There is a 5.4% performance regression in ycsb after PR#70180
- Dominant language
- Go
- Stars
- 40.5k
- Forks
- 6.2k
- PR merge metrics
- PR metrics pending
Description
## Bug Report
### 1. Minimal reproduce step (Required)
1. Deploy a cluster with 3 TiDB and 3 Tikv.
2. Run benchmark ycsb (workloada).
### 2. What did you expect to see? (Required)
No performance regression.
### 3. What did you see instead (Required)
QPS regression in benchmark ycsb (workloada) compared to baseline 44879.03 after #70180.
| description | tidb | tikv | pd | QPS |
| ----------------- | ------------------------------------------------------- | ---------------------------------------------------- | ------------------------------------------------ | ----------------------------------------------------------------------------------------------- |
| last success | pingcap/tidb@4a3336e26dcecbf425c7795713711cc91a2e8187 | tikv/tikv@8b2ddb4f1a3a48a7a752cf77551a1c0289885379 | tikv/pd@3430f76dc5a48f93cb49a14aa25f010b79af5a76 | 42829.34, 42751.54 |
| first regression | pingcap/tidb@3f3660863c5a47f8b60905e8ee4212b41956c309 | tikv/tikv@8b2ddb4f1a3a48a7a752cf77551a1c0289885379 | tikv/pd@3430f76dc5a48f93cb49a14aa25f010b79af5a76 | 42467.14, 42606.15 |
### 4. What is your TiDB version? (Required)
pingcap/tidb@3f3660863c5a47f8b60905e8ee4212b41956c309
Contributor guide
Research direction
Start by reproducing YCSB workload A on a three-TiDB, three-TiKV cluster using the commits listed in the report, then compare QPS against the stated baseline. Inspect PR #70180 and benchmark results to identify the regression; done means explaining and fixing the change without reducing workload A performance.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- go
- Domain
- databases, distributed-systems, performance
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Active
- Clarity
- Needs clarification
- Newbie friendliness
- 38/100