Hello! Has anyone run into this problem lately? Ou...
# kubernetes
r
Hello! Has anyone run into this problem lately? Our Helm charts deployed through ArgoCD(v2.10.0) on an Azure Kubernetes cluster(v1.30.3) are showing these diffs on the
resources: {}
parameter, which shows it is out of sync. This doesn't happen on other clusters with the same version, which were created identically, but earlier than the current one with the issue.
Looks like its a problem with our version of argo and kubernetes 1.30.* https://github.com/argoproj/argo-cd/issues/18764 https://github.com/argoproj/argo-cd/issues/18213 But it only affects some of our clusters even though they are identical.
t
That's really odd and I've never seen that before. I'd probably just define the resources explicitly and call it a day.
r
Yeah, we will start by upgrading to the latest argo version as we are a minor version behind. Then define the resources. Its varying stuff like MutatingWebhookConfiguration caBundle values as well. So just ignoring a bunch of stuff doesn't seem like a good long term solution
c
this looks like the kind of thing that happens when server-side apply is enabled, which can be done per-app and might explain the discrepancy
t
I think I've only got like one appset that I use server side applies for... so it makes sense I don't see this much