Selective packet sequence acceleration in a network environment

• US 8,743,690 B1
• Filed: 06/14/2011
• Issued: 06/03/2014
• Est. Priority Date: 06/14/2011
• Status: Active Grant

• Current Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Original Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Inventors: Mackie, Robert A., Tsang, Tzu-Ming, Dixon, Walter, Janarthanan, Bhagatram Yaugand
• Primary Examiner(s): Weidner, Timothy J
• Application Number:
  US13/160,426
• Time in Patent Office: 1,085 Days
• Field of Search: None
• US Class Current: 370/235
• CPC Class Codes: H04L 45/22 Alternate routing
• CPC Class Codes: H04L 45/38 Flow based routing
• CPC Class Codes: H04L 47/29 using a combination of thresholds
• CPC Class Codes: H04L 47/31 by tagging of packets, e.g. using discard eligibility [DE] bits

Abstract

At least one first frame of a first data flow is inspected by a general processing unit to at least determine whether a subsequent portion of the first data flow can be delegated to a network processing unit for accelerated processing. A deceleration trigger is defined for the first data flow from the inspection of the first frame. A first subsequent portion of the first data flow is delegated to a network processing unit for accelerated processing based, at least in part, on the defining of the first deceleration trigger. The first deceleration trigger defines at least one first condition that, when met during accelerated processing of the first data flow, prompts returning processing of the first data flow from the network processing unit to the general processing unit for further inspection of subsequent packets.

Claims

1. A method, comprising:

inspecting at least one first frame of a first data flow to at least determine whether a subsequent portion of the first data flow can be delegated to a network processing unit for accelerated processing;

delegating a first subsequent portion of the first data flow to the network processing unit for accelerated processing based, at least in part, on a first deceleration trigger for the first data flow; and

causing the at least one first frame to be sent, to the network processing unit, with acceleration instructions defining at least one first condition that prompts returning processing of the first data flow from the network processing unit to a general processing unit.

2. The method of claim 1, wherein the acceleration instructions are attached to the at least one first frame.

3. The method of claim 2, wherein the network processing unit strips the acceleration instructions from the first frame prior to forwarding the first frame to a remote computing device.

4. The method of claim 1, further comprising:

inspecting at least one particular frame of a second data flow by the general processing unit to at least determine whether a subsequent portion of the second data flow can be delegated to the network processing unit for accelerated processing;

determining, from the inspection of the particular frame, that no deceleration triggers are available for the second data flow; and

determining that the second data flow cannot be accelerated based, at least in part, on the determination that no deceleration triggers are available for the second data flow.

5. The method of claim 4, further comprising processing each subsequent frame in the second data flow using the general processing unit in response to the determination that the second data flow cannot be accelerated.

6. The method of claim 1, further comprising receiving a particular frame in the first data flow forwarded to the general processing unit by the network processing unit in response to the first deceleration trigger being met, wherein the particular frame was received by the network processing unit after delegation of the first subsequent portion of the first data flow to the network processing unit for accelerated processing.

7. The method of claim 6, further comprising inspecting at least the particular frame, using the general processing unit, to at least determine whether another subsequent portion of the first data flow can be delegated to the network processing unit for accelerated processing.

8. The method of claim 7, further comprising:

defining a second deceleration trigger for the first data flow; and

delegating another subsequent portion of the first data flow to the network processing unit for accelerated processing based, at least in part, on the defining of the second deceleration trigger.

9. The method of claim 7, further comprising:

determining that no additional deceleration triggers are available for the first data flow; and

determining that other subsequent portions of the first data flow cannot be accelerated based, at least in part, on the determination that no additional deceleration triggers are available for the first data flow.

10. The method of claim 1, wherein the general processing unit defines the first deceleration trigger.

11. The method of claim 1, wherein accelerated processing of the first subsequent portion of the first data flow includes:

receiving each frame in the first subsequent portion at the network processing unit; and

forwarding each frame in the first subsequent portion to at least one remote computing device using the network processing unit, wherein accelerated processing of the first subsequent portion results in each frame in the first subsequent portion being forwarded by the network processing unit without processing by the general processing unit.

12. The method of claim 1, wherein the first deceleration trigger is a time-based trigger.

13. The method of claim 1, wherein the first deceleration trigger is a data volume-based trigger.

14. The method of claim 1, further comprising receiving the at least one first frame from a remote computing device via the network processing unit.

15. The method of claim 1, further comprising forwarding the at least one first frame to a remote computing device, via the network processing unit, following the inspection of the at least one first frame.

16. Logic encoded in non-transitory media that includes code for execution and when executed by a processor is operable to perform operations comprising:

inspecting at least one first frame of a first data flow to at least determine whether a subsequent portion of the first data flow can be delegated to a network processing unit for accelerated processing;

delegating a first subsequent portion of the first data flow to the network processing unit for accelerated processing based, at least in part, on a first deceleration trigger for the first data flow; and

causing the at least one first frame to be sent, to the network processing unit, with acceleration instructions defining at least one first condition that prompts returning processing of the first data flow from the network processing unit to a general processing unit.

17. A system comprising:

at least one memory element for storing data;

at least one network processor configured to;

perform accelerated processing for frames in a received data flow based at least in part on acceleration instructions sent with at least one frame in the received data flow;

at least one general processor configured to;

perform a set of deep packet inspection operations on at least a particular portion of frames of the received data flow, wherein at least one deceleration trigger is provided for the received data flow; and

determine whether a subsequent portion of the received data flow can be delegated to the network processor for accelerated processing based at least in part on the at least one deceleration trigger for the received data flow.

18. The system of claim 17, wherein the system comprises a gateway router for a network.

19. The system of claim 17, further comprising:

an in-band communication link between the network processor and the general processor; and

an out-of-band communication link between the network processor and the general processor.