I have a data warehouse transactional replication environment in which
we have muliple Publisher/Distributors(100+) and a single subscriber.
All Publishers and subscribers are SQL 2000.

We recently converted the subscriber box to a Win2003
Cluster.(Publishers are Win2K for the most part, there may be a couple
remaining NT machines)

The publishers are on a very large WAN and are replicating to a central
location. Because of the large WAN environment, there are a number of
network filters in place and traffic management is of the highest
concern.

Replication IS working. However, there is one anomoly that I would
like to know about (How to control/change).

Before the Cluster, replication was moving to the DNS IP of the single
subscriber. (lets call it .14)

When the cluster was installed the SQL Cluster was given the same IP
(.14). Each of the nodes were given an IP, .72 & .70 (.72 is the active
node)

Now as replication and snapshots occur it is happening between the
publisher and the .72 node.

Why is this happening and is there a way to control this?

Thanks Rum63



*** Sent via Developersdex http://www.developersdex.com ***
Don't just participate in USENET...get rewarded for it!

Snapshots are running to a UNC file share based on the host node. You will
need to either snapshot to a clustered instance of a file share OR snapshot
to an FTP server (clustered or not is your choice).

Geoff N. Hiten
Microsoft SQL Server MVP

"rum63" <nospam (AT) devdex (DOT) com> wrote