Bibliotheca integration bad XML response (PP-1838)

[jonathangreen]

Description

Occasionally we are seeing an empty or bad XML response coming back from Bibliotheca. This causes the patron_activity.sync_patron_activity task to fail due to a lxml.etree.Error exception.

Since this is a bad / unexpected API response, I think it makes sense to wrap this in a RemoteIntegrationException. This change will mean that when we get a bad response back we will retry after #2132 goes in.

Motivation and Context

Trying to reduce the number of task failures we are seeing so that we can setup alerting when a celery task fails.

How Has This Been Tested?

• Running tests in CI.

Checklist

• I have updated the documentation accordingly.
• All new and existing tests passed.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 90.82%. Comparing base (371492c) to head (aafd727).
Report is 6 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #2135   +/-   ##
=======================================
  Coverage   90.82%   90.82%           
=======================================
  Files         352      352           
  Lines       40854    40856    +2     
  Branches     8852     8852           
=======================================
+ Hits        37104    37106    +2     
  Misses       2442     2442           
  Partials     1308     1308           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[tdilauro]

Looks good! 🚀

One question below...

[tdilauro]

Are there any cases for which we would NOT want to retry? If so, is there anything in the response that would allow us to distinguish cases where we would from those where we wouldn't? Although, RemoteIntegrationException seems like the right exception to raise, in any case.

[jonathangreen]

We may find that case in the future, but I don't see any way to distinguish or not retry right now. From what I've seen its fairly random that their API just decides to respond with a 200 status code, but an empty response body. I think retrying is a reasonable thing to do in that case.