missmatch with naming on pods from nested tasks in DAGs in the UI (fail to retrieve live logs)

[Guillermogsjc]

Pre-requisites

• [X] I have double-checked my configuration
• [X] I can confirm the issues exists when I tested with :latest
• [ ] I'd like to contribute the fix myself (see contributing guide)

What happened/what you expected to happen?

hi, from v3.4.3, the field POD_NAME in the UI is not correlated with the pod created in k8s in some cases.

As an example:

• workflow_name: foo-lg77s
• task of dag name: foo-lg77s.bar
• task of subdag of task on dag: foo-lg77s.bar.my-stage
• concrete retry pod name (because last task has retries): foo-lg77s.bar.my-stage(0)

In this example:

• reported POD_NAME in Argo UI (and the one logs are tried to be get): foo-lg77s-3278223386 .
• REAL POD_NAME created in K8S: foo-lg77s-my-stage-3278223386

Somewhere over there a mismatch between names of created pod names in task of tasks of DAGs is rising so the UI will not show the logs.

Not happening with <v3.4.3 (if i remember well)

Version

v3.4.3

Paste a small workflow that reproduces the issue. We must be able to run the workflow; don't enter a workflows that uses private images.

no wf

[Guillermogsjc]

specifically, the NAMEon the UI is kargentina-1-279wc.s2t-core-lower.wait-main-get-logs(0), the POD NAME in the UI is kargentina-1-279wc-931539070

but the pod name inside k8s is kargentina-1-279wc-wait-main-get-logs-931539070, so argo server is disconnected from what is happening in K8S (workflow controller is ok because the stage finishes so I guess it is tracking the correct pod name)

logs of workflow controller https://gist.github.com/Guillermogsjc/710bd617eae5e4eee0396dc3fc581d68 obtained through kubectl logs -n argo deploy/workflow-controller | grep kargentina-1-279wc

the most weird thing is that I have a parallel workflow created from the same templates with aligned pod name

[sarabala1979]

@JPZ13 Can you look at this issue?

[JPZ13]

@isubasinghe @rohankmr414 Can I get one of you two to look into this? I'll be available to review it Wednesday or Friday if y'all can get a fix out before then

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. If this is a mentoring request, please provide an update here. Thank you for your contributions.