[KAFKA-16541] Potential leader epoch checkpoint file corruption on OS crash

• Type: Bug
• Status: Resolved
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: 3.7.0
• Fix Version/s: 3.8.0
• Component/s: core
• Assignee: Haruki Okada
• Reporter: Haruki Okada
• Created: 12/Apr/24 01:06
• Updated: 16/Jul/24 07:18
• Resolved: 06/Jun/24 06:11

Description

Pointed out by junrao on GitHub

A patch for KAFKA-15046 got rid of fsync of leader-epoch ckeckpoint file in some path for performance reason.

However, since now checkpoint file is flushed to the device asynchronously by OS, content would corrupt if OS suddenly crashes (e.g. by power failure, kernel panic) in the middle of flush.

Corrupted checkpoint file could prevent Kafka broker to start-up

Issue Links

• links to: GitHub Pull Request #15993