Accurate traffic steering between links through sub-path path quality metrics
Some embodiments provide a method for quantifying quality of several service classes provided by a link between first and second forwarding nodes in a wide area network (WAN). At a first forwarding node, the method computes and stores first and second path quality metric (PQM) values based on packets sent from the second forwarding node for the first and second service classes. The different service classes in some embodiments are associated with different quality of service (QoS) guarantees that the WAN offers to the packets. In some embodiments, the computed PQM value for each service class quantifies the QoS provided to packets processed through the service class. In some embodiments, the first forwarding node adjusts the first and second PQM values as it processes more packets associated with the first and second service classes. The first forwarding node also periodically forwards to the second forwarding node the first and second PQM values that it maintains for the first and second service classes. In some embodiments, the second forwarding node performs a similar set of operations to compute first and second PQM values for packets sent from the first forwarding node for the first and second service classes, and to provide these PQM values to the first forwarding node periodically.
1. A method of forwarding packets along multiple physical layer 1 (L1) network links from a first forwarding node of a wide area network (WAN) to a second forwarding node of the WAN, the method comprising:
at the first forwarding node:
receiving a packet to forward to the second forwarding node along a particular service class;
receiving, from the second forwarding node, first and second path quantifying metric (PQM) values for the particular service class of first and second physical L1 network links connecting the first forwarding node to the second forwarding node, each of the first and second physical L1 network links associated with a plurality of different service classes including the particular service class, each different service class of each particular physical L1 network link associated with a different quality of service (QoS) guarantee provided by the WAN to packet flows using that service class to pass through the particular physical L1 network link, the second forwarding node being configured to compute one more PQM values for data sent by the first forwarding node;
based on the first and second PQM values, selecting the particular service class of the first physical L1 network link for the packet; and
forwarding the packet to the second forwarding node along the particular service class of the first physical L1 network link.
2. The method of claim 1 , each of the plurality of different service classes has an associated PQM value.
3. The method of claim 1 , wherein forwarding the packet comprises associating the packet with a service class sequence number that is unique to the particular service class for the first physical L1 network link.
4. The method of claim 3 , wherein different sequence numbers are used for forwarding packets along different service classes.
5. The method of claim 1 further comprising identifying the particular service class for the received packet by performing a classification operation that matches a set of header values of the packet with an identifier of a classification rule that identifies the particular service class.
6. The method of claim 1 further comprising identifying the particular service class for the received packet through a service class identifier that is contained in a header of the received packet.
7. The method of claim 1 further comprising selecting an Internet broadband connection without a plurality of service classes to forward a second packet from the first forwarding node to the second forwarding node.
8. The method of claim 1 , wherein the retrieved PQM value for each of the first and second physical L1 network links comprises one of a loss value, a delay value, and a jitter value.
9. The method of claim 1 , wherein the first and second forwarding nodes are two edge nodes connecting two different sites through a plurality of physical L1 network links associated with an MPLS (multi-path label switching) service with a plurality of service classes per physical L1 network link.
10. The method of claim 1 , wherein the selection operation is performed for different packets of one packet flow, such that one packet of the flow uses the first physical L1 network link while another packet of the flow uses the second physical L1 network link.
11. The method of claim 1 , wherein the selection operation selects the first physical L1 network link for all packets in a same flow as the received packet.
12. A non-transitory machine readable medium storing a program that when executed by at least one processing unit forwards packets along multiple physical L1 network links from a first forwarding node of a wide area network (WAN) to a second forwarding node of the WAN, the program comprising sets of instructions for:
at the first forwarding node:
receiving a packet to forward to the second forwarding node along a particular service class;
receiving, from the second forwarding node, first and second path quantifying metric (PQM) values for the particular service class of first and second physical L1 network links connecting the first forwarding node to the second forwarding node, each of the first and second physical L1 network links associated with a plurality of different service classes including the particular service class, each different service class of each particular physical L1 network link associated with a different quality of service (QoS) guarantee provided by the WAN to packet flows using that service class to pass through the particular physical L1 network link, the second forwarding node being configured to compute one more PQM values for data sent by the first forwarding node;
based on the first and second PQM values, selecting the particular service class of the first physical L1 network link for the packet; and
forwarding the packet to the second forwarding node along the particular service class of the first physical L1 network link.
13. The non-transitory machine readable medium of claim 12 , each of the plurality of different service classes has an associated PQM value.
14. The non-transitory machine readable medium of claim 12 , wherein forwarding the packet comprises associating the packet with a service class sequence number that is unique to the particular service class for the first physical L1 network link.
15. The non-transitory machine readable medium of claim 14 , wherein different sequence numbers are used for forwarding packets along different service classes.
16. The non-transitory machine readable medium of claim 12 , wherein the program further comprises a set of instructions for identifying the particular service class for the received packet by performing a classification operation that matches a set of header values of the packet with an identifier of a classification rule that identifies the particular service class.
17. The non-transitory machine readable medium of claim 12 , wherein the program further comprises a set of instructions for identifying the particular service class for the received packet through a service class identifier that is contained in a header of the received packet.
18. The non-transitory machine readable medium of claim 12 , wherein the program further comprises a set of instructions for selecting an Internet broadband connection without a plurality of service classes to forward a second packet from the first forwarding node to the second forwarding node.
19. The non-transitory machine readable medium of claim 12 , wherein the retrieved PQM value for each of the first and second physical L1 network links comprises one of a loss value, a delay value, and a jitter value.
20. The non-transitory machine readable medium of claim 12 , wherein the first and second forwarding nodes are two edge nodes connecting two different sites through a plurality of physical L1 network links associated with an MPLS (multi-path label switching) service with a plurality of service classes per physical L1 network link.