-
Notifications
You must be signed in to change notification settings - Fork 5.5k
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
Server Side diff failing for fluent-bit #17568
Comments
Just ran into a similar issue with I deleted a list from my values files and ran into this |
I'm having the same issue. I had to set
|
Same issue with {
"Version": "v2.10.5+335875d",
"BuildDate": "2024-03-28T15:02:45Z",
"GitCommit": "335875d13e018bed6e03873f4742582582964745",
"GitTreeState": "clean",
"GoVersion": "go1.21.3",
"Compiler": "gc",
"Platform": "linux/amd64",
"KustomizeVersion": "v5.2.1 2023-10-19T20:13:51Z",
"HelmVersion": "v3.14.3+gf03cc04",
"KubectlVersion": "v0.26.11",
"JsonnetVersion": "v0.20.0"
}
and |
I am now having the same issue with our custom deployment + server-side diff activated, after updating ArgoCd v2.10.4 -> v2.10.6:
Even though my container 0 has a name. How to fix that? |
Just ran into this as well with the loki-distributed, mimir-distributed and grafana-agent helm-chart. We had server side apply and server-side diff enabled for a few weeks, and now it just broke. I only modified CPU/MEM resource limits/requests. Nothing else changed. The "apply only" mentioned above didn't help either. Running the latest version, 2.11.0. |
I found a workaround for my issue. I deleted the affected StatefulSet/Deployment with Orphan, which means it doesn't delete any pods. Then ran ArgoCD Sync again, which re-created the STS/Deploy resources. |
Looks like my issue is more related to Edit: nvm, removing all ignoreDifferences didn't fix it. |
I think I found the issue related to ServerSide diff. Which is and issue inside the gitops-engine repo. Basically when you have a nested value being changed it breaks the map traversal used for doing these server side comparisons. I tried debugging it and applying a bandaid fix that I don't know if it would even be right argoproj/gitops-engine@c25fd94#diff-00282c65a618a9ea64cdb99da5137663dc5773f2c3fd8c37ed2e9a99f3d67f09L254 |
Any update on this? |
Having the same problem: Modified some containers environment variables in my spec. But the main worry in our case is that's it's kind of a silent error, the app will flap from sync to unknown for a couple of minute and then back to |
We also have this problem (Unknown Error State) and we don't use ApplicationSets with Progressive Sync. |
Update: the error happens in a portion of the code that revert webhook mutation in the diffs as stated by @STollenaar So on our side adding the Not sure what other problems can arise down the line by setting this option though... |
Unknown Error state also gone? |
Yes the unknown in this case was because of the errors while doing diffs, so there is definitely a problem in the code that "ignore the webhooks" in server side diffs, but so far it's a good workaround. Including webhooks mutation in diffs will probably cause some unwanted differences though, depending on what webhooks you have in your clusters, in our case just ignoring the
|
I'm seeing similar issues for kube-prometheus-stack
|
ArgoCD versions 2.10 and below have reached EOL. Can you upgrade and tell us if the issue is still present, please? |
Still an issue in 2.12.6 |
Observed as an issue in 2.13 too |
People experiencing this issue may want to follow my issue here, where @andrii-korotkov-verkada has been doing some digging into Kubernetes I'm thinking we're all seeing roughly the same symptoms? |
@thecosmicfrog I'm inclined to concur that you're seeing the same issue from a different perspective. Same goes for mine, #19994 , where the the block is renamed instead of removed - result being that the block is duplicated. I focused on the steps needed to consistently trigger this, but I agree that @andrii-korotkov-verkada has likely pinpointed the root cause. |
Hi all. A fix has been added to the I no longer get this error with server-side diff and server-side apply enabled. You may want to re-test. Per @andrii-korotkov-verkada, you can build the latest image by cloning https://github.com/argoproj/argo-cd and running docker image tag argocd:latest <your-docker-repo>/argocd:issue-20792-fix I then inserted this image into [...]
global:
image:
repository: <your-docker-repo>/argocd
tag: issue-20792-fix
# If uploaded to a private registry, you might need to add this:
imagePullSecrets:
- name: <secret-name>
[...] ...and Hopefully saves people a few minutes looking for the relevant values 🙂 |
Checklist:
argocd version
.Describe the bug
One of our ArgoCD instances is showing the following error when upgrading fluent-bit 0.30.4 -> 0.44.0:
To Reproduce
We use a Kyverno policy to modify images to point to a local ECR cache instead of pulling from dockerhub directly.
When upgrading the fluent-bit helm chart from 0.30.4 to 0.44.0, we get the error above.
We are using ArgoCD
v2.10.3+0fd6344
Expected behavior
Screenshots
Version
Logs
The text was updated successfully, but these errors were encountered: