Starting cluster services refers to the process of starting the HACMP for AIX daemons that enable the coordination required between nodes in a cluster. Starting cluster services on a node also triggers the execution of certain HACMP for AIX scripts that initiate the cluster. Stopping cluster services refers to stopping these same daemons on a node. This action may or may not cause the execution of additional HACMP for AIX scripts, depending on the type of shutdown you perform.
8.2.1 HACMP Daemons
The following lists the required and optional HACMP for AIX daemons.
8.2.1.1 Cluster Manager daemon (clstrmgr)
This daemon maintains the heartbeat protocol between the nodes in the cluster, monitors the status of the nodes and their interfaces, and invokes the appropriate scripts in response to node or network events. All cluster nodes must run the clstrmgr daemon.
8.2.1.2 Cluster SMUX Peer daemon (clsmuxpd)
This daemon maintains status information about cluster objects. This daemon works in conjunction with the Simple Network Management Protocol (snmpd) daemon. All cluster nodes must run the clsmuxpd daemon.
Note
The clsmuxpd daemon cannot be started unless the snmpd daemon is running.
8.2.1.3 Cluster Lock Manager daemon (cllockd)
This daemon provides advisory locking services. The cllockd daemon may be required on cluster nodes if those nodes are part of a concurrent access configuration, but this is not necessarily so. Check with your application vendor to see if it is required.
Note
If the clsmuxpd daemon or the cllockd daemon cannot be started by the Cluster Manager (e. g. the ports are already in use), the Cluster Manager logs an error message and dies.
Cluster Management and Administration 155
