Addendum

• The following note applies to the Marking Egress Packets with a DEI Value section of the S4810,
S4820T, S6000, Z9000, and MXL Switch Configuration Guides:
You cannot set the Canonical Format Identifier (CFI) bit alone in an outgoing packet because it will
cause the IEEE 802.1ad drop eligible indicator (DEI) bit in the outgoing dot1p packet to also be reset.
This behavior is expected when you configure the incoming DEI value to be honored by mapping it to
an FTOS drop precedence, and also specify the DEI marking on the ingress and egress interface. For
example, if you enter the enable dei honor 0 yellow command on the ingress interface and the
enable dei mark yellow 1 on both the ingress and egress interfaces, and send traffic with CFI bit
set to 0, the CFI bit is set to 0 instead of being set to 1 on the egress interface.
• The following note applies to the Configure Layer 2 and Layer 3 ACLs section of the Access Control
Lists (ACLs) chapter of the Configuration Guides of all of the platforms:
The egress MAC ACL counter shows zero for existing rules in a VLAN, when a new rule is added with a
sequence number that is less than the existing rule. The counter of the old rule is reset to zero. The
counter of the old rule is updated starting with zero, even though the lesser sequence number rule is
removed from the access group. The update time of the counter differs for continuous and discrete
traffic for VLAN, physical, and port channel interfaces. This behavior is seen with both standard and
extended MAC ACLs on the ingress and egress direction.
• The following additional information applies to the Important Points to Remember section of the
Virtual Link Trunking (VLT) chapter of the S4820T Configuration Guide:
Some of the IPv6 neighbors and ARP entries are learned on the VLT LAG member (physical port) and
VLTi instead of being learned on the VLT LAG after reloading the TOR device. The IP application
receives ARP and neighbor advertisement (NA) packets over the physical port. It learns ARP/NA entries
over the physical port.
• The following information regarding the enabling of DCB globally on a system applies to the dcb-
enable command topic of the Data Center bridging (DCB) chapter of the S6000 Command
Reference Guide:
On an S6000 Switch, when BFD or LACP with faster convergence is enabled and if you enter the dcb-
enable command to enable DCB globally, a protocol flap is an expected behavior.
• The following note applies to the rate-shape command topic of the QoS chapter of the S6000
Command Reference Guide:
When you configure per-queue level rate-shaping, do not use values less than 100 Mbps.
• The following note applies to the Proxy ARP section of the VLT chapter of this addendum document:
Because the proxy ARP capability depends on the peer routing settings, you must ensure the
following: You must remove the VLT domain in both the peers. You must configure the VLT peer
routing timeout by using the multicast peer-routing-timeout command in VLT DOMAIN
mode as a definitive value instead of an infinite value.
• The following note applies to the service-class dot1p-mapping command topic of the QoS
chapter of the MXL Command Reference Guide:
With default dot1p to queue mapping on an MXL Switch, ETS configuration from Cisco Nexus 7000
Series Switches not supported. In such an environment, you must add the service-class dot1p-
mapping dot1p2 2
command to the configuration file of the MXL Switch to enable the default
Cisco FCoE service policy to work when service-class dynamic dot1p is used.
• The following note applies to the Guidelines section of the FC Flex IO chapter of this addendum
document:
While storage traffic traverses through FC Flex IO modules and the Ethernet uplink port-channel
status changes (with DCB enabled on an adjacent switch), FCoE traffic is disrupted. This disruption
Documentation Updates
275