You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I was using the K8sRunLauncher and everything was working great. I thought I could leverage the CeleryK8sRunLauncher in order to be a bit more efficient for my particular use cases.
In our environment, the pod that gets created for the job run when using the K8sRunLauncher requires a few annotations to run properly as the service account. I was able to add this in to the runLauncher config as such:
This worked well and solved all my issues. However, when moving to the CeleryK8sRunLauncher, I don't see the runK8sConfig section available in the helm chart. I tried to add the annotations to runLauncher.config.celeryK8sRunLauncher.annotation, but that only gets applied to the worker pods, not the ephemeral job run pods.
I was able to get around this by specifying the configuration in code on the job:
It would be extremely helpful if we could specify the same runK8sConfig section for the CeleryK8sRunLauncher as well for all job run pods. Having to do so in the codebase is a bit tedious.
Is this something that is already supported and perhaps I am missing it? Any help is greatly appreciated here!
Ideas of implementation
Support runK8sConfig for all job run pods when using the CeleryK8sRunLauncher.
Additional information
No response
Message from the maintainers
Impacted by this issue? Give it a 👍! We factor engagement into prioritization.
The text was updated successfully, but these errors were encountered:
What's the use case?
Hi there!
I was using the
K8sRunLauncher
and everything was working great. I thought I could leverage theCeleryK8sRunLauncher
in order to be a bit more efficient for my particular use cases.In our environment, the pod that gets created for the job run when using the
K8sRunLauncher
requires a few annotations to run properly as the service account. I was able to add this in to therunLauncher
config as such:This worked well and solved all my issues. However, when moving to the
CeleryK8sRunLauncher
, I don't see therunK8sConfig
section available in the helm chart. I tried to add the annotations torunLauncher.config.celeryK8sRunLauncher.annotation
, but that only gets applied to the worker pods, not the ephemeral job run pods.I was able to get around this by specifying the configuration in code on the job:
It would be extremely helpful if we could specify the same
runK8sConfig
section for theCeleryK8sRunLauncher
as well for all job run pods. Having to do so in the codebase is a bit tedious.Is this something that is already supported and perhaps I am missing it? Any help is greatly appreciated here!
Ideas of implementation
Support
runK8sConfig
for all job run pods when using theCeleryK8sRunLauncher
.Additional information
No response
Message from the maintainers
Impacted by this issue? Give it a 👍! We factor engagement into prioritization.
The text was updated successfully, but these errors were encountered: