Make antrea-controller not tolerate Node unreachable #5521
Merged
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
0e4e4c4 to
fa5aeb4
Compare
|
/test-all |
jianjuns
previously approved these changes
Sep 23, 2023
When a Node becomes unreachable, currently it takes 5m45s+ for Kubernetes to move antrea-controller Pod to another Node. The time spent in the process includes: * 40s (default value of NodeMonitorGracePeriod) to mark a Node's Ready condition to Unknown * 5s to taint the Node with `node.kubernetes.io/unreachable:NoExecute` * 5m (default value of defaultUnreachableTolerationSeconds) to tolerate the taint The 1st duration is kind of inevitable. The 2nd duration seems a bug in kube-controller-manager, which I have opened an issue kubernetes/kubernetes#120815 and may be fixed in a future release. The 3rd duration is because Kubernetes automatically adds a default toleration for `node.kubernetes.io/unreachable:NoExecute` with tolerationSeconds of 300s if the Pod doesn't have one. This commit adds a toleration with tolerationSeconds of 0s for `node.kubernetes.io/unreachable:NoExecute` explicitly, which reduces the failover time by 5m. Signed-off-by: Quan Tian <qtian@vmware.com>
fa5aeb4 to
557a1f0
Compare
|
Fixed helm README |
antoninbas
approved these changes
Sep 25, 2023
|
/test-all |
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Outdated suggestions cannot be applied.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.
When a Node becomes unreachable, currently it takes 5m45s+ for Kubernetes to move antrea-controller Pod to another Node. The time spent in the process includes:
node.kubernetes.io/unreachable:NoExecuteThe 1st duration is kind of inevitable. The 2nd duration seems a bug in kube-controller-manager, which I have opened an issue kubernetes/kubernetes#120815 and may be fixed in a future release. The 3rd duration is because Kubernetes automatically adds a default toleration for
node.kubernetes.io/unreachable:NoExecutewith tolerationSeconds of 300s if the Pod doesn't have one.This commit adds a toleration with tolerationSeconds of 0s for
node.kubernetes.io/unreachable:NoExecuteexplicitly, which reduces the failover time by 5m.