HP Serviceguard Storage Management Suite Version A.02.00 for HP-UX 11i v2 Release Notes, September 2008

Serviceguard cluster will not start, or you will not be able to access the Veritas volumes.
update-ux does not check for software dependencies.
See “Compatibility Information and Installation Requirements” (page 17) for more information
on bundle upgrades.
Oracle Data Manager (ODM) with VxFS and LVM
• What is the limitation? ODM for VxFS file systems on LVM volumes is not supported.
• What is the workaround? Oracle Files that need ODM capability should reside on a VxFS file
system created on a VxVM/CVM volume.
Oracle Data Manager Limitation for Bundles T2776CA, T2796CA, T2777CA, and T2797CA
• What is the limitation? Single instance Oracle databases are not permitted outside of
Serviceguard packages in environments using T2776CA, T2796CA, T2777CA, and T2797CA.
If there are any Oracle instances not controlled by Serviceguard packages, a halt of the system
multi-node package SG-CFS-pkg could bring down the node or cluster.
• What is the workaround? Place all Oracle single instances in Serviceguard packages when
deploying these bundles. In addition, single instance Oracle packages using VxFS and VxVM
must have a direct dependency on the system multi-node package SG-CFS-pkg.
JAGaf82190 (SR8606422366) odm_gms_api_start_msgs fails during system startup
• What is the problem? Errors returned during system startup:
GAB WARNING V-15-1-20115 Port d registration failed, GAB not
configured vxgms: GAB_API_REGISTER error=261
ODM WARNING V-41-6-5 odm_gms_api_start_msgs fails
• What is the workaround? Ignore the messages.
HP Serviceguard Cluster File System for RAC T2777CA VxFEN JAGaf97359
• What is the limitation? Error when starting Serviceguard packages. This problem occurs when
LANG is set to a non-English locale, for example, ja_JP.SJIS. Failure causes SG-CFS-pkg to
fail to start up and causes a TOC of the box with the following messages in syslog.
VXFEND: open(): Unable to open I18N message catalog
/opt/VRTSvcs/vxfen/ (error code: 9)
• What is the workaround? Set the export VCSI18N_LANG=en or LANG=C.
JAGag06843: vxumount returns device busy on master node
• What is the problem? In some stress situations, vxumount of a VxFS filesystem might return
EBUSY if there are I/Os in progress on other nodes within the cluster. Sometimes the
vxumount can take long time because of a possible issue in the VFS layer. This is a transient
problem.
• What is the workaround? This is a transient problem and retry of the vxumount operation
should succeed.
JAGag09085: VxVM 5.0 requires more space on / directory for /etc/vx/cbr backup
• What is the problem? With VxVM 5.0, more disk space is needed in the /etc/vx/cbr directory
because of the increased metadata size. The space usage under this /etc/vx/cbr directory can
take up to 64 MB per diskgroup configured on the system.
• Workaround? This problem can be worked around by keeping enough available space in the
filesystem that has this directory (usually /). Alternatively it is possible to make this path a
symbolic link to some other directory on other filesystem.
Known Problems and Workarounds 27