Methods and apparatus to support dynamic allocation of traffic management resources in a network element

• US 20070061433A1
• Filed: 09/12/2005
• Published: 03/15/2007
• Est. Priority Date: 09/12/2005
• Status: Active Grant

• Current Assignee: Tellabs Communications Canada Ltd. (Infinera Corp.)
• Original Assignee: Zeugma Systems Canada, Inc.
• Inventors: Luft, Siegfried, Reynolds, Scott
• Granted Patent: US 7,733,891 B2
• Time in Patent Office: Days
• US Class Current: 709/223
• CPC Class Codes: H04L 41/0806 for initial configuration or provisioning, e.g. plug-and-play
• CPC Class Codes: H04L 41/0816 the condition being an adaptation, e.g. in response to network events
• CPC Class Codes: H04L 41/082 the condition being updates or upgrades of network functionality
• CPC Class Codes: H04L 41/0893 Assignment of logical groups to network elements
• CPC Class Codes: H04L 41/0894 Policy-based network configuration management
• CPC Class Codes: H04L 41/0896 Bandwidth or capacity management, i.e. automatically increasing or decreasing capacities flow or congestion control using dynamic resource allocation, e.g. in-call renegotiation, H04L47/76
• CPC Class Codes: H04L 41/5022 by giving priorities, e.g. assigning classes of service
• CPC Class Codes: H04L 41/5045 Making service definitions prior to deployment
• CPC Class Codes: H04L 43/00 Arrangements for monitoring or testing data switching networks
• CPC Class Codes: H04L 47/10 Flow control; Congestion control
• CPC Class Codes: H04L 47/20 Traffic policing
• CPC Class Codes: H04L 47/2441 relying on flow classification, e.g. using integrated services [IntServ]
• CPC Class Codes: H04L 47/2475 for supporting traffic characterised by the type of applications

Abstract

Methods and apparatus to support dynamic allocation of traffic management resources in a network element. Shared pools of traffic management resources comprising an aggregation of local line card resources distributed across the line cards or a network element maintained by apparatus software. Incoming packets are classified into subscriber flows using a hierarchical classification scheme. In view of subscriber services and flow application types, traffic management resources are dynamically allocated from the shared pools, and traffic management policies associated with the subscriber services and application types are applied to the subscriber flows via the allocated resources. In response to detecting a subscriber flow has terminated, the allocated resources are release and made available to be dynamically re-allocated to subsequent subscriber flows.

Claims

1. A method comprising:

maintaining shared pools of traffic management resources on a network element;

dynamically allocating traffic management resources from the shared pools to facilitate transport of respective subscriber flows across the network element;

applying traffic management policies on a per-subscriber flow basis using the traffic management resources allocated to that subscriber flow;

determining when a subscriber flow is terminated;

releasing the traffic management resources that were allocation for a terminated subscriber flow; and

dynamically re-allocating the released traffic management resources to new subscriber flows.

2. The method of claim 1, further comprising:

monitoring traffic management parameters; and

adjusting traffic management policies applied to subscriber flows in response to detecting a change in the traffic management parameters.

3. The method of claim 1, further comprising:

determining an application associated with a subscriber flow; and

applying an application-based traffic management policy to the subscriber flow based on the application associated with the subscriber flow.

4. The method of claim 1, further comprising:

classifying a subscriber flow by, performing a first level of classification on packets corresponding to the subscriber flow; and

performing at least a second level of classification on at least a portion of the packets; and

allocating traffic management resources based on classification of the subscriber flow.

5. The method of claim 4, further comprising:

performing an N-tuple signature match to perform the first level of classification; and

performing deep packet inspection to perform said at least a second level of classification.

6. The method of claim 5, wherein the N-tuple signature match is performed using a first processing element, and the deep packet inspection is performed using a second processing element.

7. The method of claim 1, further comprising:

receiving a first portion of packets corresponding to a unclassified subscriber flow;

employing an N-tuple signature match filter to filter the first portion of packets to effect a first filter function;

employing deep packet inspection associated with the first filter function to effect a second filter function;

classifying the subscriber flow via the first and second filter functions;

updating a signature employed by the N-tuple signature match filter to associate packets having a matching N-tuple signature with the subscriber flow; and

employing the N-tuple signature match filter to classify a subsequent portion of packets associated with the subscriber flow.

8. The method of claim 1, wherein the network element comprises a modular chassis including a plurality of traffic blades and compute blades, the method further comprising:

receiving packets at an ingress traffic blade;

performing classification of the packets to identify a subscriber flow associated with the packets; and

based on the subscriber flow, identifying an egress traffic blade to be employed for egress traffic flow operations and sending a first copy of the packets to that egress traffic blade; and

sending a second copy of the packets to one of a compute resource on the ingress traffic blade or a compute blade.

9. The method of claim 8, further comprising:

detecting initiation of a subscriber session;

classifying packets associated with the subscriber session to a subscriber flow; and

allocating traffic blade traffic management resources for the subscriber flow.

10. The method of claim 1, further comprising statistically multiplexing allocation of traffic management queues, each traffic management queue associated with a subscriber and service corresponding to a respective subscriber flow.

