Failure notification in a network having serially connected nodes

• US 8,223,660 B2
• Filed: 04/18/2008
• Issued: 07/17/2012
• Est. Priority Date: 04/18/2007
• Status: Expired due to Fees

• Current Assignee: RPX Clearinghouse LLC (RPX Corporation)
• Original Assignee: Rockstar Bidco LP (Rockstar Consortium Inc.)
• Inventors: Allan, David, Unbehagen, Paul, Ashwood Smith, Peter, Bragg, Nigel, Betts, Malcolm
• Primary Examiner(s): Yao, Kwang B
• Assistant Examiner(s): PASIA, REDENTOR M
• Application Number:
  US12/148,418
• Publication Number:
  US 20080259784A1
• Time in Patent Office: 1,551 Days
• Field of Search: 370229-231, 370/235, 370/248, 370/252, 370/312, 370/389, 370/392, 370/432, 709220-222, 709224-225, 709/238, 709/242
• US Class Current: 370/248
• CPC Class Codes: H04L 41/06 Management of faults, events, alarms or notifications
• CPC Class Codes: H04L 45/02 Topology update or discovery
• CPC Class Codes: H04L 45/03 by updating link state protocols
• CPC Class Codes: H04L 45/16 Multipoint routing
• CPC Class Codes: H04L 45/28 using route fault recovery
• CPC Class Codes: H04L 45/32 Flooding denial of service attacks H04L63/1458

Abstract

Multicast capabilities of a link state protocol controlled network are used to accelerate the flooding advertisement of topology change notifications within portions of the network. This flooding mechanism may be particularly efficient in a network with a large number of two-connected nodes such as a ring network architecture. A control plane specific multicast group address is used when flooding topology change notifications, and a process such as reverse path forwarding check is used as an additional control on forwarding of the notification to prevent looping of control plane packets. Two-connected nodes insert a forwarding entry into their FIB to enable frames containing the control message to be forwarded via the data plane on to the downstream node so that propagation of the control message along a chain of two-connected nodes may occur at data plane speeds.

Claims

1. A method of enabling rapid failure notification in a link state protocol controlled Ethernet network having dataplane multicast capability, the method comprising the steps of:

issuing, by a first node operating in the link state protocol controlled Ethernet network, a frame containing a control message signifying a local topology change, the frame being addressed to a control plane specific multicast group address;

receiving, by a second node operating in the link state protocol controlled Ethernet network, the frame containing the control message from the first node, the frame being addressed to the control plane specific multicast group address;

performing a reverse path forwarding check on the frame by the second node to determine if the frame arrived on an expected port at the second node for the first node;

discarding the frame immediately, by the second node, when the frame fails the reverse path forwarding check; and

forwarding the frame, by the second node toward a plurality of downstream nodes in the link state protocol controlled Ethernet network, according to forwarding state stored in a forwarding information base on the second node only if the frame passes the reverse path forwarding check

wherein a selection of the plurality of downstream nodes to which the frame is forwarded is determined by the steps of;

only forwarding the frame if the second node is two-connected between the first node and a third node, and

only forwarding the frame by the second node to the third node from which the frame was not received; and

processing the frame in a control plane prior to forwarding the frame if the second node is more than two-connected.

2. The method of claim 1, wherein the control message is a link state routing protocol control message.

3. The method of claim 1, where the control plane specific multicast group address is a source-specific multicast address.

4. The method of claim 1, where the selection of the plurality of downstream nodes to which the frame is forwarded is determined by a shortest path tree rooted on an originator of the message.

5. The method of claim 1, where the control plane specific multicast group address is a (*,G) multicast address common to all nodes and assigned to control message exchange.

6. The method of claim 1, further comprising the step of:

acknowledging the frame, by the second node to the first node, if the frame passes the reverse path forwarding check on the second node.

7. The method of claim 6, wherein the step of acknowledging the frame to the first node is only performed if the second node is two-connected between the first node and the third node, wherein a node is two-connected if it has two Network to Network Interface (NNI) adjacencies.

8. A method of enabling rapid failure notification in a link state protocol controlled network having dataplane multicast capability, the method comprising the steps of:

issuing, by a first node operating in the link state protocol controlled network, a frame containing a control message signifying a local topology change, the frame being addressed to a control plane specific multicast group address;

receiving, by a second node operating in the link state protocol controlled network, the frame containing the control message from the first node, the frame being addressed to the control plane specific multicast group address;

forwarding the frame, by the second node toward a plurality of downstream nodes in the link state protocol controlled network, according to forwarding state stored in a forwarding information base on the second node;

waiting for an acknowledgement of the frame from at least one of the plurality of downstream nodes;

inferring that the frame did not pass a reverse path forwarding check at a particular downstream node if the acknowledgment is not received from the particular downstream node; and

reissuing the frame containing the control message as a second frame containing a reissued control message and transmitting the second frame containing the reissued control message to the particular downstream node that did not acknowledge receipt of the control message, the reissued control message having the same format as the control message, the second frame having a source MAC address identifying a node that reissued the reissued control message as the source of the reissued control message.

9. The method of claim 8, wherein the control message is a link state routing protocol control message.

10. The method of claim 8, where the control plane specific multicast group address is a source-specific multicast address.

11. The method of claim 8, where the control plane specific multicast group address is a (*,G) multicast address common to all nodes and assigned to control message exchange.

12. The method of claim 8, where a selection of the plurality of downstream nodes to which the frame is forwarded is determined by a shortest path tree rooted on the originator of the message.

13. The method of claim 8, further comprising the step of:

acknowledging the frame, by the second node to the first node, if the frame passes the reverse path forwarding check on the second node.

14. The method of claim 13, wherein the step of acknowledging the frame to the first node is only performed if the second node is two-connected between the first node and a third node, wherein a node is two-connected if it is serially connected between two adjacent nodes.