whentimeslows/website
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Update device plugin documentation + related pages (#14331)

Browse Source 
* Update device plugin docs for Kubernetes 1.14

& related pages

This change is mainly about feature-state shortcodes & similar.

(if there's a way to get the short version string, eg "1.14" rather
than "v1.14", then that could go in place of the hard coded value
in this commit).

* Fix code block formatting

eg code blocks incorrectly marked as:
```shell

* Use glossary shortcodes where appropriate

* Hyperlink to Prometheus

* Tidy Markdown formatting

* Change example vendor domain name

Use a name inside ".example" to highlight that this is a DNS
domain name.

* Reword device plugins documentation

* Tweak headings for device plugins

* Add "what's next" to device plugins docs

* Tweak wording for device plugins docs

* Add KubeVirt device plugins
This commit is contained in:
Tim Bannister
2019-08-07 00:00:52 +01:00
committed by Kubernetes Prow Robot
parent f2c7919fc9
commit 322e88de9a
5 changed files with 133 additions and 113 deletions
Download Patch File Download Diff File Expand all files Collapse all files
content/en/docs/concepts/extend-kubernetes/compute-storage-net/device-plugins.md
+75 -51
View File
@@ -6,31 +6,33 @@ content_template: templates/concept
weight: 20
---
{{< feature-state state="beta" >}}
{{% capture overview %}}
Starting in version 1.8, Kubernetes provides a
[device plugin framework](https://github.com/kubernetes/community/blob/master/contributors/design-proposals/resource-management/device-plugin.md)
for vendors to advertise their resources to the kubelet without changing Kubernetes core code.
Instead of writing custom Kubernetes code, vendors can implement a device plugin that can
be deployed manually or as a DaemonSet. The targeted devices include GPUs,
High-performance NICs, FPGAs, InfiniBand, and other similar computing resources
that may require vendor specific initialization and setup.
{{< feature-state for_k8s_version="v1.10" state="beta" >}}
Kubernetes provides a [device plugin framework](https://github.com/kubernetes/community/blob/master/contributors/design-proposals/resource-management/device-plugin.md)
that you can use to advertise system hardware resources to the
{{< glossary_tooltip term_id="kubelet" >}}.
Instead of customising the code for Kubernetes itself, vendors can implement a
device plugin that you deploy either manually or as a {{< glossary_tooltip term_id="daemonset" >}}.
The targeted devices include GPUs, high-performance NICs, FPGAs, InfiniBand adapters,
and other similar computing resources that may require vendor specific initialization
and setup.
{{% /capture %}}
{{% capture body %}}
## Device plugin registration
The device plugins feature is gated by the `DevicePlugins` feature gate which
is disabled by default before 1.10. When the device plugins feature is enabled,
the kubelet exports a `Registration` gRPC service:
The kubelet exports a `Registration` gRPC service:
```gRPC
service Registration {
rpc Register(RegisterRequest) returns (Empty) {}
}
```
A device plugin can register itself with the kubelet through this gRPC service.
During the registration, the device plugin needs to send:
@@ -38,15 +40,15 @@ During the registration, the device plugin needs to send:
* The Device Plugin API version against which it was built.
* The `ResourceName` it wants to advertise. Here `ResourceName` needs to follow the
[extended resource naming scheme](/docs/concepts/configuration/manage-compute-resources-container/#extended-resources)
as `vendor-domain/resource`.
For example, an Nvidia GPU is advertised as `nvidia.com/gpu`.
as `vendor-domain/resourcetype`.
(For example, an NVIDIA GPU is advertised as `nvidia.com/gpu`.)
Following a successful registration, the device plugin sends the kubelet the
list of devices it manages, and the kubelet is then in charge of advertising those
resources to the API server as part of the kubelet node status update.
For example, after a device plugin registers `vendor-domain/foo` with the kubelet
For example, after a device plugin registers `hardware-vendor.example/foo` with the kubelet
and reports two healthy devices on a node, the node status is updated
to advertise 2 `vendor-domain/foo`.
to advertise that the node has 2 “Foo” devices installed and available.
Then, users can request devices in a
[Container](/docs/reference/generated/kubernetes-api/{{< param "version" >}}/#container-v1-core)
@@ -55,10 +57,11 @@ specification as they request other types of resources, with the following limit
* Extended resources are only supported as integer resources and cannot be overcommitted.
* Devices cannot be shared among Containers.
Suppose a Kubernetes cluster is running a device plugin that advertises resource `vendor-domain/resource`
on certain nodes, here is an example user pod requesting this resource:
Suppose a Kubernetes cluster is running a device plugin that advertises resource `hardware-vendor.example/foo`
on certain nodes. Here is an example of a pod requesting this resource to run a demo workload:
```yaml
---
apiVersion: v1
kind: Pod
metadata:
@@ -69,7 +72,14 @@ spec:
image: k8s.gcr.io/pause:2.0
resources:
limits:
vendor-domain/resource: 2 # requesting 2 vendor-domain/resource
hardware-vendor.example/foo: 2
#
# This Pod needs 2 of the hardware-vendor.example/foo devices
# and can only schedule onto a Node that's able to satisfy
# that need.
#
# If the Node has more than 2 of those devices available, the
# remainder would be available for other Pods to use.
```
## Device plugin implementation
@@ -107,6 +117,8 @@ If the operations succeed, the device plugin returns an `AllocateResponse` that
runtime configurations for accessing the allocated devices. The kubelet passes this information
to the container runtime.
### Handling kubelet restarts
A device plugin is expected to detect kubelet restarts and re-register itself with the new
kubelet instance. In the current implementation, a new kubelet instance deletes all the existing Unix sockets
under `/var/lib/kubelet/device-plugins` when it starts. A device plugin can monitor the deletion
@@ -114,37 +126,44 @@ of its Unix socket and re-register itself upon such an event.
## Device plugin deployment
A device plugin can be deployed manually or as a DaemonSet. Being deployed as a DaemonSet has
the benefit that Kubernetes can restart the device plugin if it fails.
Otherwise, an extra mechanism is needed to recover from device plugin failures.
You can deploy a device plugin as a DaemonSet, as a package for your node's operating system,
or manually.
The canonical directory `/var/lib/kubelet/device-plugins` requires privileged access,
so a device plugin must run in a privileged security context.
If a device plugin is running as a DaemonSet, `/var/lib/kubelet/device-plugins`
must be mounted as a
[Volume](/docs/reference/generated/kubernetes-api/{{< param "version" >}}/#volume-v1-core)
If you're deploying a device plugin as a DaemonSet, `/var/lib/kubelet/device-plugins`
must be mounted as a {{< glossary_tooltip term_id="volume" >}}
in the plugin's
[PodSpec](/docs/reference/generated/kubernetes-api/{{< param "version" >}}/#podspec-v1-core).
Kubernetes device plugin support is in beta. As development continues, its API version can
change. We recommend that device plugin developers do the following:
If you choose the DaemonSet approach you can rely on Kubernetes to: place the device plugin's
Pod onto Nodes, to restart the daemon Pod after failure, and to help automate upgrades.
## API compatibility
Kubernetes device plugin support is in beta. The API may change before stabilization,
in incompatible ways. As a project, Kubernetes recommends that device plugin developers:
* Watch for changes in future releases.
* Support multiple versions of the device plugin API for backward/forward compatibility.
If you enable the DevicePlugins feature and run device plugins on nodes that need to be upgraded to
a Kubernetes release with a newer device plugin API version, upgrade your device plugins
to support both versions before upgrading these nodes to
to support both versions before upgrading these nodes. Taking that approach will
ensure the continuous functioning of the device allocations during the upgrade.
## Monitoring Device Plugin Resources
In order to monitor resources provided by device plugins, monitoring agents need to be able to
discover the set of devices that are in-use on the node and obtain metadata to describe which
container the metric should be associated with. Prometheus metrics exposed by device monitoring
agents should follow the
[Kubernetes Instrumentation Guidelines](https://github.com/kubernetes/community/blob/master/contributors/devel/sig-instrumentation/instrumentation.md),
which requires identifying containers using `pod`, `namespace`, and `container` prometheus labels.
The kubelet provides a gRPC service to enable discovery of in-use devices, and to provide metadata
{{< feature-state for_k8s_version="v1.13" state="alpha" >}}
