Skip to content


Unable to install CLI due to certificate errors


An alternative method for downloading the CLI. Use a Linux shell to run:

kubectl port-forward -n runai svc/researcher-service 4180:4180

And in another shell run

wget --content-disposition http://localhost:4180/cli/linux

When running the CLI you get an error an invalid configuration error

When running any CLI command you get:

FATA[0000] invalid configuration: no configuration has been provided


Your machine is not connected to the Kubernetes cluster. Make sure that you have a ~/.kube directory that contains a configuration file pointing to the Kubernetes cluster.

When running the CLI you get an error: open .../.kube/config.lock: permission denied

When running any CLI command you get a permission denied error.


The user running the CLI does not have read permissions to the .kube directory.

When running 'runai logs', the logs are delayed

By default, Python buffers stdout and stderr, which is not flushed in real-time. This may cause logs to appear sometimes minutes after being buffered.


Set the env var PYTHONUNBUFFERED to any non-empty string or pass -u to Python. e.g. python -u

Runai list jobs command works but runai submit does not


(Version 2.4 or earlier) Helm utility is not installed. See Run:ai CLI Installation documentation.

Last update: June 12, 2022