golemb30
asked on
SQL2005 cluster testing "Turn off each Node" reports errors
Finished installing Windows 2003 64bit Enterprise SP2, cluster and I am following the standard tests for cluster failures. I was able to successfully do the Moving Groups test via Cluster Admin from the active node to the passive and back again. No errors in the event log on either node. The initiate failure test was also successfull by failing a resource in the cluster group. No errors in the event log on either server. The problem happens when I power down the node one (node one happens to be active). Everthing fails over to the other node find, but when node one is powered up I get an error at the login screen that one or more services failed to start. If I filp the test and make node 2 active and power it down I don't get any service failed messages and the event log is clean. The event error from node one are posted below. The first error is from the application log the second is in the system log. Node one was the node that the cluster was built on and the same one SQL2005 was installed from.
Event Type: Error
Event Source: MSSQLSERVER
Event Category: (2)
Event ID: 17058
Date: 5/21/2008
Time: 10:03:46 AM
User: N/A
Computer: XX-MSSQL001
Description:
initerrlog: Could not open error log file 'K:\Microsoft SQL Server\MSSQL.1\MSSQL\LOG\ERRORLOG'. Operating system error = 3(The system cannot find the path specified.).
Data:
0000: a2 42 00 00 10 00 00 00 ¢B......
0008: 0c 00 00 00 42 00 57 00 ....B.W.
0010: 2d 00 4d 00 53 00 53 00 -.M.S.S.
0018: 51 00 4c 00 30 00 30 00 Q.L.0.0.
0020: 31 00 00 00 00 00 00 00 1.......
Event Type: Error
Event Source: Service Control Manager
Event Category: None
Event ID: 7024
Date: 5/21/2008
Time: 10:05:14 AM
User: N/A
Computer: BW-MSSQL001-N1
Description:
The SQL Server (MSSQLSERVER) service terminated with service-specific error 17058 (0x42A2).
Well I know it seems obvious but can you tell us with what account SQL Server starts? I presume you configured it to start with localsystem or domain user with local administrative rights and not with networkservice (on both nodes).
ASKER
SQL Server starts with a dedicated domain account for SQL, the account is a member of the local Administrators group on the servers.
I think I found the problem when checking to see what account SQL was running under in the services snap-in to answer your question. On the primary node SQL Server and the SQL Agent services startup where set to automatic. The second node they are set to manual. I switch the startup option to manual and restarted the active node, everything switched over to the passive node. No errors on the primary node after it rebooted. Not sure why or how these services got set to automatic startup? I will continue testing to see if this resolve the errors in the event logs.
I think I found the problem when checking to see what account SQL was running under in the services snap-in to answer your question. On the primary node SQL Server and the SQL Agent services startup where set to automatic. The second node they are set to manual. I switch the startup option to manual and restarted the active node, everything switched over to the passive node. No errors on the primary node after it rebooted. Not sure why or how these services got set to automatic startup? I will continue testing to see if this resolve the errors in the event logs.
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.