[HBASE-27686] Recovery of BucketCache and Prefetched data after RS Crash

• Type: Improvement
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 2.6.0, 3.0.0-beta-1
• Component/s: BucketCache

Release Note

This adds a background thread into the RS process, that periodically checks if there were updates in the bucket cache. If the bucket cache has been updated since the last check, it saves the bucket cache index to the file path defined by "hbase.bucketcache.persistent.path", as well as the list of completed prefetched files into the path defined by "hbase.prefetch.file.list.path" property. The thread is named as "bucket-cache-persister", and the check interval is defined by the "hbase.bucketcache.persist.intervalinmillis" property, and it defaults to 1000 (1 second). This thread is only enabled if "hbase.bucketcache.persistent.path" is set in the configuration.

• Assignee: Shanmukha Haripriya Kota
• Reporter: Shanmukha Haripriya Kota
• Created: 03/Mar/23 17:27
• Updated: 26/Mar/24 13:40
• Resolved: 16/Mar/23 16:02

Description

HBASE-27313 introduced the ability to persist a list of hfiles for which prefetch has already been completed, so the we can avoid prefetching those files again in the event of a graceful restart, but it doesn't cover crash scenarios, as if the RS is killed or abnormally stopped, the list wouldn't be saved.

This change aims to persist the list of already prefetched from a background thread that periodically checks cache state and persists the list if updates have happened.

Issue Links

• is related to: HBASE-27743 Enhancements for the persistent cache
• is related to: HBASE-28450 BuckeCache.evictBlocksByHfileName won't work after a cache recovery from file
• links to: GitHub Pull Request #5080