Which is More Powerful: TiDB or MySQL in Terms of Performance?

Note:
This topic has been translated from a Chinese forum by GPT and might contain errors.

Original topic: tidb与MySQL性能究竟谁更强

| username: TiDBer_djgos04V

[Test Environment for TiDB]
How is TiDB stronger compared to MySQL?
Is TiDB a parallel database?
Under the same configuration on a single machine, will TiDB perform better than MySQL?

| username: tidb菜鸟一只 | Original post link

A single machine definitely makes MySQL stronger; the benefits of a distributed database only become apparent when the data volume reaches hundreds of millions.

| username: zhanggame1 | Original post link

Single-node deployment of MySQL is much stronger than TiDB.

| username: 大飞哥online | Original post link

In terms of a single machine, TiDB and MySQL are not in the same league; MySQL is much stronger. :joy:

| username: 大飞哥online | Original post link

If it can be solved with a single machine, definitely don’t use multiple machines, hahaha.

| username: Kongdom | Original post link

Here’s a simple example: TiDB can replace MySQL, but MySQL cannot replace TiDB :yum:

| username: ajin0514 | Original post link

So, who is stronger after all?

| username: TiDBer_djgos04V | Original post link

Are there any scenarios where TiDB can be tested to show performance 5-6 times higher than MySQL under the same conditions using sysbench? :cry:

| username: 有猫万事足 | Original post link

It’s quite strange to compare these two. If you want to compare, it should be with other distributed databases. The problems that MySQL can solve don’t require TiDB. Do things according to the scale.

| username: 大飞哥online | Original post link

The scenarios being addressed are different, and the business situations are also different, so it’s not appropriate to compare them this way.

| username: Fly-bird | Original post link

Distributed definitely TiDB

| username: TiDBer_djgos04V | Original post link

But I can’t help it. Now I’m required to solve the issue of TiDB performing worse than MySQL. They say the results I got are due to my own lack of rigorous testing. I’m also very helpless :cry:

| username: TiDBer_vfJBUcxl | Original post link

Single-machine MySQL is strong.

| username: 有猫万事足 | Original post link

According to the standard deployment method, the measured QPS/TPS should not differ significantly from the documentation. First, check how much the values differ from those in the documentation.

https://docs.pingcap.com/tidbcloud/v7.1.0-performance-benchmarking-with-sysbench

If you insist on comparing performance with MySQL, there’s actually no need to switch. It can still manage.
For those who can’t manage, they’ve long been unable to tolerate MySQL due to reasons like inconsistent reads, write capacity limitations, and other various issues.

| username: zhanggame1 | Original post link

Then you can use TiDB’s built-in benchmarking tool to run the same test on both MySQL and TiDB.
How to Benchmark TiDB Using Sysbench | PingCAP Documentation Center

| username: zhanggame1 | Original post link

There are also How to Perform TPC-C Testing on TiDB | PingCAP Documentation Center
How to Perform CH-benCHmark Testing on TiDB | PingCAP Documentation Center

| username: 大飞哥online | Original post link

Too absolute, hahaha.

| username: TiDBer_djgos04V | Original post link

I ran the exact same test and looked at the results, but they said my testing process had issues. TiDB couldn’t possibly be slower than MySQL, implying that my configuration was problematic. :cry:

| username: zhanggame1 | Original post link

Could you please share the completed test report and the test environment for everyone to analyze?

| username: Jellybean | Original post link

Every database is powerful within its specific applicable scope. If you want to make a comparison, you need to set business use scenarios, such as business volume, usage methods, scale, required functions, etc.

There is no silver bullet in databases; good or bad is relative.