisruptionExclusion`: 영역(zone) 장애 시 노드가 제외되지 않도록 노드 레이블 `node.kubernetes.io/exclude-disruption` + 사용을 활성화한다. - `NodeLease`: 새로운 리스(Lease) API가 노드 상태 신호로 사용될 수 있는 노드 하트비트(heartbeats)를 보고할 수 있게 한다. -- `NonPreemptingPriority`: 프라이어리티클래스(PriorityClass)와 파드에 NonPreempting 옵션을 활성화한다. +- `NonPreemptingPriority`: 프라이어리티클래스(PriorityClass)와 파드에 `preemptionPolicy` 필드를 활성화한다. +- `PVCProtection`: 파드에서 사용 중일 때 퍼시스턴트볼륨클레임(PVC)이 + 삭제되지 않도록 한다. - `PersistentLocalVolumes`: 파드에서 `local` 볼륨 유형의 사용을 활성화한다. `local` 볼륨을 요청하는 경우 파드 어피니티를 지정해야 한다. - `PodDisruptionBudget`: [PodDisruptionBudget](/docs/tasks/run-application/configure-pdb/) 기능을 활성화한다. -- `PodOverhead`: 파드 오버헤드를 판단하기 위해 [파드오버헤드(PodOverhead)](/ko/docs/concepts/scheduling-eviction/pod-overhead/) 기능을 활성화한다. -- `PodPriority`: [우선 순위](/ko/docs/concepts/configuration/pod-priority-preemption/)를 기반으로 파드의 스케줄링 취소와 선점을 활성화한다. +- `PodOverhead`: 파드 오버헤드를 판단하기 위해 [파드오버헤드(PodOverhead)](/ko/docs/concepts/scheduling-eviction/pod-overhead/) + 기능을 활성화한다. +- `PodPriority`: [우선 순위](/ko/docs/concepts/configuration/pod-priority-preemption/)를 + 기반으로 파드의 스케줄링 취소와 선점을 활성화한다. - `PodReadinessGates`: 파드 준비성 평가를 확장하기 위해 `PodReadinessGate` 필드 설정을 활성화한다. 자세한 내용은 [파드의 준비성 게이트](/ko/docs/concepts/workloads/pods/pod-lifecycle/#pod-readiness-gate)를 참고한다. - `PodShareProcessNamespace`: 파드에서 실행되는 컨테이너 간에 단일 프로세스 네임스페이스를 공유하기 위해 파드에서 `shareProcessNamespace` 설정을 활성화한다. 자세한 내용은 [파드의 컨테이너 간 프로세스 네임스페이스 공유](/docs/tasks/configure-pod-container/share-process-namespace/)에서 확인할 수 있다. -- `ProcMountType`: 컨테이너의 ProcMountType 제어를 활성화한다. -- `PVCProtection`: 파드에서 사용 중일 때 퍼시스턴트볼륨클레임(PVC)이 - 삭제되지 않도록 한다. -- `QOSReserved`: QoS 수준에서 리소스 예약을 허용하여 낮은 QoS 수준의 파드가 더 높은 QoS 수준에서 - 요청된 리소스로 파열되는 것을 방지한다(현재 메모리만 해당). +- `ProcMountType`: SecurityContext의 `procMount` 필드를 설정하여 + 컨테이너의 proc 타입의 마운트를 제어할 수 있다. +- `QOSReserved`: QoS 수준에서 리소스 예약을 허용하여 낮은 QoS 수준의 파드가 + 더 높은 QoS 수준에서 요청된 리소스로 파열되는 것을 방지한다 + (현재 메모리만 해당). +- `RemainingItemCount`: API 서버가 + [청크(chunking) 목록 요청](/docs/reference/using-api/api-concepts/#retrieving-large-results-sets-in-chunks)에 대한 + 응답에서 남은 항목 수를 표시하도록 허용한다. +- `RemoveSelfLink`: ObjectMeta 및 ListMeta에서 `selfLink` 를 사용하지 않고 + 제거한다. - `ResourceLimitsPriorityFunction` (*사용 중단됨*): 입력 파드의 CPU 및 메모리 한도 중 하나 이상을 만족하는 노드에 가능한 최저 점수 1을 할당하는 스케줄러 우선 순위 기능을 활성화한다. 의도는 동일한 점수를 가진 노드 사이의 관계를 끊는 것이다. - `ResourceQuotaScopeSelectors`: 리소스 쿼터 범위 셀렉터를 활성화한다. -- `RootCAConfigMap`: 모든 네임 스페이스에 `kube-root-ca.crt`라는 {{< glossary_tooltip text="컨피그맵" term_id="configmap" >}}을 게시하도록 kube-controller-manager를 구성한다. 이 컨피그맵에는 kube-apiserver에 대한 연결을 확인하는 데 사용되는 CA 번들이 포함되어 있다. - 자세한 내용은 [바운드 서비스 계정 토큰](https://github.com/kubernetes/enhancements/blob/master/keps/sig-auth/1205-bound-service-account-tokens/README.md)을 참조한다. +- `RootCAConfigMap`: 모든 네임스페이스에 `kube-root-ca.crt`라는 + {{< glossary_tooltip text="컨피그맵" term_id="configmap" >}}을 게시하도록 + `kube-controller-manager` 를 구성한다. 이 컨피그맵에는 kube-apiserver에 대한 연결을 확인하는 데 + 사용되는 CA 번들이 포함되어 있다. 자세한 내용은 + [바운드 서비스 계정 토큰](https://github.com/kubernetes/enhancements/blob/master/keps/sig-auth/1205-bound-service-account-tokens/README.md)을 + 참조한다. - `RotateKubeletClientCertificate`: kubelet에서 클라이언트 TLS 인증서의 로테이션을 활성화한다. 자세한 내용은 [kubelet 구성](/docs/reference/command-line-tools-reference/kubelet-tls-bootstrapping/#kubelet-configuration)을 참고한다. - `RotateKubeletServerCertificate`: kubelet에서 서버 TLS 인증서의 로테이션을 활성화한다. - 자세한 내용은 [kubelet 구성](/docs/reference/command-line-tools-reference/kubelet-tls-bootstrapping/#kubelet-configuration)을 참고한다. -- `RunAsGroup`: 컨테이너의 init 프로세스에 설정된 기본 그룹 ID 제어를 활성화한다. -- `RuntimeClass`: 컨테이너 런타임 구성을 선택하기 위해 [런타임클래스(RuntimeClass)](/ko/docs/concepts/containers/runtime-class/) 기능을 활성화한다. -- `ScheduleDaemonSetPods`: 데몬셋(DaemonSet) 컨트롤러 대신 기본 스케줄러로 데몬셋 파드를 스케줄링할 수 있다. -- `SCTPSupport`: 파드, 서비스, 엔드포인트, 엔드포인트슬라이스 및 네트워크폴리시 정의에서 _SCTP_ `protocol` 값을 활성화한다. -- `ServerSideApply`: API 서버에서 [SSA(Sever Side Apply)](/docs/reference/using-api/server-side-apply/) 경로를 활성화한다. -- `ServiceAccountIssuerDiscovery`: API 서버에서 서비스 어카운트 발행자에 대해 OIDC 디스커버리 엔드포인트(발급자 및 JWKS URL)를 활성화한다. 자세한 내용은 [파드의 서비스 어카운트 구성](/docs/tasks/configure-pod-container/configure-service-account/#service-account-issuer-discovery)을 참고한다. +- `RunAsGroup`: 컨테이너의 init 프로세스에 설정된 기본 그룹 ID 제어를 + 활성화한다. +- `RuntimeClass`: 컨테이너 런타임 구성을 선택하기 위해 [런타임클래스(RuntimeClass)](/ko/docs/concepts/containers/runtime-class/) + 기능을 활성화한다. +- `ScheduleDaemonSetPods`: 데몬셋(DaemonSet) 컨트롤러 대신 기본 스케줄러로 데몬셋 파드를 + 스케줄링할 수 있다. +- `SCTPSupport`: 파드, 서비스, 엔드포인트, 엔드포인트슬라이스 및 네트워크폴리시 정의에서 + _SCTP_ `protocol` 값을 활성화한다. +- `ServerSideApply`: API 서버에서 [SSA(Sever Side Apply)](/docs/reference/using-api/server-side-apply/) + 경로를 활성화한다. +- `ServiceAccountIssuerDiscovery`: API 서버에서 서비스 어카운트 발행자에 대해 OIDC 디스커버리 엔드포인트(발급자 및 + JWKS URL)를 활성화한다. 자세한 내용은 + [파드의 서비스 어카운트 구성](/docs/tasks/configure-pod-container/configure-service-account/#service-account-issuer-discovery)을 + 참고한다. - `ServiceAppProtocol`: 서비스와 엔드포인트에서 `AppProtocol` 필드를 활성화한다. -- `ServiceLBNodePortControl`: 서비스에서`spec.allocateLoadBalancerNodePorts` 필드를 활성화한다. +- `ServiceLBNodePortControl`: 서비스에서`spec.allocateLoadBalancerNodePorts` 필드를 + 활성화한다. - `ServiceLoadBalancerFinalizer`: 서비스 로드 밸런서에 대한 Finalizer 보호를 활성화한다. -- `ServiceNodeExclusion`: 클라우드 제공자가 생성한 로드 밸런서에서 노드를 제외할 수 있다. - "`alpha.service-controller.kubernetes.io/exclude-balancer`" 키 또는 `node.kubernetes.io/exclude-from-external-load-balancers` 로 레이블이 지정된 경우 노드를 제외할 수 있다. -- `ServiceTopology`: 서비스가 클러스터의 노드 토폴로지를 기반으로 트래픽을 라우팅할 수 있도록 한다. 자세한 내용은 [서비스토폴로지(ServiceTopology)](/ko/docs/concepts/services-networking/service-topology/)를 참고한다. -- `SizeMemoryBackedVolumes`: kubelet 지원을 사용하여 메모리 백업 볼륨의 크기를 조정한다. 자세한 내용은 [volumes](/ko/docs/concepts/storage/volumes)를 참조한다. -- `SetHostnameAsFQDN`: 전체 주소 도메인 이름(FQDN)을 파드의 호스트 이름으로 설정하는 기능을 활성화한다. [파드의 `setHostnameAsFQDN` 필드](/ko/docs/concepts/services-networking/dns-pod-service/#파드의-sethostnameasfqdn-필드)를 참고한다. -- `StartupProbe`: kubelet에서 [스타트업](/ko/docs/concepts/workloads/pods/pod-lifecycle/#언제-스타트업-프로브를-사용해야-하는가) 프로브를 활성화한다. +- `ServiceNodeExclusion`: 클라우드 제공자가 생성한 로드 밸런서에서 노드를 + 제외할 수 있다. "`node.kubernetes.io/exclude-from-external-load-balancers`"로 + 레이블이 지정된 경우 노드를 제외할 수 있다. +- `ServiceTopology`: 서비스가 클러스터의 노드 토폴로지를 기반으로 트래픽을 라우팅할 수 + 있도록 한다. 자세한 내용은 + [서비스토폴로지(ServiceTopology)](/ko/docs/concepts/services-networking/service-topology/)를 + 참고한다. +- `SizeMemoryBackedVolumes`: kubelet 지원을 사용하여 메모리 백업 볼륨의 크기를 조정한다. + 자세한 내용은 [volumes](/ko/docs/concepts/storage/volumes)를 참조한다. +- `SetHostnameAsFQDN`: 전체 주소 도메인 이름(FQDN)을 파드의 호스트 이름으로 + 설정하는 기능을 활성화한다. + [파드의 `setHostnameAsFQDN` 필드](/ko/docs/concepts/services-networking/dns-pod-service/#파드의-sethostnameasfqdn-필드)를 참고한다. +- `StartupProbe`: kubelet에서 + [스타트업](/ko/docs/concepts/workloads/pods/pod-lifecycle/#언제-스타트업-프로브를-사용해야-하는가) + 프로브를 활성화한다. - `StorageObjectInUseProtection`: 퍼시스턴트볼륨 또는 퍼시스턴트볼륨클레임 오브젝트가 여전히 사용 중인 경우 삭제를 연기한다. -- `StorageVersionHash`: API 서버가 디스커버리에서 스토리지 버전 해시를 노출하도록 허용한다. +- `StorageVersionAPI`: [스토리지 버전 API](/docs/reference/generated/kubernetes-api/{{< param "version" >}}/#storageversion-v1alpha1-internal-apiserver-k8s-io)를 + 활성화한다. +- `StorageVersionHash`: API 서버가 디스커버리에서 스토리지 버전 해시를 노출하도록 + 허용한다. - `StreamingProxyRedirects`: 스트리밍 요청을 위해 백엔드(kubelet)에서 리디렉션을 가로채서 따르도록 API 서버에 지시한다. 스트리밍 요청의 예로는 `exec`, `attach` 및 `port-forward` 요청이 있다. - `SupportIPVSProxyMode`: IPVS를 사용하여 클러스터 내 서비스 로드 밸런싱을 제공한다. 자세한 내용은 [서비스 프록시](/ko/docs/concepts/services-networking/service/#가상-ip와-서비스-프록시)를 참고한다. - `SupportPodPidsLimit`: 파드의 PID 제한을 지원한다. -- `SupportNodePidsLimit`: 노드에서 PID 제한 지원을 활성화한다. `--system-reserved` 및 `--kube-reserved` 옵션의 `pid=` 매개 변수를 지정하여 지정된 수의 프로세스 ID가 시스템 전체와 각각 쿠버네티스 시스템 데몬에 대해 예약되도록 할 수 있다. -- `Sysctls`: 각 파드에 설정할 수 있는 네임스페이스 커널 파라미터(sysctl)를 지원한다. - 자세한 내용은 [sysctl](/docs/tasks/administer-cluster/sysctl-cluster/)을 참고한다. -- `TaintBasedEvictions`: 노드의 테인트(taint) 및 파드의 톨러레이션(toleration)을 기반으로 노드에서 파드를 축출할 수 있다. - 자세한 내용은 [테인트와 톨러레이션](/ko/docs/concepts/scheduling-eviction/taint-and-toleration/)을 참고한다. -- `TaintNodesByCondition`: [노드 컨디션](/ko/docs/concepts/architecture/nodes/#condition)을 기반으로 자동 테인트 노드를 활성화한다. +- `SupportNodePidsLimit`: 노드에서 PID 제한 지원을 활성화한다. + `--system-reserved` 및 `--kube-reserved` 옵션의 `pid=` + 파라미터를 지정하여 지정된 수의 프로세스 ID가 + 시스템 전체와 각각 쿠버네티스 시스템 데몬에 대해 예약되도록 + 할 수 있다. +- `Sysctls`: 각 파드에 설정할 수 있는 네임스페이스 커널 + 파라미터(sysctl)를 지원한다. 자세한 내용은 + [sysctl](/docs/tasks/administer-cluster/sysctl-cluster/)을 참고한다. +- `TTLAfterFinished`: [TTL 컨트롤러](/ko/docs/concepts/workloads/controllers/ttlafterfinished/)가 + 실행이 끝난 후 리소스를 정리하도록 + 허용한다. +- `TaintBasedEvictions`: 노드의 테인트(taint) 및 파드의 톨러레이션(toleration)을 기반으로 + 노드에서 파드를 축출할 수 있다. + 자세한 내용은 [테인트와 톨러레이션](/ko/docs/concepts/scheduling-eviction/taint-and-toleration/)을 + 참고한다. +- `TaintNodesByCondition`: [노드 컨디션](/ko/docs/concepts/architecture/nodes/#condition)을 + 기반으로 자동 테인트 노드를 활성화한다. - `TokenRequest`: 서비스 어카운트 리소스에서 `TokenRequest` 엔드포인트를 활성화한다. -- `TokenRequestProjection`: [`projected` 볼륨](/ko/docs/concepts/storage/volumes/#projected)을 통해 서비스 어카운트 - 토큰을 파드에 주입할 수 있다. -- `TopologyManager`: 쿠버네티스의 다른 컴포넌트에 대한 세분화된 하드웨어 리소스 할당을 조정하는 메커니즘을 활성화한다. [노드의 토폴로지 관리 정책 제어](/docs/tasks/administer-cluster/topology-manager/)를 참고한다. -- `TTLAfterFinished`: [TTL 컨트롤러](/ko/docs/concepts/workloads/controllers/ttlafterfinished/)가 실행이 끝난 후 리소스를 정리하도록 허용한다. +- `TokenRequestProjection`: [`projected` 볼륨](/ko/docs/concepts/storage/volumes/#projected)을 통해 + 서비스 어카운트 토큰을 파드에 주입할 수 있다. +- `TopologyManager`: 쿠버네티스의 다른 컴포넌트에 대한 세분화된 하드웨어 리소스 + 할당을 조정하는 메커니즘을 활성화한다. + [노드의 토폴로지 관리 정책 제어](/docs/tasks/administer-cluster/topology-manager/)를 참고한다. - `VolumePVCDataSource`: 기존 PVC를 데이터 소스로 지정하는 기능을 지원한다. - `VolumeScheduling`: 볼륨 토폴로지 인식 스케줄링을 활성화하고 퍼시스턴트볼륨클레임(PVC) 바인딩이 스케줄링 결정을 인식하도록 한다. 또한 `PersistentLocalVolumes` 기능 게이트와 함께 사용될 때 [`local`](/ko/docs/concepts/storage/volumes/#local) 볼륨 유형을 사용할 수 있다. - `VolumeSnapshotDataSource`: 볼륨 스냅샷 데이터 소스 지원을 활성화한다. -- `VolumeSubpathEnvExpansion`: 환경 변수를 `subPath`로 확장하기 위해 `subPathExpr` 필드를 활성화한다. +- `VolumeSubpathEnvExpansion`: 환경 변수를 `subPath`로 확장하기 위해 + `subPathExpr` 필드를 활성화한다. +- `WarningHeaders`: API 응답에서 경고 헤더를 보낼 수 있다. - `WatchBookmark`: 감시자 북마크(watch bookmark) 이벤트 지원을 활성화한다. -- `WindowsGMSA`: 파드에서 컨테이너 런타임으로 GMSA 자격 증명 스펙을 전달할 수 있다. -- `WindowsRunAsUserName` : 기본 사용자가 아닌(non-default) 사용자로 윈도우 컨테이너에서 애플리케이션을 실행할 수 있도록 지원한다. - 자세한 내용은 [RunAsUserName 구성](/docs/tasks/configure-pod-container/configure-runasusername)을 참고한다. - `WinDSR`: kube-proxy가 윈도우용 DSR 로드 밸런서를 생성할 수 있다. - `WinOverlay`: kube-proxy가 윈도우용 오버레이 모드에서 실행될 수 있도록 한다. +- `WindowsGMSA`: 파드에서 컨테이너 런타임으로 GMSA 자격 증명 스펙을 전달할 수 있다. +- `WindowsRunAsUserName` : 기본 사용자가 아닌(non-default) 사용자로 윈도우 컨테이너에서 + 애플리케이션을 실행할 수 있도록 지원한다. 자세한 내용은 + [RunAsUserName 구성](/docs/tasks/configure-pod-container/configure-runasusername)을 + 참고한다. +- `WindowsEndpointSliceProxying`: 활성화되면, 윈도우에서 실행되는 kube-proxy는 + 엔드포인트 대신 엔드포인트슬라이스를 기본 데이터 소스로 사용하여 + 확장성과 성능을 향상시킨다. + [엔드포인트 슬라이스 활성화하기](/docs/tasks/administer-cluster/enabling-endpointslices/)를 참고한다. ## {{% heading "whatsnext" %}} diff --git a/content/ko/docs/reference/glossary/api-group.md b/content/ko/docs/reference/glossary/api-group.md index 0c27d3181e..96f32bd9ce 100644 --- a/content/ko/docs/reference/glossary/api-group.md +++ b/content/ko/docs/reference/glossary/api-group.md @@ -2,7 +2,7 @@ title: API 그룹(API Group) id: api-group date: 2019-09-02 -full_link: /ko/docs/concepts/overview/kubernetes-api/#api-groups +full_link: /ko/docs/concepts/overview/kubernetes-api/#api-그룹과-버전-규칙 short_description: > 쿠버네티스 API의 연관된 경로들의 집합. @@ -11,9 +11,9 @@ tags: - fundamental - architecture --- -쿠버네티스 API의 연관된 경로들의 집합. +쿠버네티스 API의 연관된 경로들의 집합. API 서버의 구성을 변경하여 각 API 그룹을 활성화하거나 비활성화할 수 있다. 특정 리소스에 대한 경로를 비활성화하거나 활성화할 수도 있다. API 그룹을 사용하면 쿠버네티스 API를 더 쉽게 확장할 수 있다. API 그룹은 REST 경로 및 직렬화된 오브젝트의 `apiVersion` 필드에 지정된다. -* 자세한 내용은 [API 그룹(/ko/docs/concepts/overview/kubernetes-api/#api-groups)을 참조한다. +* 자세한 내용은 [API 그룹(/ko/docs/concepts/overview/kubernetes-api/#api-그룹과-버전-규칙)을 참조한다. diff --git a/content/ko/docs/reference/glossary/quantity.md b/content/ko/docs/reference/glossary/quantity.md new file mode 100644 index 0000000000..450307841a --- /dev/null +++ b/content/ko/docs/reference/glossary/quantity.md @@ -0,0 +1,33 @@ +--- +title: 수량(Quantity) +id: quantity +date: 2018-08-07 +full_link: +short_description: > + SI 접미사를 사용하는 작거나 큰 숫자의 정수(whole-number) 표현. + +aka: +tags: +- core-object +--- + SI 접미사를 사용하는 작거나 큰 숫자의 정수(whole-number) 표현. + + + +수량은 SI 접미사가 포함된 간결한 정수 표기법을 통해서 작거나 큰 숫자를 표현한 것이다. +분수는 밀리(milli) 단위로 표시되는 반면, +큰 숫자는 킬로(kilo), 메가(mega), 또는 기가(giga) +단위로 표시할 수 있다. + + +예를 들어, 숫자 `1.5`는 `1500m`으로, 숫자 `1000`은 `1k`로, `1000000`은 +`1M`으로 표시할 수 있다. 