Debug Sessions

Most sandbox issues fall into four groups: the run is not running, the plugin is not enabled, the user does not have access, or the container environment is missing required tools.

This page continues from the Sandbox Quick Start and uses its quick-start project and RUN_UUID variable.

Check Run Status

Open the run page or inspect the run from the CLI:

polyaxon ops get -p quick-start -uid $RUN_UUID

If the run is pending, check the queue, resources, node selectors, and project limits.

Check Sandbox Health

polyaxon sandbox ping -p quick-start -uid $RUN_UUID

If the ping fails, confirm that plugins.sandbox is enabled and that the run is a supported service workload.

Inspect Logs

Use run logs for container startup issues:

polyaxon ops logs -p quick-start -uid $RUN_UUID

Start a detached sandbox command and keep the execution ID returned by the CLI:

EXEC_ID=$(polyaxon sandbox exec -p quick-start -uid $RUN_UUID --detach -- python scripts/check.py)

Read that process's logs later:

polyaxon sandbox logs -p quick-start -uid $RUN_UUID $EXEC_ID

Validate the Environment

Run simple commands before debugging application code:

polyaxon sandbox exec -p quick-start -uid $RUN_UUID -- pwd
polyaxon sandbox exec -p quick-start -uid $RUN_UUID -- printenv PATH
polyaxon sandbox exec -p quick-start -uid $RUN_UUID -- python -V

Common Fixes

• Enable the required plugin: sandbox, ssh, or tmux.
• Use a long-running service command such as sleep infinity for development sessions.
• Route the run to a queue that supports services and the requested resources.
• Move credentials into Polyaxon connections instead of the container image.
• Stop stale sandbox-enabled service runs that are holding scarce resources.