On 4/20/2012 7:44 AM, nixofortune wrote:
> So the actions should be:
> 1. Stop slaves
> 2. On Slaves: SET GLOBAL binlog_format=MIXED
> 3. Start slaves.
> What is "restart the replication from scratch with a binary ident copy
> Is it to use rsync to rsync mysql data on masters and slaves? and than run
> chnge the master to .... to start from zero ground?
> Many thanks
> On 20 April 2012 12:31, Reindl Harald<h.reindl@stripped> wrote:
>> Am 20.04.2012 13:17, schrieb nixofortune:
>>> So far I found :
>>> binlog_format different:
>>> Masters (MIXED)
>>> Slaves (STATEMENT)
>>> sync_binlog different on one of the slaves: 1024 against 0 on the rest
>> binlog format "statement" is practically unuseable
>> there are way too much possible queries which can not
>> work this way over the long
>> if i were you i would fix the configurations and restart
>> the replication from scratch with a binary ident copy (rsync)
The --binlog-format setting STATEMENT was not at play in the original
post. The user is using either MIXED or ROW.
Also, what setting the SLAVES are using only applies to the format of
the Binary Log created *by the slaves* and has no effect on any log they
process from their master.
There is nothing visibly wrong with your replication configuration at
this point. The problem most likely revolves around your table definitions.
MySQL Principal Technical Support Engineer
Oracle USA, Inc. - Hardware and Software, Engineered to Work Together.
Office: Blountville, TN