또한, 이진 표기법 접미사도 명시 가능하므로, +숫자 2048은 `2Ki`로 표기될 수 있다. + +허용되는 10진수(10의 거듭 제곱) 단위는 `m` (밀리), `k` (킬로, 의도적인 소문자), +`M` (메가), `G` (기가), `T` (테라), `P` (페타), +`E` (엑사)가 있다. + +허용되는 2진수(2의 거듭 제곱) 단위는 `Ki` (키비), `Mi` (메비), `Gi` (기비), +`Ti` (테비), `Pi` (페비), `Ei` (엑비)가 있다. + diff --git a/content/ko/docs/reference/glossary/secret.md b/content/ko/docs/reference/glossary/secret.md new file mode 100644 index 0000000000..63637adc1a --- /dev/null +++ b/content/ko/docs/reference/glossary/secret.md @@ -0,0 +1,18 @@ +--- +title: 시크릿(Secret) +id: secret +date: 2018-04-12 +full_link: /ko/docs/concepts/configuration/secret/ +short_description: > + 비밀번호, OAuth 토큰 및 ssh 키와 같은 민감한 정보를 저장한다. + +aka: +tags: +- core-object +- security +--- + 비밀번호, OAuth 토큰 및 ssh 키와 같은 민감한 정보를 저장한다. + + + +민감한 정보를 사용하는 방식에 대해 더 세밀하게 제어할 수 있으며, 유휴 상태의 [암호화](/docs/tasks/administer-cluster/encrypt-data/#ensure-all-secrets-are-encrypted)를 포함하여 우발적인 노출 위험을 줄인다. {{< glossary_tooltip text="파드(Pod)" term_id="pod" >}}는 시크릿을 마운트된 볼륨의 파일로 참조하거나, 파드의 이미지를 풀링하는 kubelet이 시크릿을 참조한다. 시크릿은 기밀 데이터에 적합하고 [컨피그맵](/docs/tasks/configure-pod-container/configure-pod-configmap/)은 기밀이 아닌 데이터에 적합하다. diff --git a/content/ko/docs/reference/glossary/storage-class.md b/content/ko/docs/reference/glossary/storage-class.md new file mode 100644 index 0000000000..63bd655b68 --- /dev/null +++ b/content/ko/docs/reference/glossary/storage-class.md @@ -0,0 +1,20 @@ +--- +title: 스토리지 클래스(Storage Class) +id: storageclass +date: 2018-04-12 +full_link: /ko/docs/concepts/storage/storage-classes +short_description: > + 스토리지클래스는 관리자가 사용 가능한 다양한 스토리지 유형을 설명할 수 있는 방법을 제공한다. + +aka: +tags: +- core-object +- storage +--- + 스토리지클래스는 관리자가 사용 가능한 다양한 스토리지 유형을 설명할 수 있는 방법을 제공한다. + + + +스토리지 클래스는 서비스 품질 수준, 백업 정책 혹은 클러스터 관리자가 결정한 임의의 정책에 매핑할 수 있다. 각 스토리지클래스에는 클래스에 속한 {{< glossary_tooltip text="퍼시스턴트 볼륨(Persistent Volume)" term_id="persistent-volume" >}}을 동적으로 프로비저닝해야 할 때 사용되는 `provisioner`, `parameters` 및 `reclaimPolicy` 필드가 있다. 사용자는 스토리지클래스 객체의 이름을 사용하여 특정 클래스를 요청할 수 있다. + + diff --git a/content/ko/docs/setup/production-environment/container-runtimes.md b/content/ko/docs/setup/production-environment/container-runtimes.md index 4b749ab7e5..f51b5f5eef 100644 --- a/content/ko/docs/setup/production-environment/container-runtimes.md +++ b/content/ko/docs/setup/production-environment/container-runtimes.md @@ -122,7 +122,7 @@ sudo apt-get update && sudo apt-get install -y containerd.io ```shell # containerd 구성 sudo mkdir -p /etc/containerd -sudo containerd config default | sudo tee /etc/containerd/config.toml +containerd config default | sudo tee /etc/containerd/config.toml ``` ```shell @@ -140,7 +140,7 @@ sudo apt-get update && sudo apt-get install -y containerd ```shell # containerd 구성 sudo mkdir -p /etc/containerd -sudo containerd config default | sudo tee /etc/containerd/config.toml +containerd config default | sudo tee /etc/containerd/config.toml ``` ```shell @@ -210,7 +210,7 @@ sudo yum update -y && sudo yum install -y containerd.io ```shell ## containerd 구성 sudo mkdir -p /etc/containerd -sudo containerd config default | sudo tee /etc/containerd/config.toml +containerd config default | sudo tee /etc/containerd/config.toml ``` ```shell diff --git a/content/ko/docs/setup/release/version-skew-policy.md b/content/ko/docs/setup/release/version-skew-policy.md index feb675f8ba..76ff7504fd 100644 --- a/content/ko/docs/setup/release/version-skew-policy.md +++ b/content/ko/docs/setup/release/version-skew-policy.md @@ -1,11 +1,18 @@ --- + + + + + + + title: 쿠버네티스 버전 및 버전 차이(skew) 지원 정책 content_type: concept weight: 30 --- -이 문서는 다양한 쿠버네티스 구성 요소 간에 지원되는 최대 버전 차이를 설명한다. +이 문서는 다양한 쿠버네티스 구성 요소 간에 지원되는 최대 버전 차이를 설명한다. 특정 클러스터 배포 도구는 버전 차이에 대한 추가적인 제한을 설정할 수 있다. @@ -19,14 +26,14 @@ weight: 30 쿠버네티스 프로젝트는 최근 세 개의 마이너 릴리스 ({{< skew latestVersion >}}, {{< skew prevMinorVersion >}}, {{< skew oldestMinorVersion >}}) 에 대한 릴리스 분기를 유지한다. 쿠버네티스 1.19 이상은 약 1년간의 패치 지원을 받는다. 쿠버네티스 1.18 이상은 약 9개월의 패치 지원을 받는다. -보안 수정사항을 포함한 해당 수정사항은 심각도와 타당성에 따라 세 개의 릴리스 브랜치로 백포트(backport) 될 수 있다. +보안 수정사항을 포함한 해당 수정사항은 심각도와 타당성에 따라 세 개의 릴리스 브랜치로 백포트(backport) 될 수 있다. 패치 릴리스는 각 브랜치별로 [정기적인 주기](https://git.k8s.io/sig-release/releases/patch-releases.md#cadence)로 제공하며, 필요한 경우 추가 긴급 릴리스도 추가한다. [릴리스 관리자](https://git.k8s.io/sig-release/release-managers.md) 그룹이 이러한 결정 권한을 가진다. 자세한 내용은 쿠버네티스 [패치 릴리스](https://git.k8s.io/sig-release/releases/patch-releases.md) 페이지를 참조한다. -## 지원되는 버전 차이 +## 지원되는 버전 차이 ### kube-apiserver @@ -133,6 +140,11 @@ HA 클러스터의 `kube-apiserver` 인스턴스 간에 버전 차이가 있으 필요에 따라서 `kubelet` 인스턴스를 **{{< skew latestVersion >}}** 으로 업그레이드할 수 있다(또는 **{{< skew prevMinorVersion >}}** 아니면 **{{< skew oldestMinorVersion >}}** 으로 유지할 수 있음). +{{< note >}} +`kubelet` 마이너 버전 업그레이드를 수행하기 전에, 해당 노드의 파드를 [드레인(drain)](/docs/tasks/administer-cluster/safely-drain-node/)해야 한다. +인플레이스(In-place) 마이너 버전 `kubelet` 업그레이드는 지원되지 않는다. +{{}} + {{< warning >}} 클러스터 안의 `kubelet` 인스턴스를 `kube-apiserver`의 버전보다 2단계 낮은 버전으로 실행하는 것을 권장하지 않는다: diff --git a/content/ko/docs/tasks/tools/install-kubectl.md b/content/ko/docs/tasks/tools/install-kubectl.md index 9d80451e1b..70a6a60409 100644 --- a/content/ko/docs/tasks/tools/install-kubectl.md +++ b/content/ko/docs/tasks/tools/install-kubectl.md @@ -1,4 +1,6 @@ --- + + title: kubectl 설치 및 설정 content_type: task weight: 10 @@ -30,33 +32,73 @@ kubectl을 사용하여 애플리케이션을 배포하고, 클러스터 리소 1. 다음 명령으로 최신 릴리스를 다운로드한다. - ``` - curl -LO "https://storage.googleapis.com/kubernetes-release/release/$(curl -s https://storage.googleapis.com/kubernetes-release/release/stable.txt)/bin/linux/amd64/kubectl" - ``` + ```bash + curl -LO "https://dl.k8s.io/release/$(curl -L -s https://dl.k8s.io/release/stable.txt)/bin/linux/amd64/kubectl" + ``` - 특정 버전을 다운로드하려면, `$(curl -s https://storage.googleapis.com/kubernetes-release/release/stable.txt)` 명령 부분을 특정 버전으로 바꾼다. + {{< note >}} +특정 버전을 다운로드하려면, `$(curl -L -s https://dl.k8s.io/release/stable.txt)` 명령 부분을 특정 버전으로 바꾼다. - 예를 들어, 리눅스에서 버전 {{< param "fullversion" >}}을 다운로드하려면, 다음을 입력한다. - ``` - curl -LO https://storage.googleapis.com/kubernetes-release/release/{{< param "fullversion" >}}/bin/linux/amd64/kubectl - ``` +예를 들어, 리눅스에서 버전 {{< param "fullversion" >}}을 다운로드하려면, 다음을 입력한다. -2. kubectl 바이너리를 실행 가능하게 만든다. + ```bash + curl -LO https://dl.k8s.io/release/{{< param "fullversion" >}}/bin/linux/amd64/kubectl + ``` + {{< /note >}} - ``` - chmod +x ./kubectl - ``` +1. 바이너리를 검증한다. (선택 사항) -3. 바이너리를 PATH가 설정된 디렉터리로 옮긴다. + kubectl 체크섬(checksum) 파일을 다운로드한다. - ``` - sudo mv ./kubectl /usr/local/bin/kubectl - ``` -4. 설치한 버전이 최신 버전인지 확인한다. + ```bash + curl -LO "https://dl.k8s.io/$(curl -L -s https://dl.k8s.io/release/stable.txt)/bin/linux/amd64/kubectl.sha256" + ``` - ``` - kubectl version --client - ``` + kubectl 바이너리를 체크섬 파일을 통해 검증한다. + + ```bash + echo "$(}} + 동일한 버전의 바이너리와 체크섬을 다운로드한다. + {{< /note >}} + +1. kubectl 설치 + + ```bash + sudo install -o root -g root -m 0755 kubectl /usr/local/bin/kubectl + ``` + + {{< note >}} + 대상 시스템에 root 접근 권한을 가지고 있지 않더라도, `~/.local/bin` 디렉터리에 kubectl을 설치할 수 있다. + + ```bash + mkdir -p ~/.local/bin/kubectl + mv ./kubectl ~/.local/bin/kubectl + # 그리고 ~/.local/bin/kubectl을 $PATH에 추가 + ``` + + {{< /note >}} + +1. 설치한 버전이 최신인지 확인한다. + + ```bash + kubectl version --client + ``` ### 기본 패키지 관리 도구를 사용하여 설치 @@ -117,29 +159,65 @@ kubectl version --client 1. 최신 릴리스를 다운로드한다. ```bash - curl -LO "https://storage.googleapis.com/kubernetes-release/release/$(curl -s https://storage.googleapis.com/kubernetes-release/release/stable.txt)/bin/darwin/amd64/kubectl" + curl -LO "https://dl.k8s.io/release/$(curl -L -s https://dl.k8s.io/release/stable.txt)/bin/darwin/amd64/kubectl" ``` - 특정 버전을 다운로드하려면, `$(curl -s https://storage.googleapis.com/kubernetes-release/release/stable.txt)` 명령 부분을 특정 버전으로 바꾼다. + {{< note >}} + 특정 버전을 다운로드하려면, `$(curl -L -s https://dl.k8s.io/release/stable.txt)` 명령 부분을 특정 버전으로 바꾼다. 예를 들어, macOS에서 버전 {{< param "fullversion" >}}을 다운로드하려면, 다음을 입력한다. - ```bash - curl -LO https://storage.googleapis.com/kubernetes-release/release/{{< param "fullversion" >}}/bin/darwin/amd64/kubectl - ``` - kubectl 바이너리를 실행 가능하게 만든다. + ```bash + curl -LO https://dl.k8s.io/release/{{< param "fullversion" >}}/bin/darwin/amd64/kubectl + ``` + + {{< /note >}} + +1. 바이너리를 검증한다. (선택 사항) + + kubectl 체크섬 파일을 다운로드한다. + + ```bash + curl -LO "https://dl.k8s.io/$(curl -L -s https://dl.k8s.io/release/stable.txt)/bin/darwin/amd64/kubectl.sha256" + ``` + + kubectl 바이너리를 체크섬 파일을 통해 검증한다. + + ```bash + echo "$(}} + 동일한 버전의 바이너리와 체크섬을 다운로드한다. + {{< /note >}} + +1. kubectl 바이너리를 실행 가능하게 한다. ```bash chmod +x ./kubectl ``` -3. 바이너리를 PATH가 설정된 디렉터리로 옮긴다. +1. kubectl 바이너리를 시스템 `PATH` 의 파일 위치로 옮긴다. ```bash - sudo mv ./kubectl /usr/local/bin/kubectl + sudo mv ./kubectl /usr/local/bin/kubectl && \ + sudo chown root: /usr/local/bin/kubectl ``` -4. 설치한 버전이 최신 버전인지 확인한다. +1. 설치한 버전이 최신 버전인지 확인한다. ```bash kubectl version --client @@ -161,7 +239,7 @@ macOS에서 [Homebrew](https://brew.sh/) 패키지 관리자를 사용하는 경 brew install kubernetes-cli ``` -2. 설치한 버전이 최신 버전인지 확인한다. +1. 설치한 버전이 최신 버전인지 확인한다. ```bash kubectl version --client @@ -178,7 +256,7 @@ macOS에서 [Macports](https://macports.org/) 패키지 관리자를 사용하 sudo port install kubectl ``` -2. 설치한 버전이 최신 버전인지 확인한다. +1. 설치한 버전이 최신 버전인지 확인한다. ```bash kubectl version --client @@ -188,30 +266,55 @@ macOS에서 [Macports](https://macports.org/) 패키지 관리자를 사용하 ### 윈도우에서 curl을 사용하여 kubectl 바이너리 설치 -1. [이 링크](https://storage.googleapis.com/kubernetes-release/release/{{< param "fullversion" >}}/bin/windows/amd64/kubectl.exe)에서 최신 릴리스 {{< param "fullversion" >}}을 다운로드한다. +1. [최신 릴리스 {{< param "fullversion" >}}](https://dl.k8s.io/release/{{< param "fullversion" >}}/bin/windows/amd64/kubectl.exe)를 다운로드한다. 또는 `curl` 을 설치한 경우, 다음 명령을 사용한다. - ```bash - curl -LO https://storage.googleapis.com/kubernetes-release/release/{{< param "fullversion" >}}/bin/windows/amd64/kubectl.exe + ```powershell + curl -LO https://dl.k8s.io/release/{{< param "fullversion" >}}/bin/windows/amd64/kubectl.exe ``` - 최신의 안정 버전(예: 스크립팅을 위한)을 찾으려면, [https://storage.googleapis.com/kubernetes-release/release/stable.txt](https://storage.googleapis.com/kubernetes-release/release/stable.txt)를 참고한다. + {{< note >}} + 최신의 안정 버전(예: 스크립팅을 위한)을 찾으려면, [https://dl.k8s.io/release/stable.txt](https://dl.k8s.io/release/stable.txt)를 참고한다. + {{< /note >}} -2. 바이너리를 PATH가 설정된 디렉터리에 추가한다. +1. 바이너리를 검증한다. (선택 사항) -3. `kubectl` 의 버전이 다운로드한 버전과 같은지 확인한다. + kubectl 체크섬 파일을 다운로드한다. - ```bash + ```powershell + curl -LO https://dl.k8s.io/{{< param "fullversion" >}}/bin/windows/amd64/kubectl.exe.sha256 + ``` + + kubectl 바이너리를 체크섬 파일을 통해 검증한다. + + - 수동으로 `CertUtil` 의 출력과 다운로드한 체크섬 파일을 비교하기 위해서 커맨드 프롬프트를 사용한다. + + ```cmd + CertUtil -hashfile kubectl.exe SHA256 + type kubectl.exe.sha256 + ``` + + - `-eq` 연산자를 통해 `True` 또는 `False` 결과를 얻는 자동 검증을 위해서 PowerShell을 사용한다. + + ```powershell + $($(CertUtil -hashfile .\kubectl.exe SHA256)[1] -replace " ", "") -eq $(type .\kubectl.exe.sha256) + ``` + +1. 바이너리를 `PATH` 가 설정된 디렉터리에 추가한다. + +1. `kubectl` 의 버전이 다운로드한 버전과 같은지 확인한다. + + ```cmd kubectl version --client ``` {{< note >}} -[윈도우용 도커 데스크톱](https://docs.docker.com/docker-for-windows/#kubernetes)은 자체 버전의 `kubectl` 을 PATH에 추가한다. -도커 데스크톱을 이전에 설치한 경우, 도커 데스크톱 설치 프로그램에서 추가한 PATH 항목 앞에 PATH 항목을 배치하거나 도커 데스크톱의 `kubectl` 을 제거해야 할 수도 있다. +[윈도우용 도커 데스크톱](https://docs.docker.com/docker-for-windows/#kubernetes)은 자체 버전의 `kubectl` 을 `PATH` 에 추가한다. +도커 데스크톱을 이전에 설치한 경우, 도커 데스크톱 설치 프로그램에서 추가한 `PATH` 항목 앞에 `PATH` 항목을 배치하거나 도커 데스크톱의 `kubectl` 을 제거해야 할 수도 있다. {{< /note >}} -### PSGallery에서 Powershell로 설치 +### PSGallery에서 PowerShell로 설치 윈도우에서 [Powershell Gallery](https://www.powershellgallery.com/) 패키지 관리자를 사용하는 경우, Powershell로 kubectl을 설치하고 업데이트할 수 있다. @@ -223,12 +326,12 @@ macOS에서 [Macports](https://macports.org/) 패키지 관리자를 사용하 ``` {{< note >}} - `DownloadLocation` 을 지정하지 않으면, `kubectl` 은 사용자의 임시 디렉터리에 설치된다. + `DownloadLocation` 을 지정하지 않으면, `kubectl` 은 사용자의 `temp` 디렉터리에 설치된다. {{< /note >}} 설치 프로그램은 `$HOME/.kube` 를 생성하고 구성 파일을 작성하도록 지시한다. -2. 설치한 버전이 최신 버전인지 확인한다. +1. 설치한 버전이 최신 버전인지 확인한다. ```powershell kubectl version --client @@ -256,32 +359,32 @@ macOS에서 [Macports](https://macports.org/) 패키지 관리자를 사용하 {{< /tabs >}} -2. 설치한 버전이 최신 버전인지 확인한다. +1. 설치한 버전이 최신 버전인지 확인한다. ```powershell kubectl version --client ``` -3. 홈 디렉터리로 이동한다. +1. 홈 디렉터리로 이동한다. ```powershell # cmd.exe를 사용한다면, 다음을 실행한다. cd %USERPROFILE% cd ~ ``` -4. `.kube` 디렉터리를 생성한다. +1. `.kube` 디렉터리를 생성한다. ```powershell mkdir .kube ``` -5. 금방 생성한 `.kube` 디렉터리로 이동한다. +1. 금방 생성한 `.kube` 디렉터리로 이동한다. ```powershell cd .kube ``` -6. 원격 쿠버네티스 클러스터를 사용하도록 kubectl을 구성한다. +1. 원격 쿠버네티스 클러스터를 사용하도록 kubectl을 구성한다. ```powershell New-Item config -type file @@ -297,13 +400,13 @@ kubectl을 Google Cloud SDK의 일부로 설치할 수 있다. 1. [Google Cloud SDK](https://cloud.google.com/sdk/)를 설치한다. -2. `kubectl` 설치 명령을 실행한다. +1. `kubectl` 설치 명령을 실행한다. ```shell gcloud components install kubectl ``` -3. 설치한 버전이 최신 버전인지 확인한다. +1. 