Troubleshooting
How to handle unauthorized issue
If you get an error 401 Unauthorized
or unauthorized: authentication required
from docker,
you may need to re-authenticate by calling eai login
.
It is possible your environment has some remaining environment variables which can impact the authentication process.
Be sure the following variables are not defined:
- EAI_TOKEN
: A sepecific token is defined and can be invalid or expired
- EAI_PROFILE
: A profile is defined and you are trying to use another Toolkit when pushing a docker image
About docker, if you called docker login
before, you can cancel it calling docker logout registry.console.elementai.com
.
You can validate it by looking at ~/.docker/config.json
, the section auths
should not contain any Toolkit registry.
To see if the docker credentials are properly set up, run the following command, it should return a JSON payload.
If it doesn’t, the symlink created above is probably not in your PATH
.
docker-credential-eai-<profile> get
How to handle forbidden issue
If you get an error 403 Forbidden
, it means you are successfully logged on Toolkit,
but you do not have the right permissions to perform the action you requested.
Be sure the following variables are not defined:
- EAI_TOKEN
: A sepecific token is defined and can have less permissions than expected
About docker, if you called docker login
before, you can cancel it calling docker logout registry.console.elementai.com
.
You can validate it by looking at ~/.docker/config.json
, the section auths
should not contain any Toolkit registry.
How can I validate permission with a role? I get no basic auth credentials
or Error: http: 401 Unauthorized
with CLI or docker commands
Check the environment var is correctly defined (don’t do that in Drone/CI to not log the secret in log):
Check you are correctly logged in with the good role ID:
Validate your role has the right permissions.
Runs Fine on a Mac but Fails in docker on the Same Machine
If you get an error like +++ killed by SIGKILL +++
, your program may be running out of memory.
Check your docker configuration for a default memory limit.
Here is a related thread on StackOverflow.