perf(replay): reduce postgres queries in recording consumer using get_from_cache

[aliu39]

@cmanallen pointed out the queries we make in _handle_breadcrumb, used to check feature flags and project options, could overload Postgres as we scale. To fix this we can do the query at the top-level and use get_from_cache

[codecov]

Codecov Report

Attention: Patch coverage is 86.66667% with 2 lines in your changes missing coverage. Please review.

✅ All tests successful. No failed tests found.

Files with missing lines	Patch %	Lines
src/sentry/replays/usecases/ingest/dom_index.py	83.33%	1 Missing and 1 partial ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##           master   #77365       +/-   ##
===========================================
- Coverage   88.40%   78.18%   -10.22%     
===========================================
  Files        2998     6959     +3961     
  Lines      186451   308986   +122535     
  Branches    30634    50556    +19922     
===========================================
+ Hits       164826   241590    +76764     
- Misses      15602    60936    +45334     
- Partials     6023     6460      +437     

[cmanallen]

How much memory will this consume?

[aliu39]

For the cache? I don't know. Got the inspiration from occurrence_consumer.py

[JoshFerge]

we already are usually caching the project for other places in ingest, so it probably won't be very much additional if any

[sentry-io]

Suspect Issues

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

• ‼️ Project.DoesNotExist: Project matching query does not exist. sentry.silo.base in override View Issue
• ‼️ AttributeError: 'list' object has no attribute 'get' sentry.replays.usecases.ingest.dom_index in cre... View Issue
• ‼️ OperationalError: OperationalError('server closed the connection unexpectedly\n\tThis probably means the server ter... sentry.models.options.project_option in reload_... View Issue

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