Random, unprovoked "Something went wrong!"

[lindhe]

Checklist:

• [x] I've searched in the docs and FAQ for my answer: https://bit.ly/argocd-faq.
• [ ] I've included steps to reproduce the bug.
• [x] I've pasted the output of argocd version.

Describe the bug

I was hanging out in the "Pods" view of an application and left my computer for a few minutes. When I was back, the window had a blank (white) background with just the text pasted below.

I think that I might have had the "App diff" drawer open.

It told me to post here, so I did. ¯\_(ツ)_/¯

To Reproduce

I have no idea.

Expected behavior

I expected to stay in the application view.

Version

argocd: v2.7.7+4650bb2
  BuildDate: 2023-07-05T20:02:35Z
  GitCommit: 4650bb2817c3c81405f40cf77e93ef2b5fb275fb
  GitTreeState: clean
  GoVersion: go1.19.10
  Compiler: gc
  Platform: linux/amd64
argocd-server: v2.7.7+4650bb2.dirty
  BuildDate: 2023-07-05T19:42:32Z
  GitCommit: 4650bb2817c3c81405f40cf77e93ef2b5fb275fb
  GitTreeState: dirty
  GoVersion: go1.19.10
  Compiler: gc
  Platform: linux/amd64
  Kustomize Version: v5.0.1 2023-03-14T01:32:48Z
  Helm Version: v3.11.2+g912ebc1
  Kubectl Version: v0.24.2
  Jsonnet Version: v0.19.1

Logs

Something went wrong!

Consider submitting an issue [here](https://github.com/argoproj/argo-cd/issues/new?labels=bug&template=bug_report.md).

Stacktrace:

bt@https://argo.example.com/main.46e660bfc1e44947837b.js:2:662469
75270/xt/<@https://argo.example.com/main.46e660bfc1e44947837b.js:2:666630
@https://argo.example.com/main.46e660bfc1e44947837b.js:2:666791
xt@https://argo.example.com/main.46e660bfc1e44947837b.js:2:667103
kt@https://argo.example.com/main.46e660bfc1e44947837b.js:2:667364
Et@https://argo.example.com/main.46e660bfc1e44947837b.js:2:667446
o@https://argo.example.com/main.46e660bfc1e44947837b.js:2:643433
value@https://argo.example.com/main.46e660bfc1e44947837b.js:2:955437
value@https://argo.example.com/main.46e660bfc1e44947837b.js:2:955730
value@https://argo.example.com/main.46e660bfc1e44947837b.js:2:957643
_r@https://argo.example.com/main.46e660bfc1e44947837b.js:2:799088
js@https://argo.example.com/main.46e660bfc1e44947837b.js:2:819123
ya@https://argo.example.com/main.46e660bfc1e44947837b.js:2:855541
pl@https://argo.example.com/main.46e660bfc1e44947837b.js:2:847018
ul@https://argo.example.com/main.46e660bfc1e44947837b.js:2:846941
il@https://argo.example.com/main.46e660bfc1e44947837b.js:2:843971
82002/qo/<@https://argo.example.com/main.46e660bfc1e44947837b.js:2:795750
90362/t.unstable_runWithPriority@https://argo.example.com/main.46e660bfc1e44947837b.js:2:870798
Uo@https://argo.example.com/main.46e660bfc1e44947837b.js:2:795459
qo@https://argo.example.com/main.46e660bfc1e44947837b.js:2:795697
$o@https://argo.example.com/main.46e660bfc1e44947837b.js:2:795630
Xa@https://argo.example.com/main.46e660bfc1e44947837b.js:2:840760
enqueueSetState@https://argo.example.com/main.46e660bfc1e44947837b.js:2:799413
17952/y.prototype.setState@https://argo.example.com/main.46e660bfc1e44947837b.js:2:1029695
value@https://argo.example.com/main.46e660bfc1e44947837b.js:2:956415
51845/e.exports</t/v</n.timer<@https://argo.example.com/main.46e660bfc1e44947837b.js:2:953823

[crenshaw-dev]

This is usually due to a simple frontend bug, like a null reference. That was the whole error message? I'd expect to see a message in addition to the stack trace.

[lindhe]

Yeah, this was the entire message.

[EsterSteinberger]

This error appeared also in my environment. Happened in ArgoCD version: v2.7.3, v2.8.4 This error appeared when I view a pod that has multiple containers. I select one of the containers that is not shown in the log view by default. I then close the log view and open another pod log view - then the error appears. A refresh fixed it. Can you issue a correction?

[Dines97]

I encountered the same bug with a 2.11.x version. I fixed it by downgrading to v2.9.6. Here's some additional information that might help someone reproduce the issue.

I was deploying one of the applications to a new environment (I added another element in the ApplicationSet list generator to deploy to a different namespace). When I went to the main screen to check all the applications (this is important—I have "Items per page" set to "all"), the error started occurring, and none of the basic fixes (different browser, incognito mode, pod restart, etc.) worked.

I then logged in using port-forward on Kubernetes, where none of the UI configurations were applied (with "Items per page" set to "10" or something similar). This restored the functionality of at least opening the main screen, but as soon as I typed the first two letters of the application's name ("ma" in this case) in the search bar, the list of displayed applications changed. Without ever displaying the newly fetched application, it threw the same error.

It seems likely that this bug is related to the application card display. The backend was working fine, and I was able to continue performing operations using the ArgoCD CLI.