Activity log for bug #1163262

Date Who What changed Old value New value Message
2013-04-02 11:55:49 Alexey Kopytov bug added bug
2013-04-02 11:56:01 Alexey Kopytov nominated for series percona-server/5.1
2013-04-02 11:56:01 Alexey Kopytov bug task added percona-server/5.1
2013-04-02 11:56:01 Alexey Kopytov nominated for series percona-server/5.5
2013-04-02 11:56:01 Alexey Kopytov bug task added percona-server/5.5
2013-04-02 11:56:01 Alexey Kopytov nominated for series percona-server/5.6
2013-04-02 11:56:01 Alexey Kopytov bug task added percona-server/5.6
2013-04-02 11:56:10 Alexey Kopytov percona-server/5.1: status New Won't Fix
2013-04-02 11:56:12 Alexey Kopytov percona-server/5.5: status New Triaged
2013-04-02 11:56:15 Alexey Kopytov percona-server/5.5: importance Undecided Medium
2013-04-02 11:56:19 Alexey Kopytov percona-server/5.6: status New Invalid
2013-04-02 11:56:25 Alexey Kopytov percona-server/5.5: assignee Alexey Kopytov (akopytov)
2013-04-02 11:56:29 Alexey Kopytov percona-server/5.5: milestone 5.5.30-30.2
2013-04-02 15:32:49 Alexey Kopytov percona-server/5.5: status Triaged In Progress
2013-04-02 15:36:30 Alexey Kopytov description log_sys->log_flush_order_mutex was introduced as an optimization to reduce contention on log_sys->mutex. On a mini-transaction commit InnoDB adds modified pages to the flush list and uses log_sys->log_flush_order_mutex to make sure pages are added to the flush list in the correct LSN order. One thing that was handled inefficient wrt. that mutex is that it was acquired even if no modifications from a mini-transaction had to be added to flush list, i.e. when an mtr has only modified dirty pages (if any). This unnecessarily increased contention on log_sys->log_flush_order_mutex and consequently on log_sys->mutex in write-intensive workloads, because the former is acquired with the latter locked. The problem is fixed in MySQL 5.6 with the following revision: http://lists.mysql.com/commits/134048 This report is to backport that change to Percona Server 5.5. log_sys->log_flush_order_mutex was introduced as an optimization to reduce contention on log_sys->mutex. On a mini-transaction commit InnoDB adds modified pages to the flush list and uses log_sys->log_flush_order_mutex to make sure pages are added to the flush list in the correct LSN order. One thing that was handled inefficiently wrt. that mutex is that it was acquired even if no modifications from a mini-transaction had to be added to flush list, i.e. when an mtr has only modified dirty pages (if any). This unnecessarily increased contention on log_sys->log_flush_order_mutex and consequently on log_sys->mutex in write-intensive workloads, because the former is acquired with the latter locked. The problem is fixed in MySQL 5.6 with the following revision: http://lists.mysql.com/commits/134048 This report is to backport that change to Percona Server 5.5.
2013-04-02 15:39:18 Alexey Kopytov percona-server/5.5: status In Progress Fix Committed
2013-04-02 15:39:23 Alexey Kopytov branch linked lp:~akopytov/percona-server/bug1163262-5.5
2013-04-03 04:21:36 Laurynas Biveinis tags xtradb
2013-04-03 11:49:28 Stewart Smith percona-server/5.5: status Fix Committed Fix Released