Add a job_history table

whabanks commented 10 months ago

Description

As a Notify Dev, I need to be able to examine the history of jobs as they move through their various states and statuses so that I can more easily investigate issues related to jobs.

WHY are we building?

Relying primarily on Cloudwatch logs to trace jobs as they move through the system is a manual and cumbersome process. A job_history table can ease this burden, allowing us to gather more information about a job during investigations or when aggregating job metrics over time.

WHAT are we building?

A history table for our jobs.

VALUE created by our solution

Less burden on devs when issues related to jobs require investigation.
A simpler means to develop semi-automated, repeatable processes (i.e. Jupyter notebooks, Blazer queries) to aid in investigating job related issues and bugs.
Additional data that can be used in future data portages and other metric gathering / data analysis initiatives.

Documentation and Artifacts

Acceptance Criteria

Given an issue arises around job processing, when investigation is required, then historical job data is available to aid investigation.

[ ] Columns that we want to preserve in the history table have been determined
[ ] A job_history table is created

QA Steps

[ ] Tested in a realistic production scenario

yaelberger-commits commented 10 months ago

Hey team! Please add your planning poker estimate with Zenhub @andrewleith @jzbahrai @whabanks

yaelberger-commits commented 4 months ago

Nice to have, not a need to have

yaelberger-commits commented 4 months ago

If we add a celery results backend, we may not need a job history table

cds-snc / notification-planning