-
Notifications
You must be signed in to change notification settings - Fork 2.7k
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
Fix eks restart pods helm #10351
Fix eks restart pods helm #10351
Conversation
Release note label not set, please set the appropriate release note. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looks good, but please add "Fixes: #" to the commit message so that we can track backports effectively.
Also add the description to the commit message, thanks! |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
it seems cilium.yaml
was added by accident
@jrajahalme I added the original issue from the first pull request |
e921115
to
70da631
Compare
Fixes issue where Helm breaks the node init script when restartPods is enabled. Fixes issue where removing container can cause issues with Kubelet scheduling. fixes: cilium#9571 Signed-off-by: Tom Hadlaw <thomas.hadlaw@hootsuite.com>
70da631
to
e611448
Compare
test-me-please |
test-me-please |
Signed-off-by: Tom Hadlaw thomas.hadlaw@hootsuite.com
Please ensure your pull request adheres to the following guidelines:
description and a
Fixes: #XXX
line if the commit addresses a particularGitHub issue.
Follow up bugfix's for my previous pull request. One problem is that helm actually breaks the docker
--format
argument value as it considers it a template value. This causes the conditional to always be false and for the container restart to never happen on EKS nodes.Secondly, upon further testing, it seems like kubelet can fail due if the container is removed: (
rpc error: code = Unknown desc = unable to inspect docker image ...
).I'm not exactly sure why this didn't happen during my first round of testing but I propose to just kill the container and let Kubelet manage the remaining container state (docker kill just sends sigkill to the container process, wheraes rm removes all the container state as well). The container does get cleaned up afterwards.
fixes: #9571
This change is