IBM SG24-5131-00 User Manual

Cluster Management and Administration 

(C-SPOC) utility can be used to start and stop cluster services on all nodes in 
cluster environments.

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. 

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.

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.

The 

clsmuxpd

 daemon cannot be started unless the 

snmpd

 daemon is 

running.

The 

clsmuxpd

 daemon cannot be started unless the 

snmpd

 daemon is 

running.

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.