May 1, 2009 at 12:33 pm
I'm rolling out a project plan to migrate several SQL 2000 servers into a beefy consolidated SQL 2008 box and using clustering in 2008 server. I'm semi-familiar with Windows clustering in using Hyper-V, but not all with SQL.
Right now, my SQL boxes have the OS, Data, Logs and backups on the SAN while the SQL files (and system db's) and page file on local RAID 1 drives.
The plan is to have multiple instances from the various upgraded servers to run on a the single SQL 2008 box while managing the memory with each instance. From what I've read, that shouldn't be a problem.
My big question is, from a clustering perspective, the best practices to make sure the performance can be at its best.
I'm thinking having OS/Page file/backups local and then have Data/Logs/SQL/System databases on the SAN. I guess I didn't know how this would work out in a active/passive cluster.
Am I on the right path or does anyone have thoughts or suggestions? I'm looking to purchase hardware soon, and I want to make sure I have the appropriate components (NICs, HBA, local storage, etc.)
May 6, 2009 at 9:27 am
I'm confused, are you looking to use HyperV and VMs for clustering?
Or just trying to lay out the drives for clustering?
There's nothing magic about SQL clustering. It sits on Windows clustering, same quorums, shared resources, etc.
For disk layout, lay out the disks as you would for a non-clustered server.
Follow me on Twitter: http://www.twitter.com/way0utwest
Forum Etiquette: How to post data/code on a forum to get the best help
My Blog: www.voiceofthedba.com
May 6, 2009 at 9:35 am
Agree with Steve plus one more question: why you are going to have backups locally? How you are going to get backup if the box is not alive? Backup should be done to SAN, then copied to tape or any other media, and then moved off site...
Alex Prusakov
May 7, 2009 at 4:51 pm
Does anyone have any instructions for setting up the Cluster on SQL 2008 from scratch. Good references for this and what is required - SAN/ IP addresses for private, public etc.
Cheers
May 7, 2009 at 6:11 pm
Hello
I am in the same boat. I found the following links useful.
Installing Clustered SQL Servers - Outline, Checklists, Document Sheets
Getting Started with SQL Server 2008 Failover Clustering
http://msdn.microsoft.com/en-us/library/ms189134.aspx
Before Installing Failover Clustering
http://msdn.microsoft.com/en-us/library/ms189910.aspx
HTH
gk
May 7, 2009 at 6:45 pm
Thanks for the links...if i find more i post them also.
May 7, 2009 at 9:18 pm
Your high level description sounds fine with the exception of putting the backups on the SAN.
If you loose your cluster node, your backups will probably be unavailable too. Remember, you can keep your backup on a lower tier of storage as it doesn't have the same performance requirement as your data and log disks.
--
Andrew Hatfield
May 7, 2009 at 9:26 pm
Steve Jones - Editor (5/6/2009)
I'm confused, are you looking to use HyperV and VMs for clustering?
In Hyper-V, Microsoft have used MSCS as their solution provider to deliver HA and Quick Migrate
--
Andrew Hatfield
May 8, 2009 at 8:42 am
May 8, 2009 at 2:21 pm
Reading
http://www.mssqltips.com/tip.asp?tip=1687
When it mentions the Nodes i am thinking this means the server name itself.
Add Target Portal enter DNS of your iSCSI Target (Is this the IP address of the Servername).
or IP of the DISK Storage.
In the iSCSI Initiator Properties page, click on the Discovery tab -- target portal, add portal
What is the Application Server Role and you install this on each servername (node).
Here is a good link on the windows clustering set up.
http://blogs.msdn.com/clustering/archive/2008/01/18/7151154.aspx
May 12, 2009 at 4:22 pm
Windows 2008 Cluster for SQL 2008 Failover.
When configuring the DISKS for active/passive.
Do you configure the DISK Management for the SAN identicial on both NODES.
i.e if have disk F, K, L do you then do this on NODE 2 the same configuration.
Viewing 11 posts - 1 through 10 (of 10 total)
You must be logged in to reply to this topic. Login to reply