설치한 버전이 최신 버전인지 확인한다. ```shell kubectl version --client @@ -381,11 +484,13 @@ source /usr/share/bash-completion/bash_completion ```bash echo 'source <(kubectl completion bash)' >>~/.bashrc ``` + - 완성 스크립트를 `/etc/bash_completion.d` 디렉터리에 추가한다. ```bash kubectl completion bash >/etc/bash_completion.d/kubectl ``` + kubectl에 대한 앨리어스(alias)가 있는 경우, 해당 앨리어스로 작업하도록 셸 완성을 확장할 수 있다. ```bash @@ -466,7 +571,6 @@ export BASH_COMPLETION_COMPAT_DIR="/usr/local/etc/bash_completion.d" ```bash echo 'source <(kubectl completion bash)' >>~/.bash_profile - ``` - 완성 스크립트를 `/usr/local/etc/bash_completion.d` 디렉터리에 추가한다. From 765e98db54b3ca8746b0c44eaa231f87f44d6170 Mon Sep 17 00:00:00 2001 From: Michael H Date: Thu, 18 Feb 2021 11:13:01 -0500 Subject: [PATCH 41/62] updated labels for cli command --- content/en/docs/tutorials/stateless-application/guestbook.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/content/en/docs/tutorials/stateless-application/guestbook.md b/content/en/docs/tutorials/stateless-application/guestbook.md index b60afcf5f8..27cc649edf 100644 --- a/content/en/docs/tutorials/stateless-application/guestbook.md +++ b/content/en/docs/tutorials/stateless-application/guestbook.md @@ -134,7 +134,7 @@ kubectl apply -f ./content/en/examples/application/guestbook/frontend-deployment 1. Query the list of Pods to verify that the three frontend replicas are running: ```shell - kubectl get pods -l app=guestbook -l tier=frontend + kubectl get pods -l app.kubernetes.io/name=guestbook -l app.kubernetes.io/component=frontend ``` The response should be similar to this: From b3d9fc873388a7a4e28a523240bfe37865908553 Mon Sep 17 00:00:00 2001 From: Arhell Date: Fri, 19 Feb 2021 00:59:11 +0200 Subject: [PATCH 42/62] [zh] fix broken link --- content/zh/docs/reference/glossary/cluster-operator.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/content/zh/docs/reference/glossary/cluster-operator.md b/content/zh/docs/reference/glossary/cluster-operator.md index 4be92284b0..0e366d76dd 100644 --- a/content/zh/docs/reference/glossary/cluster-operator.md +++ b/content/zh/docs/reference/glossary/cluster-operator.md @@ -36,10 +36,10 @@ tags: Their primary responsibility is keeping a cluster up and running, which may involve periodic maintenance activities or upgrades.
-**NOTE:** Cluster operators are different from the [Operator pattern](https://coreos.com/operators) that extends the Kubernetes API. +**NOTE:** Cluster operators are different from the [Operator pattern](https://www.openshift.com/learn/topics/operators) that extends the Kubernetes API. --> 他们的主要责任是保持集群正常运行，可能需要进行周期性的维护和升级活动。
-**注意：** 集群操作者不同于[操作者模式（Operator Pattern）](https://coreos.com/operators)，操作者模式是用来扩展 Kubernetes API 的。 +**注意：** 集群操作者不同于[操作者模式（Operator Pattern）](https://www.openshift.com/learn/topics/operators)，操作者模式是用来扩展 Kubernetes API 的。 From 1cb1ac47631a924f6badbd955ba18d00d6b83237 Mon Sep 17 00:00:00 2001 From: ydFu Date: Fri, 19 Feb 2021 11:26:48 +0800 Subject: [PATCH 43/62] [zh] Resync concepts pages for services-networking\dual-stack.md * Resync with english version in 'Dual-stack docs correction #26386' * Update in 'administer-cluster\safely-drain-node.md' #25996 * Fix git rebase in services-networking\dual-stack.md Signed-off-by: ydFu --- .../services-networking/dual-stack.md | 36 ++++++++++++++----- .../administer-cluster/safely-drain-node.md | 3 +- 2 files changed, 28 insertions(+), 11 deletions(-) diff --git a/content/zh/docs/concepts/services-networking/dual-stack.md b/content/zh/docs/concepts/services-networking/dual-stack.md index ef811874b5..f01bd78038 100644 --- a/content/zh/docs/concepts/services-networking/dual-stack.md +++ b/content/zh/docs/concepts/services-networking/dual-stack.md @@ -9,11 +9,17 @@ weight: 70 --- @@ -85,6 +91,20 @@ The following prerequisites are needed in order to utilize IPv4/IPv6 dual-stack 要启用 IPv4/IPv6 双协议栈，为集群的相关组件启用 `IPv6DualStack` [特性门控](/zh/docs/reference/command-line-tools-reference/feature-gates/)， @@ -95,8 +115,8 @@ To enable IPv4/IPv6 dual-stack, enable the `IPv6DualStack` [feature gate](/docs/ * `--service-cluster-ip-range=,` * kube-controller-manager: * `--feature-gates="IPv6DualStack=true"` - * `--cluster-cidr=,` 例如 `--cluster-cidr=10.244.0.0/16,fc00::/48` - * `--service-cluster-ip-range=,` 例如 `--service-cluster-ip-range=10.0.0.0/16,fd00::/108` + * `--cluster-cidr=,` + * `--service-cluster-ip-range=,` * `--node-cidr-mask-size-ipv4|--node-cidr-mask-size-ipv6` 对于 IPv4 默认为 /24，对于 IPv6 默认为 /64 * kubelet: * `--feature-gates="IPv6DualStack=true"` @@ -132,7 +152,7 @@ set the `.spec.ipFamilyPolicy` field to one of the following values: --> 如果你的集群启用了 IPv4/IPv6 双协议栈网络，则可以使用 IPv4 或 IPv6 地址来创建 {{< glossary_tooltip text="Service" term_id="service" >}}。 -服务的地址族默认为第一个服务集群 IP 范围的地址族（通过 kube-apiserver 的 `--service-cluster-ip-range` 参数配置） +服务的地址族默认为第一个服务集群 IP 范围的地址族（通过 kube-apiserver 的 `--service-cluster-ip-range` 参数配置）。 当你定义服务时，可以选择将其配置为双栈。若要指定所需的行为，你可以设置 `.spec.ipFamilyPolicy` 字段为以下值之一： -### LoadBalancer 类型 +### LoadBalancer 类型服务 ## 出站流量 @@ -440,6 +460,4 @@ Ensure your {{< glossary_tooltip text="CNI" term_id="cni" >}} provider supports -* [验证 IPv4/IPv6 双协议栈](/zh/docs/tasks/network/validate-dual-stack)网络 - - +* [验证 IPv4/IPv6 双协议栈](/zh/docs/tasks/network/validate-dual-stack)网络 \ No newline at end of file diff --git a/content/zh/docs/tasks/administer-cluster/safely-drain-node.md b/content/zh/docs/tasks/administer-cluster/safely-drain-node.md index 063288cb93..85de0215ca 100644 --- a/content/zh/docs/tasks/administer-cluster/safely-drain-node.md +++ b/content/zh/docs/tasks/administer-cluster/safely-drain-node.md @@ -125,7 +125,7 @@ Once it returns (without giving an error), you can power down the node If you leave the node in the cluster during the maintenance operation, you need to run --> 一旦它返回（没有报错）， -你就可以下电此节点（或者等价地，如果在云平台上，删除支持该节点的虚拟机）。 +你就可以下线此节点（或者等价地，如果在云平台上，删除支持该节点的虚拟机）。 如果要在维护操作期间将节点留在集群中，则需要运行： ```shell @@ -295,4 +295,3 @@ Kubernetes 并没有具体说明在这种情况下应该采取什么行为， --> * 执行[配置 PDB](/zh/docs/tasks/run-application/configure-pdb/)中的各个步骤， 保护你的应用 - From 837ae37271c55afc1e8bc3e11e2935479f1de49e Mon Sep 17 00:00:00 2001 From: luzg Date: Tue, 16 Feb 2021 18:09:02 +0800 Subject: [PATCH 44/62] [zh] translate blog Don't Panic: Kubernetes and Docker --- ...-12-02-dont-panic-kubernetes-and-docker.md | 210 ++++++++++++++++++ 1 file changed, 210 insertions(+) create mode 100644 content/zh/blog/_posts/2020-12-02-dont-panic-kubernetes-and-docker.md diff --git a/content/zh/blog/_posts/2020-12-02-dont-panic-kubernetes-and-docker.md b/content/zh/blog/_posts/2020-12-02-dont-panic-kubernetes-and-docker.md new file mode 100644 index 0000000000..788b28420d --- /dev/null +++ b/content/zh/blog/_posts/2020-12-02-dont-panic-kubernetes-and-docker.md @@ -0,0 +1,210 @@ +--- +layout: blog +title: "别慌: Kubernetes 和 Docker" +date: 2020-12-02 +slug: dont-panic-kubernetes-and-docker +--- + + +**作者：** Jorge Castro, Duffie Cooley, Kat Cosgrove, Justin Garrison, Noah Kantrowitz, Bob Killen, Rey Lejano, Dan “POP” Papandrea, Jeffrey Sica, Davanum “Dims” Srinivas + + +Kubernetes 从版本 v1.20 之后，[弃用 Docker](https://github.com/kubernetes/kubernetes/blob/master/CHANGELOG/CHANGELOG-1.20.md#deprecation) +这个容器运行时。 + + +**不必慌张，这件事并没有听起来那么吓人。** + + +弃用 Docker 这个底层运行时，转而支持符合为 Kubernetes 创建的 +[Container Runtime Interface (CRI)](https://kubernetes.io/blog/2016/12/container-runtime-interface-cri-in-kubernetes/) +的运行时。 +Docker 构建的镜像，将在你的集群的所有运行时中继续工作，一如既往。 + + +如果你是 Kubernetes 的终端用户，这对你不会有太大影响。 +这事并不意味着 Dockder 已死、也不意味着你不能或不该继续把 Docker 用作开发工具。 +Docker 仍然是构建容器的利器，使用命令 `docker build` 构建的镜像在 Kubernetes 集群中仍然可以运行。 + + +如果你正在使用 GKE、EKS、或 AKS +([默认使用 containerd](https://github.com/Azure/AKS/releases/tag/2020-11-16)) +这类托管 Kubernetes 服务，你需要在 Kubernetes 后续版本移除对 Docker 支持之前， +确认工作节点使用了被支持的容器运行时。 +如果你的节点被定制过，你可能需要根据你自己的环境和运行时需求更新它们。 +请与你的服务供应商协作，确保做出适当的升级测试和计划。 + + +如果你正在运营你自己的集群，那还应该做些工作，以避免集群中断。 +在 v1.20 版中，你仅会得到一个 Docker 的弃用警告。 +当对 Docker 运行时的支持在 Kubernetes 某个后续发行版（目前的计划是 2021 年晚些时候的 1.22 版）中被移除时， +你需要切换到 containerd 或 CRI-O 等兼容的容器运行时。 +只要确保你选择的运行时支持你当前使用的 Docker 守护进程配置（例如 logging）。 + + +## 那为什么会有这样的困惑，为什么每个人要害怕呢？{#so-why-the-confusion-and-what-is-everyone-freaking-out-about} + + +我们在这里讨论的是两套不同的环境，这就是造成困惑的根源。 +在你的 Kubernetes 集群中，有一个叫做容器运行时的东西，它负责拉取并运行容器镜像。 +Docker 对于运行时来说是一个流行的选择（其他常见的选择包括 containerd 和 CRI-O）， +但 Docker 并非设计用来嵌入到 Kubernetes，这就是问题所在。 + + +你看，我们称之为 “Docker” 的物件实际上并不是一个物件——它是一个完整的技术堆栈， +它其中一个叫做 “containerd” 的部件本身，才是一个高级容器运行时。 +Docker 既酷炫又实用，因为它提供了很多用户体验增强功能，而这简化了我们做开发工作时的操作， +Kubernetes 用不到这些增强的用户体验，毕竟它并非人类。 + + +因为这个用户友好的抽象层，Kubernetes 集群不得不引入一个叫做 Dockershim 的工具来访问它真正需要的 containerd。 +这不是一件好事，因为这引入了额外的运维工作量，而且还可能出错。 +实际上正在发生的事情就是：Dockershim 将在不早于 v1.23 版中从 kubelet 中被移除，也就取消对 Docker 容器运行时的支持。 +你心里可能会想，如果 containerd 已经包含在 Docker 堆栈中，为什么 Kubernetes 需要 Dockershim。 + + +Docker 不兼容 CRI， +[容器运行时接口](https://kubernetes.io/blog/2016/12/container-runtime-interface-cri-in-kubernetes/)。 +如果支持，我们就不需要这个 shim 了，也就没问题了。 +但这也不是世界末日，你也不需要恐慌——你唯一要做的就是把你的容器运行时从 Docker 切换到其他受支持的容器运行时。 + + +要注意一点：如果你依赖底层的 Docker 套接字(`/var/run/docker.sock`)，作为你集群中工作流的一部分， +切换到不同的运行时会导致你无法使用它。 +这种模式经常被称之为嵌套 Docker（Docker in Docker）。 +对于这种特殊的场景，有很多选项，比如： +[kaniko](https://github.com/GoogleContainerTools/kaniko)、 +[img](https://github.com/genuinetools/img)、和 +[buildah](https://github.com/containers/buildah)。 + + +## 那么，这一改变对开发人员意味着什么？我们还要写 Dockerfile 吗？还能用 Docker 构建镜像吗？{#what-does-this-change-mean-for-developers} + + +此次改变带来了一个不同的环境，这不同于我们常用的 Docker 交互方式。 +你在开发环境中用的 Docker 和你 Kubernetes 集群中的 Docker 运行时无关。 +我们知道这听起来让人困惑。 +对于开发人员，Docker 从所有角度来看仍然有用，就跟这次改变之前一样。 +Docker 构建的镜像并不是 Docker 特有的镜像——它是一个 +OCI（[开放容器标准](https://opencontainers.org/)）镜像。 +任一 OCI 兼容的镜像，不管它是用什么工具构建的，在 Kubernetes 的角度来看都是一样的。 +[containerd](https://containerd.io/) 和 +[CRI-O](https://cri-o.io/) +两者都知道怎么拉取并运行这些镜像。 +这就是我们制定容器标准的原因。 + + +所以，改变已经发生。 +它确实带来了一些问题，但这不是一个灾难，总的说来，这还是一件好事。 +根据你操作 Kubernetes 的方式的不同，这可能对你不构成任何问题，或者也只是意味着一点点的工作量。 +从一个长远的角度看，它使得事情更简单。 +如果你还在困惑，也没问题——这里还有很多事情； +Kubernetes 有很多变化中的功能，没有人是100%的专家。 +我们鼓励你提出任何问题，无论水平高低、问题难易。 +我们的目标是确保所有人都能在即将到来的改变中获得足够的了解。 +我们希望这已经回答了你的大部分问题，并缓解了一些焦虑！❤️ + + +还在寻求更多答案吗？请参考我们附带的 +[弃用 Dockershim 的常见问题](/zh/blog/2020/12/02/dockershim-faq/)。 From 9303a66890de4be4d580e410276388c2fd757856 Mon Sep 17 00:00:00 2001 From: caodonghui Date: Fri, 19 Feb 2021 17:33:16 +0800 Subject: [PATCH 45/62] fix typo --- .../docs/tasks/access-application-cluster/ingress-minikube.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/content/zh/docs/tasks/access-application-cluster/ingress-minikube.md b/content/zh/docs/tasks/access-application-cluster/ingress-minikube.md index 4694ac5dbc..01fba70e92 100644 --- a/content/zh/docs/tasks/access-application-cluster/ingress-minikube.md +++ b/content/zh/docs/tasks/access-application-cluster/ingress-minikube.md @@ -277,7 +277,7 @@ The following file is an Ingress resource that sends traffic to your Service via If you are running Minikube locally, you can visit hello-world.info from your browser. --> {{< note >}} - 如果你在使用本地 Minikube 环境，你可以从浏览器中访问 hellow-world.info。 + 如果你在使用本地 Minikube 环境，你可以从浏览器中访问 hello-world.info。 {{< /note >}} {{< note >}} 如果你在本地运行 Minikube 环境，你可以使用浏览器来访问 - hellow-world.info 和 hello-world.info/v2。 + hello-world.info 和 hello-world.info/v2。 {{< /note >}} ## {{% heading "whatsnext" %}} From 6d1f57ebc33eb856571fb6e8cf9f48a232fdc5a6 Mon Sep 17 00:00:00 2001 From: Maciej Filocha Date: Tue, 16 Feb 2021 08:52:32 +0100 Subject: [PATCH 46/62] Synchronize Polish localization with upstream MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Polish translation synchronized up to fa83273ca52656c3d6ad5acb67db05a82d2d3bbc. Co-authored-by: Karol Pucyński <9209870+kpucynski@users.noreply.github.com> Co-authored-by: Michał Sochoń --- .../docs/concepts/overview/kubernetes-api.md | 107 +++++++++++------- content/pl/docs/reference/_index.md | 4 +- 2 files changed, 70 insertions(+), 41 deletions(-) diff --git a/content/pl/docs/concepts/overview/kubernetes-api.md b/content/pl/docs/concepts/overview/kubernetes-api.md index cd376dab1d..731b22cacc 100644 --- a/content/pl/docs/concepts/overview/kubernetes-api.md +++ b/content/pl/docs/concepts/overview/kubernetes-api.md @@ -4,7 +4,7 @@ content_type: concept weight: 30 description: > API Kubernetesa służy do odpytywania i zmiany stanu obiektów Kubernetesa. - Sercem warstwy sterowania Kubernetesa jest serwer API i udostępniane przez niego HTTP API. Przez ten serwer odbywa się komunikacja pomiędzy użytkownikami, różnymi częściami składowymi klastra oraz komponentami zewnętrznymi. + Sercem warstwy sterowania Kubernetesa jest serwer API i udostępniane po HTTP API. Przez ten serwer odbywa się komunikacja pomiędzy użytkownikami, różnymi częściami składowymi klastra oraz komponentami zewnętrznymi. card: name: concepts weight: 30 @@ -14,13 +14,16 @@ card: Sercem {{< glossary_tooltip text="warstwy sterowania" term_id="control-plane" >}} Kubernetes jest {{< glossary_tooltip text="serwer API" term_id="kube-apiserver" >}}. Serwer udostępnia -API poprzez HTTP, umożliwiając wzajemną komunikację pomiędzy użytkownikami, częściami składowymi klastra i komponentami zewnętrznymi. +API poprzez HTTP, umożliwiając wzajemną komunikację pomiędzy użytkownikami, częściami składowymi klastra +i komponentami zewnętrznymi. -API Kubernetes pozwala na sprawdzanie i zmianę stanu obiektów (przykładowo: pody, _Namespaces_, _ConfigMaps_, _Events_). +API Kubernetesa pozwala na sprawdzanie i zmianę stanu obiektów +(przykładowo: pody, _Namespaces_, _ConfigMaps_, _Events_). Większość operacji może zostać wykonana poprzez interfejs linii komend (CLI) [kubectl](/docs/reference/kubectl/overview/) lub inne -programy, takie jak [kubeadm](/docs/reference/setup-tools/kubeadm/), które używają +programy, takie jak +[kubeadm](/docs/reference/setup-tools/kubeadm/), które używają API. Możesz też korzystać z API bezpośrednio przez wywołania typu REST. Jeśli piszesz aplikację używającą API Kubernetesa, @@ -66,54 +69,77 @@ Aby wybrać format odpowiedzi, użyj nagłówków żądania zgodnie z tabelą: -W Kubernetesie zaimplementowany jest alternatywny format serializacji na potrzeby API oparty o Protobuf, -który jest przede wszystkim przeznaczony na potrzeby wewnętrznej komunikacji w klastrze -i opisany w [design proposal](https://github.com/kubernetes/community/blob/master/contributors/design-proposals/api-machinery/protobuf.md). -Pliki IDL dla każdego ze schematów można znaleźć w pakietach Go, które definiują obiekty API. +W Kubernetesie zaimplementowany jest alternatywny format serializacji na potrzeby API oparty o +Protobuf, który jest przede wszystkim przeznaczony na potrzeby wewnętrznej komunikacji w klastrze. +Więcej szczegółów znajduje się w dokumencie [Kubernetes Protobuf serialization](https://github.com/kubernetes/community/blob/master/contributors/design-proposals/api-machinery/protobuf.md). +oraz w plikach *Interface Definition Language* (IDL) dla każdego ze schematów +zamieszczonych w pakietach Go, które definiują obiekty API. -## Zmiany API +## Przechowywanie stanu + +Kubernetes przechowuje serializowany stan swoich obiektów w +{{< glossary_tooltip term_id="etcd" >}}. + +## Grupy i wersje API + +Aby ułatwić usuwanie poszczególnych pól lub restrukturyzację reprezentacji zasobów, Kubernetes obsługuje +równocześnie wiele wersji API, każde poprzez osobną ścieżkę API, +na przykład: `/api/v1` lub `/apis/rbac.authorization.k8s.io/v1alpha1`. + +Rozdział wersji wprowadzony jest na poziomie całego API, a nie na poziomach poszczególnych zasobów lub pól, +aby być pewnym, że API odzwierciedla w sposób przejrzysty i spójny zasoby systemowe +i ich zachowania oraz pozwala na kontrolowany dostęp do tych API, które są w fazie wycofywania +lub fazie eksperymentalnej. + +Aby ułatwić rozbudowę API Kubernetes, wprowadziliśmy +[*grupy API*](https://git.k8s.io/community/contributors/design-proposals/api-machinery/api-group.md), które mogą +być [włączane i wyłączane](/docs/reference/using-api/#enabling-or-disabling). + +Zasoby API są rozróżniane poprzez przynależność do grupy API, typ zasobu, przestrzeń nazw (_namespace_, +o ile ma zastosowanie) oraz nazwę. Serwer API może przeprowadzać konwersję między +różnymi wersjami API w sposób niewidoczny dla użytkownika: wszystkie te różne wersje +reprezentują w rzeczywistości ten sam zasób. Serwer API może udostępniać te same dane +poprzez kilka różnych wersji API. + +Załóżmy przykładowo, że istnieją dwie wersje `v1` i `v1beta1` tego samego zasobu. +Obiekt utworzony przez wersję `v1beta1` może być odczytany, +zaktualizowany i skasowany zarówno przez wersję +`v1beta1`, jak i `v1`. + +## Trwałość API Z naszego doświadczenia wynika, że każdy system, który odniósł sukces, musi się nieustająco rozwijać w miarę zmieniających się potrzeb. Dlatego Kubernetes został tak zaprojektowany, aby API mogło się zmieniać i rozrastać. Projekt Kubernetes dąży do tego, aby nie wprowadzać zmian niezgodnych z istniejącymi aplikacjami klienckimi i utrzymywać zgodność przez wystarczająco długi czas, aby inne projekty zdążyły się dostosować do zmian. -W ogólności, nowe zasoby i pola definiujące zasoby API są dodawane stosunkowo często. Usuwanie zasobów lub pól -jest regulowane przez [API deprecation policy](/docs/reference/using-api/deprecation-policy/). -Definicja zmiany zgodnej (kompatybilnej) oraz metody wprowadzania zmian w API opisano w szczegółach -w [API change document](https://git.k8s.io/community/contributors/devel/sig-architecture/api_changes.md). +W ogólności, nowe zasoby i pola definiujące zasoby API są dodawane stosunkowo często. +Usuwanie zasobów lub pól jest regulowane przez +[API deprecation policy](/docs/reference/using-api/deprecation-policy/). -## Grupy i wersje API +Po osiągnięciu przez API statusu ogólnej dostępności (_general availability_ - GA), +oznaczanej zazwyczaj jako wersja API `v1`, bardzo zależy nam na utrzymaniu jej zgodności w kolejnych wydaniach. +Kubernetes utrzymuje także zgodność dla wersji _beta_ API tam, gdzie jest to możliwe: +jeśli zdecydowałeś się używać API w wersji beta, możesz z niego korzystać także później, +kiedy dana funkcjonalność osiągnie status stabilnej. -Aby ułatwić usuwanie poszczególnych pól lub restrukturyzację reprezentacji zasobów, Kubernetes obsługuje -równocześnie wiele wersji API, każde poprzez osobną ścieżkę API, na przykład: `/api/v1` lub -`/apis/rbac.authorization.k8s.io/v1alpha1`. - -Rozdział wersji wprowadzony jest na poziomie całego API, a nie na poziomach poszczególnych zasobów lub pól, aby być pewnym, -że API odzwierciedla w sposób przejrzysty i spójny zasoby systemowe i ich zachowania i pozwala -na kontrolowany dostęp do tych API, które są w fazie wycofywania lub fazie eksperymentalnej. - -Aby ułatwić rozbudowę API Kubernetes, wprowadziliśmy [*grupy API*](https://git.k8s.io/community/contributors/design-proposals/api-machinery/api-group.md), -które mogą być [włączane i wyłączane](/docs/reference/using-api/#enabling-or-disabling). - -Zasoby API są rozróżniane poprzez przynależność do grupy API, typ zasobu, przestrzeń nazw (_namespace_, -o ile ma zastosowanie) oraz nazwę. Serwer API może obsługiwać -te same dane poprzez różne wersje API i przeprowadzać konwersję między -różnymi wersjami API w sposób niewidoczny dla użytkownika. Wszystkie te różne wersje -reprezentują w rzeczywistości ten sam zasób. Załóżmy przykładowo, że istnieją dwie -wersje `v1` i `v1beta1` tego samego zasobu. Obiekt utworzony przez -wersję `v1beta1` może być odczytany, zaktualizowany i skasowany zarówno przez wersję -`v1beta1`, jak i `v1`. +{{< note >}} +Mimo, że Kubernetes stara się także zachować zgodność dla API w wersji _alpha_, zdarzają się przypadki, +kiedy nie jest to możliwe. Jeśli korzystasz z API w wersji alfa, przed aktualizacją klastra do nowej wersji +zalecamy sprawdzenie w informacjach o wydaniu, czy nie nastąpiła jakaś zmiana w tej części API. +{{< /note >}} Zajrzyj do [API versions reference](/docs/reference/using-api/#api-versioning) -po szczegółowe informacje, jak definiuje się poziomy wersji API. +po szczegółowe definicje różnych poziomów wersji API. + + ## Rozbudowa API -API Kubernetesa można rozbudowywać (rozszerzać) na dwa sposoby: +API Kubernetesa można rozszerzać na dwa sposoby: -1. [Definicje zasobów własnych](/docs/concepts/extend-kubernetes/api-extension/custom-resources/) - pozwalają deklaratywnie określać, jak serwer API powinien dostarczać wybrane zasoby API. +1. [Definicje zasobów własnych (_custom resources_)](/docs/concepts/extend-kubernetes/api-extension/custom-resources/) + pozwalają deklaratywnie określać, jak serwer API powinien dostarczać wybrane przez Ciebie zasoby API. 1. Można także rozszerzać API Kubernetesa implementując [warstwę agregacji](/docs/concepts/extend-kubernetes/api-extension/apiserver-aggregation/). @@ -121,6 +147,9 @@ API Kubernetesa można rozbudowywać (rozszerzać) na dwa sposoby: - Naucz się, jak rozbudowywać API Kubernetesa poprzez dodawanie własnych [CustomResourceDefinition](/docs/tasks/extend-kubernetes/custom-resources/custom-resource-definitions/). -- [Controlling API Access](/docs/reference/access-authn-authz/controlling-access/) opisuje +- [Controlling Access To The Kubernetes API](/docs/concepts/security/controlling-access/) opisuje sposoby, jakimi klaster zarządza dostępem do API. -- Punkty dostępowe API _(endpoints)_, typy zasobów i przykłady zamieszczono w [API Reference](/docs/reference/kubernetes-api/). +- Punkty dostępowe API _(endpoints)_, typy zasobów i przykłady zamieszczono w + [API Reference](/docs/reference/kubernetes-api/). +- Aby dowiedzieć się, jaki rodzaj zmian można określić jako zgodne i jak zmieniać API, zajrzyj do + [API changes](https://git.k8s.io/community/contributors/devel/sig-architecture/api_changes.md#readme). diff --git a/content/pl/docs/reference/_index.md b/content/pl/docs/reference/_index.md index bfc120218c..cf461cf7f2 100644 --- a/content/pl/docs/reference/_index.md +++ b/content/pl/docs/reference/_index.md @@ -8,13 +8,13 @@ content_type: concept -Tutaj znajdziesz dokumentację źródłową Kubernetes. +Tutaj znajdziesz dokumentację źródłową Kubernetesa. ## Dokumentacja API -* [Dokumentacja źródłowa API Kubernetesa {{< latest-version >}}](/docs/reference/generated/kubernetes-api/{{< latest-version >}}/) +* [Dokumentacja źródłowa API Kubernetesa {{< param "version" >}}](/docs/reference/generated/kubernetes-api/{{< param "version" >}}/) * [Using The Kubernetes API](/docs/reference/using-api/) - ogólne informacje na temat API Kubernetesa. ## Biblioteki klientów API From 49a6df0dbca3d8c5e3cc074e27d9a95ffa9a4ee9 Mon Sep 17 00:00:00 2001 From: Jailton Lopes Date: Fri, 19 Feb 2021 15:22:24 -0300 Subject: [PATCH 47/62] [pt] Add Explore You App in Portuguese (#26556) * Translate Explore You App in Portuguese * Update Explore Your App link in Learn Kubernetes Basics _index.html * Update link to Explore Interactive page translation in Portuguese * Fix typo and change the translation of the word *worker*. Signed-off-by: Jailton Lopes * Fix typo Signed-off-by: Jailton Lopes --- .../tutorials/kubernetes-basics/_index.html | 4 +- .../deploy-app/deploy-interactive.html | 2 +- .../kubernetes-basics/explore/_index.md | 4 + .../explore/explore-interactive.html | 41 +++++ .../explore/explore-intro.html | 143 ++++++++++++++++++ 5 files changed, 191 insertions(+), 3 deletions(-) create mode 100644 content/pt/docs/tutorials/kubernetes-basics/explore/_index.md create mode 100644 content/pt/docs/tutorials/kubernetes-basics/explore/explore-interactive.html create mode 100644 content/pt/docs/tutorials/kubernetes-basics/explore/explore-intro.html diff --git a/content/pt/docs/tutorials/kubernetes-basics/_index.html b/content/pt/docs/tutorials/kubernetes-basics/_index.html index b4a247b3c6..90e0592c18 100644 --- a/content/pt/docs/tutorials/kubernetes-basics/_index.html +++ b/content/pt/docs/tutorials/kubernetes-basics/_index.html @@ -70,9 +70,9 @@ card:
- +
-
3. Explore seu aplicativo
 +
3. Explore seu aplicativo
diff --git a/content/pt/docs/tutorials/kubernetes-basics/deploy-app/deploy-interactive.html b/content/pt/docs/tutorials/kubernetes-basics/deploy-app/deploy-interactive.html index cbc22b0d60..96f73e9250 100644 --- a/content/pt/docs/tutorials/kubernetes-basics/deploy-app/deploy-interactive.html +++ b/content/pt/docs/tutorials/kubernetes-basics/deploy-app/deploy-interactive.html @@ -37,7 +37,7 @@ weight: 20
- Continue para o Módulo 3› + Continue para o Módulo 3›
diff --git a/content/pt/docs/tutorials/kubernetes-basics/explore/_index.md b/content/pt/docs/tutorials/kubernetes-basics/explore/_index.md new file mode 100644 index 0000000000..c95e536676 --- /dev/null +++ b/content/pt/docs/tutorials/kubernetes-basics/explore/_index.md @@ -0,0 +1,4 @@ +--- +title: Explore seu aplicativo +weight: 30 +--- diff --git a/content/pt/docs/tutorials/kubernetes-basics/explore/explore-interactive.html b/content/pt/docs/tutorials/kubernetes-basics/explore/explore-interactive.html new file mode 100644 index 0000000000..08dd5736b2 --- /dev/null +++ b/content/pt/docs/tutorials/kubernetes-basics/explore/explore-interactive.html @@ -0,0 +1,41 @@ +--- +title: Tutorial Interativo - Explorando seu aplicativo +weight: 20 +--- + + + + + + + + + + + +
+ +
+ +
+
+ +
+ Para interagir com o Terminal, por favor, use a versão para desktop ou table. +
+ +
+
+
+
+
+ Continue para o Módulo 4› +
+
+ +
+ +
+ + + diff --git a/content/pt/docs/tutorials/kubernetes-basics/explore/explore-intro.html b/content/pt/docs/tutorials/kubernetes-basics/explore/explore-intro.html new file mode 100644 index 0000000000..c9720995dc --- /dev/null +++ b/content/pt/docs/tutorials/kubernetes-basics/explore/explore-intro.html @@ -0,0 +1,143 @@ +--- +title: Visualizando Pods e Nós (Nodes) +weight: 10 +--- + + + + + + + + + + +
+ +
+ +
+ +
+

