Harden exponential backoff: clamp visibility >= 0, don't mask origina…

[mensfeld]

…l error

Two defensive fixes in the exponential backoff retry middleware:

1. next_visibility_timeout could return a negative value when a job ran longer than the 12h SQS ceiling (max_timeout goes below zero), and change_message_visibility rejects a negative timeout. Clamp to 0.

2. handle_failure called change_visibility unguarded. If rescheduling raised (e.g. an expired receipt handle), that exception propagated out of the rescue in #call and masked the original worker error, hiding the real failure from exception handlers/notifiers. Rescue it, log, and report 'not retried' so the original error is re-raised and the message falls back to the queue's default visibility timeout.

[coderabbitai]

Important

Review skipped

Draft detected.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: 8ea762d7-614e-45b3-b89c-1f10a5e8e612

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch fix/b11-backoff-visibility-guard

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}