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
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
.
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®