Comments on: Crash-resistant replication: How to avoid MySQL replication errors https://www.percona.com/blog/crash-resistant-replication-how-to-avoid-mysql-replication-errors/ Mon, 01 Dec 2014 09:57:41 +0000 hourly 1 https://wordpress.org/?v=6.5.2 By: das-ich https://www.percona.com/blog/crash-resistant-replication-how-to-avoid-mysql-replication-errors/#comment-10255861 Mon, 01 Dec 2014 09:57:41 +0000 https://www.percona.com/blog/?p=16403#comment-10255861 Hi.
relay-log-recovery not workig for me and mysql stop slave with error Last_Errno: 1032.

mysql> show global variables like ‘relay_log_recovery’;
+——————–+——-+
| Variable_name | Value |
+——————–+——-+
| relay_log_recovery | ON |
+——————–+——-+
1 row in set (0,00 sec)

What else can I do?

]]> By: Muhammad Irfan https://www.percona.com/blog/crash-resistant-replication-how-to-avoid-mysql-replication-errors/#comment-2014039 Tue, 16 Jul 2013 13:27:01 +0000 https://www.percona.com/blog/?p=16403#comment-2014039 Patryk,

It’s absolutely fine to use innodb_overwrite_relay_log_info in a master-master setup.

]]> By: Andrew Shieh https://www.percona.com/blog/crash-resistant-replication-how-to-avoid-mysql-replication-errors/#comment-2010503 Tue, 16 Jul 2013 00:32:40 +0000 https://www.percona.com/blog/?p=16403#comment-2010503 And beware this bug: https://bugs.launchpad.net/percona-server/+bug/1092593

Run Percona 5.5.31 or greater.

]]> By: Mark Callaghan https://www.percona.com/blog/crash-resistant-replication-how-to-avoid-mysql-replication-errors/#comment-2008104 Mon, 15 Jul 2013 16:36:04 +0000 https://www.percona.com/blog/?p=16403#comment-2008104 AFAIK, it can improve performance on the slave. With rpl_transaction_enabled from the Google/Facebook patches putting slave state into InnoDB means that fsync on commit is not needed by the slave SQL thread. Slave SQL thread state and InnoDB recover to the same point in time whether or not fsync is done. If a few commits are lost from InnoDB on the crash then they are also “lost” from the slave state and the slave SQL thread begins where it should begin.

]]> By: Jacky Leung https://www.percona.com/blog/crash-resistant-replication-how-to-avoid-mysql-replication-errors/#comment-2006539 Mon, 15 Jul 2013 12:51:04 +0000 https://www.percona.com/blog/?p=16403#comment-2006539 This is a nice features, but I wonder will this cause performance problem on slave? Because in theory if we are putting replication data into innodb log file transactionally (which I assume it is innodb log buffer is being used). In this case will there more flushing being done?

]]> By: Patryk Pomykalski https://www.percona.com/blog/crash-resistant-replication-how-to-avoid-mysql-replication-errors/#comment-2006217 Mon, 15 Jul 2013 12:05:23 +0000 https://www.percona.com/blog/?p=16403#comment-2006217 Is it ok to use innodb_overwrite_relay_log_info in a master-master setup?

]]>