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
Thanks for opening the issue.
We're using a constant polling of 100ms which can be a little too low for cloud task runners like GcpBatch that are anyway long to start.
We will discuss internally what to do to improve this.
Describe the issue
I have 5 concurrent jobs running for about 40-50 minutes, and I am seeing 28k requests to
google.cloud.batch.v1.BatchService.GetJob
.Here is requests per second screenshot over last hour:
Here is total requests picture:
My thoughts are that there could be a setting to set 'poll between' time or potentially just lessen the requests overall.
Let me know if you need more context/have any questions.
Thank you Kestra team 🙏 !
Environment
The text was updated successfully, but these errors were encountered: