DISTRIBUTED SERVICE FUNCTION FORWARDING SYSTEM

• US 20170353387A1
• Filed: 08/10/2016
• Published: 12/07/2017
• Est. Priority Date: 06/07/2016
• Status: Active Grant

• Current Assignee: Electronics and Telecommunications Research Institute
• Original Assignee: Electronics and Telecommunications Research Institute
• Inventors: KWAK, Ji Young, KANG, Sae Hoon, BAE, Hyun Joo
• Granted Patent: US 10,063,482 B2
• Time in Patent Office: Days
• CPC Class Codes: H04L 45/28 using route fault recovery
• CPC Class Codes: H04L 45/38 Flow based routing
• CPC Class Codes: H04L 47/125 by balancing the load, e.g. traffic engineering
• CPC Class Codes: H04L 47/18 End to end

Abstract

Provided is a distributed service function (SF) forwarding system that applies the corresponding service function chain (SFC) to traffic classified by a plurality of service network (SN) controller instances based on an SN overlay structure. Therefore, by selectively combining and executing necessary network functions (SFs) according to a path and traffic made up of defined component services, it is possible to dynamically configure and control one network service.

Claims

1. A distributed service function (SF) forwarding system comprising:

service network (SN) controller instances configured to control SNs of a network overlay structure in units of service function chains (SFCs) obtained by abstracting network functions necessary for a service and application sequences of the network functions, and generate service graphs which are service function paths (SFPs); and

a software defined network (SDN) infrastructure controller including an SFC handling classifier function of mapping an SN capable of generating a service graph of an SFC requested for generated traffic to an SN controller instance to resolve a traffic bottleneck caused according to a location of an SFC classifier on which data traffic is concentrated and a number of instances.

2. The distributed SF forwarding system of claim 1, wherein, when there is an SN appropriate for an SFC requested by traffic or a new SN is configured, the SDN infrastructure controller includes SFC-identification (ID) tag information in a request for an SFP of the traffic and transmits the request to an SN controller instance controlling the corresponding SN.

3. The distributed SF forwarding system of claim 1, wherein the SN controller instance configures several SFPs by selectively combining ordered logical SF components.

4. The distributed SF forwarding system of claim 3, wherein the SN controller instance generates the service graph according to an ordered SFP included in the SFC with reference to SF sequences.

5. The distributed SF forwarding system of claim 1, wherein, when an amount of traffic generated in the SN exceeds a requested amount of network resources, the SN controller instance distributes a generated traffic load by resetting multiple SFPs for the SFC according to a multipath forwarding method based on minimum costs dependent on currently available network resources.

6. The distributed SF forwarding system of claim 5, wherein, when resetting a path of the service graph according to the SFC of the SN, the SN controller instance applies the minimum costs-based multipath forwarding method, and distributes a part of generated traffic by selecting an SFP having a smallest number of SF instances constituting an ordered SFP from among a plurality of generated SFPs and setting a replacement path for the path or multiple paths.

7. The distributed SF forwarding system of claim 6, wherein, when a processing overload occurs at a particular SF instance or the SF instance attains a failure state and it is not possible to support the requested SFP during a process of transmitting traffic generated on the SFP constituted by selectively combining ordered logical SF components of the SFC supported by the SN, the SN controller instance provides an uninterrupted SFP function by copying a candidate SF instance capable of replacing the SF instance determined to be in the failure state, in consideration of currently available computing resources.

8. The distributed SF forwarding system of claim 7, wherein, when it is determined that it is difficult to copy the SF instance in the failure state, the SN controller instance determines whether SF instances currently in a stop state among SF instances running for an SFC supported by another SN are of a type capable of replacing a failed SF, and adjusts a configuration of the SN to include a candidate SF instance capable of replacing the failed SF.

9. The distributed SF forwarding system of claim 8, wherein the SN controller instance selects the candidate SF instance for the failed SF from among SF instances included in the SFC of the other SN, in consideration of an increase in communication cost caused by a roundabout path of an SFP passing through the candidate SF instance.

10. The distributed SF forwarding system of claim 9, wherein, when it is determined that it is not possible to select a candidate SF instance, the SN controller instance selects another SFC capable of supporting an SFP function related to the failed SF.

11. The distributed SF forwarding system of claim 10, wherein the SN controller instance selects the other SFC after determining whether there are available network resources and computing resources besides a plurality of SFPs currently running.