The job is always failing because producer trino is down at the time it tries to run.
The easier solution might be to migrate the job to dagster and reuse all the resources available.
OSO-1243