Datasheet

206
Release Notes for Cisco IOS Release 12.1E on the Catalyst 6500 and Cisco 7600 Supervisor Engine and MSFC
OL-2310-11
Caveats
• Deny ACEs that do not specify any Layer 4 ports incorrectly do not deny fragmented packets. This
problem is resolved in Release 12.1(20)E. (CSCeb04343)
• An IEEE 802.Q trunking Gigabit EtherChannel formed with interfaces on different DFC-equipped
switching modules might drop some traffic that is Layer 3 switched in hardware or that is routed in
software. This problem is resolved in Release 12.1(20)E. (CSCeb05464)
• When the maximum number of RLB sticky subscribers for a real server (SSG1) are exceeded, the
RLB does not pass AcctStop packets (from GGSN) to SSG1 for the existing host objects, but to the
next SSG in the round-robin pool. The SSG proxies the AcctStop packets to the AAA server, which
then closes the corresponding RADIUS sessions. This situation leads to stale host objects on the first
SSG. This problem is resolved in Release 12.1(20)E. (CSCeb09340)
• Multicast shortcuts take a long time to install during dense mode fallback, which causes latency in
hardware switching. This problem is resolved in Release 12.1(20)E. (CSCeb14435)
• With a Supervisor Engine 2, the IOS ARP and adjacency entries for the next-hop IP address
configured for a static route might not be created. This problem is resolved in Release 12.1(20)E.
(CSCeb38062)
• Supervisor Engine 1 does not have RP-SP inband channel communication monitoring. This problem
is resolved in Release 12.1(20)E. (CSCeb46610)
• On WS-X6548-GE-TX and WS-X6548V-GE-TX modules, CEF-switched Ethernet egress packets
that are less than 64-bytes long are not padded correctly. This problem is resolved in
Release 12.1(20)E. (CSCeb47640)
• A memory leak might occur with Layer 2 aging and EtherChannels that include ports on different
DFC-equipped modules configured. This problem is resolved in Release 12.1(20)E. (CSCeb48732)
• With RPR redundancy configured, the MSFC and OSMs might incorrectly reload. This problem is
resolved in Release 12.1(20)E. (CSCeb49134)
• The CEF entries for traffic from a directly connected Layer 3 address are removed and recreated
randomly, which causes Unicast traffic loss for the affected entries. This problem is resolved in
Release 12.1(20)E. (CSCeb53542)
• There might be OSPF neighbor drops and HSRP flaps when QoS is enabled on a Supervisor
Engine 1 and MSFC2. This problem is resolved in Release 12.1(20)E. (CSCeb55271)
• With a redundant supervisor engine installed, the configuration of EtherChannels that are
reconfigured from Layer 2 to Layer 3 is not synchronized to the redundant supervisor engine. This
problem is resolved in Release 12.1(20)E. (CSCeb56353)
• Memory usage when handling route flaps is not optimal, which causes the route-flap handling
process to hold memory longer than necessary and which can cause out-of-memory conditions when
routes flap continuously. This problem is resolved in Release 12.1(20)E. (CSCeb57465)
• Incorrect VTP pruning might occur if you delete or rename VLANs in VLAN database mode. This
problem is resolved in Release 12.1(20)E. (CSCeb60262)
• The 64-bit SNMP counters on OSM-4GE-WAN and OSM-2+4GE-WAN+ modules behave like
32-bit counters. This problem is resolved in Release 12.1(20)E. (CSCeb60961)
• When the multicast traffic level exceeds the Layer 3 hardware switching capacity, the excess
multicast traffic might be dropped instead of being routed in software on the MSFC. This problem
is resolved in Release 12.1(20)E. (CSCeb62692)
• If you enter the shutdown command and then the no shutdown command on an interface that is
handling a high volume of Layer 3 hardware switched multicast traffic, some of the multicast traffic
is routed in software on the MSFC instead of being Layer 3 switched in hardware when the interface
comes back up. This problem is resolved in Release 12.1(20)E. (CSCeb67996)