pingcap / pingcap/tiflash

Improve the split read tasks logic

Open
#4,407 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

type/enhancement
Dominant language
C++
Stars
1k
Forks
423
Avg merge
1d 15h
Merged PRs (30d)
24

Description

Enhancement

The current trySplitReadTasks cannot handle the case: if the expected stream num is less than segment num, the segment will become the basic unit for IO scheduling. And if there are some really large segments, the read data cannot be distributed evenly among the read threads which may cause the performance degrade.
What's more, we may also use region range as the basic IO scheduling unit in some other cases. And after dynamic region is supported, this may not continue to be a good choice. So we need to test it later after dynamic region is supported.

Contributor guide

No contributing guide indexed for this repository

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 with trySplitReadTasks in dbms/src/Storages/DeltaMerge/SegmentReadTaskPool.cpp and examine how expected stream count and segment count determine IO scheduling. Define and validate a scheduling approach that distributes large-segment reads more evenly, while considering region ranges and the future dynamic-region behavior; the issue does not specify tests or a final design.

Written by the indexing model from the issue text.

Assessment

Tech stack
cpp
Domain
databases, performance
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.