ServiceGuard test

[Jim Mulshine]

What should be the expected result of the following cluster test?

Node1 has one FC adapter connected to one controller port of a VA7100 array.

Node2 has one FC adapter connected to the other port of the VA7100 array.

Package1 is running on node1.

The fiber cable connecting node1 to the VA7100 is removed.

Will package1 move to node2?
(I've tried it and it doesn't. Replacing the cable resulted in package1 continuing to run normally on node1 as if nothing happened.)

How does one go about getting package1 to be moved by ServiceGuard over to node2 when node1 loses fiber contact with the VA7100?

[Geoff Wild]

What do you see in your syslog when you pull the cable?

Rgds...Geoff

[Carsten Krege]

SG doesn't monitor disk connectivity. The availability of your disks and access to it has to be provided by other means (e.g. LVM pvlinks, MirrorDisk/UX, Disk Arrays which provide RAID).

However, you could configure EMS resources as package dependencies. Doing this, SG monitors these EMS resources and switches the package to an adoptive node when the resource goes down.

The SG manual states:

"If you are using LVM, you can configure disk monitoring to detect a failed mechanism by using the disk monitor capabilities of the EMS HA Monitors, available as a separate product (B5735AA). Monitoring can be set up to trigger a package failover or to report disk failure events to a ServiceGuard, to another application, or by email. For more information,
refer to the manual Using High Availability Monitors (B5736-90042),available at http://docs.hp.com -> High Availability."

Alternatively to this HA Monitor (diskmond) that monitors disk configured into LVM, you could use EMS Hardware Monitors provided with the Online Diagnostic to monitor disk status with disk_em monitor and other. Refer to http://docs.hp.com/hpux/diag (EMS Hardware Monitors (for HP 9000)).

Carsten