Building Disaster Recovery Serviceguard Solutions Using Metrocluster with Continuous Access XP P9000 for Linux B.12.00.00

When using the paircreate command to create PVOL/SVOL Continuous Access pairs,
specify the -c 15 switch to ensure the fastest data copy from PVOL to SVOL.
Pair creation of Journal Groups
The Continuous Access Journal has the same characteristic as Continuous Access Asynchronous
such that Raid Manager controls Continuous Access Journal similar to Continuous Access
Asynchronous.
The Raid Manager configuration of the device group pair for Continuous Access Journal is exactly
the same as the configuration of the Continuous Access Asynchronous device group pair. In the
/etc/horcm0.conf file, do not specify any journal volumes or journal group number. Only
data volumes (device group and its devices) must be in the configuration file.
Creating Continuous Access Journal Pair
To create a journal group pair, use the paircreate command:
# paircreate -g <device_group> -f async -vl -c 15 -jp <id> -js <id>
Similar to Continuous Access Asynchronous, the fence “async ” must be assigned to the command
with two additional options -jp and -js.
Table 3
DescriptionOptions
This option is to specify a journal group ID for PVOL-jp <id>
This option is to specify a journal group ID for SVOL-js <id>
The -jp and -js options are required if the device group is configured to use Continuous Access
Journal. The <id> used with -jp and -js option must not be the same.
Notes on the Raid Manager configuration
• A single XP or P9000 device group must be defined for every package on every host that is
connected to the XP or P9000 series disk array.
• Device groups are defined in the Raid Manager configuration file under the heading
HORCM_DEV or HORCM_LDEV.
• The disk target IDs and LUNs for all Physical Volumes (PVs) defined in Volume Groups (VGs)
or disk groups (DGs) that belong to the package must be defined in one XP or P9000 device
group on every host system that might run one or more Metrocluster packages.
• The device group name (dev_group) is user-defined and must be the same on every host in
the Metrocluster that accesses the XP or P9000 disk array. The device group name (dev_group)
must be unique within the cluster; it must be a name that is easily associated with the application
name or Serviceguard package name.
• The TargetID and LU# fields for every device name might be different on different hosts in the
clusters, to allow for different hardware I/O paths on different hosts.
Configuring automatic Raid Manager startup
After editing the Raid Manager configuration files and installing them on the nodes that are attached
to the XP or P9000 Series disk arrays, you must configure automatic Raid Manager startup on the
same nodes. This is done by editing the rc script /etc/mcxpca/raidmgr. Set the
START_RAIDMGR parameter to 1, and define RAIDMGR_INSTANCE as the number of the Raid
Manager instances being used. By default, this is zero (0).
See “Sample file for configuring automatic Raid Manager startup” (page 67) for a sample edited
configuration file for XP Raid Manager Instance.
Setting up the replication 19