Document overprovisioning for node autoscaling #46635
Draft
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
k8s-ci-robot
added
cncf-cla: yes
✅ Pull request preview available for checkingBuilt without sensitive environment variables
To edit notification comments on pull requests, go to your Netlify site configuration. |
sftim
reviewed
Jun 3, 2024
There was a problem hiding this comment.
This is explaining a task so I'd expect it to be inside http://k8s.io/docs/tasks/
sftim
reviewed
Jun 3, 2024
Comment on lines
27
to
31
| 1. **Create a Placeholder Deployment:** Create a Deployment of pods that run a minimal process, like a pause container. These pods should have very low priority (using Priority Classes in Kubernetes) to ensure they are preempted easily. | ||
|
|
||
| 2. **Adjust Resource Requests and Limits:** Configure the placeholder pods' resource requests and limits to define the amount of overprovisioned resources you want to maintain. For example, requesting a small amount of CPU and memory will reserve those resources without consuming them fully. | ||
|
|
||
| 3. **Set the Desired Replica Count:** Determine the number of placeholder pods you need to create to achieve your desired level of overprovisioning. You can start with a small number and gradually increase it to find the right balance between resource reservation and cost. |
There was a problem hiding this comment.
You should document how to perform each of these steps.
|
|
||
| ## How Node Overprovisioning Works | ||
|
|
||
| The most common way to implement node overprovisioning is to use a _placeholder_ deployment running pods with very low priority and a minimal resource footprint. These pods consume resources that could otherwise be used by regular application pods. However, due to their low priority, they are easily preempted _evicted_ when higher-priority pods need those resources. |
There was a problem hiding this comment.
Suggested change
| The most common way to implement node overprovisioning is to use a _placeholder_ deployment running pods with very low priority and a minimal resource footprint. These pods consume resources that could otherwise be used by regular application pods. However, due to their low priority, they are easily preempted _evicted_ when higher-priority pods need those resources. | |
| The most common way to implement node overprovisioning is to use a _placeholder_ deployment running pods with very low priority and a minimal resource footprint. These pods consume resources that could otherwise be used by regular application pods. However, due to their low priority, they are easily preempted (evicted) when higher-priority pods need those resources. |
- Fixed redundant space.
- Corrected the grammar: Regardless of whether you italicized one, using "preempted" and "evicted" right next to each other without parenthesis is wrong because an eviction occurs when a preemption logic is triggered. The previous sentence assumes that each word is a different action.
|
/hold I want to emphasize the points in #46635 (review) OK to unhold once this PR is adding task documentation. |
k8s-ci-robot
added
the
do-not-merge/hold
|
[APPROVALNOTIFIER] This PR is NOT APPROVED This pull-request has been approved by: The full list of commands accepted by this bot can be found here.
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
sftim
requested changes
Jun 10, 2024
There was a problem hiding this comment.
Thanks, but:
- you haven't yet documented the steps for the task
- you should list any prerequisites, as task pages typically do
- (less important)
we'd also like you to follow our https://kubernetes.io/docs/contribute/style/style-guide/
| @@ -115,3 +115,4 @@ in the cluster. | |||
| ## {{% heading "whatsnext" %}} | |||
|
|
|||
| - Read about [workload-level autoscaling](/docs/concepts/workloads/autoscaling/) | |||
| - Read about [node overprovisioning](/docs/concepts/cluster-administration/node-overprovisioning/) | |||
There was a problem hiding this comment.
Suggested change
| - Read about [node overprovisioning](/docs/concepts/cluster-administration/node-overprovisioning/) | |
| - Read about [node overprovisioning](/docs/tasks/cluster-administration/node-overprovisioning/) | |
| - ``` |
There was a problem hiding this comment.
Please move this to the cluster adminstration section.
There was a problem hiding this comment.
This should be in the cluster adminstration section.
Comment on lines
+27
to
+37
| ## Create a Placeholder Deployment | ||
|
|
||
| Create a Deployment of pods that run a minimal process, like a pause container. These pods should have very low priority (using Priority Classes in Kubernetes) to ensure they are preempted easily. | ||
|
|
||
| ## Adjust Resource Requests and Limits | ||
|
|
||
| Configure the placeholder pods' resource requests and limits to define the amount of overprovisioned resources you want to maintain. For example, requesting a small amount of CPU and memory will reserve those resources without consuming them fully. | ||
|
|
||
| ## Set the Desired Replica Count | ||
|
|
||
| Determine the number of placeholder pods you need to create to achieve your desired level of overprovisioning. You can start with a small number and gradually increase it to find the right balance between resource reservation and cost. |
There was a problem hiding this comment.
Thank you. There is far too little detail here for us to be able to accept this documentation. Please write a page that explains what someone must do, and details what they need to know in order to perform each step.
|
/hold cancel |
k8s-ci-robot
removed
the
do-not-merge/hold
k8s-ci-robot
added
the
do-not-merge/work-in-progress
Ritikaa96
reviewed
Jun 27, 2024
|
|
||
| ## How Node Overprovisioning Works | ||
|
|
||
| The most common way to implement node overprovisioning is to use a _placeholder_ deployment running pods with very low priority and a minimal resource footprint. These pods consume resources that could otherwise be used by regular application pods. However, due to their low priority, they are easily preempted (evicted) when higher-priority pods need those resources. |
There was a problem hiding this comment.
Please wrap the lines and also brief up the content as per requirement
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Fixes: #45850