Use attempt_threshold to skip reporting on first N attempts

[benhutton]

This is an implementation for #1632.

I am bringing over the concept of attempt_threshold from Honeybadger.

Retry counting is a bit tricky, and so I included a comment based on a GH comment from Honeybadger.

There are two relatively arbitrary decisions I made here that you should validate:

1. I am 1-indexing attempts, not 0-indexing
2. I am skipping until the number of attempts equals the attempt_threshold. Alternatively, we could skip until number of attempts is greater than attempt_threshold.

Thanks for your consideration!

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 98.18%. Comparing base (bf8d8c5) to head (536166e).
Report is 2 commits behind head on master.

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #2503   +/-   ##
=======================================
  Coverage   98.18%   98.18%           
=======================================
  Files         128      128           
  Lines        4837     4845    +8     
=======================================
+ Hits         4749     4757    +8     
  Misses         88       88           

Components	Coverage Δ
sentry-ruby	98.57% <100.00%> (ø)
sentry-rails	97.07% <100.00%> (+0.15%)	⬆️
sentry-sidekiq	97.09% <100.00%> (-0.33%)	⬇️
sentry-resque	92.85% <ø> (ø)
sentry-delayed_job	95.65% <ø> (ø)
sentry-opentelemetry	99.31% <ø> (ø)

Files with missing lines	Coverage Δ
sentry-sidekiq/lib/sentry/sidekiq/error_handler.rb	94.73% <100.00%> (-1.93%)	⬇️

... and 1 file with indirect coverage changes

[solnic]

Thank you, this LGTM. Could you update the CHANGELOG.md (top level one) too so that we can get a green build?

[benhutton]

@solnic sorry for the delay. Updated!