Improve manager logging of task information

[bennybp]

Description

Improve printing of task status and information in compute manager logs

Information now is printed as tables and now include record_id, which is distinct from task_id. In general, task_id is rarely useful, and often confused with record_id, so we print both explicitly.

Now, two tables are printed. The first is when computations are found to be finished, and includes executor, walltime, and error information.

  task id    record id  executor    walltime    status
---------  -----------  ----------  ----------  --------------------
        3            4  local       00:00:2.74  success
        4            5  local       00:00:2.74  success
        5            6  local       00:00:2.52  error: unknown_error

The second displays the status of records being sent back to the server (whether they are accepted, rejected, or deferred due to server downtime/network issues).

  task id    record id  status    reason
---------  -----------  --------  -------------------------------------
        3            4  sent
        4            5  sent
        5            6  rejected  Task does not exist in the task queue

Changelog description

Improve manager logging of task information

Status

• [X] Code base linted
• [x] Ready to go

[bennybp]

Just wondering if you have any comments @j-wags or @dotsdl. Any additional info you would find useful? You are some of the few people to run a manager :)

[j-wags]

This looks really slick! I'd be happy with it. I'll ping @dotsdl directly in case he (like me) didn't see the notification.