-
Notifications
You must be signed in to change notification settings - Fork 10
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
feat: remove CPU resource limit during the time of a boost #57
Comments
Hi @mikouaj, I am keen to understand the implications of the proposed change. If the CPU resource limit is removed during a boost, could this potentially cause CPU resource contention/starvation for other applications running on the node, thereby impacting their stability and performance? |
Hi @AnoshanJ, Thank you for your question. On a high level, when there is no CPU limit for a container, the container can use all of the resources available on a node. When multiple containers request for such excess (above the requests) CPU resources, the node's scheduler will try to distribute them (Linux uses CFS - completely fair scheduler - by default). To sum up, only requested CPU resources are guaranteed. CPU resources that excess the requested ones are given on a best-effort only, distributed by the node's scheduler when available. Therefore, removing the CPU limits from a container is a safe operation and in most cases even a recommended practice. |
Thanks @mikouaj. That clears up my question. |
Community Note
to the original issue to help the community and maintainers prioritize this request
not help prioritize the request
Description
As proposed by @yyvess in #36 (comment) the container's CPU limits, if present, can be removed during the boost instead of being increased. This will allow to allocate even more CPU time if available. The behavior should be configuration driven with "always-on" default.
References
The text was updated successfully, but these errors were encountered: