Re: Binlog errors

Do you have any other applications which work the SAN as hard as MySQL does? I"m not familiar with the EMC but have you run it through through a stress test? I agree that load sohuldn't cause corruption but, sad to say, in some cases it does. We have had similar difficulties with certain SANs under high stress.

• michael dykman

On 6/27/07, John Mancuso <John.Mancuso@americanhm.com> wrote:
> Ext3 fs on an emc SAN. Rhel 4 box with 4 cpus and tons of memory. I have
> no problems with other types of files- only binlogs. I am using mostly
> innodb tables.
>
> It is a high load server but that's no excuse to corrupt all the binlogs
>
>
> John Mancuso
> Linux Administrator/MySQL DBA
> IT Infrastructure
> American Home Mortgage
> w: 631-622-6382
> c: 516-652-2475
>
> -----Original Message-----
> From: Michael Dykman [mailto:mdykman@gmail.com]
> Sent: Wednesday, June 27, 2007 11:33 AM
> To: John Mancuso
> Subject: Re: Binlog errors
>
> Not that any ideas are jumping out at me but what file system are you
> employing here? and what types of MySQL tables are you using? I note
> from the listing above that you seem to be taking extremely high volumn
> (>1G in barely over 3 hours on a couple occasions). It seems a bit
> spooky to me that the majority of your logs are capped at a level just
> over 1 Gig (2^30 = 1073741824).
>
> MyISAM has been known to spontaneously corrupt it's data files under
> extremely high load although I have not heard of this happening at the
> binlog level.
>
> - michael dykman
>
> On 6/27/07, John Mancuso <John.Mancuso@americanhm.com> wrote:
> > This seems to be a frequent occurrence:
> >
> > [root@mexxxx01 dblog]# mysqlbinlog mysql.000025
> > /*!40019 SET @@session.max_insert_delayed_threads=0*/;
> > ERROR: Error in Log_event::read_log_event(): 'Found invalid event in
> > binary log', data_len: 94, event_type: 15 Could not read entry at
> > offset 4:Error in log format or read error
> >
> > This is happening to all my binlogs (other servers as well) all the
> > time. There is nothing wrong with the drive- I have other data on
> > there that never gets corrupted.
> >
> > [root@mexxxx01 dblog]# ll
> > total 4676748
> > drwx------ 2 mysql mysql 16384 May 29 11:50 lost+found
> > -rw-rw---- 1 mysql mysql 1073746319 Jun 24 00:43 mysql.000020
> > -rw-rw---- 1 mysql mysql 1073784914 Jun 24 03:50 mysql.000021
> > -rw-rw---- 1 mysql mysql 1073754672 Jun 24 06:57 mysql.000022
> > -rw-rw---- 1 mysql mysql 1073743683 Jun 26 10:24 mysql.000023
> > -rw-rw---- 1 mysql mysql 111863423 Jun 26 11:41 mysql.000024
> > -rw-rw---- 1 mysql mysql 377289867 Jun 27 10:53 mysql.000025
> > -rw-rw---- 1 mysql mysql 120 Jun 26 23:00 mysql.index
> > -rw-rw---- 1 mysql mysql 12418 Jun 26 23:47 mysql-slow.log
> > -rw-rw---- 1 mysql mysql 117 Jun 13 17:08 relay.000001
> > -rw-rw---- 1 mysql mysql 20 Jun 13 17:08 relay.index
> > -rw-r--r-- 1 root root 55 Jun 27 10:53 test.txt
> >
> >
> > What is going on here?
> >
> > John Mancuso
> > Linux Administrator/MySQL DBA
> > IT Infrastructure
> > American Home Mortgage
> > w: 631-622-6382
> > c: 516-652-2475
> >
> >
>
>
> --
> - michael dykman
> - mdykman@gmail.com
>
> - All models are wrong. Some models are useful.
>
> --
> MySQL General Mailing List
> For list archives: http://lists.mysql.com/mysql
> To unsubscribe: http://lists.mysql.com/mysql?unsub=mdykman@gmail.com
>
>

-- 
 - michael dykman
 - mdykman@gmail.com

 - All models are wrong.  Some models are useful.

-- 
MySQL General Mailing List
For list archives: 
http://lists.mysql.com/mysql
To unsubscribe:    
http://lists.mysql.com/mysql?unsub=lists@pantek.com