HP XP Cluster Extension Software Administrator Guide

Creating RAID Manager device/copy groups
A device/copy group is the unit in which the failover/failback operation is performed. A device/
copy group can contain several volume groups.
Configure a single device/copy group for a failover cluster group (MSFC) . This device/copy
group must include all disks being used for the application service.
The RAID Manager configuration file (horcmX.conf) is used to map device/copy groups to the
internal disk array disks. A device/copy group is the common unit for failover operations initiated
from the server side.
Network considerations
Because RAID Manager is essential to XP Cluster Extension, HP recommends that you use the
heartbeat network (a private network) for RAID Manager communications. Alternative network
paths are highly recommended. Configure the networks, RAID Manager uses for each device/
copy group in the HORCM_INST part of the RAID Manager configuration file.
Starting and stopping the RAID Manager instances
Start the RAID Manager instances for XP Cluster Extension at system boot time to provide the fastest
access to disk status information.
XP Cluster Extension provides a service to integrate the RAID Manager instance startup into the
system startup process. This feature reduces resource group failover times because the XP Cluster
Extension resource does not need to start the RAID Manager instances. If the system cannot
automatically start and monitor RAID Manager instances, you can start and stop RAID Manager
with the following commands:
horcmstart instance_numbers
horcmshutdown instance_numbers
Starting RAID Manager without specifying an instance number will start instance 0 with the
associated horcm.conf file. For this reason, zero (0) is not recommended as an instance number
for a XP Cluster Extension RAID Manager instance.
Test takeover function
After configuring RAID Manager for the device/copy groups used by XP Cluster Extension, verify
that each device fails over correctly from each server in the cluster. The device/copy group must
be in PAIR state.
CAUTION: RAID Manager keeps configuration data of the disk array in system memory. Therefore,
you must stop and restart RAID Manager instances on all servers if a configuration change is
applied to any disk array.
To test the correct failover and failback behavior, log in to each server used with XP Cluster Extension
and invoke the following commands if the local disk is the secondary (S-VOL) disk:
set HORCMINST=instance_number
pairdisplay –g device_group_name –fx –CLI
horctakeover –g device_group_name –t timeout
The output of the pairdisplay command indicates whether the local disk is the secondary (S-VOL)
disk and if so, the horctakeover command shows a SWAP-takeover as a result. If pairdisplay
shows the local disk as the primary (P-VOL) disk, log in to a system connected to the secondary
(S-VOL) disk and invoke the horctakeover command there. If the horctakeover command
does not result in a SWAP-takeover, see “Recovery sequence” (page 46) to resolve the issue.
The –t option of the horctakeover command is used only for fence level ASYNC (both Async
and Journal).
Planning for XP Cluster Extension 11