HP StorageWorks Fabric OS V3.2.1c Release Notes (AA-RUQYH-TE, June 2007)

• Using Web Tools, the message NO PERFORMANCE LICENSE was received during basic
monitoring.
If you are not upgrading to this latest Fabric OS release, the workaround is to install the Advanced
Performance Monitoring (APM) license.
• The config upload command failed on switches supporting a large Zone database.
If you are not upgrading to this latest Fabric OS release, the workaround is to save the config
database using the CLI.
• The SAN Switch 2/16 running Fabric OS 3.2.0 panics during zoning updates in interop mode.
• SAN Switch 2/16 m odels used incorrect bu ffer-to-buffer (BB) credits when the switch BB credit
was set less than the device BB credit.
If you are not upgrading to this latest Fabric OS release, the workaround is to increase the
switch BB credit limit to >= the device BB credit.
• A short burst of bad primitives on the link causes an rx_under flow error .
• If two or more ports are zoned together on a SAN Switch 2/16, a ZONEGROUPADDFAIL
message is displayed.
• E-mail configured via Fabric Watch was not received on certain third-party mail servers.
• The PLOGI accept frame (ACC) from storage was not being forwarded by Fabric OS 3.2.x during
resetofmultiplestoragenodes. Thehostcouldnotseethetarget.
If you are not upgrading to this latest Fabric OS release, the workaround is to run the
portDisable and then portEnable commands to reset the port.
• If a switch is running Fabric OS 3.2.0 in Interop mode, an E_por t may segment when connecting
to specific third-party switches.
If you are not upgrading to this latest Fabric OS release, the workaround is to use an older version
of code on the third-party switch that supports version 2 exchange fabric parameters (EFP).
• The end-to-end acceptance of the n_port login (PLOGI ACC) is not getting through on a SAN
Switch 2/16.
If you are not upgrading to this latest Fabric OS release, the workaround is to run
portDisable/portEnable on the p ort for the host that is losing access to the device.
• A SAN Switch 2/16 running Fabric OS 3.2.0a dropped the n_port login (PLOGI ACC) with a
device running at 1 Gbps.
• Intermittent loss of target occurred on a switch running Fabric OS 3.2.0x.
• A SAN Switch 2/16 was repor ted to have flooded a host by sending n_por t logins (PLOGI)
every millisecond.
• A SAN Switch 2/16 displayed a stack overflow message and panicked during normal operation.
• On SAN Switch 2/16 only, after using Fabric Manager 5.0.0 to change the admin password,
the switch reverts to an old adm in password when rebooted.
If you are not upgrading to this latest Fabric OS release, the workaround is to use the C LI
command passwd to change the original admin password.
• The Management server did not provide the correct port data when running extended Port
Identifier (PID) format. Fabric OS 3.2.x running under PID format 2 gives SMI-A no connected
N_Port data.
Fabri
cOS3.2.1cfixes
Thefollowingdefectshavebeenfixed since the release of Fabric OS 3.2.1b:
• Frames were dropped on cfgenable when the final number of zone members for a port hits
the exact CAM limit of 64.
• Duri
ng Fabric Watch initialization, if a p ort is in a transient state and marked as FAULT Y, its
status will not be updated, when it does come online without a Fabric Watch license. The
switchStatusShow command will indicate Faulty ports monitor DOWN.
• Asw
itch will not take two different IP addresses for SNMP trap recipients on two different
community strings for index 0 and 3 o n HP StorageWorks MSA SAN Switch 2/8 (Brocade
3000). An enhancement in Fabric OS 3.2.1c adds the root command snmpBRSetNonDefault
4