11. The method of claim 1, further comprising:

performing sub-sampling of packets associated with a subscriber flow, the subscriber flow being identified via a first level of classification;

performing at least a second level of classification on the sub-sampling of packets; and

modifying traffic management operations corresponding to the subscriber flow in view of information obtained via the at least a second level of classification.

12. The method of claim 1, further comprising:

implementing a multi-level classification hierarchy mechanism to classify packets into subscriber flows.

13. The method of claim 12, wherein the multi-level classification hierarchy mechanism classifies subscribers flows based on identification of a subscriber and an application.

14. The method of claim 12, wherein the multi-level classification hierarchy mechanisms comprises:

a first-level filtering mechanism implemented at line rate;

a packet bifurcation mechanism to bifurcate processing of at least a portion of packets classified by the first-level filtering mechanism to a second-level filtering mechanism,

wherein the packet bifurcation mechanisms introduces no jitter or delay in processing of the first-level filtering mechanism.

15. The method of claim 12, further comprising installing new filter rules into lower levels of the classification hierarchy in response to subscriber flow classification information derived for the multi-level classification hierarchy mechanism.

16. An apparatus, comprising:

a chassis having a plurality of slots and including a backplane providing a mesh interconnect between the slots;

a plurality of traffic blades, each installed in the chassis in a respective slot and including a backplane interface coupled to the backplane, the plurality of traffic blades including ingress traffic blades and egress traffic blades;

a plurality of compute blades, each installed in the chassis in a respective slot and including a backplane interface coupled to the backplane; and

software components distributed across the plurality of traffic blades and compute blades, the software components to execute on processing elements hosted by the traffic blades and compute blades to perform operations including,

managing shared pools of traffic management resources comprising an aggregation of local traffic management resources hosted by respective traffic blades;

performing ingress processing for packets received at an ingress traffic blade, the ingress processing including performing classification of the packets to identify a subscriber flow associated with the packets;

dynamically allocating traffic management resources from the shared pools to facilitate transport of the subscriber flow across the network element; and

applying traffic management policies on a per-subscriber flow basis using the traffic management resources allocated to each subscriber flow.

17. The apparatus of claim 16, wherein execution of the software performs further operations comprising:

implementing a multi-level classification hierarchy mechanism to classify packets into subscriber flows.

18. The apparatus of claim 17, wherein the multi-level classification hierarchy mechanisms comprises:

a first-level filtering mechanism implemented at line rate by a network processor unit on an ingress traffic blade;

a packet bifurcation mechanism to bifurcate processing of at least a portion of packets classified by the first-level filtering mechanism to a second-level filtering mechanism;

wherein the second-level filtering mechanism is implemented via execution of a portion of the distributed software on a host processor for the ingress traffic blade.

19. The apparatus of claim 16, wherein execution of the software performs further operations comprising:

statistically multiplexing allocation of traffic management queues, each traffic management queue associated with a subscriber and service corresponding to a respective subscriber flow.

20. The apparatus of claim 16, wherein execution of the software performs further operations comprising:

classifying a subscriber flow by, performing a first level of classification on packets corresponding to the subscriber flow; and

performing at least a second level of classification on at least a portion of the packets; and

allocating traffic management resources based on classification of the subscriber flow.

21. The apparatus of claim 16, wherein execution of the software performs further operations comprising:

performing classification of the packets to classify the subscriber flow as a particular application traffic type; and

managing egress traffic flow operations corresponding to the subscriber flow based on its application traffic type.

22. The apparatus of claim 21, wherein an ingress traffic blade includes a network processor unit (NPU) and a host processor, and wherein classification of packets into particular application traffic types is performed by:

employing the NPU to perform a first level classification using the NPU; and

employing the host processor to perform at least one additional level of classification including deep packet inspection.

23. The apparatus of claim 16, wherein execution of the software performs further operations comprising:

effecting a distributed traffic management database in which traffic management data is stored, the distributed traffic management database including local instances of the database hosted on respective traffic and compute blades;

updating a local instance of the traffic management database with traffic management data generated by its host blade; and

propagating the update to other local instances of the traffic management database hosted by other blades.

24. The apparatus of claim 16, wherein the chassis comprises an Advanced Telecommunication and Computing Architecture (ATCA) chassis.

25. A network element, comprising:

a chassis in which a plurality of line cards are installed means for forwarding IP datagrams received at an ingress interface using a plurality of managed traffic flows having different service levels;

means for associating IP datagrams with a subscriber flow;

means for assigning a subscriber flow to a managed traffic flow;

means for maintaining a shared pool of traffic management resources; and

means for dynamically allocating traffic management resources from the shared pool to a managed traffic flow,

wherein the shared pool of traffic management resources is distributed across the plurality of line cards.

26. The network element of claim 25, further comprising:

means for identifying an application associated with the subscriber flow.

27. The network element of claim 25, further comprising:

means for implementing a multi-level hierarchy classification scheme.

28. The network element of claim 25, further comprising:

means for determining when a subscriber flow has terminated;

means for re-allocating traffic management resources allocated to the terminated subscriber flow to a new subscriber flow.