Monitoring for replica placement and request distribution

• US 20060190602A1
• Filed: 11/30/2005
• Published: 08/24/2006
• Est. Priority Date: 02/23/2005
• Status: Active Grant

• Current Assignee: AT&T Corporation (AT&T, Inc.)
• Original Assignee: AT&T Corporation (AT&T, Inc.)
• Inventors: Fisher, Stephen, Spatscheck, Oliver, Gerber, Alexandre, Xiao, Zhen, Rabinovich, Michael, Canali, Claudia
• Granted Patent: US 7,941,556 B2
• Time in Patent Office: Days
• US Class Current: 709/226
• CPC Class Codes: H04L 61/4511 using domain name system [DNS]
• CPC Class Codes: H04L 67/1001 for accessing one among a plurality of replicated servers
• CPC Class Codes: H04L 67/1008 based on parameters of servers, e.g. available memory or workload monitoring of computer activity G06F11/30
• CPC Class Codes: H04L 67/101 based on network conditions
• CPC Class Codes: H04L 67/1017 based on a round robin mechanism
• CPC Class Codes: H04L 67/1021 based on client or server locations
• CPC Class Codes: H04L 67/1029 using data related to the state of servers by a load balancer
• CPC Class Codes: H04L 67/1031 Controlling of the operation of servers by a load balancer, e.g. adding or removing servers that serve requests
• CPC Class Codes: H04L 67/1034 Reaction to server failures by a load balancer
• CPC Class Codes: H04L 67/63 Routing a service request depending on the request content or context
• CPC Class Codes: H04L 69/329 in the application layer [OSI layer 7]
• CPC Class Codes: H04L 9/40 Network security protocols

Abstract

A platform that may be used to dynamically reallocation resources to support an Internet application is disclosed. In an embodiment, the platform may include two layers of distribution, one at the DNS layer and one at an Internet data center layer. The platform may include a dynamic resource allocation manager that causes instances of applications to be initiated in Internet data centers in response to dynamic conditions and characteristics of the application. A monitoring system module may be provided to keep the dynamic resource allocation manager informed as the health and utilization of instances of the application.

Claims

1. A method of adjusting a routing of requests for an Internet application on a content distribution network, comprising:

(a) measuring a front-end proximity of a client cluster to an Internet data center running an instance of the Internet application;

(b) measuring a back-end proximity of the Internet data center to a back-end resource;

(c) measuring a first volume of traffic specific to the Internet application on the front-end and measuring a second volume of traffic specific to the Internet application on the back-end;

(d) calculating at least one performance metric for the Internet application with respect to the client cluster; and

(e) routing the requests for the Internet application from the client cluster to the Internet data center in response to the calculated at least one performance metric.

2. The method of claim 1, wherein the measuring in (a) comprises:

(i) passively collecting the round-trip time of at least one client that is member of the client cluster.

3. The method of claim 1, wherein the measuring in (a) comprises:

(i) collecting a plurality of round trip times for a plurality of clients that are members of the client cluster; and

(ii) aggregating the plurality of round trip times for the plurality of clients.

4. The method of claim 1, wherein the measuring in (b) comprises:

(i) actively probing to determine the proximity of the Internet data center to the back-end resource.

5. The method of claim 1, wherein the back end resource is a first back-end resource and the and the measuring in (c) comprises:

(i) passively measuring the volume of traffic between the Internet data center and the client cluster for the application; and

(ii) passively measuring the volume of traffic between the Internet data center and the first back-end resource.

6. The method of claim 5, wherein the Internet application includes a second back-end resource and measuring in (c) further comprises:

(iii) measuring the volume of traffic between the Internet data center and the second back-end resource.

7. The method of claim 1, wherein the calculating in (d) comprises:

(i) multiplying the measured front-end proximity by the first volume of traffic specific to the Internet application;

(ii) multiplying the measured back-end proximity by the second volume of traffic specific to the Internet application;

(iii) dividing the sum of the (i) and (ii) by the sum of the first volume and the second volume; and

(iv) ranking the proximity metric of two or more Internet data centers.

8. The method of claim 1, wherein the Internet data center is a first Internet data center, the method further comprising:

(f) initiating an instance of the Internet application in a second Internet data center in response to the calculating in (d); and

;

(g) routing the request the client cluster to the instance of the Internet application in the second Internet data center.

9. A method of providing a client with an Internet application, comprising:

(a) using a load balancing DNS module to allocate a client'"'"'s request associated with the Internet application to an Internet data center running an instance of the Internet application, the client being positioned in a first client cluster;

(b) using a load balancing switch module in the Internet data center to route the client'"'"'s request to a first application cluster in the Internet data center that is running the instance of the Internet application and has availability; and

(c) monitoring the performance of the Internet application with respect to the first client cluster.

10. The method of claim 9, further comprising:

(d) updating the load balancing DNS module with the performance of the application obtained in (c).

11. The method of claim 9, further comprising:

(d) initiating a second application cluster in the Internet data center; and

(e) updating the load balancing switch module with the status of the second application cluster

12. The method of claim 9, wherein the Internet data center is a first Internet data center and the method further comprises:

(d) initiating an application cluster in a second Internet data center; and

(e) updating the load balancing DNS module regarding the initiation of the application cluster in the second Internet data center.

13. The method of claim 9, wherein the monitoring in (c) comprises:

(i) measuring front-end proximity of the client cluster to the Internet data center running the instance of the Internet application;

(ii) measuring back-end proximity of the Internet data center to a back-end resource; and

(iii) calculating a performance metric for the Internet application with respect to the client cluster and the Internet data center.

14. The method of claim 9, wherein the monitoring in (c) comprises:

(i) using a network sniffer to passively determine round-trip time for a client.

15. The method of claim 9, further comprising:

(d) monitoring the health and utilization of the application cluster; and

(e) providing seamless failover in the event that a server in the application cluster fails.

21. The method of claim 15, wherein the redirecting in (d) comprises:

(i) receiving the client'"'"'s request at a load balancing switch module in an Internet data center;

(ii) determining that the primary server has failed based on information provided by a monitoring system module; and

(iii) redirecting the client request to the secondary server identified in a cookie previously provided to the client.

16. A method of providing for failover for an Internet application, comprising:

(a) creating a client session on a primary server positioned in an Internet data center in response to a request received from a client;

(b) replicating the session on a secondary server;

(c) providing an identify of the secondary server to the client; and

(d) if the primary server fails, redirecting the request to the secondary server based on the identity provided to the client.

17. The method of claim 16, further comprising:

(e) monitoring the health and utilization of the primary server with a monitoring system module; and

(f) updating a load balancing switch module on the health and utilization of the primary server.

18. The method of claim 16, further comprising:

(e) initiating an application cluster in the Internet data center in response to request for additional resources from a dynamic resource allocation manager.

19. The method of claim 18, further comprising:

(f) providing an update to a load balancing DNS module regarding the initiation of the application cluster, whereby client requests may be routed to the application cluster.

20. The method of claim 16, wherein the providing the identity in (c) comprises:

(i) sending a cookie to the client with the identification of the second server provided in the cookie.