hello All,
From internet some community post, ext4 + InnoDB may cause jbd2 (journal block device) high I/O usage.
From MySQL documentation, InnoDB, by default, commit every transaction into file system (i.e. innodb_flush_log_at_trx_commit = 1 , default)
If set innodb_flush_log_at_trx_commit to 0 or 2 , shall it have better performance, is it?
Does it affect existing data?
Can anyone share your experience?
(one of our server running MySQL 5.5 on ext4, and some InnoDB database, it has some load average. So looking for ways to reduce it)
ref.:
http://dev.mysql.com/doc/refman/5.5/en/innodb-parameters.html#sysvar_innodb_flush_log_at_trx_commit
http://themech.net/2011/10/struggling-with-mysql-performance-problems-under-linux/
http://forums.devshed.com/mysql-help-4/effect-filesystems-mysql-performance-948013.html
Thank you very much for your kind attention.
Regards
From internet some community post, ext4 + InnoDB may cause jbd2 (journal block device) high I/O usage.
From MySQL documentation, InnoDB, by default, commit every transaction into file system (i.e. innodb_flush_log_at_trx_commit = 1 , default)
If set innodb_flush_log_at_trx_commit to 0 or 2 , shall it have better performance, is it?
Does it affect existing data?
Can anyone share your experience?
(one of our server running MySQL 5.5 on ext4, and some InnoDB database, it has some load average. So looking for ways to reduce it)
ref.:
http://dev.mysql.com/doc/refman/5.5/en/innodb-parameters.html#sysvar_innodb_flush_log_at_trx_commit
http://themech.net/2011/10/struggling-with-mysql-performance-problems-under-linux/
http://forums.devshed.com/mysql-help-4/effect-filesystems-mysql-performance-948013.html
Thank you very much for your kind attention.
Regards