HP StoreAll OS 6.5.1 Release Notes

• Updated the Red Hat Linux kernel to accommodate system configurations in which vLANs do not
have an IP address configured at the bond level. Previously, this situation would cause the node
to fail.
• Updated the StoreAll OS software to correctly report the bond status for tagged VLAN bonds.
Previously, no bond status would be reported if the bond became degraded.
• Adjusted the timeout value in the FMT to ensure that discovery of SAS switches does not time out
with the error code 0x1A13.
• If an interruption occurs during an upgrade to 6.3 or later, the HP-SMB service will no longer fail
during a retry of the upgrade, causing the upgrade to fail. Previously, the interruption caused the
HP-SMB registry settings file to become corrupted. Because the HP-SMB service could not read
this corrupted file during startup, the upgrade failed. The HP-SMB registry settings are now backed
up so that the backup file can be used if the original file is corrupted.
• If you attempt to create an SMB share that contains "C$" or "c$" in the share name, an error
message now displays, indicating the share name is invalid. This error message displays whether
you are using the GUI or CLI to create the SMB share. Previously, the SMB share would be created
but no additional management operations could be performed from the StoreAll OS software.
• The code for tuning the segmentSpaceUsedEventThreshold parameter using the
ibrix_fm_tune command has been fixed so that when the value of this parameter is changed
from its default of 90%, email and SNMP notifications are sent when the new threshold is exceeded.
Previously, the user-defined value was ignored and notifications were only sent when the default
threshold was exceeded.
• ACL blocks no longer get accidentally corrupted, which ensures all parts of a file system remain
accessible to all applications.
• Added an important note to the "Creating and mounting file systems" chapter of the HP StoreAll
OS User Guide that says:
IMPORTANT: You cannot use the ibrix_lwmount -a command to mount the file system from
the Linux client if the file system was unmounted using ibrix_lwumount. You must use
ibrix_lwmount only (without the -a option) to mount the file system. This behavior is controlled
by a flag in the ifstab.txt file, located in /etc/ibrix.
• When a node failover occurs, the IP address, netmask, and bcast address of the backup NIC is
now the same as the NIC that is failing over. Previously, if the information was different, it caused
the backup NIC to fail and could only be resolved by manually updating the details on the backup
NIC.
• The Segment Preferences tab of the Modify Filesystem Properties wizard now includes a scroll
bar for the list of available hosts to which a segment can be preferred. Previously, only the first
four hosts were visible and you could only display others by changing the sort order.
• When using the ibrix_exportfs command, if the host name included in the command contains
wildcard characters (such as *), the command will now execute successfully. Previously, the
command did not work if wildcards were included in the host name.
• Updated hpspmon to ensure that all information reported for temperature sensors is accurate and
complete for the system on which hpspmon is running. Previously, some sensors would display
blank information in the Temp column and there was no indication that the affected sensor was
not applicable to the system on which hpspmon was running.
• When executing the ibrix_online_quotacheck command, evacuated segments are no
longer checked by this command. Previously, evacuated segments were included when this
command was executed, resulting in unnecessary errors being reported.
• Updated the StoreAll OS software to obtain the NIC status for each node from the node running
the active Fusion Manager when executing the ibrix_crrhealth command. This ensures that
10 6.5