tidb: TiDB cluster is much slower using sysbench
Please answer these questions before submitting your issue. Thanks!
-
What did you do? 参考文档:https://github.com/pingcap/docs-cn/blob/master/benchmark/sysbench.md 设置了:2 tidb server/3 pd/12 tikv(3台虚拟机,每个上4个tikv实例) 测试脚本:https://github.com/pingcap/tidb-bench/tree/cwen/not_prepared_statement/sysbench new_oltp.sh 测试结果:并发1、2、4、8、16、32、164、128、256,最高tps 530,qps:10000,而虚拟机上一个单实例mysql 并发256时候tps 约3000,qps约50000,是tidb的5倍多,和你们的测试结果差异很大,测试期间io wait很低,io不是瓶颈,会是什么原因?
-
What did you expect to see?
-
What did you see instead?
-
What version of TiDB are you using (
tidb-server -V
)? 1.0
About this issue
- Original URL
- State: closed
- Created 7 years ago
- Comments: 19 (6 by maintainers)
You data is not too much. MySQL could cache most of the data in memory. So it could beat any distributed database system on this scenario.