Speed up ResumeImpl
Open
- Dominant language
- C++
- Stars
- 32.1k
- Forks
- 6.9k
- Avg merge
- 32m
- Merged PRs (30d)
- 1
Description
- We use rocksdb on a remote distributed file system. Sometimes, an IOError during `WAL` may cause the db instance to go into read-only mode，so we call `DB::Resume()` manually. However, `WaitForBackgroundWork` is required here, which may take a long time. Is it necessary in this case? Can we skip this step? Thanks !
Contributor guide
Research direction
Start by reading the DB::Resume() implementation and its WaitForBackgroundWork path, then trace how WAL I/O errors transition a database into read-only mode. Determine whether skipping the wait preserves safe recovery for remote distributed filesystems, and validate the chosen behavior with the relevant recovery tests if available.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- cpp
- Domain
- database
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100