I have a couple of questions now that I installed a multi instance SQL 2K
failover cluster on Win2003 . In my case 2 instances

1) How does one apply SQL service packs or Windows service packs/security
patches to the 2 instances ?

2) If i need to restart the SQL instances, how do i do it ? I hear we should
not do it from the services or thru Service Manager ? Is that right ? If so
, how do we do it ?

3) I have manually moved and initiated failovers thru the cluster
administrator but now to get a lil more into it and kinda master it before
we go live, what do I need to do to prepare myself in supporting one. The
installation was simple and things look manageable thru Cluster Admin, but
Im sure theres lot more than that when it comes to live troubleshooting.

4) How does clustering support a db that is also a publisher and is part of
transactional replication ? anything that i need to be aware of especially
when it failsover ?

5) Do all of you'll that have setup the cluster stick to the default
settings when you look at the properties for the cluster/resource/group with
regards to thresholds,etc..

Thanks a lot in advance..

Comments Inline

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

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

[quoted text, click to view]
Read the Readme file. Service packs and hotfixes run much like the SQL
installer and handle installing to all nodes at once for a particular
instance. Therefore, you will need to apply SPs and Hotfixes to each
instance. OS security patches are obviously applied to each node
independantly.
[quoted text, click to view]
Service manager, Enterprise Mangler, or the Cluster adminstrator tool can be
used to stop and start SQL instances. Do not use the Services applet.
[quoted text, click to view]
Yep. Event logs and the cluster log are your friend when it comes to
cluster troubleshooting.

[quoted text, click to view]
Make sure the log reader task restarts, especially if your publisher and
distributer are on different instances or machines. Otherwise it is the
same as any other SQL service restart.
[quoted text, click to view]
Mostly. I changed the startup timeouts because I have hosts with 32GB of
RAM and I assign 28GB to SQL. That takes a while to initialize so I
adjusted accordingly. Otherwise, I have left the default settings alone.
[quoted text, click to view]

1. They applied on an instance by instance basis for SQL Server. Windows
goes down on each instance individually.

2. I use one of two things. Cluster Administrator or Enterprise Manager

3. That is the main thing. Basics of troubleshooting is that you do it in
the exact order the cluster was installed. That means you trouble shoot SQL
Server LAST.

4. Exactly the same as it does on a stand alone server.

5. The defaults, always.

--

Mike
Principal Mentor
Solid Quality Learning
"More than just Training"
SQL Server MVP
http://www.solidqualitylearning.com
http://www.mssqlserver.com