Understanding How Layer 3 Switching Works
These sections describe Layer 3 switching:
Understanding Hardware Layer 3 Switching
Hardware Layer 3 switching allows the PFC and DFCs, instead of the MSFC, to forward IP unicast traffic between subnets. Hardware Layer 3 switching provides wire-speed forwarding on the PFC and DFCs, instead of in software on the MSFC. Hardware Layer 3 switching requires minimal support from the MSFC. The MSFC routes any traffic that cannot be hardware Layer 3 switched.
Hardware Layer 3 switching supports the routing protocols configured on the MSFC. Hardware Layer 3 switching does not replace the routing protocols configured on the MSFC.
Hardware Layer 3 switching runs equally on the PF3 and DFCs to provide IP unicast Layer 3 switching locally on each module. Hardware Layer 3 switching provides the following functions:
- Hardware access control list (ACL) switching for policy-based routing (PBR)
- Hardware NetFlow switching for TCP intercept, reflexive ACL forwarding decisions
- Hardware Cisco Express Forwarding (CEF) switching for all other IP unicast traffic
Hardware Layer 3 switching on the PFC supports modules that do not have a DFC. The MSFC forwards traffic that cannot be Layer 3 switched.
Traffic is hardware Layer 3 switched after being processed by access lists and quality of service (QoS).
Hardware Layer 3 switching makes a forwarding decision locally on the ingress-port module for each packet and sends the rewrite information for each packet to the egress port, where the rewrite occurs when the packet is transmitted from the Cisco 7600 series router.
Hardware Layer 3 switching generates flow statistics for Layer 3-switched traffic. Hardware Layer 3 flow statistics can be used for NetFlow Data Export (NDE). (See Chapter 56, “Configuring NetFlow and NDE”.)
Understanding Layer 3-Switched Packet Rewrite
When a packet is Layer 3 switched from a source in one subnet to a destination in another subnet, the Cisco 7600 series router performs a packet rewrite at the egress port based on information learned from the MSFC so that the packets appear to have been routed by the MSFC.
Packet rewrite alters five fields:
- Layer 2 (MAC) destination address
- Layer 2 (MAC) source address
- Layer 3 IP Time to Live (TTL)
- Layer 3 checksum
- Layer 2 (MAC) checksum (also called the frame checksum or FCS)
Note
Packets are rewritten with the encapsulation appropriate for the next-hop subnet.
If Source A and Destination B are in different subnets and Source A sends a packet to the MSFC to be routed to Destination B, the router recognizes that the packet was sent to the Layer 2 (MAC) address of the MSFC.
To perform Layer 3 switching, the router rewrites the Layer 2 frame header, changing the Layer 2 destination address to the Layer 2 address of Destination B and the Layer 2 source address to the Layer 2 address of the MSFC. The Layer 3 addresses remain the same.
In IP unicast and IP multicast traffic, the router decrements the Layer 3 TTL value by 1 and recomputes the Layer 3 packet checksum. The router recomputes the Layer 2 frame checksum and forwards (or, for multicast packets, replicates as necessary) the rewritten packet to Destination B’s subnet.
A received IP unicast packet is formatted (conceptually) as follows:
|
|
|
|
Destination |
Source |
Destination |
Source |
TTL |
Checksum |
|
|
MSFC MAC |
Source A MAC |
Destination B IP |
Source A IP |
n |
calculation1 |
After the router rewrites an IP unicast packet, it is formatted (conceptually) as follows:
|
|
|
|
Destination |
Source |
Destination |
Source |
TTL |
Checksum |
|
|
Destination B MAC |
MSFC MAC |
Destination B IP |
Source A IP |
n-1 |
calculation2 |
Hardware Layer 3 Switching Examples
Figure 29-1 shows a simple network topology. In this example, Host A is on the Sales VLAN (IP subnet 171.59.1.0), Host B is on the Marketing VLAN (IP subnet 171.59.3.0), and Host C is on the Engineering VLAN (IP subnet 171.59.2.0).
When Host A initiates an HTTP file transfer to Host C, Hardware Layer 3 switching uses the information in the local forwarding information base (FIB) and adjacency table to forward packets from Host A to Host C.
Figure 29-1 Hardware Layer 3 Switching Example Topology
Default Hardware Layer 3 Switching Configuration
Table 29-1 shows the default hardware Layer 3 switching configuration.
Table 29-1 Default Hardware Layer 3 Switching Configuration
|
|
Hardware Layer 3 switching enable state |
Enabled (cannot be disabled) |
Cisco IOS CEF enable state on MSFC |
Enabled (cannot be disabled) |
Cisco IOS dCEF enable state on MSFC |
Enabled (cannot be disabled) |
Configuration Guidelines and Restrictions
Follow these guidelines and restrictions when configuring hardware Layer 3 switching:
- Hardware Layer 3 switching supports the following ingress and egress encapsulations:
–
Ethernet V2.0 (ARPA)
–
802.3 with 802.2 with 1 byte control (SAP1)
–
802.3 with 802.2 and SNAP
Configuring Hardware Layer 3 Switching
Note
For information on configuring unicast routing on the MSFC, see Chapter22, “Configuring Layer 3 Interfaces”
Hardware Layer 3 switching is permanently enabled. No configuration is required.
To display information about Layer 3-switched traffic, perform this task:
|
|
Router# show interface {{ type slot/port } | { port-channel number }} | begin L3 |
Displays a summary of Layer 3-switched traffic. |
This example shows how to display information about hardware Layer 3-switched traffic on Fast Ethernet port 3/3:
Router# show interface fastethernet 3/3 | begin L3
L3 in Switched: ucast: 0 pkt, 0 bytes - mcast: 12 pkt, 778 bytes mcast
L3 out Switched: ucast: 0 pkt, 0 bytes - mcast: 0 pkt, 0 bytes
4046399 packets input, 349370039 bytes, 0 no buffer
Received 3795255 broadcasts, 2 runts, 0 giants, 0 throttles
Note
The Layer 3 switching packet count is updated approximately every five seconds.
Cisco IOS CEF and dCEF are permanently enabled. No configuration is required to support hardware Layer 3 switching.
With a PFC (and DFCs, if present), hardware Layer 3 switching uses per-flow load balancing based on IP source and destination addresses. Per-flow load balancing avoids the packet reordering that can be necessary with per-packet load balancing. For any given flow, all PFC- and DFC-equipped switches make exactly the same load-balancing decision, which can result in nonrandom load balancing.
The Cisco IOS CEF ip load-sharing per-packet, ip cef accounting per-prefix, and ip cef accounting non-recursive commands are not supported on the C7600 series routers.
For information about Cisco IOS CEF and dCEF on the MSFC, refer to these publications:
- The “Cisco Express Forwarding” sections at this URL:
http://www.cisco.com/univercd/cc/td/doc/product/software/ios122/122cgcr/fswtch_c/swprt1/index.htm
- The Cisco IOS Switching Services Command Reference publication at this URL:
http://www.cisco.com/en/US/docs/ios/12_2/switch/command/reference/fswtch_r.html
Displaying Hardware Layer 3 Switching Statistics
Hardware Layer 3 switching statistics are obtained on a per-VLAN basis.
To display hardware Layer 3 switching statistics, perform this task:
|
|
Router# show interfaces {{ type slot/port } | { port-channel number }} |
Displays hardware Layer 3 switching statistics. |
This example shows how to display hardware Layer 3 switching statistics:
Router# show interfaces gigabitethernet 9/5 | include Switched
L2 Switched: ucast: 8199 pkt, 1362060 bytes - mcast: 6980 pkt, 371952 bytes
L3 in Switched: ucast: 0 pkt, 0 bytes - mcast: 0 pkt, 0 bytes mcast
L3 out Switched: ucast: 0 pkt, 0 bytes - mcast: 0 pkt, 0 bytes
To display adjacency table information, perform this task:
|
|
Router# show adjacency [{{ type slot/port } | { port-channel number }} | detail | internal | summary ] |
Displays adjacency table information. The optional detail keyword displays detailed adjacency information, including Layer 2 information. |
This example shows how to display adjacency statistics:
Router# show adjacency gigabitethernet 9/5 detail
Protocol Interface Address
IP GigabitEthernet9/5 172.20.53.206(11)
Note
Adjacency statistics are updated approximately every 60 seconds.