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
This helm chart is setting fsGroup. When fsGroup is used, most of the times unless there are reasons not to - fsGroupChangePolicy: "OnRootMismatch" must be set. This is especially relevant for Jenkins with its CPS/survivability implementation. My Jenkins instance has 31 million xml files. To attach a volume - takes 15+ minutes, and that's super performant nvme ssd based volume. Probably could take an hour on a regular EBS or something like that.
Version of Helm and Kubernetes
- Helm:
- Kubernetes:
Chart version
irrelevant
What happened?
1.
2.
3.
...
What you expected to happen?
No response
How to reproduce it
No response
Anything else we need to know?
No response
The text was updated successfully, but these errors were encountered:
Yes I can send a PR when I get a minute. @mohanisch-sixt I am currently using podSecurityContextOverride as a workaround, however
Chart is setting fsGroup by default with no user input, which without fsGroupChangePolicy to go with it results in a non-optimal settings. For unexperienced user, this may not be evident what the problem is and why it takes 15+ minutes to start Jenkins.
podSecurityContextOverride currently requires me to set a bunch of other options that I would like to leave default. If this chart tomorrow decides to change runAsUser, I should not be required to inspect every commit line by line on the off chance my override needs to change as well.
Describe the bug
This helm chart is setting
fsGroup
. WhenfsGroup
is used, most of the times unless there are reasons not to -fsGroupChangePolicy: "OnRootMismatch"
must be set. This is especially relevant for Jenkins with its CPS/survivability implementation. My Jenkins instance has 31 million xml files. To attach a volume - takes 15+ minutes, and that's super performant nvme ssd based volume. Probably could take an hour on a regular EBS or something like that.Version of Helm and Kubernetes
Chart version
irrelevant
What happened?
What you expected to happen?
No response
How to reproduce it
No response
Anything else we need to know?
No response
The text was updated successfully, but these errors were encountered: