Granted Patent B2

US 8,605,726 · App. 12/972,391 · Granted Dec 10, 2013

Inventors: Donald B. Grosser, Jr. (Apex, NC); Edward J. Rovner (Chapel Hill, NC)

Assignee: Extreme Networks, Inc.

Abstract

Claims (39)

1. A system for next hop scaling with link aggregation, the system comprising:

a packet forwarding device including:

a plurality of packet processors for performing the next hop and link aggregation group (LAG) selection operations;

wherein ingress packet processors of the plurality of packet processors are configured to perform a link aggregation group (LAG) member selection operation for packets that arrive from external ports of the packet forwarding device and are further configured to indicate, for received packets that have a next hop on a different packet processor, that an egress next hop selection operation is needed;

wherein an egress packet processor of the plurality of packet processors is configured to, for the packets for which the egress next hop selection operation is indicated, perform the egress next hop and member selection operations and to limit forwarding of the packets to active LAG group members local to the egress packet processor;

a memory associated with each of the plurality of packet processors stores relay entries for next hops connected to a remainder of the plurality of packet processors; and

wherein the memory associated with each of the plurality of packet processors stores a single relay entry for each of the remainder of the plurality of packet processors.

2. The system of claim 1 wherein receipt of a packet that matched a relay entry on an ingress packet processor triggers an egress packet processor to perform an egress next hop lookup.

3. The system of claim 1 wherein the relay entries do not store MAC addresses for hosts connected to the remainder of the packet processors.

4. The system of claim 1 wherein a memory associated with each of the plurality of packet processors stores MAC addresses only for next hops that are directly connected thereto such that next hop bindings for hosts connected to the packet forwarding device are distributed among memories associated with all of the plurality of packet processors.

5. The system of claim 1 wherein each memory stores a layer 3 host or layer 3 longest prefix match route table that maps IP address information to next hop information and a next hop table that maps the next hop information to layer 2 packet forwarding information.

6. The system of claim 5 wherein an available memory for the layer 3 host or longest prefix match route table and the next hop table is shared.

7. The system of claim 6 wherein each memory stores a MAC forwarding database that stores MAC and VLAN information for hosts connected to the packet forwarding device.

8. The system of claim 1 wherein memories associated with the plurality of packet processors store next hop bindings for real and virtual hosts.

9. The system of claim 8 wherein the next hop bindings comprise address resolution protocol (ARP) bindings.

10. The system of claim 8 wherein the next hop bindings comprise IPv6 neighbor discovery bindings.

11. A method for next hop scaling, the method comprising:

distributing next hop information among memories associated with different packet processors in a packet forwarding device including a plurality of packet processors such that the memories associated with the different packet processors store different next hop bindings;

receiving a packet at an ingress packet processor, the packet having a next hop associated with a link aggregation group (LAG);

performing a link aggregation group (LAG) member selection operation for packets that arrive from external ports of the packet forwarding device;

for received packets that have a next hop on a different packet processor, indicating that an egress next hop selection operation is needed and relaying the packet to an egress packet processor;

for the packets for which the egress next hop selection operation is indicated, performing the egress next hop and link aggregation group (LAG) member selection operations, wherein the link aggregation group (LAG) member selection operation limits forwarding of the packets to active members of the link aggregation group (LAG) that are local to the egress packet processor;

a memory associated with each of the plurality of packet processors stores relay entries for next hops connected to a remainder of the plurality of packet processors; and

wherein the memory associated with each of the plurality of packet processors stores a single relay entry for each of the remainder of the plurality of packet processors.

12. The method of claim 11 wherein receipt of a packet that matched a relay entry on the ingress packet processor triggers the egress packet processor to perform an egress next hop lookup.

13. The method of claim 11 wherein the relay entries do not store MAC addresses for hosts connected to the remainder of the packet processors.

14. The method of claim 11 wherein the memory associated with each of the plurality of packet processors stores MAC addresses only for next hops that are directly connected to each of the plurality of packet processors such that next hop bindings for hosts connected to the packet forwarding device are distributed among the memories associated with all of the plurality of packet processors.

15. The method of claim 11 wherein each memory stores a layer 3 host or longest prefix match route table that maps IP address information to next hop information and a next hop table that maps the next hop information to layer 2 packet forwarding information.

16. The method of claim 15 wherein an available memory for the layer 3 host or longest prefix match route table and the next hop table is shared.

17. The method of claim 16 wherein each of the memories stores a MAC forwarding database that stores MAC and VLAN information for hosts connected to the packet forwarding device.

18. The method of claim 11 wherein the next hop bindings comprise address resolution protocol (ARP) bindings.

19. The method of claim 11 wherein the next hop bindings comprise IPv6 neighbor discovery bindings.

20. A non-transitory computer readable medium having stored thereon executable instructions that when executed by a processor of a computer control the computer to perform steps comprising:

receiving a packet at an ingress packet processor of a packet forwarding device, the packet having a next hop associated with a link aggregation group (LAG);

performing a link aggregation group (LAG) member selection operation for packets that arrive from external ports of the packet forwarding device;

for received packets that have a next hop on a different packet processor, indicating that an egress next hop selection operation is needed and relaying the packet to an egress packet processor of the packet forwarding device;

for the packets for which the egress next hop selection operation is indicated, performing the egress next hop and link aggregation group (LAG) member selection operations, wherein the link aggregation group (LAG) member selection operation limits forwarding of the packets to active members of the link aggregation group (LAG) that are local to the egress packet processor;

a memory associated with each of the plurality of packet processors stores relay entries for next hops connected to a remainder of the plurality of packet processors; and

wherein the memory associated with each of the plurality of packet processors stores a single relay entry for each of the remainder of the plurality of packet processors.

Assignments (10)

Continuity (2)