HP StorageWorks 4/8 - SAN Switch Release Note page 29

Very rare occurrence of a switch panic when supportsave reads
invalid data on G_Pports without FC FastWrite enabled.
After upgrading one CP from Fabric OS 5.1x to 5.3x, if the other CP
has not yet been upgraded and there is another evmd listening event,
the switch may panic.
IFCCs are seen with class 2 ELS timeouts, due to F_RJT not being re-
turned for class 2 RNID when source and destination are excluded by
hard zoning.
Reduced I/O seen in large fabric testing when core blade removal
and insertion cycles are executed on 8-Gb directors.
If an HP StorageWorks Multi-Protocol Router blade installed in a dir-
ector fails to initialize (bad blade), it could cause the system to assert
and reboot due to FastWrite attempting to configure.
Switch panic can occur during a rare occurrence of a timed-out ex-
change being accessed.
With default zoning set to All Access, a race condition may cause
ACK frame with FC-type Extended Link to be dropped.
RSCN transmitted by the switch did not transmit all the port IDs in the
payload after HPF key on or off.
New ARB (FF) protocol required to be fully compliant with the latest
FC standard (FC-PI-4) for 8–Gb FC ports.
The portcfgspeed command with ax option on a switch running
Fabric OS 6.1.x returns Error: Invalid speed level.
Occasionally during firmware upgrade process, the 1250 processor
is still rebooting when FCR routes are being processed, which results
in missing FCR routes and an inability to route traffic across VEX_Ports.
This leads to I/O running through VEX_Ports to be dropped after
firmware upgrade.
Management server drops IPC message from CALD module causing
it to wait indefinitely, leading to CALD queue full message. May cause
problem with management application, but traffic still passes without
any problem.
The fabstateshow log wraps, overwriting useful fabstate data
on the B-series switch when interoperating with McDATA switches.
When failover occurs, password-less SCP cannot be used from the
switch, because they are not copied during HA failover. SCP with
passwords can be used.
Access Gateways connected to McDATA switches (EOSc firmware)
are not shown in DCFM.
Fixed in Fabric OS 6.1.2
Workaround: Make sure both CPs
are upgraded together.
Fixed in Fabric OS 6.1.2
Fixed in Fabric OS 6.1.2
Fixed in Fabric OS 6.1.2
Fixed in Fabric OS 6.1.2
Fixed in Fabric OS 6.1.2
Fixed in Fabric OS 6.1.2
Fixed in Fabric OS 6.1.2
Fixed in Fabric OS 6.1.2
Fixed in Fabric OS 6.1.2
Fixed in Fabric OS 6.1.2
Fixed in Fabric OS 6.1.2
Fixed in Fabric OS 6.1.2
Fixed in Fabric OS 6.1.2
Fixed in Fabric OS 6.1.2
HP StorageWorks Fabric OS 6.1.2b release notes 29