pingcap / pingcap/tiflash

Disagg: evict read node local cache when the file is not used anymore

Open
#7,306 3 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

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

Description

Bug Report

Please answer these questions before submitting your issue. Thanks!

1. Minimal reproduce step (Required)

insert some table and select count(*) using tiflash. deploy tiflash with s3 using minio
set read node remote cache 800G

2. What did you expect to see? (Required)

The remote cache can be evict and tiflash run stable

3. What did you see instead (Required)

The remote cache can't not be evicted

[2023/04/18 13:29:38.322 +08:00] [WARN] [PathCapacityMetrics.cpp:306] ["No available space for path: /data0/tiflash/cache, capacity: 800.00 GiB, used: 802.70 GiB"] [thread_id=2313]
[2023/04/18 13:29:48.323 +08:00] [WARN] [PathCapacityMetrics.cpp:306] ["No available space for path: /data0/tiflash/cache, capacity: 800.00 GiB, used: 802.70 GiB"] [thread_id=2313]
[2023/04/18 13:29:58.324 +08:00] [WARN] [PathCapacityMetrics.cpp:306] ["No available space for path: /data0/tiflash/cache, capacity: 800.00 GiB, used: 802.70 GiB"] [thread_id=2313]
[2023/04/18 13:30:08.325 +08:00] [WARN] [PathCapacityMetrics.cpp:306] ["No available space for path: /data0/tiflash/cache, capacity: 800.00 GiB, used: 802.70 GiB"] [thread_id=2313]
[2023/04/18 13:30:18.326 +08:00] [WARN] [PathCapacityMetrics.cpp:306] ["No available space for path: /data0/tiflash/cache, capacity: 800.00 GiB, used: 802.70 GiB"] [thread_id=2313]

TiFlash seems hang when the remote cache capacity is full
image

4. What is your TiFlash version? (Required)

master

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

Reproduce the full-cache case with TiFlash, S3 and MinIO using the reported remote-cache settings, then inspect PathCapacityMetrics.cpp around line 306. Confirm that unused local cache data is evicted when capacity is reached and that TiFlash continues running without repeated no-space warnings.

Written by the indexing model from the issue text.

Assessment

Tech stack
cpp
Domain
databases, performance
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
38/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.