-
Notifications
You must be signed in to change notification settings - Fork 3.2k
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
Multi-step workflow does not terminate (wait container does not exist with Docker executor in v3.0) #6064
Comments
I suspect this is a setup issue. Your easiest solution would be to try the pjs executor and see if that works. If that works, it points to a bug rather than a setup issue. |
I checked https://github.com/argoproj/argo-workflows/blob/master/docs/workflow-executors.md but I don't see this pjs executor that you mentioned. Do you mean pns executor, by any chance? Also, how should I go about configuring this executor? Because even the hello-world workflow example above also failed to terminate cleanly. |
Sorry. Autocorrect. PNS |
@ongchinkai I've created a dev build for you to test |
Sorry if this is a stupid question but how do I go about testing this on DigitalOcean? I tried |
Just edit wherever you have configured the executor. You might just be able to search and replace in you manifests. |
I downloaded the contents of This is the
Following this, I try again to run the hello-world but the workflow still does not terminate even after the pod has completed execution.
|
I can confirm this issue occurs with pns too Edit: I misread this issue description, what I experience actually is this issue: |
@ongchinkai can you please attach the wait container logs? |
Signed-off-by: Alex Collins <alex_collins@intuit.com>
Installed using Installation output:
The wait logs still look the same as OP. Here's a TLDR:
|
Can you add the workflow pod YAML too? |
I hope I didn't misunderstand what you were asking for.
|
No. Not that pod, the pod that the workflow runs. It'll be labeled with |
Sorry, I think this is what you are asking for.
|
It looks to me that your main container exited quickly, <1s? |
Maybe it's because it's simply executing a hello-world statement. The reason I'm using this to test is because it's much simpler than the original workflow I was using. In that workflow I was encountering the same issue, in which the workflow would stop at a step and not proceed to the next. |
I've just pushed a change to |
This is the
These are the first few lines for the
In case you need the yaml for the pod again...
Hope this helps! |
Thank you. The logs you’ve attached appear to be from a different version as they do not contain the additional diagnostic output. You’ll need to change your image pull policy. |
Not sure if this is what you're referring to, but I notice in the logs I've attached that there's this setting |
I've just pushed |
I am using 3.2.6 version and I get the same behavior. |
Summary
Executing a multi-step workflow does not terminate or proceed to the next step even after the pod has terminated.
The issue started at https://github.com/hyfen-nl/PIVT/issues/106, where the developer identified this as a potential Argo issue. The logs below are based on the example at https://argoproj.github.io/argo-workflows/examples/#steps.
Diagnostics
What Kubernetes provider are you using? Digital Ocean
What version of Argo Workflows are you running? v3.0.7
Message from the maintainers:
Impacted by this bug? Give it a 👍. We prioritise the issues with the most 👍.
The text was updated successfully, but these errors were encountered: