lance-format / lance-format/lance

Provide some way to bulk query a scalar index

Open
#4,523 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

performance
Dominant language
Rust
Stars
7.1k
Forks
852
Avg merge
3d 18h
Merged PRs (30d)
272

Description

For indexes that have pages / partitions, like BTree, each time we run a query we load the index part from cache.

This means if we have a tight loop, like during an indexed join, we keep reloading the same partition from the IndexCache. This is a lot of wasted allocations and CPU cycles.

https://github.com/lancedb/lance/blob/b3a41fdb38b0123c70256beb30ee2505acb2911f/rust/lance-index/src/scalar/btree.rs#L815-L834

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by reading the referenced section of rust/lance-index/src/scalar/btree.rs around lines 815-834 to understand how index partitions are loaded for individual queries. Trace the scalar index query entry points and existing tests, if present, before deciding how a bulk-query interface should fit. Done means tight-loop queries can reuse loaded partitions instead of repeatedly reloading them.

Written by the indexing model from the issue text.

Assessment

Tech stack
rust
Domain
databases, performance
Issue type
Feature
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.