Monday, July 1, 2013

Update SQL Server Collation for a Clustered Installation

I been in a situation with one of my clients where I been implementing and Active/Active clustering setup over SQL Server 2008 R2 as my customer is still using GP 10.0, after finalizing and testing everything I noticed that the code page I must use is 1256 while I have used 1250!!!

Thanks god for having a utilities to update the collation without having to destroy the cluster! I have followed the steps below and it worked like a charm:

1. Backup your objects! As this process will drop logins, linked servers, jobs, and anything related to the instant itself plus will de-attach your database.

2. Make sure that your current server is online and carrying your SQL Instant.

3. Make sure that SQL Services is offline.

4. Run the following command:


E:\SQL\setup.exe will need to be replaced with the locations of the SQL Server 2008 R2 setup.exe path.

MSSQLSERVER represents the default instance, if you have a named instance this must be changed to hold the instance name.

P@ssW0rd: is the new SA user password, it must be strong password or otherwise the repair will fail.

Arabic_CI_AI: is the new collation name.

DAOUD-PC\DAOUD need to be replaced with the local administrator of the server or PC you are installing the SQL on.

Paste above command into CMD and wait for 3-4 minutes, you will have your SQL Collation changed.

Warning: this operation will recreate your master database, were all existing setting will be reset, if you had databases attached it will be de-attached and will be found in the database folder.


Mohammad R. Daoud MVP - MCT
+962 - 79 - 999 65 85


Ascya Bishara said...

Hi Mohammad

That's an interesting read on the clustered installation. I am more of an infrastructure guy and am preparing a HA SQL environment to host multiple GP Databases as part of centralizing.
I will be running a Hyper-V 2012 R2 Clustered Hosts with VMs including 2 Clustered SQL servers, terminal server for GP and others, Virtual switching for some performance.
For this comment I am more concerned about the SQL Cluster and your post is the only post I have seen with an Active/Active cluster. Technically I would think it could be done as long as the shared volume is accessible by both nodes and the GP install files are also on a shared volume accessible to both nodes.
Is another key part the cluster group and resources you add to ensure a successful failover.

Look forward to hearing from you.

Ascya Bishara

xj900 said...


The following error occurred:
The SQL Server resource group must be owned by the current cluster node. Setup requires access to shared cluster resources.

Error result: -2067920928
Result facility code: 1214
Result error code: 2016

Please review the summary.txt log for further details

Related Posts:

Related Posts with Thumbnails