Method and system for wide area network load balancing

• US 6,671,259 B1
• Filed: 03/30/1999
• Issued: 12/30/2003
• Est. Priority Date: 03/30/1999
• Status: Expired due to Fees

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: Taniguchi, Tomohiko, He, Jingsha
• Primary Examiner(s): Chin, Wellington
• Assistant Examiner(s): HO, CHUONG T
• Application Number:
  US09/281,359
• Time in Patent Office: 1,736 Days
• Field of Search: 370/400-408, 370/389, 370/392, 370/466, 370/468, 370/399, 370/237, 370/238, 370/238.1, 370/395.5, 370/401, 370/235-236, 709/105, 709/244, 709/229, 709/228, 709/224, 709/225, 455/453
• US Class Current: 370/238
• 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 67/1038 Load balancing arrangements to avoid a single path through a load balancer

Abstract

A system and method of load balancing a network having a plurality of client systems and servers. The load balancing system and method distributes requests from the client systems to different servers on a wide area network by selecting the most optimal server for a specific client request. The client request for a server is received from a client system and the destination point of the client request is resolved in part by a load balancing server selector. The client request is sent to a load balancing server based on the resolved destination point and one server out of a subset of the plurality of servers is selected based on a predetermined criteria, such as network traffic congestion or server load, and a conduit to the client system and the selected server for transmission of the client request is provided.

Claims

1. A method of load balancing for a network, the network having a plurality of client systems and a plurality of servers, comprising:

receiving a client request for a server from a client system;

resolving the destination point of the client request;

sending the client request to a load balancing server based on the resolved destination point;

selecting one server out of a subset of the plurality of servers based on a predetermined criteria;

providing a conduit to the client system and the selected server for transmission of the client request;

receiving network measurements, the network measurements being a total amount of bytes of data received by each of the subset of the plurality of servers from the plurality of client systems and a total amount of bytes of data sent by each of the subset of the plurality of servers to the plurality of client systems over a predetermined time period and the network measurements received is the predetermined criteria; and

dynamically switching between one of two modes based on load experienced by the load balancing server, a first mode being macro-mode and a second mode being micro-mode; and

wherein the load balancing server, in micro-mode, monitors each request from the plurality of client systems and, in macro-mode, causes requests from the plurality of client systems to bypass the load balancing server and be sent directly from the plurality of client systems to the selected server.

2. The method of claim 1 wherein the predetermined criteria further comprises a load measurement of each server among the subset of the plurality of servers.

3. The method of claim 1 wherein the predetermined criteria further comprises a network measurement of network traffic experienced by each server among the subset of the plurality of servers.

4. The method of claim 1 further comprising forming groups of servers, each group of servers including a different subset of servers out of the plurality of servers.

5. The method of claim 1 further comprising determining the load on each server within the subset of the plurality of servers.

6. The method of claim 1 further comprising sending requests from the client system to the selected server.

7. The method of claim 1 wherein the request is a domain name resolution request, requesting to resolve a logical name to an IP address.

8. A load balancing system for a network, the network having a plurality of client systems and a plurality of servers, comprising:

at least one load balancing server adapted to be coupled with the network; and

at least one load balancing server selector coupled with the at least one load balancing server, the load balancing server selector adapted to receive requests from the client systems and to forward the requests to the at least one load balancing server, the at least one load balancing server selecting one server out of a subset of the plurality of servers based on a predetermined criteria and the requests from the client systems;

wherein the at least one load balancing server is further configured to operate in one of two modes, a first mode being macro-mode and a second mode being micro-mode and the at least one load balancing server in micro-mode monitors the requests from the client systems and a portion of the requests are session specific.

9. The load balancing system of claim 8 wherein the predetermined criteria is a load measurement of each server among the subset of the plurality of servers.

10. The load balancing system of claim 8 wherein the predetermined criteria is an amount of network traffic experienced by each server among the subset of the plurality of servers.

11. The load balancing system of claim 8 wherein the requests from the client systems are domain name resolution requests, requesting to resolve a logical name to an IP address.

12. The load balancing system of claim 8 further comprising a plurality of client domain servers, and the load balancing server comprises a processor configured with a look up table of a series of IP addresses for the plurality of client domain name servers having a one to one correspondence with at least one IP address for the at least one load balancing server.

13. The load balancing system of claim 8 wherein the load balancing server selector comprises a processor configured with a look up table of a series of IP addresses for the plurality of servers having a one to one correspondence with a series of IP addresses for a plurality of load balancing servers.

14. The load balancing system of claim 8 wherein the load balancing server selector comprises a processor configured with a look up table of a series of IP addresses allocated to the at least one load balancing server having a one to one correspondence with a series of IP addresses for a plurality of servers.

15. The load balancing system of claim 8 further comprising a client domain name server, such that the requests from the client systems are sent to the client domain name server to resolve a logical name to an IP address.

