HP StorageWorks Fabric OS 6.3.1c Release Notes (5697-0509 September 2010 - includes all 6.3.1x versions)

Fixed in Fabric OS 6.3.1a
Workaround prior to upgrade:
Bounce ports to recover
After upgrade from Fabric OS 6.2.x to 6.3.x, RSCNs are not sent to
all devices the target is zoned with, so hosts are unable to see the target
until all host ports are bounced.
Fabric OS 6.3.1b fixes
Table 13 lists defects closed in the Fabric OS 6.3.1b firmware release.
Table 13 Fabric OS 6.3.1b closed defects
SolutionClosed defect summary
Fixed in Fabric OS 6.3.1b
On HP StorageWorks 1606 Extension SAN switch and Multiprotocol
extension SAN blade running Fabric OS 6.3.1a, with emulated tape
and tape pipeline, if there is a frame drop on the back end, read tape
pipeline does not handle the error and causes tape access to hang.
Fixed in Fabric OS 6.3.1b
On HP StorageWorks 2408 FCoE converged network switch with
factory installed Fabric OS 6.3.1a, switch will panic during subsequent
firmware download. HP did not factory install 6.3.1a or any other 6.3.x
version of Fabric OS in this switch, so this is not an issue for HP. Not
an issue on any field firmware upgrades to 6.3.1a, but only factory
installed upgrade.
Fixed in Fabric OS 6.3.1b
Workaround prior to upgrade:
Hafailover/hareboot and then toggle
the EX_Port to recover or make sure
base switch is configured before
hafailover/hareboot.
After going through warm recovery on switch with VF and FCR con-
figured, but no base switch configured, devices are not getting imported
after adding base switch configuration. Switch stays in initializing state
in an end to end FCR fabric, causing devices not to get imported. This
affects HP StorageWorks DC SAN and DC04 Directors, 8/40 and
8/80 SAN switches only.
Fixed in Fabric OS 6.3.1b
During hot code load under a very rare condition, the blade processor
in the 400 multiprotocol router loses communication with the CP, causing
the firmware download to fail with raslog SULB-1035. The switch will
come up faulty.
Fixed in Fabric OS 6.3.1b
HP multiprotocol extension blade initialization fails and panics after
being hot plugged into the director if an SFP is installed in one of the
GE ports when the blade is hot plugged.
Optional RAS features were added
in 6.3.1b that can be turned on to
help detect such problems, as well
as allowing setting of different hold
times at the edge fabric to alleviate
congestion, if appropriate.
Identification of slow drain devices, stuck Virtual Channels, and credit
loss conditions that lead to sluggish fabric performance and fabric wide
frame drop are difficult to determine.
Fixed in Fabric OS 6.3.1b
Race condition encountered during extensive writes to compact flash
causing long HA recovery times during hot code load. This can lead to
link reset or frame drops when credit is not returned in time. Affects
virtual fabric environment and connecting switch operating under single
virtual channel credit model more extensively, and can happen on both
director and switch platforms.
Fixed in Fabric OS 6.3.1b
Workaround prior to upgrade:
Manually re-enable ports.
All ports on 48 port 8Gb blade become persistently disabled after up-
grading from Fabric OS 6.2.x to 6.3.x with virtual fabrics disabled and
fmsmode enabled.
HP StorageWorks Fabric OS 6.3.1c Release Notes 87