Re: Innodb logfiles timestamp question
Yes that makes sense, thanks Heikki. I monitored the ib_logfiles some more and see the cycling between pairs logfile0+logfile1 and logfile0+logfile2. Heikki Tuuri wrote: John, - Original Message - From: John Thorpe [EMAIL PROTECTED] I am running 4.0.4 using innodb tables on a linux box. My innodb config is set-variable = innodb_mirrored_log_groups=1 set-variable = innodb_log_files_in_group=3 set-variable = innodb_log_file_size=500M set-variable = innodb_log_buffer_size=30M innodb_flush_log_at_trx_commit=1 innodb_log_archive=0 My question has to do with innodb's usage of the redo log files. Currently they are: 524288000 Mar 11 11:19 ib_logfile0 524288000 Mar 3 08:59 ib_logfile1 524288000 Mar 11 11:19 ib_logfile2 I always see them timestamped like this, with two having identical times, or the same within a minute or two. The particular pair of the three having the same timestamp varies. the checkpoint stamp fields are in the first ib_logfile. Does that explain the observed phenomenon? I was under the impression that mysql would cycle through these logfiles - e.g. write to logfile0 until it is full, then switch to logfile1 until full, then logfile2.., then logfile0, etc. Thanks, John Best regards, Heikki Tuuri -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]
Re: Innodb logfiles timestamp question
John Thorpe wrote: 524288000 Mar 11 11:19 ib_logfile0 524288000 Mar 3 08:59 ib_logfile1 524288000 Mar 11 11:19 ib_logfile2 I would venture a random guess that ib_logfile1 was in use until Mar 3, then ib_logfile2 began and that around Mar 11 it switched back up to ib_logfile0 where it still was when you looked at it (since you posted your message on the 12th). Given more time, I'd presume ib_logfile0 to eventually show Mar 20 or so and ib_logfile1 to show the same as it has moved on to the next file ... -- Michael T. Babcock -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]
Re: Innodb logfiles timestamp question
John, - Original Message - From: John Thorpe [EMAIL PROTECTED] Newsgroups: mailing.database.myodbc Sent: Thursday, March 11, 2004 7:34 PM Subject: Innodb logfiles timestamp question Hi, I am running 4.0.4 using innodb tables on a linux box. My innodb config is innodb_data_file_path = ibdata1:1800M;ibdata2:1800M;...ibdata10:1800M set-variable = innodb_mirrored_log_groups=1 set-variable = innodb_log_files_in_group=3 set-variable = innodb_log_file_size=500M set-variable = innodb_log_buffer_size=30M set-variable = innodb_buffer_pool_size=800M set-variable = innodb_additional_mem_pool_size=150M My question has to do with innodb's usage of the redo log files. Currently they are: 524288000 Mar 11 11:19 ib_logfile0 524288000 Mar 3 08:59 ib_logfile1 524288000 Mar 11 11:19 ib_logfile2 I always see them timestamped like this, with two having identical times, or the same within a minute or two. The particular pair of the three having the same timestamp varies. the checkpoint stamp fields are in the first ib_logfile. Does that explain the observed phenomenon? I was under the impression that mysql would cycle through these logfiles - e.g. write to logfile0 until it is full, then switch to logfile1 until full, then logfile2.., then logfile0, etc. This doesn't appear to be the case. Has anyone run across this before? (My goal is to reduce the size of the logfiles so that the time between switching is on the order of a couple hours rather than days.) Thanks, John Best regards, Heikki Tuuri Innobase Oy Foreign keys, transactions, and row level locking for MySQL InnoDB Hot Backup - a hot backup tool for InnoDB which also backs up MyISAM tables http://www.innodb.com/order.php Register now for the 2004 MySQL Users Conference! http://www.mysql.com/events/uc2004/index.html -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]