16. The load balancing system of claim 15 further comprising a server domain name server, such that the requests from the client domain name server are sent to a server domain name server based upon a predetermined condition to resolve a logical name with an IP address.

17. The load balancing system of claim 16 wherein one of the client systems is a member of a domain;

and

the predetermined condition is a determination of whether the request from the client system is to a server that is not a member of the domain or not.

18. The load balancing system of claim 16 wherein the predetermined condition is a determination whether the client domain name server is able to resolve the requests from the client systems.

19. A load balancing server for a network, the network having a plurality of client systems and a plurality of servers, comprising:

a processor configured to select one server out of a subset of the plurality of servers based on a predetermined criteria and requests from the client systems;

a network interface adapted to be coupled with the network and the processor;

wherein the network interface receives network measurements, the network measurements being a total amount of bytes of data received by each of the subset of the plurality of servers from the plurality of client systems and a total amount of bytes of data sent by each of the subset of the plurality of servers to the plurality of client systems over a predetermined time period and the network measurements received is the predetermined criteria used by the processor; and

wherein the processor is further configured to operate in one of two modes, a first mode being macro-mode and a second mode being micro-mode and the processor in micro-mode monitors the requests from the plurality of client systems and a portion of the requests are session specific.

20. The load balancing server of claim 19 further comprising a load balancing server selector coupled with the network, and the network interface receives requests from the load balancing server selector.

21. The load balancing server of claim 19 wherein the processor is configured to send an IP address to the client systems representing one of the plurality of servers.

22. The load balancing server of claim 21 wherein the processor is configured to monitor a total number of requests from the plurality of client systems received by the load balancing server over a predetermined time period, such that the processor switches configuration to send a different IP address to one of the plurality of the client systems, and the different IP address represents a path from the client system to the load balancing server to one of the plurality of servers.

23. The load balancing server of claim 19 wherein the processor is configured to send an IP address to one of the plurality of the client systems, and the IP address represents a path from the client system to the load balancing server to one of the plurality of servers.

24. The load balancing server of claim 23 wherein the processor is configured to monitor a total number of requests from the plurality of client systems received by the load balancing server over a predetermined time period, and the processor switches configuration to send a different IP address to the client systems representing one of the plurality of servers.

25. A load balancing system for a network, the network having client systems and servers, the system comprising:

at least one load balancing server adapted to be coupled with the network;

at least one load balancing server selector coupled with the at least one load balancing server, the load balancing server selector adapted to receive requests from the client systems and to forward the requests to the at least one load balancing server, the at least one load balancing server selecting one server out of a subset of the servers based on a predetermined criteria and the requests from the client systems;

wherein the at least one load balancing server is configured to dynamically switch between one of two modes based on load experienced by the at least one load balancing server, a first mode being macro-mode and a second mode being micro-mode; and

wherein the at least one load balancing server, in micro-mode, monitors each request from the client systems and, in macro-mode, causes requests from the client systems to bypass the load balancing server and be sent directly from the client systems to the selected server.

26. The system of claim 25 wherein the load experienced comprises network measurements, the network measurements being a total amount of bytes of data received by each of the subset of the servers from the client systems and a total amount of bytes of data sent by each of the subset of the servers to the client systems over a predetermined time period.

27. The system of claim 26 wherein the load experienced comprises computation processing based on a total number of requests from the client systems received by the load balancing server over a predetermined time period to be processed by the load balancing server.

28. The system of claim 27 wherein the load balancing server, in macro mode, is configured to send an IP address that represents one of the servers to one of the client systems.

29. The system of claim 25 wherein the load balancing server is configured to switch from micro-mode to macro-mode when the load experienced by the load balancing server exceeds a predetermined limit.

30. The system of claim 29 wherein the predetermined limit is based on network measurements received by the load balancing server, the network measurements being a total amount of bytes of data received by each of the subset of the servers from the client systems and a total amount of bytes of data sent by each of the subset of the servers to the client systems over a predetermined time period.

31. The system of claim 30 wherein the predetermined limit is based on a total number of requests from the client systems received by the load balancing server over a predetermined time period to be processed by the load balancing server.

32. The system of claim 25 wherein the load balancing server is configured to switch from macro-mode to micro-mode when the load experienced by the load balancing server is less then or corresponds to a predetermined limit.

33. The system of claim 32 wherein the predetermined limit is based on an amount of network measurements received by the load balancing server, the network measurements being a total amount of bytes of data received by each of the subset of the servers from the client systems and a total amount of bytes of data sent by each of the subset of the servers to the client systems over a predetermined time period.

34. The system of claim 33 wherein the predetermined limit is based on a total number of requests from the client systems received by the load balancing server over a predetermined time period to be processed by the load balancing server.