3.5.1 Matrix Server Command Reference

Chapter 2: mx Commands 85

[--frequency <seconds>]

The interval of time at which the monitor probes the service. For CUSTOM

service monitors, the default is 60 seconds. For all other service monitors,

the default is 30 seconds.

[--probetype single|multiple]

This argument applies only to CUSTOM monitors.

[--probeSeverity nofailover|autorecover|noautorecover]

The failover behavior for the service monitor. nofailover prevents failover

of the virtual host when the monitored service fails. autorecover fails over

the virtual host and, when the service is restored, fails the virtual host back

to the original network interface. noautorecover fails over the virtual host

but does not fail it back after the service is restored. autorecover is the

default.

[--parameters <parameters>]

For an HTTP monitor, the parameter is the URL for the service. For an NFS

monitor, the parameters are proto (either udp or tcp), rpcname (typically

nfs), and rpcvers (either 2 or 3). For an NIS monitor, the parameters are

proto (either udp or tcp), rpcname (typically ypserv), rpcvers (either 1 or 2),

and domain (the NIS domain to be monitored). For a custom monitor, the

parameter is a probe script.

[--ordering serial|parallel]

Whether Matrix Server enforces a strict ordering sequence when it runs

Start or Stop scripts. The default is serial, the strict ordering sequence.

[--recoveryScript <script>]

Runs after a monitor probe failure is detected, in an attempt to restore the

service.

[--recoveryTimeout <seconds>]

The amount of time to wait for the script to complete.

[--startScript <script>]

When Matrix Server selects the active server for a monitored service, the

Start script runs on that server.

[--startTimeout <seconds>]

The amount of time to wait for the Start script to complete.