Replacing disk system in cluster environment.

Hi!
We are planning to replace our disk system in our SQL Server cluster
environment. Our system consist of two cluster nodes and two SQL Server 2000
instances. There are only SQL Server database files including master
database in this disk system we are replacing.
Can we just shut down sql server resource from cluster administrator and
copy database files to a safety. Then replace the disk system and configure
it so that drive letters are the same as before and configure cluster to see
these drives. Then copy database files back to their old position on this
new disk system and start sql server from cluster administration. What
happens? Does this work or do we have to consider reinstalling sql server?
That would just take too much time...
Thanks
Mikko
Mikko,
Just replacing the drive letters would not be enough to have MSCS recognize
the replaced disk. You should follow the procedure outlined in the following
KB article:
http://support.microsoft.com/?id=305793
Regards,
John
"Mikko Rantonen" <mikko.rantonen@.mloy.fi> wrote in message
news:er7DvFbtEHA.2864@.TK2MSFTNGP09.phx.gbl...
> Hi!
> We are planning to replace our disk system in our SQL Server cluster
> environment. Our system consist of two cluster nodes and two SQL Server
2000
> instances. There are only SQL Server database files including master
> database in this disk system we are replacing.
> Can we just shut down sql server resource from cluster administrator and
> copy database files to a safety. Then replace the disk system and
configure
> it so that drive letters are the same as before and configure cluster to
see
> these drives. Then copy database files back to their old position on this
> new disk system and start sql server from cluster administration. What
> happens? Does this work or do we have to consider reinstalling sql server?
> That would just take too much time...
> Thanks
> Mikko
>
>
|||Thanks for your link.
Can we do it so that we first stop sql server cluster resource, then drop
all disks that we are replacing from cluster services and then when new
disks are attached just add those new drives to cluster services?
When everything is in it's place, cluster sees new drives correctly and SQL
server database files are in right places.. will it work?
Thanks
Mikko
"John Toner [MVP]" <jtoner@.DIE.SPAM.DIE.mvps.org> wrote in message
news:uzKDWoctEHA.2804@.TK2MSFTNGP14.phx.gbl...
> Mikko,
> Just replacing the drive letters would not be enough to have MSCS
recognize
> the replaced disk. You should follow the procedure outlined in the
following[vbcol=seagreen]
> KB article:
> http://support.microsoft.com/?id=305793
> Regards,
> John
> "Mikko Rantonen" <mikko.rantonen@.mloy.fi> wrote in message
> news:er7DvFbtEHA.2864@.TK2MSFTNGP09.phx.gbl...
> 2000
> configure
> see
this[vbcol=seagreen]
server?
>
|||Hi
Where is the Quorum drive? It has to be on some SAN, this one?
This may affect your planned change even more.
Regards
Mike
"Mikko Rantonen" wrote:

