Cloudera Docs

Upgrading Agent Studio

Upgrade the Agent Studio application, by updating the ML Runtime image for Agent Studio in an air-gapped environment.

In air-gapped environments, the new ML Runtime image is published through the ML Runtime repository file. You must manually synchronize these assets to make the upgrade available in your local catalog.

Docker must be running in your environment.

  1. Synchronize the ML Runtimerepository file with your self-hosted ML Runtime repository file.
    1. Download ML Runtimes from the locations specified in the repo-assembly.json manifest file.
      https://archive.cloudera.com/ml-runtimes/[***RUNTIMES VERSION***]/artifacts/repo-assembly.json
    2. Locate the list of images for a particular release within the file using the ${ml_runtime_version} variable.
    3. Create a tar.gz file to copy to your air-gapped environment using the docker pull and docker save commands for each image in the repo-assembly.json manifest file.
  2. When the new ML Runtime image is available in the Runtime catalog, click the upgrade alert link in the Agent Studio application.
    The existing ML Runtime image is replaced and the application restarts.