Notes on the cluster, Notes on the automatic failback, Notes on the cluster -16 – HP XP Racks User Manual

Page 78: Notes on the automatic failback -16

Advertising
background image

Notes on the Cluster

When you use HDLM in a cluster configuration, you must install the same

version of HDLM on all the nodes that comprise the cluster. If different

versions of HDLM are installed, the cluster system may not operate

correctly. If the HDLM Version and Service Pack Version, which are

displayed by executing the following command, are the same, the

versions of HDLM are the same:

# /usr/DynamicLinkManager/bin/dlnkmgr view -sys

When you use PowerHA, GPFS, Oracle RAC 10g, Oracle RAC 11g, or VCS,

settings such as script registration or the reservation policy setting are

required. For details about the settings used for each cluster, see the

following:
For PowerHA:

Settings for Using PowerHA on page 3-87

.

For GPFS:

Settings for Using GPFS on page 3-90

.

For Oracle RAC 10g or Oracle RAC 11g:

Settings for Using Oracle RAC 10g

or Oracle RAC 11g on page 3-90

.

For VCS:

Settings for Using VCS on page 3-93

.

If you are using PowerHA, note that HDLM does not support non-

concurrent volume groups due to limitations in PowerHA. For details,

contact IBM.

If all of the following conditions are met, HDLM can be migrated without

stopping PowerHA services:

¢

The shared volume groups used by PowerHA are concurrent volume

groups.

¢

A custom disk method is specified in the PowerHA script.

¢

A virtual I/O server environment is not used on the host.

Notes on the Automatic Failback

If the automatic failback processing starts before a path failure has been

recovered from, the cluster system switchover time might become longer. To

prevent this, for the automatic failback checking interval, specify a value that

is greater than the value obtained from the following formula:

checking-interval (minutes)

= maximum-number-of-paths-among-the-LUs × number-of-connected-

storage-devices

Notes on the queue_depth Parameter Value for an HDLM-Managed

device

Hitachi recommends that you set the queue_depth parameter of an hdisk for

an HDLM-managed device to a value greater than or equal to the number of

paths that are connected to the hdisk. For details on how to specify the
queue_depth parameter, see the AIX documentation.

3-16

Creating an HDLM Environment

Hitachi Dynamic Link Manager User Guide for AIX®

Advertising
This manual is related to the following products: