Re: MySQL 5.7 Innodb performans issue
Found the cause. sync_binlog was set to 1. I suspect that the default value is 5.5 was 0 and that is has changed to 1 sometime after that. Setting it to 0 boosted the performance back to normal (4x speed) and the HD LED indicated much lower stress on the hard disk. Found this after I found out what caused it: https://www.percona.com/blog/2009/01/21/beware-ext3-and-sync-binlog-do-not-play-well-together/ I suspect that this also apply to ext4, or? -- Jørn Dahl-Stamnes homepage: http://photo.dahl-stamnes.net/ -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe:http://lists.mysql.com/mysql
Re: MySQL 5.7 Innodb performans issue
Hi Jørn, Found this after I found out what caused it: https://www.percona.com/blog/2009/01/21/beware-ext3-and-sync-binlog-do-not-play-well-together/ I suspect that this also apply to ext4, or? I would go more specific and say that sync_binlog=1 does not play well with single-threaded workloads. MySQL 5.6 and above can group commit to the binary log (reducing a lot of IO if transactions arrive at the same time). I have an explanation on how it works here: http://www.tocker.ca/2014/12/30/an-easy-way-to-describe-mysqls-binary-log-group-commit.html -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe:http://lists.mysql.com/mysql
