FWIW, we have been running SQL server replication for a couple of years to a database in China (relatively slow line between systems). We have not been really happy with it but don't have another solution. It works but requires significant looking after. Just recently, it was "rebuilt" - hopefully to work better and resolve the errors .
________________________________
From: Nathan Andelin <nandelin@xxxxxxxxx>
To: Midrange Systems Technical Discussion <midrange-l@xxxxxxxxxxxx>
Sent: Monday, March 11, 2013 4:43 PM
Subject: Re: Bi-directional transactional replication on IBM i
Not sure how you would mirror disks located in two separate
datacenters :-)
http://tinyurl.com/b2pd9qh
That's a link to cross-site mirroring in the IBM i Information Center, if interested.
We need a share nothing type of solution like what is shown
in those diagrams I showed in the MS SQL examples.
Those workflows look like a kluge for a base runtime (MS SQL Server) which can't scale vertically. So they try to come up with a workflow that scales horizontally. Are you sure you "need" that?
User issues INSERT statement on SERVER A,
it gets replicated to SERVER B.
All IBM i HA products do that; via remote journaling; via local or remote disk mirroring.
If user issues INSERT on SERVER B, it gets replicated to SERVER A.
AFAIK the only reason you might need this might be for a horizontal scaling. The IBM i answer is to add more cores and scale vertically. NOBODY really likes duplicating database tables across servers. Again, IBM i scales vertically, without the hassle.
-Nathan.
