What needs to happen?
We are fundamentally changing how we're using the job_status collection with our dedup_hash usage and recipe passing changes. As a result, we probably want to retain data in the job_status collection for longer than 24 hours (which is our current retention policy).
We should also determine if we want to add an additional timestamp to job_status to track the last time an entry was accessed and possibly use that for the retention policy instead of when the entry was created.
Once we determine the ideal retention policy, update the code accordingly to implement it.