> Thanks for your link.
> Can we do it so that we first stop sql server cluster resource, then drop
> all disks that we are replacing from cluster services and then when new
> disks are attached just add those new drives to cluster services?
> When everything is in it's place, cluster sees new drives correctly and SQL
> server database files are in right places.. will it work?
> Thanks
> Mikko
>
> "John Toner [MVP]" <jtoner@.DIE.SPAM.DIE.mvps.org> wrote in message
> news:uzKDWoctEHA.2804@.TK2MSFTNGP14.phx.gbl...
> recognize
> following
> this
> server?
>
>
|||Quorum drive stays as it is! So drives that are been replaced contain only
sql server database files.
Mikko
"Mike Epprecht (SQL MVP)" <mike@.epprecht.net> wrote in message
news:39CB13C4-876F-4DD6-8E81-996DB884D67D@.microsoft.com...[vbcol=seagreen]
> Hi
> Where is the Quorum drive? It has to be on some SAN, this one?
> This may affect your planned change even more.
> Regards
> Mike
> "Mikko Rantonen" wrote:
drop[vbcol=seagreen]
SQL[vbcol=seagreen]
Server[vbcol=seagreen]
and[vbcol=seagreen]
cluster to[vbcol=seagreen]
What[vbcol=seagreen]
|||I have done this in the past. Replacing the drive letters will work, but
you must recreate the dependancy path in the cluster group as well. Since
there wil be new disk resources with the old letters, you will have to do
this manually.
Geoff N. Hiten
Microsoft SQL Server MVP
Senior Database Administrator
Careerbuilder.com
I support the Professional Association for SQL Server
www.sqlpass.org
"Mikko Rantonen" <mikko.rantonen@.mloy.fi> wrote in message
news:er7DvFbtEHA.2864@.TK2MSFTNGP09.phx.gbl...
> Hi!
> We are planning to replace our disk system in our SQL Server cluster
> environment. Our system consist of two cluster nodes and two SQL Server
2000
> instances. There are only SQL Server database files including master
> database in this disk system we are replacing.
> Can we just shut down sql server resource from cluster administrator and
> copy database files to a safety. Then replace the disk system and
configure
> it so that drive letters are the same as before and configure cluster to
see
> these drives. Then copy database files back to their old position on this
> new disk system and start sql server from cluster administration. What
> happens? Does this work or do we have to consider reinstalling sql server?
> That would just take too much time...
> Thanks
> Mikko
>
>

Replace SQL server in multiple server network

Hello all DBAs
what is the best methodolgy to replace a server with a new server in a
multiple sql server environment.. Is is using back up/restore to
another/bridge type server then rename servers or is it to bring in new
server then tranfer the datafiles. Does tranfering the data files
include logins and jobs ?? Or is ther a better way that I have not
realised?

Bill

*** Sent via Developersdex http://www.developersdex.com ***
Don't just participate in USENET...get rewarded for it!"William Logan" <kwiso@.msn.com> wrote in message
news:4152f0c3$0$26150$c397aba@.news.newsgroups.ws.. .
> Hello all DBAs
> what is the best methodolgy to replace a server with a new server in a
> multiple sql server environment.. Is is using back up/restore to
> another/bridge type server then rename servers or is it to bring in new
> server then tranfer the datafiles. Does tranfering the data files
> include logins and jobs ?? Or is ther a better way that I have not
> realised?

Funny you should ask that.

I'm doing that right now. The complication is one of our larger databases
has file corruption that is preventing me from doing backups.

So... for all but one user database the plan will be to do a full backup
from the source machine and then restore to the new machine.

Then, at the proper time, shut down the source machine, perform a
transaction backup, then restore that to the destination machine.

As for user logins, jobs, etc. There are DTS packages that will help.
(though the DTS job to move logins does NOT move the sid, so you either have
to script that by hand or you have to do a sp_change_users_login for each
user in each database.

MS has several KB articles you can search for that outline some of the
steps.

So far the transfer looks to be fairly straightforward (other than that one
DB I've got to do by hand.)

> Bill
> *** Sent via Developersdex http://www.developersdex.com ***
> Don't just participate in USENET...get rewarded for it!|||Thanx that is what I thought and told my cio I was told to do it another
way like copying the files over by some else I like my way better thanx
Bill

*** Sent via Developersdex http://www.developersdex.com ***
Don't just participate in USENET...get rewarded for it!|||If you can afford the downtime a simple detach and reattach should do
the trick pretty quickly with minimal overhead. You just have to wait
for the files to copy from one server to the next.

Its usually much faster than a backup and restore.

Then run some scripts to re-permission the users, You can then use DTS
packages to move your jobs and DTS packages to the new server.

Either way you need to figure out how to repermission and move jobs...
why spend the extra time on a backup and restore?|||"Bill" <nospam@.devdex.com> wrote in message
news:4154423e$0$26123$c397aba@.news.newsgroups.ws.. .
> Thanx that is what I thought and told my cio I was told to do it another
> way like copying the files over by some else I like my way better thanx

Well, it depends.

If I could afford the downtime, I'd probably detach the databases from the
primary server and copy them to the secondary and then re-attach them. Much
simpler.

But I can't afford the downtime that would require in my case.

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