November 20, 2009 at 2:22 pm
In order to install SQL Server 2008 Fail over you need to Slipstream SQL Server 2008 w/SP1
1. Copy your original SQL Server 2008 source media to c:\SQLServer2008_FullSP1
2. Download Service Pack 1 from http://www.microsoft.com/downloads/details.aspx?FamilyID=66ab3dbb-bf3e-4f46-9559-ccc6a4f9dc19. The three architectures of Service Pack 1 should be included, the package names are as follows:
SQLServer2008SP1-KB968369-IA64-ENU.exe
SQLServer2008SP1-KB968369-x64-ENU.exe
SQLServer2008SP1-KB968369-x86-ENU.exe
3. Extract the packages as follows:
SQLServer2008SP1-KB968369-IA64-ENU.exe /x:c:\SQLServer2008_FullSP1\PCU
SQLServer2008SP1-KB968369-x64-ENU.exe /x:c:\SQLServer2008_FullSP1\PCU
SQLServer2008SP1-KB968369-x86-ENU.exe /x:c:\SQLServer2008_FullSP1\PCU
Ensure you complete this step for all architectures to ensure the original media is updated correctly.
4. Copy Setup.exe and Setup.rll from the PCU extracted location to original source media location
robocopy C:\SQLServer2008_FullSP1\PCU c:\SQLServer2008_FullSP1 Setup.exe
robocopy C:\SQLServer2008_FullSP1\PCU c:\SQLServer2008_FullSP1 Setup.rll
5. Copy all files not the folders, except the Microsoft.SQL.Chainer.PackageData.dll, in c:\SQLServer2008_FullSP1\PCU\<architecture> to C:\SQLServer2008_FullSP1 \<architecture> to update the original files.
robocopy C:\SQLServer2008_FullSP1\pcu\x86 C:\SQLServer2008_FullSP1\x86 /XF Microsoft.SQL.Chainer.PackageData.dll
robocopy C:\SQLServer2008_FullSP1\pcu\x64 C:\SQLServer2008_FullSP1\x64 /XF Microsoft.SQL.Chainer.PackageData.dll
robocopy C:\SQLServer2008_FullSP1\pcu\ia64 C:\SQLServer2008_FullSP1\ia64 /XF Microsoft.SQL.Chainer.PackageData.dll
6. Determine if you have a defaultsetup.ini at the following locations:
C:\SQLServer2008_FullSP1\x86
C:\SQLServer2008_FullSP1\x64
C:\SQLServer2008_FullSP1\ia64
If you have a defaultsetup.ini, add PCUSOURCE="{Full path}\PCU".
NOTE: The {Full path} needs to be the absolute path to the PCU folder. If you will just be running from local folder it would be C:\SQLServer2008_FullSP1. If you will eventually share this folder out, {Full path} would be \\ServerA\SQLServer2008_FullSP1.
See question #11 here if you would like to use a relative path.
;SQLSERVER2008 Configuration File
[SQLSERVER2008]
...
PCUSOURCE="{Full path}\PCU"
If you do NOT have a defaultsetup.ini, create one with the following content:
;SQLSERVER2008 Configuration File
[SQLSERVER2008]
PCUSOURCE="{full path}\PCU"
7. run the setup.exe
8. Add Nodes to cluster then your done.
November 29, 2009 at 3:36 pm
December 3, 2009 at 3:16 pm
support.sql (11/29/2009)
HiWhat exactly is your question?
My post isn't a question, its information to the community for those going to windows server 08 R2 and clustering sql 08
December 18, 2009 at 8:44 pm
Scott,
Thanks for your post. We are planning our SQL Server 2008 installation, and this should help get that effort started. By the way, do you know if there is a practical limitation to the number of instances that can be installed on a cluster?
At this point, we are planning:
- A Development/Test cluster with 7 instances (3 instances for Development, 3 instances for Test, and 1 DBA Sandbox instance)
- A Production cluster with 3 instances
Our team is fairly new to clustering, and we would like to avoid any obvious pitfalls that we do not currently foresee.
J
December 18, 2009 at 8:52 pm
That's good Scott. Appreciate for taking time and sharing your experience.
Blog -- LearnSQLWithBru
Join on Facebook Page Facebook.comLearnSQLWithBru
Twitter -- BruMedishetty
December 22, 2009 at 8:41 am
shew (12/18/2009)
Scott,Thanks for your post. We are planning our SQL Server 2008 installation, and this should help get that effort started. By the way, do you know if there is a practical limitation to the number of instances that can be installed on a cluster?
At this point, we are planning:
- A Development/Test cluster with 7 instances (3 instances for Development, 3 instances for Test, and 1 DBA Sandbox instance)
- A Production cluster with 3 instances
Our team is fairly new to clustering, and we would like to avoid any obvious pitfalls that we do not currently foresee.
