The SAServer agent controls the failover of a SQL Anywhere database server to another node in the cluster.
Shut down all SQL Anywhere database servers running on nodes in the cluster.
Choose a node in the cluster and create a directory named SAServer under the %VCS_HOME%\bin directory on that node. You will see other Veritas Cluster Server agents within this folder (such as NIC and IP).
Copy the following files from the %SQLANY12%\VCSAgent\SAServer directory to the SAServer directory you created in Step 2:
Copy the file %VCS_HOME%\bin\VCSdefault.dll into the %VCS_HOME%\bin\SAServer directory and rename it to SAServer.dll.
Copy the file %SQLANY12%\VCSAgent\SAServer\SAServerTypes.cf into the %VCS_HOME%\conf\config directory.
Repeat Steps 1-5 for all other nodes in the cluster.
Start the Veritas Cluster Server Manager and enter your user name and password to connect to the cluster.
Add the SAServer agent:
Click File » Import Types.
Navigate to %VCS_HOME%\conf\config\SAServerTypes.cf, and then click Import.
Start the Veritas Cluster Server Manager and enter your user name and password to connect.
Add SAServer as a resource to a service group:
Click Edit » Add » Resource.
In the Resource Type list, click SAServer.
On Windows, if SAServer does not appear in the Resource Type list under Windows, you may have to add the SAServer.xml file to the %VCS_ROOT%\cluster manager\attrpool\Win2K\400 and restart the cluster services.
In the Resource Name field, type a name.
Add the following attribute values to the following attributes:
cmdStart dbsrv12 -x tcpip database-file-on-shared-disk -n server-name
cmdMonitor dbping -c "Server=server-name"
cmdStop dbstop -c user-id,password -y
This indicates that the resource is ready to be used.
Ensure that the resource dependencies are configured correctly. There are other resources that must be started and grouped together before SAServer can be started, such as the shared disk resources and the IP address resources.
Right-click the service group and click Online » node-name, where node-name is the name of the computer in the cluster on which you want the resource to run.
The service group is now online.
Discuss this page in DocCommentXchange.
|Copyright © 2012, iAnywhere Solutions, Inc. - SQL Anywhere 12.0.1|