Objetivos

+
    +
  • Aprenda sobre Pods do Kubernetes.
    • +
  • Aprenda sobre Nós do Kubernetes.
    • +
  • Solucionar problemas de aplicativos implantados no Kubernetes.
    • +
+
+ +
+

Kubernetes Pods

+

Quando você criou um Deployment no Módulo 2, o Kubernetes criou um Pod para hospedar a instância do seu aplicativo. Um Pod é uma abstração do Kubernetes que representa um grupo de um ou mais contêineres de aplicativos (como Docker) e alguns recursos compartilhados para esses contêineres. Esses recursos incluem:

+
    +
  • Armazenamento compartilhado, como Volumes
    • +
  • Rede, como um endereço IP único no cluster
    • +
  • Informações sobre como executar cada contêiner, como a versão da imagem do contêiner ou portas específicas a serem usadas
    • +
+

Um Pod define um "host lógico" específico para o aplicativo e pode conter diferentes contêineres que, na maioria dos casos, são fortemente acoplados. Por exemplo, um Pod pode incluir o contêiner com seu aplicativo Node.js, bem como um outro contêiner que alimenta os dados a serem publicados pelo servidor web Node.js. Os contêineres de um Pod compartilham um endereço IP e intervalo de portas; são sempre localizados, programados e executam em um contexto compartilhado no mesmo Nó.

+ +

Pods são a unidade atômica na plataforma Kubernetes. Quando criamos um Deployment no Kubernetes, esse Deployment cria Pods com contêineres dentro dele (em vez de você criar contêineres diretamente). Cada Pod está vinculado ao nó onde está programado (scheduled) e lá permanece até o encerramento (de acordo com a política de reinicialização) ou exclusão. Em caso de falha do nó, Pods idênticos são programados em outros nós disponíveis no cluster.

+ +
+
+
+

Sumário:

+
    +
  • Pods
    • +
  • Nós (Nodes)
    • +
  • Principais comandos do Kubectl
    • +
+
+
+

+ Um Pod é um grupo de um ou mais contêineres de aplicativos (como Docker) que inclui armazenamento compartilhado (volumes), endereço IP e informações sobre como executá-los. +

+
+
+
+
+ +
+
+

Visão geral sobre os Pods

+
+
+ +
+
+

+
+
+
+ +
+
+

Nós (Nodes)

+

Um Pod sempre será executando em um Nó. Um Nó é uma máquina de processamento em um cluster Kubernetes e pode ser uma máquina física ou virtual. Cada Nó é gerenciado pelo Control Plane. Um Nó pode possuir múltiplos Pods e o Control Plane do Kubernetes gerencia automaticamente o agendamento dos Pods nos nós do cluster. Para o agendamento automático dos Pods, o Control Plane leva em consideração os recursos disponíveis em cada Nó.

+ +

Cada Nó do Kubernetes executa pelo menos:

+
    +
  • Kubelet, o processo responsável pela comunicação entre o Control Plane e o Nó; gerencia os Pods e os contêineres rodando em uma máquina.
    • +
  • Um runtime de contêiner (por exemplo o Docker) é responsável por baixar a imagem do contêiner de um registro de imagens (por exemplo o Docker Hub), extrair o contêiner e executar a aplicação.
    • +
+ +
+
+
+

Os contêineres só devem ser agendados juntos em um único Pod se estiverem fortemente acoplados e precisarem compartilhar recursos, como disco e IP.

+
+
+
+ +
+ +
+
+

Visão Geral sobre os Nós

+
+
+ +
+
+

+
+
+
+ +
+
+

Solucionar problemas usando o comando kubectl

+

No Módulo 2, você usou o comando Kubectl. Você pode continuar utilizando o Kubectl no Módulo 3 para obter informação sobre Deployment realizado e seus recursos. As operações mais comuns podem ser realizadas com os comandos abaixo:

+
    +
  • kubectl get - listar recursos
    • +
  • kubectl describe - mostrar informações detalhadas sobre um recurso
    • +
  • kubectl logs - mostrar os logs de um container em um Pod
    • +
  • kubectl exec - executar um comando em um contêiner em um Pod
    • +
+ +

Você pode usar esses comandos para verificar quando o Deployment foi realizado, qual seu status atual, ondes os Pods estão rodando e qual são as suas configurações.

+ +

Agora que sabemos mais sobre os componentes de um cluster Kubernetes e o comando kubectl, vamos explorar a nossa aplicação.

+ +
+
+
+

Um nó é uma máquina operária do Kubernetes e pode ser uma VM ou máquina física, dependendo do cluster. Vários Pods podem ser executados em um nó.

+
+
+
+
+ +
+
+ Inciar o Tutorial Interativo› +
+
+ +
+ +
+ + + From 1077ed896f7aa6874c244159c1373b096c1aa255 Mon Sep 17 00:00:00 2001 From: Jailton Lopes Date: Fri, 19 Feb 2021 17:42:50 -0300 Subject: [PATCH 48/62] =?UTF-8?q?Add=20glossary=20tooltip=20for=20?= =?UTF-8?q?=E2=80=9CAPI=20server=E2=80=9D=20to=20Resource=20Quotas=20conce?= =?UTF-8?q?pt?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: Jailton Lopes --- content/en/docs/concepts/policy/resource-quotas.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/content/en/docs/concepts/policy/resource-quotas.md b/content/en/docs/concepts/policy/resource-quotas.md index d5c94d99cf..bc5eca8754 100644 --- a/content/en/docs/concepts/policy/resource-quotas.md +++ b/content/en/docs/concepts/policy/resource-quotas.md @@ -58,7 +58,7 @@ Neither contention nor changes to quota will affect already created resources. ## Enabling Resource Quota Resource Quota support is enabled by default for many Kubernetes distributions. It is -enabled when the API server `--enable-admission-plugins=` flag has `ResourceQuota` as +enabled when the {{< glossary_tooltip text="API server" term_id="kube-apiserver" >}} `--enable-admission-plugins=` flag has `ResourceQuota` as one of its arguments. A resource quota is enforced in a particular namespace when there is a From 27013dfbf5aaf006b3a3a2467bf8094dcf6a828f Mon Sep 17 00:00:00 2001 From: June Yi Date: Sat, 20 Feb 2021 12:13:36 +0900 Subject: [PATCH 49/62] Describe helper scripts for localization --- content/en/docs/contribute/localization.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/content/en/docs/contribute/localization.md b/content/en/docs/contribute/localization.md index 48c48b6ed3..2d89ee7e74 100644 --- a/content/en/docs/contribute/localization.md +++ b/content/en/docs/contribute/localization.md @@ -267,7 +267,7 @@ Teams must merge localized content into the same release branch from which the c An approver must maintain a development branch by keeping it current with its source branch and resolving merge conflicts. The longer a development branch stays open, the more maintenance it typically requires. Consider periodically merging development branches and opening new ones, rather than maintaining one extremely long-running development branch. -At the beginning of every team milestone, it's helpful to open an issue [comparing upstream changes](https://github.com/kubernetes/website/blob/master/scripts/upstream_changes.py) between the previous development branch and the current development branch. +At the beginning of every team milestone, it's helpful to open an issue comparing upstream changes between the previous development branch and the current development branch. There are two scripts for comparing upstream changes. [`upstream_changes.py`](https://github.com/kubernetes/website/tree/master/scripts#upstream_changespy) is useful for checking the changes made to a specific file. And [`diff_l10n_branches.py`](https://github.com/kubernetes/website/tree/master/scripts#diff_l10n_branchespy) is useful for creating a list of outdated files for a specific localization branch. While only approvers can open a new development branch and merge pull requests, anyone can open a pull request for a new development branch. No special permissions are required. From 9f4eda2569ef1847fc8c7ff21c68a590c775c103 Mon Sep 17 00:00:00 2001 From: Maciej Filocha Date: Sat, 20 Feb 2021 14:00:46 +0100 Subject: [PATCH 50/62] Synchronize Polish localization with upstream Polish translation synchronized up to 544afc899952bbcd025c82d3aca2f13525e47a85 --- content/pl/docs/concepts/overview/what-is-kubernetes.md | 2 +- content/pl/docs/reference/_index.md | 3 ++- content/pl/docs/reference/tools.md | 2 +- content/pl/docs/tutorials/_index.md | 2 +- content/pl/docs/tutorials/kubernetes-basics/_index.html | 2 +- .../kubernetes-basics/deploy-app/deploy-intro.html | 4 +--- .../tutorials/kubernetes-basics/expose/expose-intro.html | 6 ------ 7 files changed, 7 insertions(+), 14 deletions(-) diff --git a/content/pl/docs/concepts/overview/what-is-kubernetes.md b/content/pl/docs/concepts/overview/what-is-kubernetes.md index db8ea18b70..d28c841553 100644 --- a/content/pl/docs/concepts/overview/what-is-kubernetes.md +++ b/content/pl/docs/concepts/overview/what-is-kubernetes.md @@ -42,7 +42,7 @@ Kontenery działają w sposób zbliżony do maszyn wirtualnych, ale mają mniejs Kontenery zyskały popularność ze względu na swoje zalety, takie jak: * Szybkość i elastyczność w tworzeniu i instalacji aplikacji: obraz kontenera buduje się łatwiej niż obraz VM. -* Ułatwienie ciągłego rozwoju, integracji oraz wdrażania aplikacji (*Continuous development, integration, and deployment*): obrazy kontenerów mogą być budowane w sposób wiarygodny i częsty. Wycofanie zmian jest łatwe i szybkie (ponieważ obrazy są niezmienne). +* Ułatwienie ciągłego rozwoju, integracji oraz wdrażania aplikacji (*Continuous development, integration, and deployment*): obrazy kontenerów mogą być budowane w sposób wiarygodny i częsty. Wycofywanie zmian jest skuteczne i szybkie (ponieważ obrazy są niezmienne). * Rozdzielenie zadań *Dev* i *Ops*: obrazy kontenerów powstają w fazie *build/release*, oddzielając w ten sposób aplikacje od infrastruktury. * Obserwowalność obejmuje nie tylko informacje i metryki z poziomu systemu operacyjnego, ale także poprawność działania samej aplikacji i inne sygnały. * Spójność środowiska na etapach rozwoju oprogramowania, testowania i działania w trybie produkcyjnym: działa w ten sam sposób na laptopie i w chmurze. diff --git a/content/pl/docs/reference/_index.md b/content/pl/docs/reference/_index.md index cf461cf7f2..d598908162 100644 --- a/content/pl/docs/reference/_index.md +++ b/content/pl/docs/reference/_index.md @@ -14,7 +14,8 @@ Tutaj znajdziesz dokumentację źródłową Kubernetesa. ## Dokumentacja API -* [Dokumentacja źródłowa API Kubernetesa {{< param "version" >}}](/docs/reference/generated/kubernetes-api/{{< param "version" >}}/) +* [Kubernetes API Reference](/docs/reference/kubernetes-api/) +* [One-page API Reference for Kubernetes {{< param "version" >}}](/docs/reference/generated/kubernetes-api/{{< param "version" >}}/) * [Using The Kubernetes API](/docs/reference/using-api/) - ogólne informacje na temat API Kubernetesa. ## Biblioteki klientów API diff --git a/content/pl/docs/reference/tools.md b/content/pl/docs/reference/tools.md index 5d60370ee3..2ec66964ed 100644 --- a/content/pl/docs/reference/tools.md +++ b/content/pl/docs/reference/tools.md @@ -18,7 +18,7 @@ Kubernetes zawiera różne wbudowane narzędzia służące do pracy z systemem: ## Minikube -[`minikube`](https://minikube.sigs.k8s.io/docs/) to narzędzie do łatwego uruchamiania lokalnego klastra Kubernetes na twojej stacji roboczej na potrzeby rozwoju oprogramowania lub prowadzenia testów. +[`minikube`](https://minikube.sigs.k8s.io/docs/) to narzędzie do uruchamiania jednowęzłowego klastra Kubernetes na twojej stacji roboczej na potrzeby rozwoju oprogramowania lub prowadzenia testów. ## Pulpit *(Dashboard)* diff --git a/content/pl/docs/tutorials/_index.md b/content/pl/docs/tutorials/_index.md index e9f8ed32d8..c55fd9c3ff 100644 --- a/content/pl/docs/tutorials/_index.md +++ b/content/pl/docs/tutorials/_index.md @@ -32,7 +32,7 @@ Przed zapoznaniem się z samouczkami warto stworzyć zakładkę do * [Exposing an External IP Address to Access an Application in a Cluster](/docs/tutorials/stateless-application/expose-external-ip-address/) -* [Example: Deploying PHP Guestbook application with Redis](/docs/tutorials/stateless-application/guestbook/) +* [Example: Deploying PHP Guestbook application with MongoDB](/docs/tutorials/stateless-application/guestbook/) ## Aplikacje stanowe *(Stateful Applications)* diff --git a/content/pl/docs/tutorials/kubernetes-basics/_index.html b/content/pl/docs/tutorials/kubernetes-basics/_index.html index 39d8bf63c9..e27a3ad6bf 100644 --- a/content/pl/docs/tutorials/kubernetes-basics/_index.html +++ b/content/pl/docs/tutorials/kubernetes-basics/_index.html @@ -41,7 +41,7 @@ card:

Co Kubernetes może dla Ciebie zrobić?

-

Użytkownicy oczekują od współczesnych serwisów internetowych dostępności non-stop, a deweloperzy chcą móc instalować nowe wersje swoich serwisów kilka razy dziennie. Używając kontenerów można przygotowywać oprogramowanie w taki sposób, aby mogło być instalowane i aktualizowane łatwo i nie powodując żadnych przestojów. Kubernetes pomaga uruchamiać te aplikacje w kontenerach tam, gdzie chcesz i kiedy chcesz i znajdować niezbędne zasoby i narzędzia wymagane do ich pracy. Kubernetes może działać w środowiskach produkcyjnych, jest otwartym oprogramowaniem zaprojektowanym z wykorzystaniem nagromadzonego przez Google doświadczenia w zarządzaniu kontenerami, w połączeniu z najcenniejszymi ideami społeczności.

+

Użytkownicy oczekują od współczesnych serwisów internetowych dostępności non-stop, a deweloperzy chcą móc instalować nowe wersje swoich serwisów kilka razy dziennie. Używając kontenerów można przygotowywać oprogramowanie w taki sposób, aby mogło być instalowane i aktualizowane nie powodując żadnych przestojów. Kubernetes pomaga uruchamiać te aplikacje w kontenerach tam, gdzie chcesz i kiedy chcesz i znajdować niezbędne zasoby i narzędzia wymagane do ich pracy. Kubernetes może działać w środowiskach produkcyjnych, jest otwartym oprogramowaniem zaprojektowanym z wykorzystaniem nagromadzonego przez Google doświadczenia w zarządzaniu kontenerami, w połączeniu z najcenniejszymi ideami społeczności.

diff --git a/content/pl/docs/tutorials/kubernetes-basics/deploy-app/deploy-intro.html b/content/pl/docs/tutorials/kubernetes-basics/deploy-app/deploy-intro.html index 5b90420c5a..c879aa82b9 100644 --- a/content/pl/docs/tutorials/kubernetes-basics/deploy-app/deploy-intro.html +++ b/content/pl/docs/tutorials/kubernetes-basics/deploy-app/deploy-intro.html @@ -91,9 +91,7 @@ weight: 10

- Na potrzeby pierwszej instalacji użyjesz aplikacji na Node.js zapakowaną w kontener Docker-a. (Jeśli jeszcze nie próbowałeś stworzyć - aplikacji na Node.js i uruchomić za pomocą kontenerów, możesz spróbować teraz, kierując się instrukcjami samouczka - Hello Minikube). + Na potrzeby pierwszej instalacji użyjesz aplikacji hello-node zapakowaną w kontener Docker-a, która korzysta z NGINXa i powtarza wszystkie wysłane do niej zapytania. (Jeśli jeszcze nie próbowałeś stworzyć aplikacji hello-node i uruchomić za pomocą kontenerów, możesz spróbować teraz, kierując się instrukcjami samouczka Hello Minikube).

Teraz, kiedy wiesz, czym są Deploymenty, przejdźmy do samouczka online, żeby zainstalować naszą pierwszą aplikację!

diff --git a/content/pl/docs/tutorials/kubernetes-basics/expose/expose-intro.html b/content/pl/docs/tutorials/kubernetes-basics/expose/expose-intro.html index 4ad9a7be76..f9f9134e4a 100644 --- a/content/pl/docs/tutorials/kubernetes-basics/expose/expose-intro.html +++ b/content/pl/docs/tutorials/kubernetes-basics/expose/expose-intro.html @@ -64,12 +64,6 @@ weight: 10
-
-
-

-
-
-

Serwis kieruje przychodzący ruch do grupy Podów. Serwisy są obiektami abstrakcyjnymi, dzięki którym pody mogą się psuć i być zastępowane przez Kubernetes nowymi bez ujemnego wpływu na działanie twoich aplikacji. Detekcją nowych podów i kierowaniem ruchu pomiędzy zależnymi podami (takimi, jak składowe front-end i back-end w aplikacji) zajmują się Serwisy Kubernetes.

From f381ec8dc3ac60f5afa9d928d76fe17737ff6d25 Mon Sep 17 00:00:00 2001 From: Jailton Lopes Date: Sat, 20 Feb 2021 16:04:00 -0300 Subject: [PATCH 51/62] Update docs/tutorials/kubernetes-basics/public/images/module_01_cluster.svg to to use the "control plane" terminology Signed-off-by: Jailton Lopes --- .../public/images/module_01_cluster.svg | 633 +++++++++++++----- 1 file changed, 460 insertions(+), 173 deletions(-) diff --git a/content/en/docs/tutorials/kubernetes-basics/public/images/module_01_cluster.svg b/content/en/docs/tutorials/kubernetes-basics/public/images/module_01_cluster.svg index e1f92dace0..b183377467 100644 --- a/content/en/docs/tutorials/kubernetes-basics/public/images/module_01_cluster.svg +++ b/content/en/docs/tutorials/kubernetes-basics/public/images/module_01_cluster.svg @@ -1,6 +1,32 @@ - - - From 25a21cb87abd250e2f832ef6f5619ac7f270f1cb Mon Sep 17 00:00:00 2001 From: Jailton Lopes Date: Sat, 20 Feb 2021 16:17:31 -0300 Subject: [PATCH 52/62] Update content/en/docs/tutorials/kubernetes-basics/create-cluster/cluster-intro.html to to use the "control plane" terminology Signed-off-by: Jailton Lopes --- .../create-cluster/cluster-intro.html | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/content/en/docs/tutorials/kubernetes-basics/create-cluster/cluster-intro.html b/content/en/docs/tutorials/kubernetes-basics/create-cluster/cluster-intro.html index 5ac682d7af..47a2629feb 100644 --- a/content/en/docs/tutorials/kubernetes-basics/create-cluster/cluster-intro.html +++ b/content/en/docs/tutorials/kubernetes-basics/create-cluster/cluster-intro.html @@ -33,7 +33,7 @@ weight: 10

A Kubernetes cluster consists of two types of resources:

    -
  • The Master coordinates the cluster
    • +
  • The Control Plane coordinates the cluster
  • Nodes are the workers that run applications

@@ -71,22 +71,22 @@ weight: 10
-

The Master is responsible for managing the cluster. The master coordinates all activities in your cluster, such as scheduling applications, maintaining applications' desired state, scaling applications, and rolling out new updates.

