HP B-series Fabric OS 7.0.2d Release Notes (5697-2822, August 2013--includes all 7.0.x versions)

Table 13 Fabric OS 7.0.1 closed defects (continued)
SolutionClosed fixes summary
releases, users were required to reseat the blade to recover when backend
credit was lost.
Fixed in Fabric OS 7.0.1.On 16 Gbps platforms, configdefault may lead to an invalid port
speed configuration if ports in the same group are in different logical
switches. If configdefault is executed in the default switch,
portcfgoctetspeedcombo settings get set to default on all logical
switches.
Fixed in Fabric OS 7.0.1.Devices are unable to communicate with peer devices after a logout/login
sequence. When devices send LOGO and then FLOGI, without a port
toggle, the route is not set up, causing the PLOGIs sent to peer devices to
be dropped.
Fixed in Fabric OS 7.0.1.On 4 Gbps switches and blades, initialization failure may occur causing
them to fail to go online. This may be experienced in high temperature
conditions while running portloopback test.
Workaround prior to upgrade: Disable
IPsec on the FCIP tunnel.
After modifying the metric on a standby FCIP Circuit of a crossport with
IPSec enabled, the tunnels for that DP complex go down due to a software
fault on the DP.
Fixed in Fabric OS 7.0.1.
Fixed in Fabric OS 7.0.1.If users disable multiple circuits in an FCIP tunnel, traffic may not be evenly
distributed across the remaining circuits.
Fixed in Fabric OS 7.0.1.HP SN8000B 4-Slot SAN Backbone Director Switch gets inundated with
[PS-5011] internal messages in the performance monitor.
Fixed in Fabric OS 7.0.1.There is a pdmd memory increase on the standby CP of an HP SN8000B
8-Slot SAN Backbone Director Switch with max LS creation and fully
populated EX_Ports after a long period of time with I/O running.
Fixed in Fabric OS 7.0.1.The Brocade FCoE LACP PDU length is non-standard and is being discarded
at the remote end. The Brocade switch is unable to form a dynamic LAG
with a third-party switch.
Fixed in Fabric OS 7.0.1.The severity level of SCN-1002 RASLOG message is erroneously set and
displays as a warning instead of an error.
Fixed in Fabric OS 7.0.1.During a reboot of the HP 1606 SAN Extension Switch or the HP DC SAN
Director Multiprotocol Extension Blade, customers may see their Ethernet
switch ports fenced offline due to having received a BDPU packet forwarded
by the switch/blade if STP/BDPU guard is enabled on the Ethernet ports.
Fixed in Fabric OS 7.0.1.Tape devices may drop frames when connected to an encryption switch.
The tape device driver fails with a check condition due to an illegal
command while trying to retrieve a drive dump.
Workaround prior to upgrade: Use LSAN
zones or ensure that devices do not
contain stale proxy IDs in static memory.
Under rare conditions in FCR fabrics, a cold recovery may occur when
devices that are not LSAN zoned send ELS frames that are not properly
freed.
Fixed in Fabric OS 7.0.1.
Fixed in Fabric OS 7.0.1.The FFDC file has no information because it does not pull kernel data during
an Out Of Memory (OOM) condition.
Workaround prior to upgrade: Execute
the portcfgdefault CLI command on
the GE_Ports.
A switchshow command erroneously displays the GE_Port as an FCIP
copper on an HP StorageWorks 400 Multi-protocol Router and the HP
StorageWorks B-series Multi-protocol Router Blade. This is a display error
only and does not affect functional operations.
Fixed in Fabric OS 7.0.1.
Workaround prior to upgrade: Avoid
sending LOGO with SID 0 to Access
Gateway.
Access Gateway may crash upon receiving bad frames from devices. The
switch panic occurs when devices send ELS_LOGO with SID 0 and DID as
Login server.
52