feat(worker): add dynamic Sentry sampling for high-volume expected errors

[drazisil-codecov]

Implement a before_send hook to sample (not suppress) known high-volume errors from specific repositories, reducing Sentry noise by 99.9% while maintaining visibility.

Summary

• Add SAMPLED_ERROR_PATTERNS config dict for repo/error type sample rates
• Add deterministic sampling using event ID hash (per code review feedback)
• Add before_send hook to filter LockError, LockRetry, MaxRetriesExceededError from square-web at 0.1% sample rate
• Add comprehensive tests for the sampling logic

Problem

We have high-volume Sentry errors that are expected behavior for certain high-traffic repositories:

Error Type	Primary Source	Volume
LockError	square-web	~4M/month
LockRetry	square-web	High
MaxRetriesExceededError	square-web	High

These errors pollute our Sentry error counts, make it harder to identify real issues, and consume Sentry quota unnecessarily.

Solution

Sample these errors at 0.1% (keeping ~4K/month instead of 4M) using a deterministic hash-based approach that ensures consistent sampling decisions.

Test plan

• Unit tests for before_send hook
• Unit tests for _should_sample_event helper
• Tests for deterministic sampling behavior
• Tests for edge cases (missing tags, non-matching repos/errors)

Closes CCMRG-2010

---

Note

Implements deterministic Sentry error sampling to reduce noise while preserving visibility.

• Add SAMPLED_ERROR_PATTERNS and _should_sample_event (MD5-based) to sample specific errors per repo (e.g., LockError, LockRetry, MaxRetriesExceededError for square-web at 0.1%) in helpers/sentry.py
• New before_send hook filters events based on exception type and repo_name tag (supports tags as dict or list); passes through when data is missing
• Wire before_send into initialize_sentry and keep before_send_transaction (now filtering UploadBreadcrumb by short and fully-qualified names)
• Add unit tests covering init wiring, transaction filtering, sampling behavior, determinism, and edge cases in helpers/tests/unit/test_sentry.py

^{Written by Cursor Bugbot for commit 5fcb7b1. This will update automatically on new commits. Configure here.}

[linear]

CCMRG-2010 [Proposal] Implement Dynamic Sentry Sampling for High-Volume Expected Errors

[sentry]

Codecov Report

❌ Patch coverage is 97.67442% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 92.45%. Comparing base (6ce1874) to head (5fcb7b1).
⚠️ Report is 1 commits behind head on main.
✅ All tests successful. No failed tests found.

Files with missing lines	Patch %	Lines
apps/worker/helpers/sentry.py	97.67%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #665      +/-   ##
==========================================
- Coverage   92.98%   92.45%   -0.53%     
==========================================
  Files        1294     1295       +1     
  Lines       47327    47640     +313     
  Branches     1592     1592              
==========================================
+ Hits        44005    44047      +42     
- Misses       3013     3284     +271     
  Partials      309      309              

Flag	Coverage Δ
workerintegration	58.01% <11.62%> (-1.13%)	⬇️
workerunit	89.74% <97.67%> (-1.53%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

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

[codecov-notifications]

Codecov Report

❌ Patch coverage is 97.67442% with 1 line in your changes missing coverage. Please review.
✅ All tests successful. No failed tests found.

Files with missing lines	Patch %	Lines
apps/worker/helpers/sentry.py	97.67%	1 Missing ⚠️

📢 Thoughts on this report? Let us know!

[cursor]

Cursor Bugbot has reviewed your changes and found 1 potential issue.

^{Bugbot Autofix is OFF. To automatically fix reported issues with Cloud Agents, enable Autofix in the Cursor dashboard.}

[cursor]

Exception sampling checks wrong exception in chains

Medium Severity

The code extracts error_type from event["exception"]["values"][0]["type"], but Sentry's exception values array is ordered from oldest to newest (innermost to outermost). For chained exceptions, values[0] is the root cause exception, while values[-1] is the primary outermost exception that users see. If a LockError wraps another exception, the sampling would incorrectly check the inner exception type instead of LockError, causing those events to bypass sampling entirely.