Hello - We are considering implementing a clustered SQL scenario over 2
different datacentre locations. At the moment we are thinking of probably
going down the hot standby situation where we run the primary server in one
datacentre and the failover in the other with some replication software
running on both servers to keep the SQL files the same in realtime. What
we'd really like to do is to have both servers live at the same time in an
Active/Active cluster but are not sure of the best way to get around the
replication issue. I know that shared storage is the norm for SQL clusters
but we really seperate storage in 2 different physical locations. Does
anyone have any advice on the best way or any options we have in tackling
this?

Thanks in advance, ian.

Hi

Veritas offers clustering that allows for Geographically dispersed nodes. We
are not using it at the moment, but might be investigated in the next few
months.

We use a 2+1 cluster. 2 Nodes of a cluster in one location, and with EMC
SRDF, we keep the 2nd location real-time updated. The 2nd location has a
single server, setup as a cluster, and if location 1 goes down, we do a
SPLIT at EMC level, and bring up the cluster. We will not loose any data.

Don't forget, with a SAN, if one node writes garbage onto the drives, your
other node is not much use to you either.

Log shipping is generally the least intrusive form of DR as replication at
data level does impose requirements on the DB and table structure

SQL Server 2005 will offer another solution, called database mirroring.

Regards
--------------------------------
Mike Epprecht, Microsoft SQL Server MVP
Zurich, Switzerland

IM: mike (AT) epprecht (DOT) net

MVP Program: http://www.microsoft.com/mvp

Blog: http://www.msmvps.com/epprecht/

"Ian Rogers" <ianr0gers (AT) hotmail (DOT) comm> wrote