You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
-- someone who has experienced this issue can test and let us know if this is what's needed.
Describe the bug
Helm Controller and Notification Controller were reported by a user on EKS to be in crashloopbackoff.
I am not sure why I haven't seen this issue on any of my clusters before, but I noticed that error often surfaces when fsGroup setting is missing from the securityContext and when I checked my deployments, I noticed those two controllers were the only ones that did not have this block in their deployment config:
securityContext:
fsGroup: 1337
Seems too congruous to be a coincidence! Should we add those in the config of helm-controller and notification-controller or are they omitted on purpose?
We tried some things to bisect the issue, but I believe this is likely the problem.
Expected behavior
Not crashloopbackoff
Screenshots and recordings
No response
OS / Distro
EKS
Flux version
0.27.3
Flux check
N/A
Git provider
No response
Container Registry provider
No response
Additional context
I'm happy to submit the PRs for this if we're agreed this is what's needed (I'll start them now, so they can be merged straight away if that's the case).
Code of Conduct
I agree to follow this project's Code of Conduct
The text was updated successfully, but these errors were encountered:
Maybe fixed by:
-- someone who has experienced this issue can test and let us know if this is what's needed.
Describe the bug
Helm Controller and Notification Controller were reported by a user on EKS to be in crashloopbackoff.
I am not sure why I haven't seen this issue on any of my clusters before, but I noticed that error often surfaces when
fsGroup
setting is missing from thesecurityContext
and when I checked my deployments, I noticed those two controllers were the only ones that did not have this block in their deployment config:Seems too congruous to be a coincidence! Should we add those in the config of helm-controller and notification-controller or are they omitted on purpose?
Steps to reproduce
I don't have details for a reproduction, I only know that this issue was reported by an EKS user who said they didn't do anything special to their cluster. (Thread: https://cloud-native.slack.com/archives/CLAJ40HV3/p1647269384516989)
We tried some things to bisect the issue, but I believe this is likely the problem.
Expected behavior
Not crashloopbackoff
Screenshots and recordings
No response
OS / Distro
EKS
Flux version
0.27.3
Flux check
N/A
Git provider
No response
Container Registry provider
No response
Additional context
I'm happy to submit the PRs for this if we're agreed this is what's needed (I'll start them now, so they can be merged straight away if that's the case).
Code of Conduct
The text was updated successfully, but these errors were encountered: