1

I would like to access google cloud storage buckets and github repo from Vertex AI workbench JupyterLab (which is a user-managaed notebook) on google cloud.

I build the notebook with a customized image that install some python packages. But, I found that most of the basic functions do not perform well. For example, I cannot access google cloud storage buckets and my github repo from the notebook.

I have tried to use the commands from a terminal of the jupyterlab:

 gcsfuse gs://MY_BUCKETS /local_bucket.  # error, daemonize.Run: readFGromProcess: sub-process : mountWithArgsL mountWithConn: Mount : mount: running /usr/bin/fusermount: exit status 1

 git clone MY_GITHUB_REPO # error, unable to access the repo, Failed to connect ti github,com port 443, could not connect to server

If I built a Vertex AI workbench jupyter notebook with a pre-built image, everything works well. How to config the notebook correctly if it is built with a customized image ?

This post Accessing Google Cloud Storage from local Jupyter and Notebooks in Google AI Platform / Vertex AI does not work because I need to enable the google cloud storage access in the Jupyter Lab UI (e.g. the buttons on the side bar) not in a notebook. I clicked the buttons, but nothing happened, no gcs buckets show up.

user3448011
  • 1,469
  • 1
  • 17
  • 39
  • When you say "I clicked the buttons, but nothing happened, no gcs buckets show up" I believe you're referring in managed notebook? For user-managed notebook and answer mentioned in the Stackex thread should work. – Sakshi Gatyan Mar 13 '23 at 05:40

0 Answers0