-

A node is a VM or a physical computer that serves as a worker machine in a Kubernetes cluster. Each node has a Kubelet, which is an agent for managing the node and communicating with the Kubernetes master. The node should also have tools for handling container operations, such as containerd or Docker. A Kubernetes cluster that handles production traffic should have a minimum of three nodes.

+

The Control Plane is responsible for managing the cluster. The Control Plane coordinates all activities in your cluster, such as scheduling applications, maintaining applications' desired state, scaling applications, and rolling out new updates.

+

A node is a VM or a physical computer that serves as a worker machine in a Kubernetes cluster. Each node has a Kubelet, which is an agent for managing the node and communicating with the Kubernetes control plane. The node should also have tools for handling container operations, such as containerd or Docker. A Kubernetes cluster that handles production traffic should have a minimum of three nodes.

-

Masters manage the cluster and the nodes that are used to host the running applications.

+

Control Planes manage the cluster and the nodes that are used to host the running applications.

-

When you deploy applications on Kubernetes, you tell the master to start the application containers. The master schedules the containers to run on the cluster's nodes. The nodes communicate with the master using the Kubernetes API, which the master exposes. End users can also use the Kubernetes API directly to interact with the cluster.

+

When you deploy applications on Kubernetes, you tell the control plane to start the application containers. The control plane schedules the containers to run on the cluster's nodes. The nodes communicate with the control plane using the Kubernetes API, which the control plane exposes. End users can also use the Kubernetes API directly to interact with the cluster.

-

A Kubernetes cluster can be deployed on either physical or virtual machines. To get started with Kubernetes development, you can use Minikube. Minikube is a lightweight Kubernetes implementation that creates a VM on your local machine and deploys a simple cluster containing only one node. Minikube is available for Linux, macOS, and Windows systems. The Minikube CLI provides basic bootstrapping operations for working with your cluster, including start, stop, status, and delete. For this tutorial, however, you'll use a provided online terminal with Minikube pre-installed.

+

A Kubernetes cluster can be deployed on either physical or virtual machines. To get started with Kubernetes development, you can use Minikube. Minikube is a lightweight Kubernetes implementation that creates a VM on your local machine and deploys a simple cluster containing only one node. Minikube is available for Linux, macOS, and Windows systems. The Minikube CLI provides basic bootstrapping operations for working with your cluster, including start, stop, status, and delete. For this tutorial, however, you'll use a provided online terminal with Minikube pre-installed.

Now that you know what Kubernetes is, let's go to the online tutorial and start our first cluster!

