fix(sidekiq): Report error when retry limit is below attempt_threshold

When a job's `retry` option is lower than its `attempt_threshold` (e.g.
`retry: 1` with `attempt_threshold: 3`), the error handler skipped every
attempt and the job was silently dropped — it died before ever reaching
the threshold.

Cap the threshold at the job's final attempt so these jobs still report
before entering the dead set. Jobs whose retry count meets or exceeds
the threshold behave exactly as before.

Author: marcboquet

CHANGELOG.md

@@ -1,3 +1,9 @@
+## Unreleased
+
+### Bug Fixes 🐛
+
+- (sidekiq) Report final attempt when retry limit is below `attempt_threshold` by @marcboquet in [#XXXX](https://github.com/getsentry/sentry-ruby/pull/XXXX)
+
 ## 6.5.0
 
 ### New Features ✨

sentry-sidekiq/lib/sentry/sidekiq/error_handler.rb

@@ -43,8 +43,10 @@ def call(ex, context, sidekiq_config = nil)
           # attempt 2 - this is your first retry so retry_count is 0
           # attempt 3 - you have retried once, retry_count is 1
           attempt = retry_count.nil? ? 1 : retry_count.to_i + 2
+          # Cap at the final attempt so jobs with fewer retries than the threshold still report.
+          effective_threshold = [attempt_threshold, retry_limit(context, sidekiq_config) + 1].min
 
-          return if attempt < attempt_threshold
+          return if attempt < effective_threshold
         end
 
         Sentry::Sidekiq.capture_exception(

sentry-sidekiq/spec/sentry/sidekiq_spec.rb

@@ -141,6 +141,17 @@ def retry_last_failed_job
       retry_last_failed_job
       expect(transport.events.count).to eq(0)
     end
+
+    it "reports on the final attempt when retry limit is below the threshold" do
+      worker = Class.new(SadWorker)
+      worker.sidekiq_options attempt_threshold: 3, retry: 1
+
+      execute_worker(processor, worker)
+      expect(transport.events.count).to eq(0)
+
+      retry_last_failed_job
+      expect(transport.events.count).to eq(1)
+    end
   end
 
   context "with config.report_only_dead_jobs = true" do