HP A7533A - Brocade 4Gb SAN Switch Base Release Note page 75

Switch panic occurs when management application performs repeat
FCP LUN emulation queries. If the device gives an inappropriate re-
sponse, or no response, to an FCP LUN request, FCP Daemon retries
the query and the thread becomes stuck. Any new FCP LUN request
that is issued from the management application generates a new thread.
Eventually FCP Daemon generates up to the maximum number of allowed
threads, leading to the switch panic.
The host failed to discover tapes due to a VI/VT (Virtual Initiator/Virtual
Target) database mismatch between the switches that are part of the
same Encryption Group (EG). Therefore, the host cannot discover tape
devices that are part of the same EG.
Common Access Library Daemon (CALD) detects a large memory in-
crease on the HP StorageWorks 2408 FCoE switch when running mul-
tiple SMIA clients over FCoE.
The loginshow command on an FCoE switch displays unexpected
values when an LP21000 or LP21002 CNA is moved between switches.
In a remote corner case, an IPC failure message appears on one of the
nodes when using the cryptocfg --show -groupmember -all
command.
Running Fabric OS commands on the default switch with virtual fabrics
enabled, after performing multiple cfgenable commands with TI zones
enabled, causes a No space left on device error, and no further
Fabric OS commands can be executed.
In a multi-circuit FCIP tunnel configuration with 1606 Extension SAN
switch, or multiprotocol extension blade, circuit failover on 1 Gb tunnels
may not fail over to other available circuits with two rate limited circuits.
Due to slow read track performance when XRC emulation is enabled,
MVS I/O errors are generated when an invalid selective reset frame is
sent to the controller, causing Abort Sequences later on.
Need additional information of EE (Encryption Engine) status for the
entire EG (Encryption Group) in a single command. Currently we have
to invoke cryptocfg --show -groupcfg", cryptocfg --show
-groupmember -all, or cryptocfg --show -localEE on all
nodes to know the overall state of EG including EE state across the EG.
.
When auditing is on for security class, a race condition can occur in
raslogd, which leads to a switch panic or hafailover.
IFCCs are encountered when processing Read Channel Extender Cap-
abilities with XRC Emulation (IPLing systems with SDM LPARs).
HP StorageWorks Fabric OS 6.3.0c Release Notes
Fixed in Fabric OS 6.3.0c
Fixed in Fabric OS 6.3.0c
Workaround prior to upgrade: Issue
cryptocfg --commit again to
sync up the VT/VI database across
nodes in the EG.
Fixed in Fabric OS 6.3.0c
Fixed in Fabric OS 6.3.0c
Fixed in Fabric OS 6.3.0c
Workaround prior to upgrade: Issue
cryptocfg --show -groupcfg
and cryptocfg --show
-localEE commands to show the
health of the EG (Encryption Group).
Fixed in Fabric OS 6.3.0c
Fixed in Fabric OS 6.3.0c
Fixed in Fabric OS 6.3.0c
Workaround prior to upgrade:
Disable FICON XRC Emulation.
Fixed in Fabric OS 6.3.0c
Fixed in Fabric OS 6.3.0c
Workaround prior to upgrade:
Disable auditing via auditcfg
-disable.
Fixed in Fabric OS 6.3.0c
Workaround prior to upgrade:
Disable FICON XRC emulation.
75