fix(aci milestone 3): clean up orphaned objects from workflow engine

[mifu67]

• Find orphaned data condition groups that are not connected to any detector or workflow.
• Delete orphaned actions that are connected to the orphaned data condition groups. This deletes the AARTA and DCGA lookup entries as well.
• Delete the orphaned data condition groups, which also deletes their associated data conditions.

After running this cleanup, we will fix an issue where some trigger actions had multiple entries in the AARTA table, which preventing users from editing alerts with this error.

[mifu67]

Marked post-deploy, but it probably doesn't have to be.

[github-actions]

This PR has a migration; here is the generated SQL for src/sentry/workflow_engine/migrations/0054_clean_up_orphaned_metric_alert_objects.py

for 0054_clean_up_orphaned_metric_alert_objects in workflow_engine

--
-- Raw Python operation
--
-- THIS OPERATION CANNOT BE WRITTEN AS SQL

[codecov]

Codecov Report

Attention: Patch coverage is 64.28571% with 20 lines in your changes missing coverage. Please review.

✅ All tests successful. No failed tests found.

Files with missing lines	Patch %	Lines
...est_0054_clean_up_orphaned_metric_alert_objects.py	37.93%	18 Missing ⚠️
...ons/0054_clean_up_orphaned_metric_alert_objects.py	92.59%	2 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff            @@
##           master   #90796    +/-   ##
========================================
  Coverage   87.80%   87.81%            
========================================
  Files       10283    10285     +2     
  Lines      583510   583651   +141     
  Branches    22569    22569            
========================================
+ Hits       512338   512510   +172     
+ Misses      70742    70711    -31     
  Partials      430      430            

[wedamija]

How many rows are going to be affected here?

[wedamija]

Typically we prefer to not run large batch deletions like this, since it could potentially run for a long time and will time out. Usually we prefer to take the slower route of just iterating through the rows individually and deleting.

[mifu67]

How many rows are going to be affected here?

@wedamija ~700 data condition groups and ~250 actions. I can change it to iterate and delete.

[wedamija]

I'd also recommend not wrapping this entire operation in a transaction, because all involved rows will be locked for the whole transaction. Maybe it doesn't matter too much since they're orphaned, but better to be safe.

If you want to make sure the rows related to each other are in a transaction that's fine, but you'll just need to associate them and have a smaller transaction for each set of related rows. I think it's probably safe enough to just skip it though.

[wedamija]

Let's comment this out now that it has proved this runs fine, just to make sure ci doesn't end up breaking.

[sentry-io]

Suspect Issues

This pull request was deployed and Sentry observed the following issues:

• ‼️ ProgrammingError: UndefinedTable('relation "workflow_engine_action" does not exist\nLINE 1: SELECT COUNT(*) AS "__count" FROM "workflow_engine_action" W...\n ^\n') sentry.workflow_engine.migrations.0054_clean_up... View Issue

_{Did you find this useful? React with a 👍 or 👎}