Jan Wieck jan at wi3ck.info
Thu Oct 9 15:34:59 PDT 2014
On 10/09/2014 06:34 PM, Dave Cramer wrote:
> Jan,
>
> But as you said they all have to be in the same cluster, correct ?

Yes.


Jan


>
> Dave Cramer
>
> On 9 October 2014 18:31, Jan Wieck <jan at wi3ck.info
> <mailto:jan at wi3ck.info>> wrote:
>
>     On 10/09/2014 05:26 PM, Glyn Astill wrote:
>      >> From: Granthana Biswas <granthana.biswas at gmail.com
>     <mailto:granthana.biswas at gmail.com>>
>      >>To: Glyn Astill <glynastill at yahoo.co.uk
>     <mailto:glynastill at yahoo.co.uk>>
>      >>Sent: Thursday, 9 October 2014, 16:34
>      >>Subject: Re: [Slony1-general] Error with Slony replication from
>     another slony cluster slave
>      >>
>      >>
>      >>
>      >>Hi Glyn,
>      >>
>      >>
>      >>In my case I have two clusters:
>      >>
>      >>
>      >>Cluster1 ->  replicating from DB1 -> DB2
>      >>Cluster2 ->  replicating from DB1 -> DB3
>      >>
>      >>
>      >>Can I stop Cluster2 and add DB3 to Cluster1 with DB2 as its
>     master? Or do I have to delete the data first in DB3?
>      >>
>      >>
>      >
>      > You'll want to run DROP NODE against each node in Cluster2, (or
>     if on 2.0+ you can get away with just DROP SCHEMA _Cluster2 CASCADE)
>     and stop the slons for Cluster2.
>
>
>     Cascading, forwarding and all that is one of the core concepts of Slony.
>     If you create one single setup looking like this:
>
>           DB1 -> DB2 -> DB3
>
>     then you have a lot more flexibility and functionality than is obvious
>     at first. Aside from being able to fail over to DB2.
>
>     Slony will let you "MOVE" the master role from DB1 to DB2. That
>     operation will not just make DB2 the master, but at the same time DB1
>     becomes a replica that doesn't need an initial sync, so your setup now
>     would look like this:
>
>           DB1 <- DB2 -> DB3
>
>     This is useful if you need to perform some maintenance on DB1. At this
>     point DB2 is your master an you would just stop the slon process for
>     DB1, do whatever you need to do, and start the slon process again. Once
>     DB1 has caught up, you just transfer the master role back and everything
>     is as it was.
>
>     What also works is that if you need to perform maintenance on DB2, it
>     doesn't mean that DB3 has to fall behind too. You can easily change the
>     data provider for DB3 to be DB1, so your configuration changes to
>
>           DB1 -> DB2
>            |
>            V
>           DB3
>
>     At this point you stop the slon process for DB2, perform what you need
>     to do on DB2, start the slon process and after DB2 has caught up, make
>     it the data provider for DB3 again.
>
>
>     Regards,
>     Jan
>
>     --
>     Jan Wieck
>     Senior Software Engineer
>     http://slony.info
>     _______________________________________________
>     Slony1-general mailing list
>     Slony1-general at lists.slony.info <mailto:Slony1-general at lists.slony.info>
>     http://lists.slony.info/mailman/listinfo/slony1-general
>
>


-- 
Jan Wieck
Senior Software Engineer
http://slony.info


More information about the Slony1-general mailing list