[quoted text, click to view]

Why don't you just add the two new servers into the existing cluster,
install SQL on them, test, and then evict the old nodes?


--
Russ Kaufmann
MVP - Windows Server - Clustering
ClusterHelp.com, a Microsoft Certified Gold Partner
Web http://www.clusterhelp.com
Blog http://msmvps.com/clusterhelp

The next ClusterHelp classes are:
Dec 10 - 13 in Denver
Jan 21 - 24 in New York

We have 1 current Windows 2003/SQL 2000 Cluster (CL1).
We are purchasing 2 more servers that are targeted to replace (CL1) for
performance reasons.

We have 2 goals.
1) Initially, bring up the 2 new servers as a new cluster (CL2) and migrate
all data from (CL1) in order to properly test performance.

2) Ultimately, bring up the 2 new servers as (CL1) and migrate all data from
current
(CL1) in order to go-live production.

Question: In scenario 1, after we install SQL 2000 with SQL Virtual as CL2
and install all needed Service Packs, patches, etc., can we move all (CL1)'s
system and
use databases over to the new cluster by taking CL1 services offline, and
file copying all system and user mdf/ldf to the new (CL2)? I'm afraid of any
differences
that might exist between the old server being known as CL1 and the new
server being known as CL2 and being started with CL1's master,msdb, etc.
Obviously,
any 4 part naming convention which includes the old server name (CL1) in any
views, sp, triggers, UDF's would need to be addressed. But what about system
databases.

Question: In scenario 2, If we shut-down (CL1), then install SQL2000 virtual
on the 2 new servers with the CL1 virtual name, and we copy the files over the
same way as in scenario 1, would we be good to go?

Are there any other suggestions as to how to accomplish our 2 goals.

Don't servers in the same cluster point to the same data on the SAN?
We have created additional RAID arrays (more disks) to improve performance
in addition to the beefed up servers. We plan on copying the old data to the
new
arrays for performance, and also we would like a week or so to test.

If all 4 servers were in the same cluster, and we failed over to one of the
new servers, we would be messing with production data.


[quoted text, click to view]