Get hands-on experience with 20+ free Google Cloud products and $300 in free credit for new customers.

Unable to Schedule Vertex AI Workbench notebooks using a custom container

Hi! I would like to schedule notebook execution using a custom container with a custom default kernel and a custom least-privilege service account. It seems the standard base container excludes nbexecutor.

I attempted to FROM the default workbench container (currently 

us-docker.pkg.dev/deeplearning-platform/colab-on-workbench/workbench-agent:20250327-2003-rc0) but this failed due to IAM_PERMISSION_DENIED.
 
Anyone have an accessible base container that includes nbexecutor?
 
The error I have encountered is as follows:
Custom job failed with error message: The replica workerpool0-0 exited with a non-zero status of 127. Termination reason: Error. To find out more about why your job exited please check the logs: https://console.cloud.google.com/logs/viewer?project=830895911586&resource=ml_job%2Fjob_id%2F7799194158907785216&advancedFilter=resource.type%3D%22ml_job%22%0Aresource.labels.job_id%3D%227799194158907785216%22
 
and the linked instance logs show the following:
 
/opt/nvidia/nvidia_entrypoint.sh: line 67: exec: nbexecutor: not found
 
thanks!
dfry
0 0 19
0 REPLIES 0