DimitriK's (dim) Weblog

MySQL Performance: 5.5.4 @dbSTRESS

Today we have something to celebrate - the MySQL 5.5.4 was announced! And a huge number of performance improvement made in this version make me very happy :-))

I've got an immense pleasure to participate in this story and seek for the most optimal solutions together with all these talented engineers from MySQL/Sun/InnoDB/Oracle teams (and in few months all Oracle :-)) - and I think it was a huge step ahead to see how well we may all work together :-))

And now I'm happy to share with you my results obtained with MySQL 5.5.4 on dbSTRESS benchmark. All tests were executed on the 32cores Intel server running Fedora 10 and having 128GB RAM and absolutely fast internal SSD drives (and if from the memory I've used only 16GB for the buffer pools, the SSD drives helped a lot to reduce any dependency on the I/O activity).

Test scenario :

• Workload(s): Read-Only, Read-Write
• Sessions: 1, 2, 4, 8, 16, 32, 64, 128, 256
• Think time: 0 sec. (non-stop)

Tested InnoDB engine implementations :

• MySQL 5.5.4
• MySQL 5.1.45 with innodb plugin 1.0.6 (I did not test the integrated innodb because we've already demonstrated it last year  that 5.1 with integrated innodb is way slower comparing to any other InnoDB engines implementation (including innodb plugin :-))
• XtraDB-9.1

Abbreviations used in graphs :

• concurrency: innodb_thread_concurrency settings
• suffix -prg1: means engine was started with a one purge thread (to see why separated purge thread is needed and what are the benefits you may read my benchmark report when we were the first to propose and published the purge thread patch.
• X axis: number of sessions
• Y axis: number of transactions per second (TPS)

Read-Only Workload

Read+Write Workload

Relaxed Read+Write Workload (10 Reads per Write)

I'll not go too much in details for the moment, but you may already see that:

• MySQL 5.5.4 is reaching higher TPS levels than others
• MySQL 5.5.4 is better prepared now to scale up to 32 cores

Of course it's not perfect yet, but there is already a huge progress was made! Even get a look on the last year's graphs and you'll see what I mean ;-)

More data will come, so stay tuned ;-)