HP A7533A - Brocade 4Gb SAN Switch Base Release Note page 44

Closed defect summary
HP StorageWorks DC Switch Encryption blade comes up faulty after
firmwaredownload -sf and reboot of both CPs at the same time.
After upgrade to Fabric OS 6.x, rsh using user credential gets ifmod-
eshow eth0: ioctl (ETHTOOL_GSET) failed error in response
to ifmodeshow command. Telnet and rsh with root credential work
fine.
Backup job stopped for an extended period of time when running
through HP StorageWorks Encryption switch or encryption blade after
bouncing 1 of the 2 connected ISLs.
The 4Gb SAN Switch for HP c-Class BladeSystem AG may login to
DC SAN Director with wrong PWWN.
During director switch initialization from power on, reboot, or
hafailover, daemons fail to load causing switch to auto reboot and
ffdc HAM-1008 and HAM-1007 errors are observed.
When running a third-party host scan, the http daemon
(weblinker.fcg) terminates and restarts several times, causing a
temporary management service interruption.
In Fabric OS 6.2, OID for connUnitType no longer returns the correct
access gateway mode status after upgrade from Fabric OS 6.1.1d to
6.2.0f.
Fabric binding does not work in 239 domain mode in Interopmode 3
with FCR on 8Gb switches. EX port rejects it.
Fabric Watch sometimes logs messages with values above 100%.
Most likely occurs during first Fabric Watch polling cycle after slot-
statsclear command is issued.
The following message may display on the console while firmware
commit is in progress: ssnGet: Unable to get WWN info.sw=
0 rc=-5.
While doing a disruptive CP reboot on a director class switch (reboot
active CP when no standby CP is present), if simultaneous fabric update
causes a large volume of F-Class traffic, a potential exists for an edge
switch to be faulted due to lack of response from the core director.
In a rare circumstance with a switch running Fabric Watch, a panic
can occur after fwd terminated with exit code:134, exit sig:17, and
parent sig:0 when IPC buffer is overrun on a busy system.
Telnet banner and switchname are not consistent after a new switch-
name is defined without doing multiple hafailover commands. Can
also occur after a standby CP is replaced where the new CP hosts file
does not get synchronized with active CP without two hafailover
commands.
44
Solution
Fixed in Fabric OS 6.2.2
Workaround prior to update:
nl
Reseat the blade.
Fixed in Fabric OS 6.2.2
Workaround prior to update:
nl
Set suid permissions for ethmode:
chmod +s /fabos/libexec/
ethmode
Fixed in Fabric OS 6.2.2
Fixed in Fabric OS 6.2.2
Workaround prior to update:
nl
Reboot AG.
Fixed in Fabric OS 6.2.2
Fixed in Fabric OS 6.2.2
Fixed in Fabric OS 6.2.2
Fixed in Fabric OS 6.2.2
Fixed in Fabric OS 6.2.2
Fixed in Fabric OS 6.2.2
Fixed in Fabric OS 6.2.2
Fixed in Fabric OS 6.2.2
Fixed in Fabric OS 6.2.2
Workaround prior to update:
nl
Use the hostname xyz command
when a switchname xyz is used.
The problem does not occur with SSH;
it does not use the banner.