Closed mjnagel closed 3 months ago
Noting that this behavior appears to have been reverted in 5.7.1 so this may be a non-issue in that version+. Still worth evaluating testing/documenting version support!
User reports they were on v5.7.1. Reverting to v5.6.3 allowed them to continue with the deployment.
@unicorn-kp in testing I'm unable to reproduce the issue on 5.7.1. I did encounter this problem on 5.7.0. Let me know if you are able to reproduce and we can reopen but I believe this issue is tied to upstream changes and seems to have been reverted.
Also noting that our docs do note the pre-req of k3d and link the versions from the uds-common repo. Currently our CI is testing with 5.7.1 and just confirmed latest CI runs look OK on #92.
@mjnagel I ran a few deployments with 5.7.1 this weekend and didn't encounter issues so I think we can close with the upstream changes as long as users aren't on 5.7.0.
Environment
k3d 5.7.0
Issue
k3d 5.7.0 introduced a change to core-dns that now leverages the coredns-custom configmap by default. When trying to deploy our uds-dev-stack this results in an error due to helm ownership mismatching.
Severity/Priority
Low/Medium, downgrading to pre-5.7.0 does work.
Additional Context
This should be kept up to date + tested in the uds-k3d CI (and likely documented as "current tested/validated version(s)).