Skip to content

Conversation

ttlequals0
Copy link
Owner

Summary

  • Fixed Phase 2 scan getting stuck with 0 files processed
  • Fixed UI showing incorrect counts for corrupted and warning files
  • Fixed regression where files with visual corruption were marked as warnings

Details

1. Phase 2 Stuck Issue

  • Root cause: Database had deep_scan column with NOT NULL constraint after v2.2.79 removal
  • Fix: Created migration to handle schema mismatch

2. UI Count Regression

  • Root cause: SQL query excluded files with both corruption AND warnings from corrupted count
  • Fix: Updated counting logic - corrupted files counted as corrupted regardless of warnings

3. Visual Corruption Classification

  • Root cause: Same as UI counts - visual corruption files often have warnings too
  • Fix: Ensured corrupted status takes precedence over warning status

Test Plan

  • All tests pass
  • Deploy to test environment
  • Verify Phase 2 processes files correctly
  • Verify UI counts are accurate
  • Verify visual corruption shows in corrupted filter

TTLequals0 and others added 10 commits September 1, 2025 19:03
- Added temporary deep_scan column back to model until migration runs
- Fixed bulk insert missing has_warnings and deep_scan fields
- Fixed pending files count showing 0 (removed unnecessary is_corrupted IS NULL condition)
- Migration script to properly handle deep_scan column removal
- Migration runs automatically on startup
- No manual intervention required
- Fixes deep_scan constraint issue automatically
…rupted

- Detect truncation errors in PIL load test
- Mark files with 'bytes not processed' as corrupted
- Properly categorize incomplete image data as corruption
- Add PostgreSQL advisory lock to prevent race conditions
- Multiple workers no longer deadlock during migration
- Critical fix for startup hang issue
- Use worker ID check instead of advisory locks
- Run DDL in separate transactions
- Handle errors gracefully
- Critical fix for startup deadlock
- Disabled problematic v2.2.90 deep_scan migration that was causing startup hang
- Fixed create_performance_indexes to use separate transactions per index
- Properly handle migration errors without blocking startup
- Tested locally and confirmed app starts successfully

This resolves the app hanging on startup issue reported in production.

🤖 Generated with Claude Code

Co-Authored-By: Claude <[email protected]>
Updated version number from 2.2.92 to 2.2.93 for the startup hang fix release.

🤖 Generated with Claude Code

Co-Authored-By: Claude <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant