My understanding is that relay-info.log stores Exec_Master_Log_Pos so
in order to only use master.info in a snapshot you need to make sure
that the slave is caught up when the snapshot is taken. It's common
for the I/O thread (controls master.info and downloads logs files) to
be a few queries ahead of the SQL thread (controls relay-log.info and
If I'm wrong here please correct me. If you have Exec_Master_Log_Pos
you can use it for Read_Master_Log_Pos on change master command on the
new slave so it starts downloading relay logs at the proper position.
On Mon, 14 Feb 2005 15:55:49 -0800 (PST), Atle Veka
> On Mon, 14 Feb 2005, Eric Bergen wrote:
> > Did you copy the relay-log.info and relay logs from the seed slave?
> > Relay_Master_Log_file is line 3 in the relay-log.info file. This file
> > is managed by the SQL thread and helps the slave keep track of what it
> > has execute in the relay logs.
> > http://dev.mysql.com/doc/mysql/en/slave-logs.html
> > -Eric
> Hi Eric, I do not include the relay logs/relay-log.info when I seed slaves
> as from what I understand the new slave will start populating new relay
> logs starting at the position from master.info.