Through various cases and environments now I've noticed that often the core_task and core_progressreport table can get out of sync with one another.

For example;
───────────────────────────────────────────────────────────────────────
[root@satellite ~]# echo "select pulp_id, name, state, worker_id from core_task where state = 'running';" | su - postgres -c "psql pulpcore"
 pulp_id | name | state | worker_id 
---------+------+-------+-----------
(0 rows)

[root@satellite ~]# echo "select pulp_id, pulp_created, pulp_last_updated, message, total, done, state from core_progressreport where state in ('running');" | sudo -u pulp PULP_SETTINGS='/etc/pulp/settings.py' pulpcore-manager dbshell | wc
     47     675   10093
───────────────────────────────────────────────────────────────────────

The discrepancy has no impact on the Satellite or Pulp as far as I can tell. Really my only concern is the reliability of the progressreport table - seeing running tasks there and believing there are running tasks without checking the tasks table first.

It would be good to have some process of cleaning the progressreport table if it doesn't already exist. Perhaps a process to reference core_task and remove entries in progressreport if the pulp_ids no longer exist, or correct the state if the state of the ID doesn't match between the tables.

Ultimately, however, I'm not familiar enough with the progressreport table and it's significance to know if these questions make sense to address so please correct me there if needed, as well.

Thanks