PRIMECLUSTER Installation and Administration Guide 4.1 (for Solaris(TM) Operating System)

G.1 Registering a Cluster Service

If you used the CUI to set up userApplication, you need to register a SynfinityCluster-compatible cluster service and a cluster service instance resource to the Resource Database for the following product:

• Interstage Traffic Director

• If you used the GUI to set up userApplication, the procedure below is unnecessary.
• If you are using the CUI to set up SynfinityCluster-compatible userApplication, see the following:

"/opt/SMAW/SMAWRrms/htdocs.solaris/wizards.en/wizards/CRM.htm"

Describe below is the procedure for registering a SynfinityCluster-compatible cluster service and a cluster service instance resource.

Operation Procedure:

Carry out Steps 1 and 2 in any one of the nodes that make up userApplication. The nodes refer to all nodes that were set to "Machines" in "(5)" of "CRM.htm."

1. Log in with the system administrator authority to any node of the cluster system.
2. Execute the "claddrsc(1)" command in the format shown below. If the registration is successful, the resource ID of the cluster service is output to standard output.

   # /etc/opt/FJSVcluster/sys/claddrsc -c ClusterService -k ApplicationName-of-userApplication *  -p 0 -t nocontroler -s -1 -d self -e service

   * For "ApplicationName-of-userApplication," specify the application name of the user application that was set to "(4)" of "CRM.htm."

   Example) If the application name of userApplication is "generic."

   # /etc/opt/FJSVcluster/sys/claddrsc -c ClusterService -k generic -p 0 -t nocontroler -s -1 -d self -e service

   100

   

   Carry out Steps 3 to 5 in all nodes that make up userApplication.

3. Log in with the system administrator authority to the node.
4. Obtain the node identification number of the local node.

   Execute the clgetnode(1) command in the format shown below, and check the node identification number of the local node. If clgetnode(1) execution is successful, the node identification number of the local node is output to standard output.

   # /etc/opt/FJSVcluster/bin/clgetnode -i

   In the following example, the node identification number of the local node is "0."

   # /etc/opt/FJSVcluster/bin/clgetnode -i

   0

5. Register the cluster service instance resource.

   Execute the claddrsc(1) command in the format shown below. If the registration is successful, the resource ID of the cluster service instance is output to standard output.

   # /etc/opt/FJSVcluster/sys/claddrsc -c AndBinding -k ApplicationName-of-userApplication (*)  -p resource-ID-of-cluster-service (**)  -t nocontroler -s node-identification-number-of-local-node (***)  -d self -e service

   *For "ApplicationName-of-userApplication," specify the application name of the user application that was set to "(4)" of "CRM.htm."

   **For "resource-ID-of-cluster-service," specify the resource ID of the cluster service that was registered in Step 2.

   ***For "node-identification-number-of-local-node," specify the node identification number that was obtained in Step 4 for the local node.

   Example) If the application name of userApplication is "generic," the resource ID of the cluster service is "100," and the node identification number of the local node is "0."

   # /etc/opt/FJSVcluster/sys/claddrsc -c AndBinding -k generic -p 100 -t nocontroler -s 0 -d self -e service

   101

   

   Carry out Steps 6 and 7 in any one of the nodes that make up userApplication.

6. Log in with the system administrator authority to any one of the nodes that make up the cluster system.
7. Set up the operation sequence of the cluster service instances.

   Execute the clsetinsorder(1M) command in the format shown below, and set the operation sequence of the cluster service instances.

   # /etc/opt/FJSVcluster/bin/clsetinsorder -s resource-ID-of-cluster-service(*) -r resource-ID-of-cluster-service-instance ...(**) -S

   *For "resource-ID-of-cluster-service," specify the resource ID of the cluster service that was registered in Step 2.

   **For "resource-ID-of-cluster-service-instance," specify the resource ID of the cluster service instance that was registered in Step 5 for each node. Specify the resource IDs of the cluster service instances in the "Machines" sequence that was set in "(5)" of "CRM.htm."

   Example) If the resource ID of the cluster service is "100," and the Machines sequence is as follows: The resource ID of the cluster service instance in the narcissus node is "101" and the resource ID of the cluster service instance in the sweetpea node is "102."

   Machines[0]  narcissusRMS

   Machines[1]  sweetpeaRMS

   # /etc/opt/FJSVcluster/bin/clsetinsorder -s 100 -r 101,102 -S

8. Check if cluster services and cluster service instance resources are registered using the clgettree(1) command as follows:

   # /etc/opt/FJSVcluster/bin/clgettree -s

   Check the following:

   • The cluster services are displayed.

     The cluster service resource class is "Standby" or "Scalable".

   • The cluster service instances are displayed.

     The cluster service instance resource class is "AndBinding".

   • The order of operating cluster service instances is displayed.

     The order is "()" in the cluster service instance resources. The first cluster service instance will be "RUN", and the remaining instances will be "WAIT".

   Example) In the following example, the cluster service resource ID is "100", and the cluster service instance resource IDs are "101 and "102".

     # /etc/opt/FJSVcluster/bin/clgettree -s

     Standby 100 generic STOP REJECT

           AndBinding 101 genelic STOP UNDEFINED(RUN) narcissus

           AndBinding 102 genelic STOP UNDEFINED(WAIT) sweetpea