quickwit-oss / quickwit-oss/quickwit
Indexer and DocProcessor run on one thread [tokio issue]
Nobody has claimed this yet.
- Dominant language
- Rust
- Stars
- 11.7k
- Forks
- 597
- Avg merge
- 2d 22h
- Merged PRs (30d)
- 37
Description
Indexer and DocProcessor run on one thread, due to bad tokio scheduling (lifo slot).
Quickwit has 2 runtimes, one for blocking and one for non-blocking operations. The blocking runtime in this scenario has 14 threads assigned, which don't get utilized.
[quickwit-common/src/runtimes.rs:86] &config = RuntimesConfig {
num_threads_non_blocking: 2,
num_threads_blocking: 14,
}
disable_lifo_slot fixes the issue, but it's an tokio_unstable feature
Hotspot image, with filter on the single thread. Doc processing and indexing happens serialized on one thread (Blue/Orange pattern).
tokio-console shows DocProcessor get's barely polled.
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start in quickwit-common/src/runtimes.rs around the RuntimesConfig at line 86, then review the Tokio runtime setup and the disable_lifo_slot limitation described in the issue. Use the Hotspot and tokio-console observations as the baseline; done means indexing and document processing are no longer serialized on one thread and the configured blocking threads are utilized.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- rust
- Domain
- backend, performance
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 45/100