The nhwdtd daemon implements the Watchdog Timer service.
The nhwdtd daemon monitors the hardware watchdog
on the nodes of the cluster, at the lights-off management (LOM) level. If
you are using CompactPCI servers, do not use the Watchdog Timer service provided
by the Foundation Services. The CompactPCI hardware watchdogs operate at the OpenBoot
PROM level and are monitored by the platform's software.
On hardware watchdogs that operate at the LOM level, the nhwdtd daemon monitors a node for operating system hang, but does
not monitor the boot process.
The nhwdtd daemon can be configured differently
on each node depending on your requirements. For information on how to configure
the nhwdtd daemon, see the nhfs.conf(4) man page.
The nhwdtd daemon is monitored by the nhpmd daemon. If the nhwdtd fails, it is relaunched
three times. If the nhwdtd daemon fails a third time, the
node is rebooted.
The nhwdtd daemon operates at a lower priority than
the nhpmd daemon, but at a higher priority than the other Foundation Services
daemons.