Also, it won't help you much as it's only prototype code but it looks like this feature
might have been on interest...
http://www.clusterdb.com/mysql-replication/software-preview-mysql-scriptable-replication/
Any feedback on whether this would have fit the bill would be appreciated (especially if
there are tweaks that you'd suggest).
Regards, Andrew.
> -----Original Message-----
> From: Andrew Morgan
> Sent: 14 September 2010 08:10
> To: Ritesh Nadhani
> Cc: replication@stripped
> Subject: RE: Filtered data replication
>
> How about actually storing the data in different tables and then
> provide a consolidated view on the master by creating a view? You could
> then control which slave applied the changes from which of these
> tables.
>
> Regards, Andrew.
>
> > -----Original Message-----
> > From: Johan De Meersman [mailto:vegivamp@stripped]
> > Sent: 14 September 2010 08:06
> > To: Ritesh Nadhani
> > Cc: replication@stripped
> > Subject: Re: Filtered data replication
> >
> > On Tue, Sep 14, 2010 at 8:30 AM, Ritesh Nadhani <riteshn@stripped>
> > wrote:
> >
> > > On Mon, Sep 13, 2010 at 11:29 PM, Johan De Meersman
> > <vegivamp@stripped>
> > > wrote:
> > > > Well, traditionally, partitioning :-)
> > > >
> > > > Bandwidth, storage, ... ?
> > > >
> > >
> > > Well, without getting into much details, it mostly about security.
> > >
> >
> >
> > Unless you're worried about the actual DBAs, I see no issue in
> granting
> > a
> > user select only on a single partition instead of the global table.
> > Don't
> > forget, a "partitioned table'" is just a layer, and the partitions
> are
> > actual, real tables that may be queried directly.
> >
> > If you *are* worried about your DBAs, you have bigger problems than
> > partitioning :-)
> >
> > --
> > Bier met grenadyn
> > Is als mosterd by den wyn
> > Sy die't drinkt, is eene kwezel
> > Hy die't drinkt, is ras een ezel