Redis: X Autoclaim Slowdown Fix

A single focused fix addresses a major X Autoclaim regression where sparse pending lists over wide stream entries ran up to 23 times slower. Pull request 15893 replaces the costly last-entry scan with a constant-cost index lookup.

Duration: PT1M59S

Episode overview

This episode is a short developer briefing from Redis.

It explains recent repository work in plain language.

  • Show: Redis
  • Published: 2026-10-01T13:17:01Z
  • Audio duration: PT1M59S

Transcript excerpt

This excerpt keeps the crawler page concise. Listen to the episode or use the RSS feed for the full update.

Good morning, it's October 1st, 2026, and this is your Redis developer briefing.

No broader pattern today — just one high-impact performance fix for streams.

The takeaway is that X auto claim recovery gets its speed back when pending lists are sparse and stream entries are large.

As background, linked issue 15505 added a check to see if a pending entry target was past the current list pack node by reading that node's last entry ID. The problem, flagged in pull request 15893, is that the edge lookup walks backward over every field of that last entry. If your pending list only touches a few…

The fix reframes the boundary test. Instead of inspecting the last entry in the current node, it looks ahead to the first ID in the next index node. That costs one index step per node, independent of how large entries are. In other words, sparse recovery no longer pays a penalty for wide data.

For developers, the impact is reliability and latency in consumer group failover. X auto claim is often on the critical path when reclaiming idle messages, and a 23-fold regression there can stall lagging consumers and cascade into timeouts. This change restores predictable iteration cost, especially for…

Nearby episodes from Redis

  1. Edge-Case Correctness, Faster ZSCAN, OpenSSL 4 Fix
  2. Clock Races, Restart Safety, and CLI Fixes
  3. Clean Reconnects and Louder Timeouts
  4. Weekly Recap - Test Harness Overhaul and Reliability Hardening
  5. Cluster Test Consolidation
  6. Array and Hash Correctness Fixes
  7. Hot-Path Performance Cuts and Sync Fixes
  8. Silent Failures and AOF Reliability Fixes