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

Merge pull request #6006 from scjane/patch-11

Browse Source 
Update scheduling-hugepages.md
This commit is contained in:
Joseph Heck
2017-10-22 18:58:13 -07:00
committed by GitHub
parent d5b0c2490f 29c5cf0d0b
commit aab17c16a2
1 changed files with 6 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/tasks/manage-hugepages/scheduling-hugepages.md
+6 -6
View File
@@ -8,7 +8,7 @@ title: Manage HugePages
{% include feature-state-alpha.md %} {% include feature-state-alpha.md %}
Kubernetes supports the allocation and consumption of pre-allocated huge pages Kubernetes supports the allocation and consumption of pre-allocated huge pages
by applications in a Pod as an **alpha** feature. This page describes how users by applications in a Pod as an **alpha** feature. This page describes how users
can consume huge pages and the current limitations. can consume huge pages and the current limitations.
{% endcapture %} {% endcapture %}
@@ -16,7 +16,7 @@ can consume huge pages and the current limitations.
{% capture prerequisites %} {% capture prerequisites %}
1. Kubernetes nodes must pre-allocate huge pages in order for the node to report 1. Kubernetes nodes must pre-allocate huge pages in order for the node to report
its huge page capacity. A node may only pre-allocate huge pages for a single its huge page capacity. A node may only pre-allocate huge pages for a single
size. size.
1. A special **alpha** feature gate `HugePages` has to be set to true across the 1. A special **alpha** feature gate `HugePages` has to be set to true across the
system: `--feature-gates="HugePages=true"`. system: `--feature-gates="HugePages=true"`.
@@ -32,9 +32,9 @@ schedulable resource.
Huge pages can be consumed via container level resource requirements using the Huge pages can be consumed via container level resource requirements using the
resource name `hugepages-<size>`, where size is the most compact binary notation resource name `hugepages-<size>`, where size is the most compact binary notation
using integer values supported on a particular node. For example, if a node using integer values supported on a particular node. For example, if a node
supports 2048KiB page sizes, it will expose a schedulable resource supports 2048KiB page sizes, it will expose a schedulable resource
`hugepages-2Mi`. Unlike CPU or memory, huge pages do not support overcommit. `hugepages-2Mi`. Unlike CPU or memory, huge pages do not support overcommit.
```yaml ```yaml
apiVersion: v1 apiVersion: v1
@@ -60,14 +60,14 @@ spec:
medium: HugePages medium: HugePages
``` ```
- Huge page requests must equal the limits. This is the default if limits are - Huge page requests must equal the limits. This is the default if limits are
specified, but requests are not. specified, but requests are not.
- Huge pages are isolated at a pod scope, container isolation is planned in a - Huge pages are isolated at a pod scope, container isolation is planned in a
future iteration. future iteration.
- EmptyDir volumes backed by huge pages may not consume more huge page memory - EmptyDir volumes backed by huge pages may not consume more huge page memory
than the pod request. than the pod request.
- Applications that consume huge pages via `shmget()` with `SHM_HUGETLB` must - Applications that consume huge pages via `shmget()` with `SHM_HUGETLB` must
run with a supplemental group that matches `proc/sys/vm/hugetlb_shm_group` run with a supplemental group that matches `proc/sys/vm/hugetlb_shm_group`.
## Future ## Future