From d042a1a687c7e801d5d15f1e4c11b0d812f43929 Mon Sep 17 00:00:00 2001 From: Will Hanstedt <39772580+Billy-The-Squid@users.noreply.github.com> Date: Sat, 20 Feb 2021 23:57:53 -0500 Subject: [PATCH 53/62] Fix broken link Attempt to resolve issue #26641. Redirected the "CoreOS' original article" link to an archived version of the article. --- content/en/docs/concepts/extend-kubernetes/operator.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/content/en/docs/concepts/extend-kubernetes/operator.md b/content/en/docs/concepts/extend-kubernetes/operator.md index 0e9d227a53..64b9ed3eb8 100644 --- a/content/en/docs/concepts/extend-kubernetes/operator.md +++ b/content/en/docs/concepts/extend-kubernetes/operator.md @@ -124,6 +124,6 @@ that can act as a [client for the Kubernetes API](/docs/reference/using-api/clie you implement yourself * using the [Operator Framework](https://operatorframework.io) * [Publish](https://operatorhub.io/) your operator for other people to use -* Read [CoreOS' original article](https://coreos.com/blog/introducing-operators.html) that introduced the Operator pattern +* Read [CoreOS' original article](https://web.archive.org/web/20170129131616/https://coreos.com/blog/introducing-operators.html) that introduced the Operator pattern (this is an archived version of the original article). * Read an [article](https://cloud.google.com/blog/products/containers-kubernetes/best-practices-for-building-kubernetes-operators-and-stateful-apps) from Google Cloud about best practices for building Operators From ad5bebe2aae2c208b6935445a92ff27a565bcfeb Mon Sep 17 00:00:00 2001 From: Tim Bannister Date: Sun, 21 Feb 2021 11:03:43 +0000 Subject: [PATCH 54/62] Add (stub) localization guide for Spanish (#24477) * Add (stub) localization guide for Spanish Co-authored-by: Victor Morales * docs: Add brief page introduction * docs: use archivo as discussed with the team Co-authored-by: Victor Morales Co-authored-by: Rael Garcia --- content/es/docs/contribute/_index.md | 4 ++ content/es/docs/contribute/localization_es.md | 43 +++++++++++++++++++ 2 files changed, 47 insertions(+) create mode 100644 content/es/docs/contribute/localization_es.md diff --git a/content/es/docs/contribute/_index.md b/content/es/docs/contribute/_index.md index 46157bab39..24851d8db0 100644 --- a/content/es/docs/contribute/_index.md +++ b/content/es/docs/contribute/_index.md @@ -83,3 +83,7 @@ para proporcionar un punto de partida. - Proponer mejoras al sitio web de Kubernetes y otras herramientas +## {{% heading "whatsnext" %}} + +También puedes leer la +[guía de localización para español](/es/docs/contribute/localization_es/). diff --git a/content/es/docs/contribute/localization_es.md b/content/es/docs/contribute/localization_es.md new file mode 100644 index 0000000000..22602d3ed9 --- /dev/null +++ b/content/es/docs/contribute/localization_es.md @@ -0,0 +1,43 @@ +--- +title: Contribuir a la documentación de Kubernetes en español +content_type: concept +--- + + + +¡Bienvenido(a)! + +En esta página encontrarás información sobre convenciones utilizadas en la documentación en castellano y un glosario de términos con sus traducciones. + + + +## Glosario de terminología {#terminologia} + +| English | Español | Género | Commentarios | +| ----------------- | ---------------------- | ------------ | ------------------------ | +| availability zone | zona de disponibilidad | femenino | | +| bearer token | bearer token | masculino | | +| built-in | incorporados | masculino | | +| conditions | condiciones | masculino | para node conditions | +| container | contenedor | masculino | | +| controller | controlador | masculino | | +| deploy | desplegar | | | +| Deployment | Deployment | masculino | objeto Kubernetes | +| Endpoints | Endpoints | masculino | objeto Kubernetes | +| file | archivo | masculino | | +| frontend | frontend | masculino | | +| healthy | operativo | | | +| high availability | alta disponibilidad | | | +| hook | hook | masculino | | +| instance | instancia | femenino | | +| Lease | Lease | masculino | objeto Kubernetes | +| Pod | Pod | masculino | objeto Kubernetes | +| ratio | ritmo | | | +| runtime | motor de ejecución | masculino | Container Runtime | +| scheduler | planificador | masculino | | +| Secret | Secret | masculino | objeto Kubernetes | +| secret | secreto | masculino | información confidencial | +| shell | terminal | femenino | | +| stateless | stateless | | | +| taint | contaminación | | | +| worker node | nodo de trabajo | masculino | | From 0f277a5c87a11d0e61667deaac18b4bf63529161 Mon Sep 17 00:00:00 2001 From: seokho-son Date: Mon, 22 Feb 2021 19:43:11 +0900 Subject: [PATCH 55/62] ko: fix typo in cloud-controller-manager --- .../ko/docs/reference/glossary/cloud-controller-manager.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/content/ko/docs/reference/glossary/cloud-controller-manager.md b/content/ko/docs/reference/glossary/cloud-controller-manager.md index 20121a9371..ebfa3d926c 100644 --- a/content/ko/docs/reference/glossary/cloud-controller-manager.md +++ b/content/ko/docs/reference/glossary/cloud-controller-manager.md @@ -5,7 +5,7 @@ date: 2018-04-12 full_link: /ko/docs/concepts/architecture/cloud-controller/ short_description: > 쿠버네티스를 타사 클라우드 공급자와 통합하는 컨트롤 플레인 컴포넌트. -aka: +aka: tags: - core-object - architecture @@ -13,7 +13,7 @@ tags: --- 클라우드별 컨트롤 로직을 포함하는 쿠버네티스 {{< glossary_tooltip text="컨트롤 플레인" term_id="control-plane" >}} 컴포넌트이다. -클라우트 컨트롤러 매니저를 통해 클러스터를 클라우드 공급자의 API에 연결하고, +클라우드 컨트롤러 매니저를 통해 클러스터를 클라우드 공급자의 API에 연결하고, 해당 클라우드 플랫폼과 상호 작용하는 컴포넌트와 클러스터와 상호 작용하는 컴포넌트를 분리할 수 있다. From 7373ebbb38254cd97ac00e99711cb4d2a1a76592 Mon Sep 17 00:00:00 2001 From: Guilherme Zanelato Date: Sat, 20 Feb 2021 07:57:14 -0300 Subject: [PATCH 56/62] Add content/pt/docs/concepts/configuration/organize-cluster-access-kubeconfig.md --- .../organize-cluster-access-kubeconfig.md | 131 ++++++++++++++++++ 1 file changed, 131 insertions(+) create mode 100644 content/pt/docs/concepts/configuration/organize-cluster-access-kubeconfig.md diff --git a/content/pt/docs/concepts/configuration/organize-cluster-access-kubeconfig.md b/content/pt/docs/concepts/configuration/organize-cluster-access-kubeconfig.md new file mode 100644 index 0000000000..4b431b486f --- /dev/null +++ b/content/pt/docs/concepts/configuration/organize-cluster-access-kubeconfig.md @@ -0,0 +1,131 @@ +--- +title: Organizando o acesso ao cluster usando arquivos kubeconfig +content_type: concept +weight: 60 +--- + + + +Utilize arquivos kubeconfig para organizar informações sobre clusters, usuários, namespaces e mecanismos de autenticação. A ferramenta de linha de comando `kubectl` faz uso dos arquivos kubeconfig para encontrar as informações necessárias para escolher e se comunicar com o serviço de API de um cluster. + + +{{< note >}} +Um arquivo que é utilizado para configurar o acesso aos clusters é chamado de *kubeconfig*. Esta á uma forma genérica de referenciamento para um arquivo de configuração desta natureza. Isso não significa que existe um arquivo com o nome `kubeconfig`. +{{< /note >}} + +Por padrão, o `kubectl` procura por um arquivo de nome `config` no diretório `$HOME/.kube` + +Você pode especificar outros arquivos kubeconfig através da variável de ambiente `KUBECONFIG` ou adicionando a opção [`--kubeconfig`](/docs/reference/generated/kubectl/kubectl/). + +Para maiores detalhes na criação e especificação de um kubeconfig, veja o passo a passo em [Configurar Acesso para Múltiplos Clusters](/docs/tasks/access-application-cluster/configure-access-multiple-clusters). + + + + +## Suportando múltiplos clusters, usuários e mecanismos de autenticação + +Imagine que você possua inúmeros clusters, e seus usuários e componentes se autenticam de várias formas. Por exemplo: + +- Um kubelet ativo pode se autenticar utilizando certificados +- Um usuário pode se autenticar através de tokens +- Administradores podem possuir conjuntos de certificados os quais provém acesso aos usuários de forma individual. + +Através de arquivos kubeconfig, você pode organizar os seus clusters, usuários, e namespaces. Você também pode definir contextos para uma fácil troca entre clusters e namespaces. + + +## Contexto + +Um elemento de *contexto* em um kubeconfig é utilizado para agrupar parâmetros de acesso em um nome conveniente. Cada contexto possui três parâmetros: cluster, namespace, e usuário. + +Por padrão, a ferramenta de linha de comando `kubectl` utiliza os parâmetros do _contexto atual_ para se comunicar com o cluster. + +Para escolher o contexto atual: + +```shell +kubectl config use-context +``` + +## A variável de ambiente KUBECONFIG + +A variável de ambiente `KUBECONFIG` possui uma lista dos arquivos kubeconfig. Para Linux e Mac, esta lista é delimitada por vírgula. No Windows, a lista é delimitada por ponto e vírgula. A variável de ambiente `KUBECONFIG` não é um requisito obrigatório - caso ela não exista o `kubectl` utilizará o arquivo kubeconfig padrão localizado no caminho `$HOME/.kube/config`. + +Se a variável de ambiente `KUBECONFIG` existir, o `kubectl` utilizará uma configuração que é o resultado da combinação dos arquivos listados na variável de ambiente `KUBECONFIG`. + +## Combinando arquivos kubeconfig + +Para inspecionar a sua configuração atual, execute o seguinte comando: + +```shell +kubectl config view +``` + +Como descrito anteriormente, a saída poderá ser resultado de um único arquivo kubeconfig, ou poderá ser o resultado da junção de vários arquivos kubeconfig. + +Aqui estão as regras que o `kubectl` utiliza quando realiza a combinação de arquivos kubeconfig: + +1. Se o argumento `--kubeconfig` está definido, apenas o arquivo especificado será utilizado. Apenas uma instância desta flag é permitida. + + Caso contrário, se a variável de ambiente `KUBECONFIG` estiver definida, esta deverá ser utilizada como uma lista de arquivos a serem combinados, seguindo o fluxo a seguir: + + * Ignorar arquivos vazios. + * Produzir erros para aquivos cujo conteúdo não for possível desserializar. + * O primeiro arquivo que definir um valor ou mapear uma chave determinada, será o escolhido. + * Nunca modificar um valor ou mapear uma chave. + Exemplo: Preservar o contexto do primeiro arquivo que definir `current-context`. + Exemplo: Se dois arquivos especificarem um `red-user`, use apenas os valores do primeiro `red-user`. Mesmo se um segundo arquivo possuir entradas não conflitantes sobre a mesma entrada `red-user`, estas deverão ser descartadas. + + Para um exemplo de definição da variável de ambiente `KUBECONFIG` veja [Definido a variável de ambiente KUBECONFIG](/docs/tasks/access-application-cluster/configure-access-multiple-clusters/#set-the-kubeconfig-environment-variable). + + Caso contrário, utilize o arquivo kubeconfig padrão encontrado no diretório `$HOME/.kube/config`, sem qualquer tipo de combinação. + +1. Determine o contexto a ser utilizado baseado no primeiro padrão encontrado, nesta ordem: + + 1. Usar o conteúdo da flag `--context` caso ela existir. + 1. Usar o `current-context` a partir da combinação dos arquivos kubeconfig. + + + Um contexto vazio é permitido neste momento. + + +1. Determinar o cluster e o usuário. Neste ponto, poderá ou não existir um contexto. + Determinar o cluster e o usuário no primeiro padrão encontrado de acordo com a ordem à seguir. Este procedimento deverá executado duas vezes: uma para definir o usuário a outra para definir o cluster. + + 1. Utilizar a flag caso ela existir: `--user` ou `--cluster`. + 1. Se o contexto não estiver vazio, utilizar o cluster ou usuário deste contexto. + + O usuário e o cluster poderão estar vazios neste ponto. + +1. Determinar as informações do cluster atual a serem utilizadas. Neste ponto, poderá ou não existir informações de um cluster. + + Construir cada peça de informação do cluster baseado nas opções à seguir; a primeira ocorrência encontrada será a opção vencedora: + + 1. Usar as flags de linha de comando caso existirem: `--server`, `--certificate-authority`, `--insecure-skip-tls-verify`. + 1. Se algum atributo do cluster existir a partir da combinação de kubeconfigs, estes deverão ser utilizados. + 1. Se não existir informação de localização do servidor falhar. + +1. Determinar a informação atual de usuário a ser utilizada. Construir a informação de usuário utilizando as mesmas regras utilizadas para o caso de informações de cluster, exceto para a regra de técnica de autenticação que deverá ser única por usuário: + + 1. Usar as flags, caso existirem: `--client-certificate`, `--client-key`, `--username`, `--password`, `--token`. + 1. Usar os campos `user` resultado da combinação de arquivos kubeconfig. + 1. Se existirem duas técnicas conflitantes, falhar. + +1. Para qualquer informação que ainda estiver ausente, utilizar os valores padrão e potencialmente solicitar informações de autenticação a partir do prompt de comando. + + +## Referências de arquivos + +Arquivos e caminhos referenciados em um arquivo kubeconfig são relativos à localização do arquivo kubeconfig. + +Referências de arquivos na linha de comando são relativas ao diretório de trabalho vigente. + +No arquivo `$HOME/.kube/config`, caminhos relativos são armazenados de forma relativa, e caminhos absolutos são armazenados de forma absoluta. + +## {{% heading "whatsnext" %}} + + +* [Configurar Accesso para Multiplos Clusters](/docs/tasks/access-application-cluster/configure-access-multiple-clusters/) +* [`kubectl config`](/docs/reference/generated/kubectl/kubectl-commands#config) + + + + From 13379c974cf079296c5e7f0f7d86f586b771529a Mon Sep 17 00:00:00 2001 From: BITLIU Date: Tue, 23 Feb 2021 21:37:58 +0800 Subject: [PATCH 57/62] Fix Markdown Error --- content/zh/docs/concepts/architecture/nodes.md | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/content/zh/docs/concepts/architecture/nodes.md b/content/zh/docs/concepts/architecture/nodes.md index 7c6000268f..dab5e341ff 100644 --- a/content/zh/docs/concepts/architecture/nodes.md +++ b/content/zh/docs/concepts/architecture/nodes.md @@ -457,8 +457,7 @@ of the node heartbeats as the cluster scales. #### 心跳机制 {#heartbeats} Kubernetes 节点发送的心跳（Heartbeats）有助于确定节点的可用性。 -心跳有两种形式：`NodeStatus` 和 [`Lease` 对象] -(/docs/reference/generated/kubernetes-api/{{< param "version" >}}/#lease-v1-coordination-k8s-io)。 +心跳有两种形式：`NodeStatus` 和 [`Lease` 对象](/docs/reference/generated/kubernetes-api/{{< param "version" >}}/#lease-v1-coordination-k8s-io)。 每个节点在 `kube-node-lease`{{< glossary_tooltip term_id="namespace" text="名字空间">}} 中都有一个与之关联的 `Lease` 对象。 `Lease` 是一种轻量级的资源，可在集群规模扩大时提高节点心跳机制的性能。 From 0e2aea53be53539a1124cebaafb37ae5ba7a1372 Mon Sep 17 00:00:00 2001 From: Dominic Yin Date: Wed, 24 Feb 2021 14:26:18 +0800 Subject: [PATCH 58/62] remove period at end of a list line #25656 --- content/zh/docs/concepts/overview/components.md | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/content/zh/docs/concepts/overview/components.md b/content/zh/docs/concepts/overview/components.md index 1e42cdcbaa..090468282a 100644 --- a/content/zh/docs/concepts/overview/components.md +++ b/content/zh/docs/concepts/overview/components.md @@ -92,10 +92,10 @@ These controllers include: --> 这些控制器包括: -* 节点控制器（Node Controller）: 负责在节点出现故障时进行通知和响应。 -* 副本控制器（Replication Controller）: 负责为系统中的每个副本控制器对象维护正确数量的 Pod。 -* 端点控制器（Endpoints Controller）: 填充端点(Endpoints)对象(即加入 Service 与 Pod)。 -* 服务帐户和令牌控制器（Service Account & Token Controllers）: 为新的命名空间创建默认帐户和 API 访问令牌. +* 节点控制器（Node Controller）: 负责在节点出现故障时进行通知和响应 +* 副本控制器（Replication Controller）: 负责为系统中的每个副本控制器对象维护正确数量的 Pod +* 端点控制器（Endpoints Controller）: 填充端点(Endpoints)对象(即加入 Service 与 Pod) +* 服务帐户和令牌控制器（Service Account & Token Controllers）: 为新的命名空间创建默认帐户和 API 访问令牌 Kubernetes 通过将容器放入在节点（Node）上运行的 Pod 中来执行你的工作负载。 -节点可以是一个虚拟机或者物理机器，取决于所在的集群配置。每个节点由 -{{< glossary_tooltip text="控制面" term_id="control-plane" >}} 负责管理， -并包含运行 {{< glossary_tooltip text="Pods" term_id="pod" >}} 所需的服务。 +节点可以是一个虚拟机或者物理机器，取决于所在的集群配置。 +每个节点包含运行 {{< glossary_tooltip text="Pods" term_id="pod" >}} 所需的服务， +这些 Pods 由 {{< glossary_tooltip text="控制面" term_id="control-plane" >}} 负责管理。 通常集群中会有若干个节点；而在一个学习用或者资源受限的环境中，你的集群中也可能 只有一个节点。 @@ -121,7 +120,7 @@ register itself with the API server. This is the preferred pattern, used by mos For self-registration, the kubelet is started with the following options: --> -### 节点自注册 +### 节点自注册 {#self-registration-of-nodes} 当 kubelet 标志 `--register-node` 为 true（默认）时，它会尝试向 API 服务注册自己。 这是首选模式，被绝大多数发行版选用。 @@ -171,7 +170,7 @@ When you want to create Node objects manually, set the kubelet flag `--register- You can modify Node objects regardless of the setting of `--register-node`. For example, you can set labels on an existing Node, or mark it unschedulable. --> -### 手动节点管理 +### 手动节点管理 {#manual-node-administration} 你可以使用 {{< glossary_tooltip text="kubectl" term_id="kubectl" >}} 来创建和修改 Node 对象。 From 6b092d754823a03576abb8aa73b19ece989a8e9e Mon Sep 17 00:00:00 2001 From: BITLIU Date: Wed, 24 Feb 2021 18:10:28 +0800 Subject: [PATCH 60/62] Update deprecated URL update https://www.haproxy.org/download/1.5/doc/proxy-protocol.txt to https://www.haproxy.org/download/1.8/doc/proxy-protocol.txt --- content/en/docs/tutorials/services/source-ip.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/content/en/docs/tutorials/services/source-ip.md b/content/en/docs/tutorials/services/source-ip.md index 2d938f3031..4ed574e158 100644 --- a/content/en/docs/tutorials/services/source-ip.md +++ b/content/en/docs/tutorials/services/source-ip.md @@ -412,7 +412,7 @@ protocol between the loadbalancer and backend to communicate the true client IP such as the HTTP [Forwarded](https://tools.ietf.org/html/rfc7239#section-5.2) or [X-FORWARDED-FOR](https://en.wikipedia.org/wiki/X-Forwarded-For) headers, or the -[proxy protocol](https://www.haproxy.org/download/1.5/doc/proxy-protocol.txt). +[proxy protocol](https://www.haproxy.org/download/1.8/doc/proxy-protocol.txt). Load balancers in the second category can leverage the feature described above by creating an HTTP health check pointing at the port stored in the `service.spec.healthCheckNodePort` field on the Service. From 3a7c478d65604662d174519a2b55070bf49144e4 Mon Sep 17 00:00:00 2001 From: BITLIU Date: Wed, 24 Feb 2021 18:39:13 +0800 Subject: [PATCH 61/62] Fix Helm install instruction URL https://github.com/kubernetes/helm/blob/master/docs/install.md is deprecated --- .../tasks/service-catalog/install-service-catalog-using-helm.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/content/zh/docs/tasks/service-catalog/install-service-catalog-using-helm.md b/content/zh/docs/tasks/service-catalog/install-service-catalog-using-helm.md index ecc001e62d..66f0dd1e1e 100644 --- a/content/zh/docs/tasks/service-catalog/install-service-catalog-using-helm.md +++ b/content/zh/docs/tasks/service-catalog/install-service-catalog-using-helm.md @@ -39,7 +39,7 @@ Use [Helm](https://helm.sh/) to install Service Catalog on your Kubernetes clust * 如果你正在使用 `hack/local-up-cluster.sh`，请确保设置了 `KUBE_ENABLE_CLUSTER_DNS` 环境变量，然后运行安装脚本。 * [安装和设置 v1.7 或更高版本的 kubectl](/zh/docs/tasks/tools/install-kubectl/)，确保将其配置为连接到 Kubernetes 集群。 * 安装 v2.7.0 或更高版本的 [Helm](https://helm.sh/)。 - * 遵照 [Helm 安装说明](https://github.com/kubernetes/helm/blob/master/docs/install.md)。 + * 遵照 [Helm 安装说明](https://helm.sh/docs/intro/install/)。 * 如果已经安装了适当版本的 Helm，请执行 `helm init` 来安装 Helm 的服务器端组件 Tiller。 From fd227a8405149614ca89c95d1293b9195ee2bfff Mon Sep 17 00:00:00 2001 From: Ricardo Katz Date: Thu, 25 Feb 2021 10:49:24 -0300 Subject: [PATCH 62/62] Add EndpointSlice blog post in pt language (#26708) * Add EndpointSlice blog post in pt language * Apply corrections into endpointslice blog post translation Co-authored-by: Jhon Mike Co-authored-by: Jhon Mike --- ...ng-kubernetes-networking-endpointslices.md | 47 +++++++++++++++++++ 1 file changed, 47 insertions(+) create mode 100644 content/pt/blog/_posts/2020-09-02-scaling-kubernetes-networking-endpointslices.md diff --git a/content/pt/blog/_posts/2020-09-02-scaling-kubernetes-networking-endpointslices.md b/content/pt/blog/_posts/2020-09-02-scaling-kubernetes-networking-endpointslices.md new file mode 100644 index 0000000000..7440689d5f --- /dev/null +++ b/content/pt/blog/_posts/2020-09-02-scaling-kubernetes-networking-endpointslices.md @@ -0,0 +1,47 @@ +--- +layout: blog +title: 'Escalando a rede do Kubernetes com EndpointSlices' +date: 2020-09-02 +slug: scaling-kubernetes-networking-with-endpointslices +--- + +**Autor:** Rob Scott (Google) + +EndpointSlices é um novo tipo de API que provê uma alternativa escalável e extensível à API de Endpoints. EndpointSlices mantém o rastreio dos endereços IP, portas, informações de topologia e prontidão de Pods que compõem um serviço. + +No Kubernetes 1.19 essa funcionalidade está habilitada por padrão, com o kube-proxy lendo os [EndpointSlices](/docs/concepts/services-networking/endpoint-slices/) ao invés de Endpoints. Apesar de isso ser uma mudança praticamente transparente, resulta numa melhoria notável de escalabilidade em grandes clusters. Também permite a adição de novas funcionalidades em releases futuras do Kubernetes, como o [Roteamento baseado em topologia.](/docs/concepts/services-networking/service-topology/). + +## Limitações de escalabilidade da API de Endpoints +Na API de Endpoints, existia apenas um recurso de Endpoint por serviço (Service). Isso significa que +era necessário ser possível armazenar endereços IPs e portas para cada Pod que compunha o serviço correspondente. Isso resultava em recursos imensos de API. Para piorar, o kube-proxy rodava em cada um dos nós e observava qualquer alteração nos recursos de Endpoint. Mesmo que fosse uma simples mudança em um Endpoint, todo o objeto precisava ser enviado para cada uma das instâncias do kube-proxy. + +Outra limitação da API de Endpoints era que ela limitava o número de objetos que podiam ser associados a um _Service_. O tamanho padrão de um objeto armazenado no etcd é 1.5MB. Em alguns casos, isso poderia limitar um Endpoint a 5,000 IPs de Pod. Isso não chega a ser um problema para a maioria dos usuários, mas torna-se um problema significativo para serviços que se aproximem desse tamanho. + +Para demonstrar o quão significante se torna esse problema em grande escala, vamos usar de um simples exemplo: Imagine um _Service_ que possua 5,000 Pods, e que possa causar o Endpoint a ter 1.5Mb . Se apenas um Endpoint nessa lista sofra uma alteração, todo o objeto de Endpoint precisará ser redistribuído para cada um dos nós do cluster. Em um cluster com 3.000 nós, essa atualização causará o envio de 4.5Gb de dados (1.5Mb de Endpoints * 3,000 nós) para todo o cluster. Isso é quase que o suficiente para encher um DVD, e acontecerá para cada mudança de Endpoint. Agora imagine uma atualização gradual em um _Deployment_ que resulte nos 5,000 Pods serem substituídos - isso é mais que 22Tb (ou 5,000 DVDs) de dados transferidos. + +## Dividindo os endpoints com a API de EndpointSlice +A API de EndpointSlice foi desenhada para resolver esse problema com um modelo similar de _sharding_. Ao invés de rastrar todos os IPs dos Pods para um _Service_, com um único recurso de Endpoint, nós dividimos eles em múltiplos EndpointSlices menores. + +Usemos por exemplo um serviço com 15 pods. Nós teríamos um único recurso de Endpoints referente a todos eles. Se o EndpointSlices for configurado para armazenar 5 _endpoints_ cada, nós teríamos 3 EndpointSlices diferentes: +![EndpointSlices](/images/blog/2020-09-02-scaling-kubernetes-networking-endpointslices/endpoint-slices.png) + +Por padrão, o EndpointSlices armazena um máximo de 100 _endpoints_ cada, podendo isso ser configurado com a flag `--max-endpoints-per-slice` no kube-controller-manager. + +## EndpointSlices provê uma melhoria de escalabilidade em 10x +Essa API melhora dramaticamente a escalabilidade da rede. Agora quando um Pod é adicionado ou removido, apenas 1 pequeno EndpointSlice necessita ser atualizado. Essa diferença começa a ser notada quando centenas ou milhares de Pods compõem um único _Service_. + +Mais significativo, agora que todos os IPs de Pods para um _Service_ não precisam ser armazenados em um único recurso, nós não precisamos nos preocupar com o limite de tamanho para objetos armazendos no etcd. EndpointSlices já foram utilizados para escalar um serviço além de 100,000 endpoints de rede. + +Tudo isso é possível com uma melhoria significativa de performance feita no kube-proxy. Quando o EndpointSlices é usado em grande escala, muito menos dados serão transferidos para as atualizações de endpoints e o kube-proxy torna-se mais rápido para atualizar regras do iptables ou do ipvs. Além disso, os _Services_ podem escalar agora para pelo menos 10x mais além dos limites anteriores. + +## EndpointSlices permitem novas funcionalidades +Introduzido como uma funcionalidade alpha no Kubernetes v1.16, os EndpointSlices foram construídos para permitir algumas novas funcionalidades arrebatadoras em futuras versões do Kubernetes. Isso inclui serviços dual-stack, roteamento baseado em topologia e subconjuntos de _endpoints_. + +Serviços Dual-stack são uma nova funcionalidade que foi desenvolvida juntamente com o EndpointSlices. Eles irão utilizar simultâneamente endereços IPv4 e IPv6 para serviços, e dependem do campo addressType do Endpointslices para conter esses novos tipos de endereço por família de IP. + +O roteamento baseado por topologia irá atualizar o kube-proxy para dar preferência no roteamento de requisições para a mesma região ou zona, utilizando-se de campos de topologia armazenados em cada endpoint dentro de um EndpointSlice. Como uma melhoria futura disso, estamos explorando o potencial de subconjuntos de endpoint. Isso irá permitir o kube-proxy apenas observar um subconjunto de EndpointSlices. Por exemplo, isso pode ser combinado com o roteamento baseado em topologia e assim, o kube-proxy precisará observar apenas EndpointSlices contendo _endpoints_ na mesma zona. Isso irá permitir uma outra melhoria significativa de escalabilidade. + +## O que isso significa para a API de Endpoints? +Apesar da API de EndpointSlice prover uma alternativa nova e escalável à API de Endpoints, a API de Endpoints continuará a ser considerada uma funcionalidade estável. A mudança mais significativa para a API de Endpoints envolve começar a truncar Endpoints que podem causar problemas de escalabilidade. + +A API de Endpoints não será removida, mas muitas novas funcionalidades irão depender da nova API EndpointSlice. Para obter vantágem da funcionalidade e escalabilidade que os EndpointSlices provém, aplicações que hoje consomem a API de Endpoints devem considerar suportar EndpointSlices no futuro.