cmond(1m)

NAME
     cmond - Cluster Monitor daemon

SYNOPSIS
     /usr/cluster/bin/cmond [option]...

DESCRIPTION
     The cmond daemon provides a framework for starting, stopping and
     monitoring process groups. Processes groups are made up of one or more
     processes that can be started, stopped and restarted together. cmond
     restarts daemons if they crash.  FailSafe daemons are classified into
     separate process groups and are monitored by cmond.

     Process groups are specified by a configuration file in the
     /var/cluster/cmon/process_groups directory. The name of this file is same
     as name of the process group. The file for each process group has three
     entries:

     TYPE              = process group type

     PROCESSES         = list of processes belonging to this group

     ACTIONS           = list of actions cmond is allowed to perform on this
                       process group, like start, stop, restart, attach,
                       detach, status. This entry is optional. A missing
                       ACTIONS entry means, cmond is allowed to perform all
                       actions.

     Process group types supported and various daemons under each type are:

     Type              FailSafe Daemons

     cluster_admin     cad

     cluster_control   crsd

     cluster_hainfra   ha_cmsd, ha_gcd ha_srmd

     cluster_failsafe  ha_fsd

     cluster_agent     ha_ifd

     The dependencies of various process group types are:

     Type               Depends on

     cluster_admin     none

     cluster_control   none
     cluster_hainfra   cluster_admin and cluster_control

     cluster_failsafe  cluster_admin, cluster_control and cluster_hainfra

     cluster_agent     cluster_hainfra Any agents that users must be of this
                       type.

OPTIONS
     cmond takes several options:

     -L log level      Set logging level to loglevel.  The log levels are:
                       NULL, normal, critical, error, warning, info, frequent,
                       all.

     -d                Run in debug mode.

     -l                Lazy mode where cmond does not validate its connection
                       to cluster database.

     -t nap interval   The time interval in milliseconds after which cmond
                       checks for liveliness of process groups it is
                       monitoring.

     cmond reads the file /etc/config/cmond.options, if it exists, for a
     default set of options.

FILES
     /var/cluster/cmon/process_groups Directory used to hold the process group
     configuration files.
     /etc/config/cmond.options File read to obtain default options

SEE ALSO