HP H-series SN6000 and 8/20q Fibre Channel Switch Release Notes

Fixes
The following issues are fixed in this release (since the release of firmware version 8.0.14.03.00).
• By using a hidden account, there is a possibility of an Ethernet vulnerability when using the
switch in an unintended manner. This is resolved and associated contributing function is
disabled.
• Received unexpected ELP (Exchange Link Parameter) revision, received revision 3 instead of
revision 2. The ELP revision alarm is due to incorrect check for ELP revision.
• In QuickTools, an Emulex NPIV WWN does not turn bold when it is added or applied to a
zone.
• When configuring the switch for an SSL certificate, a DOD CA Certificate (dod-ca27) could
not be added.
• Rejects for a failed response to a nameserver request are not cleared properly. This might
result in running out of allocated memory.
• In setting up the Transparent Router function, if an activation request is received for devices
already bridged together, the internal process responds with the request ID of the first request,
and the nameserver does not respond to the new request. For example, when a device is
moved from one H-series switch to another but the original mapping is not removed.
• When the devices logged in on a given port are in eight or more zones, the hard zoning for
the devices in more than eight zones is not configured correctly. This impacts the flow of data
frames to and from those devices.
• RX_ID is assigned by the switch to Fibre Channel frames with Domain ID of 239 or less, where
some connected devices could drop or reject frames with an incorrectly assigned RX_ID field.
• Some internal data transfers could exceed buffer length by one byte and cause an unexpected
error.
• The internal management server (mserver) application fabric ping command (FPNG) contained
a stack buffer overrun, which could cause the mserver to become unresponsive. The size of
the destination character string is increased to accommodate the formatted arguments.
• If a request for fabric topology through the entry switch is made before completing the previous
request, this could cause the internal management application to stop. If this occurs, the switch
continues to function in the fabric but it is not manageable and requires a disruptive reset to
recover.
• NPIV correction is made to enable IOSteamGuard for QLogic HBAs. When using a QLogic
HBA, NPIV logins after the initial login might not have a QLogic WWN. When this occurs,
IOSteamGuard is disabled. Now, if the first device to log into a port has a QLogic WWN,
IOSteamGuard is enabled for the initial login and any subsequent NPIV logins on that port.
If the first device to log into a port does not have a QLogic WWN, IOSteamGuard is disabled
for the initial login and any subsequent NPIV logins on that port.
• Duplicate Port WWN Alarms are seen in switch logs. This is caused by simultaneous logout
at one point in the fabric and login at another, where propagation delay of the updated
information through the fabric could cause nameserver to issue an alarm.
• IOStreamGuard is incorrectly disabled for ports connected to QLogic Host Bus Adapters with
a WWN of xx:xx:00:24:ff:xx:xx:xx.
• Denial of Service list is not updated correctly after device logged in properly or after the zoning
is corrected.
• FC frames with same D_ID and S_ID are dropped, when users try to mirror data between
LUNs. This is now permitted using the Command Line Interface, entering advanced mode,
config edit, set config zoning, and setting the AllowMirroring parameter to TRUE.
8