Validators can manually execute jobs either from the admin panel or via cli
To trigger a job through the cli:
first ssh into the machine running the validator and into the container:
ssh user@validator-host
docker exec -it valdiator-celery-worker-1 /bin/bash
then trigger the job with the following command:
python3 manage.py debug_run_organic_job --miner_hotkey 5EHGR... --timeout 60 --docker_image python:3.11-slim --raw_scrip "print('hello world')"
or alternatively to the raw script you can provide cmd_args, as well as
python3 manage.py debug_run_organic_job --miner_hotkey 5EHGR... --timeout 99 --docker_image python:3.11-slim --cmd_args "python3 --version"
The job progress can be monitored via the generated logs or alternatively in the admin panel by checking the AdminJobRequest and OrganicJob views
To trigger a job through the admin panel:
Access the validator admin panel by tunneling the port to your local machine, then access the admin panel at localhost:8000/admin:
ssh -L 8000:localhost:8000 user@validator-host
First, create a AdminJobRequest object - this will trigger the creation and execution of a new OrganicJob with the specified uuid.
Then, go to the OrganicJob details page to see the job status as well as stdout generated by the job execution.
Any errors from the job execution can be checked in the OrganicJob view, in the stderr field. Likewise details about the job failing to trigger or errors previous to the job execution will be in the AdminJobRequest view in the admin panel.