There was sort of a bug which was fixed in MySQL 5.5 with replication heartbeat. Before
the replication heartbeat, a new relay log file would be created after every
"slave_net_timeout". It doesn't have any negative impact though.
Hope that helps.
From: Akshay Suryavanshi <akshay.suryavanshi50@stripped>
To: Wayne Leutwyler <wleutwyl@stripped>
Sent: Wednesday, January 9, 2013 1:42 AM
Subject: Re: Relay log Question
Also, you may want to see, if at all new file is really getting every hour
exactly, if any cron'd script runs, which executes "flush logs" on the
slave server. That will also rotate relay log.
On Wed, Jan 9, 2013 at 1:35 AM, Akshay Suryavanshi <
> Please re-phrase your question. The relay logs are created as and when
> required by the Slave_SQL thread. Once all the events in the relay logs are
> executed the relay log would be purged by the Slave_SQL thread.
> By setting relay_log_purge=0 you are disabling this automatic purge
> option. So the new relay log files will be created however the older ones
> would not be deleted.
> Creation or rotation of relay or binary logs is not time-based, for some
> situations it might create in one hour, however for many others it might
> create after 1 day, week, or even months.
> Usually the relay logs in your case should be of definite size, check out
> this setting "max_relay_log_size" and "max_binlog_size", the latter would
> come into picture if the prior one is disabled.
> Akshay S
> On Wed, Jan 9, 2013 at 1:21 AM, Wayne Leutwyler
>> Hello List,
>> Quick question. I am running MySQL 5.1.66, and I noticed that every hour,
>> my slave is creating a new relay log. I added this:
>> relay_log_purge = 0
>> to the my.cnf and it did not solve the issue.
>> What am I missing? Never seen this issue before.
>> "Courage is being scared to death, but saddling up anyway." --John Wayne