J
Depends on what edition of sql your installing, i believe standard is 2 nodes only & enterprise has no limit. If you want or need more info just PM me and i'll get back in touch with you.
We just moved to SQL 2008 active/active cluster in our prod on windows server r2.. in a few weeks we will be adding a 3rd node for olap 😀 but also adding more instances so I am playing the memory number game also :w00t:
Just make sure you set the min/max memory you want sql server to use per instance or your servers are gonna be fighting for memory, in the event of a failover and you have all instances on one server which could happen, you will need to manually change the min/max memory of sql server so you dont kill that one box in the event of a hardware/server failure.
example - this allows some breathing room incase of a failure memory wise, you will need to adjust per your environment specs I.E oltp or olap I/O cpu ect..
Node a 100gb Ram
Node b 100gb Ram
Node c 100gb Ram
Node a
10GB - OS
SQL instance 30gb
SQL instance 30gb
SQL instance 20gb
Node b
10GB - OS
SQL instance 30gb
SQL instance 30gb
Node c
10GB - OS
SQL instance 30gb
SQL instance 30gb
January 12, 2010 at 8:03 am
Have any of you guys installed SQL 2008 (SP1 slipstreamed) onto a 2008 R2 VM?
I've already installed a SQL 2008 cluster on 2 physical 2008 R2 server. Now I'm getting 'The network address is invalid.' error when running through this same setup on my VMware Workstation 7 2008 R2 setup.
There is a bug filed on the ms connect site:
https://connect.microsoft.com/SQLServer/feedback/ViewFeedback.aspx?FeedbackID=520498&wa=wsignin1.0
January 12, 2010 at 1:27 pm
tsquillario (1/12/2010)
I've already installed a SQL 2008 cluster on 2 physical 2008 R2 server. Now I'm getting 'The network address is invalid.' error when running through this same setup on my VMware Workstation 7 2008 R2 setup.
There is a bug filed on the ms connect site:
I've tested slipstreaming on a VM, but it was vsphere 4.0 - didnt have any issues
January 12, 2010 at 1:33 pm
I thought installs were meant to be getting simpler!
🙂
---------------------------------------------------------------------
January 14, 2010 at 11:18 am
I have created a 5 (hopefully soon to be 6 if I reclaim my other node from networking) node cluster. If we are talking Enterprise SQL then the limit for instances is quite high.
I already have 7 instance on my cluster (Sharepoint, SystemCenter, WebSense etc..).
Another route you could choose would be to build one slightly larger cluster and combine everything into it. With the ability to confine certain instances to certain nodes, as well as the ability to have some control in the choices the instances make during failover , including the ability to have certain instances not end up together depending on how many node failures you might have etc...
Unless you are splitting the clusters for the purposes of windows updates etc and want to keep things separated for those reasons.
A test instance by itself on a node or with development instances are completely separate from other nodes and do not affect them if issues arise.
I am not saying this route is better just another option that allows for another node utilized.
db
January 20, 2010 at 5:29 pm
I followed this and worked without problem. Same instructions from msdn; http://blogs.msdn.com/petersad/archive/2009/02/25/sql-server-2008-creating-a-merged-slisptream-drop.aspx, with pictures.
If it's of help, the reason I needed to do this was I was getting this error "The following error has occurred: There was an error setting private property 'RequireKerberos' to value '1' for resource 'SQL Network Name'. Error: Value does not fall within the expected range." as in this post; http://www.sqlservercentral.com/Forums/Topic723728-391-1.aspx This was encountered while completing a prep'd failover cluster with "Advanced Cluster Completion" option from the installer with SQL Server 2008 Enterprise on Windows Server 2008 R2 Enterprise.
Ken
-Ken
February 15, 2010 at 12:07 am
Hi guys,
Just wondering if anyone can help me with this.
I am trying to moving a node from one cluster to another over iSCSI, and the setup was like
Cluster A Cluster B
node 1 node 3
node 2 node 4
desired situation
Cluster A Cluster B
node 1 node 3
node 4 node 2
right now
Cluster B Cluster A
node 3 node 1
node 2
node 4
so far I have got node 4 with SQL uninstalled, evicted from the cluster, iSCSI disconnected from Cluster A SAN (Data, MSDTC/Quarum) and connected to the Cluster B Data, MSDTC, Quarum partition (appearing as offline due to policy)
From node 3, I have added node 4 as a new node for Cluster B
Seems all ready for "add a cluster node installation", but there are basically two issues.
1. server validate, I can skip the warning for non-MS support
2. no shared drive detected -> can not continue with install because no shared storage... <- what does this mean and how do I fix this?
Can anyone please help with what should be done to get my work continue from here? I can see the storage has now got all three node (2/3/4) as potential owner.
The SAN is iSCSI EqualLogic PX-6000.
Thanks!!
Viewing 12 posts - 1 through 11 (of 11 total)
You must be logged in to reply to this topic. Login to reply