End-to-end acceleration of dynamic content

• US 8,819,187 B1
• Filed: 10/29/2013
• Issued: 08/26/2014
• Est. Priority Date: 10/29/2013
• Status: Active Grant

• Current Assignee: Limelight Networks, Inc.
• Original Assignee: Limelight Networks, Inc.
• Inventors: Hofmann, Jason
• Primary Examiner(s): WANG, LIANG CHE A
• Application Number:
  US14/066,415
• Time in Patent Office: 301 Days
• Field of Search: 709/217, 709/218, 709/219
• US Class Current: 709/219
• CPC Class Codes: H04L 2101/668 Internet protocol [IP] address subnets
• CPC Class Codes: H04L 47/12 Avoiding congestion; Recovering from congestion
• CPC Class Codes: H04L 61/5007 Internet protocol [IP] addresses
• CPC Class Codes: H04L 65/612 for unicast
• CPC Class Codes: H04L 65/80 Responding to QoS
• CPC Class Codes: H04L 67/1097 for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS]
• CPC Class Codes: H04L 67/563 Data redirection of data network streams
• CPC Class Codes: H04L 67/568 Storing data temporarily at an intermediate stage, e.g. caching
• CPC Class Codes: H04L 67/5681 Pre-fetching or pre-delivering data based on network characteristics
• CPC Class Codes: H04L 69/14 Multichannel or multilink protocols

Abstract

A method of accelerating dynamic content between Points of Presence (POPs) and front-loading latency bottlenecks in a Content Delivery Network (CDN) may include receiving, at a first POP, a content request from a user device, where content responsive to the content request includes a first content portion that is not stored in the first POP. The method may also include determining that the first content portion is available through a second POP. The method may additionally include transmitting the first content portion through a persistent connection mesh from the second POP to the first POP, wherein the persistent connection mesh is maintained between the first POP and the second POP prior to the request for content being received by the first POP. The method may further include delivering the content responsive to the content request to the user device.

Claims

1. A system for accelerating dynamic content between Points of Presence (POPs) and front-loading latency bottlenecks in a Content Delivery Network (CDN), the system comprising:

a first POP in the CDN configured to receive a content request from a user device, wherein;

the first POP comprises a first plurality of edge servers that store and distribute content in response to user requests; and

content responsive to the content request comprises a first content portion that is not stored in the first POP;

a second POP in the CDN configured to receive a content request from a user device, wherein;

the second POP comprises a second plurality of edge servers that store and distribute content in response to user requests; and

the first content portion is available through the second POP;

a persistent connection mesh between the first POP and the second POP, wherein;

the persistent connection mesh is maintained between the first POP and the second POP prior to the request for content being received by the first POP; and

the first content portion is transmitted through the persistent connection mesh from the second POP to the first POP.

2. The system for accelerating dynamic content between POPs and front-loading latency bottlenecks in a CDN of claim 1, wherein the persistent connection mesh is maintained without requiring a user request.

3. The system for accelerating dynamic content between POPs and front-loading latency bottlenecks in a CDN of claim 1, wherein the persistent connection mesh eliminates TCP handshake requirements for the content request.

4. The system for accelerating dynamic content between POPs and front-loading latency bottlenecks in a CDN of claim 1, wherein:

the persistent communication mesh comprises a first connection;

the first connection is maintained until it becomes stale or unreliable; and

in response to the first connection becoming stale or unreliable, the first connection is refreshed or renewed.

5. The system for accelerating dynamic content between POPs and front-loading latency bottlenecks in a CDN of claim 1, wherein the first content portion is dynamic content.

6. The system for accelerating dynamic content between POPs and front-loading latency bottlenecks in a CDN of claim 1, wherein the first content portion is retrieved from an origin server through the second POP.

7. The system for accelerating dynamic content between POPs and front-loading latency bottlenecks in a CDN of claim 1, wherein:

the content responsive to the content request comprises a second content portion;

the second content portion comprises static content; and

the static content is stored in the first POP.

8. The system for accelerating dynamic content between POPs and front-loading latency bottlenecks in a CDN of claim 1, wherein the first POP comprises one or more concentration servers that maintain the persistent connection mesh.

9. The system for accelerating dynamic content between POPs and front-loading latency bottlenecks in a CDN of claim 8, wherein the one or more concentration servers send the content request through the persistent connection mesh to at least one concentration server in the second POP.

10. The system for accelerating dynamic content between POPs and front-loading latency bottlenecks in a CDN of claim 1, wherein the persistent connection mesh is pre-warmed to increase a congestion window.

11. A method of accelerating dynamic content between Points of Presence (POPs) and front-loading latency bottlenecks in a Content Delivery Network (CDN), the method comprising:

receiving, at a first POP, a content request from a user device;

wherein;

the first POP comprises a first plurality of edge servers that store and distribute content in response to user requests; and

content responsive to the content request comprises a first content portion that is not stored in the first POP;

determining that the first content portion is available through a second POP, wherein the second POP comprises a second plurality of edge servers that store and distribute content in response to user requests;

transmitting the first content portion through a persistent connection mesh from the second POP to the first POP, wherein the persistent connection mesh is maintained between the first POP and the second POP prior to the request for content being received by the first POP; and

delivering the content responsive to the content request to the user device.

12. The method of accelerating dynamic content between POPs and front-loading latency bottlenecks in a CDN of claim 11, wherein the persistent connection mesh is maintained without requiring a user request.

13. The method of accelerating dynamic content between POPs and front-loading latency bottlenecks in a CDN of claim 11, wherein the persistent connection mesh eliminates TCP handshake requirements for the content request.

14. The method of accelerating dynamic content between POPs and front-loading latency bottlenecks in a CDN of claim 11, further comprising:

detecting that a first connection in the persistent connection mesh has become stale or unreliable; and

in response to the first connection becoming stale or unreliable, refreshing or renewing the first connection.

15. The method of accelerating dynamic content between POPs and front-loading latency bottlenecks in a CDN of claim 11, wherein the first content portion is dynamic content.

16. The method of accelerating dynamic content between POPs and front-loading latency bottlenecks in a CDN of claim 11, further comprising retrieving the first content from an origin server through the second POP.

17. The method of accelerating dynamic content between POPs and front-loading latency bottlenecks in a CDN of claim 11, wherein:

the content responsive to the content request comprises a second content portion;

the second content portion comprises static content; and

the static content is stored in the first POP.

18. The method of accelerating dynamic content between POPs and front-loading latency bottlenecks in a CDN of claim 11, wherein the first POP comprises one or more concentration servers that maintain the persistent connection mesh.

19. The method of accelerating dynamic content between POPs and front-loading latency bottlenecks in a CDN of claim 18, wherein the one or more concentration servers send the content request through the persistent connection mesh to at least one concentration server in the second POP.

20. The system for accelerating dynamic content between POPs and front-loading latency bottlenecks in a CDN of claim 11, wherein the persistent connection mesh is pre-warmed to increase a congestion window.