PRIMECLUSTER Global Link Services Configuration and Administration Guide: Multipath Function 4.2 - for Solaris(TM) Operating System -

B.2.1 /opt/FJSVmpnet/daemon/mpnetpolld.conf

This file defines basic behavior of the Line Monitoring Daemon.

The default values are beforehand defined as this file. Please change values, if needed.

The form of the entry is shown below.

TIME

This indicates the monitoring interval (sec) of the online NIC. The monitoring interval can be set in the range of 2 to 300. By default, this value is 5 seconds.

COUNT

This indicates the monitoring count (count) of the transmission path. The monitoring count can be set in the range of 1 to 300 times. By default, this value is 2 times.

In the fault monitoring, if this daemon fails in monitoring continuously "monitoring count" times, the communication path is switched.

PATROL_MODE

This indicates the fault monitoring of standby NICs. The following values can be specified:

• 1 : Monitoring the standby NICs (default)
• 0 : Not monitoring the standby NICs.

If any other value is specified, this daemon is not monitored the standby NICs.

PATROL_TIME

This indicates the monitoring interval (sec) of standby NICs. The monitoring interval of standby NICs can be set in the range of TIME value to 30000 seconds. By default, this value is 300 seconds.

WAIT_TIME

This indicates the monitoring wait time (sec) until this daemon starts the fault monitoring, after this daemon is started. This value can be set in range of 0 to 300 seconds. The default value is 0 second.

AUTO_RECOVER_MODE

This indicates recovery mode of transmission paths. The following values can be specified:

• 1 : automatic recovery mode (default)
• 0 : manual recovery mode

The automatic recovery function is valid only for the transmission path that became "fail" state by Link Down, such as a cable disconnection. The transmission path that became "fail" state by causes other than Link Down is not recovered automatically.

The configuration example is shown below. In configuration example, the monitoring interval is 3 seconds, monitoring count is 3 times, monitoring interval of standby NICs is 600 seconds, and monitoring wait time is 10 seconds.

• Behavior of this daemon cannot be configured for every logical interface.
• When the load of CPU or a transmission path is high, the failure in a transmission path may be detected by this daemon. If the failure is detected, please increase the monitoring interval (TIME value).
• When changing parameter values in this file, please change enough in consideration of relation with other applications etc.