We currently have active/passive clustering and we are planning to tes
active/active clustering. I understand that active/active require
atleast 2 nodes and each node has an sql instance running on it. Bot
the nodes are totally independent and do not share any data o
processing what so ever. If one fails it switches over to the othe
node and now you have 2 Instances running on the same node instead o
one on each node.

My question is :

1. how the data will be insync between the nodes ? If node1 is faile
and it is not online for couple of days how will it catchup with al
the updates from Node2.
2. How is logshipping works on active/active ? If a node become
unavailable and if logshipping is setup from that node, do we need t
setup log shipping from other node ?
3. We currently have 15 databases on production server. Do we need t
have all the databases on both the nodes ?


I hope someone can help me on these issues.


Thanks,
Gopa


-
gopa
-----------------------------------------------------------------------
Posted via http://www.mcse.m
-----------------------------------------------------------------------
View this thread: http://www.mcse.ms/message441592.htm

Active-Active is a bit of a misnomer. It is a holdover from the SQL 7.0
clustering configuration. Under SQL 2000, such a system is known as
multi-instance. Each SQL instance in a cluster is an independant virtual
server, much like when you install multiple instances of SQL on a
stand-alone system. There is no sharing of database access between multiple
instances on a cluster. Each instance must have its own disk, IP, network
name, and SQL server resources. It is up to you to decide which node(s)
hold which instances under normal operating conditions. There are some
memory issues when hosting multiple instances on a single node, but
otherwise they are pretty much independant.

--
Geoff N. Hiten
Microsoft SQL Server MVP
Senior Database Administrator
Careerbuilder.com

I support the Professional Association for SQL Server
www.sqlpass.org

"gopal" <gopal.12i0z8 (AT) mail (DOT) mcse.ms> wrote

The major goal of having a cluster is to achieve high availability. Having one SQL Instance logship to another SQL Instance on the same cluster does not go very far in achieving this. If you want to use logshipping to protect your data against an SQL Instance failure, a better method is to logship from one cluster to another (ie have 2 separate clusters). This way if you loose a quorum disk or something equally bad happens on your primary cluster, you can perform a role change on the second cluster and get back into business within 30mins rather than have to wait for the first cluster to be rebuilt or repaired.