Fix context propagation for DisabledBatcher

[bogdandrutu]

Finishes the implementation started in #12225

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 91.30%. Comparing base (4ee2b50) to head (f0c4433).
Report is 1 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main   #12231   +/-   ##
=======================================
  Coverage   91.30%   91.30%           
=======================================
  Files         464      464           
  Lines       25662    25662           
=======================================
  Hits        23430    23430           
  Misses       1819     1819           
  Partials      413      413           

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

[sfc-gh-sili]

Thank you, Bogdan!
I agree that using the callback function is a cleaner approach compared to using an index. However, I'm not sure if we can handle partial error handling with a callback.

For example, consider this scenario: we start with a batch that contains request x, then we received another request y and this resulted in a split. Then, we have

• Batch A that contains request x and part of request y
• Batch B that contains the other part of request y

If batch A was sent out successfully but batch B failed, we still propagate the failure back to the queue?

[dmitryax]

@bogdandrutu please rebase

[bogdandrutu]

@dmitryax done