On 2/19/2014 10:07 AM, Johan De Meersman wrote:
> ----- Original Message -----
>> From: "shawn l.green" <shawn.l.green@stripped>
>> Subject: Re: Mysql replication setup
>> Consider this case.
> I see. That does make sense, yes. I feel that this is a bit of an edge case, though -
> replication will still halt if the slave receives a ROW update for a nonexisting table, I
Under a normal replication setup, without any filters, you are correct.
However, the original poster was setting up filters that would only
allow events affecting certain tables.
This means that those filters will ignore any ROW-formatted events
destined to tables not on his list (--replicate-do-table). When
filtering ROW-formatted changes, the only the destination of the change
is evaluated. When filtering STATEMENT-formatted changes, the context of
the change is evaluated.
MySQL Senior Principal Technical Support Engineer
Oracle USA, Inc. - Hardware and Software, Engineered to Work Together.
Office: Blountville, TN