Provide current IP count for kubernetes scheduling
cskinfill opened this issue · 6 comments
What would you like to be added:
Provide the kube api with the current number of pooled IPs so that these resources can be used for scheduling decisions. There are a number of issues that seem related to this (#1160 and #26)
Using the "Device Plugins" feature, the aws cni (or a sidecar) can provide to the kubelet the number of IP addresses. The kubelet will then report these resources back to the kube api which the scheduler can then use for scheduling decisions.
Why is this needed:
To ensure that a pod isn't scheduled to a node that has sufficient resources but NOT sufficient IP addresses available.
This would be massively helpful generally for operation of the CNI.
This issue is stale because it has been open 60 days with no activity. Remove stale label or comment or this will be closed in 14 days
Not stale
Closing in favor of aws/containers-roadmap#2189. While I do not think Device Plugins is a good fit here, we are working on other approaches to prevent a pod from ever being scheduled on a node that does not have enough IP addresses
This issue is now closed. Comments on closed issues are hard for our team to see.
If you need more assistance, please either tag a team member or open a new issue that references this one.