SQL 2008 R2 Cluster installation fails

Question

I have a 2 node cluster running on Windows 2008 R2 64 bit. There are currently 10 SQL 2008 R2 Clusters running on this MSCS cluster with no problem. Some are running SP1 and others RTM. I've never had a problem adding another SQL cluster to this MSCS cluster until today. 

The installation proceeded normally on node 1 until the very end when I got a message that said the installation failed because "the resource could not be brought online". MSCS showed the group as offline but there are no errors in the SQL errorlog file. There is nothing that I can find in the installer log files either indicating a problem other than the resource not coming online. In fact it states that SQL was online and only stopped when  it received a stop request from the service control manager.

What went wrong? I followed the same procedure I've done several other times. I use the RTM media to install the new cluster then apply SP1.

Here is the output from the installer.

Description:
  SQL Server 2008 R2 Setup has encountered an error. 

Problem signature:
  Problem Event Name:	SQL100Exception
  Problem Signature 01:	SQL Server 2008 R2@RTM@
  Problem Signature 02:	0xC9090D30
  Problem Signature 03:	0xF2CBEAFB
  Problem Signature 04:	0x1C2074D8@1216@1
  Problem Signature 05:	0x26D62EE1
  Problem Signature 06:	SqlEngineConfigAction_install_startup
  Problem Signature 07:	0x26DE8310
  Problem Signature 08:	Unknown
  Problem Signature 09:	Unknown
  Problem Signature 10:	Unknown
  OS Version:	6.1.7601.2.1.0.274.10
  Locale ID:	1033

Additional information about the problem:
  LCID:	1033

Here is the final line from the error log right after getting the above...

2012-03-12 10:23:24.05 Server      SQL Server is now ready for client connections. This is an informational message; no user action is required.

The only thing I see in the detail.txt installer log that's relevant is this...
2012-03-12 10:22:49 Slp: The cluster resource 'SQL Server (BISGSQL12)' could not be brought online.  Error: The group or resource is not in the correct state to perform the requested operation. (Exception from HRESULT: 0x8007139F)
2012-03-12 10:22:49 Slp: Watson bucket for exception based failure has been created
2012-03-12 10:23:10 Slp: Error: Action "SqlEngineConfigAction_install_startup_Cpu64" failed during execution.
2012-03-12 10:23:10 Slp: Completed Action: SqlEngineConfigAction_install_startup_Cpu64, returned False

I subsequently uninstalled the SQL Cluster. Now the installer will not allow me to reinstall without rebooting the node. Can I avoid this somehow?

---

Chuck

Answer 1

 

Hello,

This issue may be related to SQL Server Browser.

http://support.microsoft.com/kb/950599

The following post may help also:

http://blogs.msdn.com/b/karthick_pk/archive/2009/03/27/installation-of-sqlserver2008-cluster-fails-on-windows2008-the-group-or-resource-is-not-in-the-correct-state-to-perform-the-requested-operation-exception-from-hresult-0x8007139f.aspx

Hope this helps.

Regards,

Alberto Morillo
SQLCoffee.com

Answer 2

If the problem were related to the browser service wouldn't all my connections to existing SQL Servers on this cluster be failing? They all need to use the browser service to get the port # right? Connections to the 10 other clustered SQL instances are working.

As to the other article, it mentions it's related to SQL 2008. Is it also applicable to SQL 2008 R2? I would think that I would have encountered this error before too if it were related to SPNs. 10 other SQL Servers were created on this cluster without an issue. Also the SQL Service account has domain and local admin privileges so it should be able to create SPNs as the article states.

I'm still at a loss as to why suddenly I cannot install a new SQL Server on this cluster.

---

Chuck

Answer 3

Figured it out. It's being caused by a DNS problem. The new SQL Server's name is not resolving to the correct address from the node of the cluster where I was installing it.

---

Chuck