Cisco Cisco Gigabit Ethernet Switch Module (CGESM) for HP

When the rate of received DHCP requests

exceeds 2,000 packets per minute for a long time, the response 

time might be slow when you are using the console. The workaround is to use rate limiting on DHCP 
traffic to prevent a denial of service attack from occurring. (CSCeb59166)

After you change the access VLAN on a port that has 802.1x enabled, the IP Phone address is 
removed. Because learning is restricted on 802.1x capable ports, it takes approximately 30 seconds 
before the address is relearned. No workaround is necessary. (CSCea85312)

If the number of multicast routes and Internet Group Management Protocol (IGMP) groups are more 
than the maximum number specified by the show sdm prefer global configuration command, the 
traffic received on unknown groups is flooded in the received VLAN even though the show ip igmp 
snooping multicast-table privileged EXEC command output shows otherwise. The workaround is 
to reduce the number of multicast routes and IGMP snooping groups to less than the maximum 
supported value. (CSCdy09008)

IGMP filtering is applied to packets that are forwarded through hardware. It is not applied to packets 
that are forwarded through software. Hence, with multicast routing enabled, the first few packets are 
sent from a port even when IGMP filtering is set to deny those groups on that port. There is no 
workaround. (CSCdy82818)

If an IG MP report packet has two multicast group records, the switch removes or adds interfaces 
depending on the order of the records in the packet:

If the ALLOW_NEW_SOURCE record is before the BLOCK_OLD_SOURCE record, the 
switch removes the port from the group. 

If the BLOCK_OLD_SOURCE record is before the ALLOW_NEW_SOURCE record, the 
switch adds the port to the group. 

There is no workaround. (CSCec20128)

When IGMP snooping is disabled and you enter the switchport block multicast interface 
configuration command, IP multicast traffic is not blocked. 

The switchport block multicast interface configuration command is only applicable to non-IP 
multicast traffic.

There is no workaround. (CSCee16865)

Some switch queues are disabled if the buffer size or threshold level is set too low with the mls qos 
queue-set output global configuration command. The ratio of buffer size to threshold level should 
be greater than 10 to avoid disabling the queue. The workaround is to choose compatible buffer sizes 
and threshold levels. (CSCea76893)

When auto-QoS is enabled on the switch, priority queuing is not enabled. Instead, the switch uses 
shaped round robin (SRR) as the queuing mechanism. The auto-QoS feature is designed on each 
platform based on the feature set and hardware limitations, and the queuing mechanism supported 
on each platform might be different. There is no workaround. (CSCee22591)