It doesn't seem communication problem. The master server never stopped. Here is a couple
of binlog files from the master:
Below is the end of binlog 000084
# at 50000851
#110720 12:52:40 server id 3 end_log_pos 50000891 Rotate to binlog.000085 pos: 4
DELIMITER ;
# End of log file
ROLLBACK /* added by mysqlbinlog */;
/*!50003 SET COMPLETION_TYPE=@OLD_COMPLETION_TYPE*/;
Below is the beginning of binlog 000085
/*!40019 SET @@session.max_insert_delayed_threads=0*/;
/*!50003 SET @OLD_COMPLETION_TYPE=@@COMPLETION_TYPE,COMPLETION_TYPE=0*/;
DELIMITER /*!*/;
# at 4
#110720 12:52:40 server id 3 end_log_pos 106 Start: binlog v 4, server v
5.1.44-community-log created 110720 12:52:40
# Warning: this binlog was not closed properly. Most probably mysqld crashed writing it.
Any ideas?
Thanks,
Lisa
--- On Wed, 7/20/11, Johan De Meersman <vegivamp@stripped> wrote:
> From: Johan De Meersman <vegivamp@stripped>
> Subject: Re: Slave receives end packet from master
> To: database100@stripped
> Cc: "Replication" <replication@stripped>
> Date: Wednesday, July 20, 2011, 11:44 AM
> ----- Original Message -----
> > From: "Database System" <database100@stripped>
> >
> > The replication keeps on and off (Slave I/O thread
> value changes
> > between Yes and No automatically). It comes back on
> its own, but
> > since it being off so often, and doesn't have enough
> time to catch
>
> Wiggle the cable :-)
>
> More specifically, this sounds like the slave easily loses
> connection - set up network throughput graphs; have a look
> in your system logs if any down/ups appear there; see if
> there isn't a pattern in the disconnects; check for
> duplicate IPs (some host trying to steal the slave's IP);
> and do, indeed, check if the cables are securely in their
> socket.
>
>
> --
> Bier met grenadyn
> Is als mosterd by den wyn
> Sy die't drinkt, is eene kwezel
> Hy die't drinkt, is ras een ezel
>