In order to monitor resources provided by device plugins, monitoring agents need to be able to
discover the set of devices that are in-use on the node and obtain metadata to describe which
container the metric should be associated with. [Prometheus](https://prometheus.io/) metrics
exposed by device monitoring agents should follow the
[Kubernetes Instrumentation Guidelines](https://github.com/kubernetes/community/blob/master/contributors/devel/sig-instrumentation/instrumentation.md),
identifying containers using `pod`, `namespace`, and `container` prometheus labels.
The kubelet provides a gRPC service to enable discovery of in-use devices, and to provide metadata
for these devices:
```gRPC
@@ -155,31 +174,36 @@ service PodResourcesLister {
}
```
The gRPC service is served over a unix socket at `/var/lib/kubelet/pod-resources/kubelet.sock`.
Monitoring agents for device plugin resources can be deployed as a daemon, or as a DaemonSet.
The canonical directory `/var/lib/kubelet/pod-resources` requires privileged access, so monitoring
agents must run in a privileged security context. If a device monitoring agent is running as a
DaemonSet, `/var/lib/kubelet/pod-resources` must be mounted as a
[Volume](/docs/reference/generated/kubernetes-api/{{< param "version" >}}/#volume-v1-core)
in the plugin's
The gRPC service is served over a unix socket at `/var/lib/kubelet/pod-resources/kubelet.sock`.
Monitoring agents for device plugin resources can be deployed as a daemon, or as a DaemonSet.
The canonical directory `/var/lib/kubelet/pod-resources` requires privileged access, so monitoring
agents must run in a privileged security context. If a device monitoring agent is running as a
DaemonSet, `/var/lib/kubelet/pod-resources` must be mounted as a
{{< glossary_tooltip term_id="volume" >}} in the plugin's
[PodSpec](/docs/reference/generated/kubernetes-api/{{< param "version" >}}/#podspec-v1-core).
Support for the "PodResources service" is in beta, and is enabled by default.
## Examples
## Device plugin examples {#examples}
For examples of device plugin implementations, see:
Here are some examples of device plugin implementations:
* The official [NVIDIA GPU device plugin](https://github.com/NVIDIA/k8s-device-plugin)
* Requires [nvidia-docker 2.0](https://github.com/NVIDIA/nvidia-docker) which allows you to run GPU enabled docker containers.
* A detailed guide on how to [schedule NVIDIA GPUs](/docs/tasks/manage-gpus/scheduling-gpus) on k8s.
* The [NVIDIA GPU device plugin for COS base OS](https://github.com/GoogleCloudPlatform/container-engine-accelerators/tree/master/cmd/nvidia_gpu)
* The [AMD GPU device plugin](https://github.com/RadeonOpenCompute/k8s-device-plugin)
* The [Intel device plugins](https://github.com/intel/intel-device-plugins-for-kubernetes) for Intel GPU, FPGA and QuickAssist devices
* The [KubeVirt device plugins](https://github.com/kubevirt/kubernetes-device-plugins) for hardware-assisted virtualization
* The [NVIDIA GPU device plugin](https://github.com/NVIDIA/k8s-device-plugin)
* Requires [nvidia-docker](https://github.com/NVIDIA/nvidia-docker) 2.0, which allows you to run GPU-enabled Docker containers.
* The [NVIDIA GPU device plugin for Container-Optimized OS](https://github.com/GoogleCloudPlatform/container-engine-accelerators/tree/master/cmd/nvidia_gpu)
* The [RDMA device plugin](https://github.com/hustcat/k8s-rdma-device-plugin)
* The [Solarflare device plugin](https://github.com/vikaschoudhary16/sfc-device-plugin)
* The [AMD GPU device plugin](https://github.com/RadeonOpenCompute/k8s-device-plugin)
* The [SRIOV Network device plugin](https://github.com/intel/sriov-network-device-plugin)
* The [Intel device plugins](https://github.com/intel/intel-device-plugins-for-kubernetes) for GPU, FPGA and QuickAssist devices
* The [SR-IOV Network device plugin](https://github.com/intel/sriov-network-device-plugin)
* The [Xilinx FPGA device plugins](https://github.com/Xilinx/FPGA_as_a_Service/tree/master/k8s-fpga-device-plugin/trunk) for Xilinx FPGA devices
{{% /capture %}}
{{% capture whatsnext %}}
* Learn about [scheduling GPU resources](/docs/tasks/manage-gpus/scheduling-gpus/) using device plugins
* Learn about [advertising extended resources](/docs/tasks/administer-cluster/extended-resource-node/) on a node
* Read about using [hardware acceleration for TLS ingress](https://kubernetes.io/blog/2019/04/24/hardware-accelerated-ssl-tls-termination-in-ingress-controllers-using-kubernetes-device-plugins-and-runtimeclass/) with Kubernetes
{{% /capture %}}