System and method for load balancing

• US 6,389,448 B1
• Filed: 05/05/2000
• Issued: 05/14/2002
• Est. Priority Date: 12/06/1999
• Status: Expired due to Fees

• Current Assignee: Warp Solutions Inc. (Halo Technology Holdings, Inc.)
• Original Assignee: Warp Solutions Inc. (Halo Technology Holdings, Inc.)
• Inventors: Volovich, Gene R., Gnip, John, Primak, Leonard
• Primary Examiner(s): VU, VIET DUY
• Application Number:
  US09/565,259
• Time in Patent Office: 739 Days
• Field of Search: 709/104, 709/105, 709/203, 709/217, 709/219, 709/223, 709/225, 709/226, 709/227, 709/229, 709/232, 709/238
• US Class Current: 718/105
• CPC Class Codes: G06F 9/505 considering the load
• 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 67/1029 using data related to the state of servers by a load balancer

Abstract

A system for distributing load between multiple servers where more than one server in a sever cluster receives a request for connection from a client and each server makes a determination of whether or not to respond to the request. Software modules running on the servers monitor and communicate relative abilities of each server to respond to client requests. Each server responding to a percentage of client requests corresponding to its relative ability to respond.

Claims

1. A system for distributing load within a network, comprising:

at least one cluster having a plurality of servers connected to each other, each server in said cluster being addressable by a common network address; and

a load balancing module in each server for calculating a connection value for each connection request addressed to said network address and for accepting a connection request as a function of available capacity of said respective server with respect to overall available capacity of said cluster and the connection value associated with said connection request.

2. The system of claim 1, wherein said connection value being within a predefined range of values.

3. The system of claim 2, wherein said module in said each server transmits the available capacity of said each server to other servers in said cluster and determines the overall available capacity from the available capacity of said plurality of servers in said cluster.

4. The system of claim 3, wherein said module in said each server determines a sub-range within said predefined range of values as a function of the available capacity of said each server with respect to the overall available capacity, each sub-range not overlapping with another sub-range.

5. The system of claim 4, wherein said connection value being a pseudo-random number;

and wherein said module accepts only connection request having a connection value within the sub-range of said associated server.

6. The system of claim 2, further comprising a coordinating device;

wherein said module in said each server transmits the available capacity of said each server to said coordinating device which determines the overall available capacity from the available capacity of said plurality of servers in said cluster.

7. The system of claim 6, wherein said coordinating device determines for said each server a sub-range within said predefined range of values as a function of the available capacity of said each server with respect to the overall available capacity, each sub-range not overlapping with another sub-range.

8. The system of claim 7, wherein said connection value being a pseudo-random number;

and wherein said module accepts only connection request having a connection value within the sub-range of said associated server.

9. The system of claim 1, wherein said network is either a TCP/IP network or a system network architecture (SNA) network.

10. The system of claim 1, wherein said available capacity of said each server is function of at least one of the following:

server'"'"'s processing or CPU capacity, CPU load, number of concurrent processes or tasks being performed, and number of existing connections.

11. A system for distributing load within a network, comprising:

at least one cluster having a plurality of servers connected to each other, each server in said cluster being addressable by a common network address;

a plurality of routers for routing connection requests to said plurality of servers, each server in said cluster connected to said at least one router; and

a load balancing module in each server for calculating a connection value for each connection request addressed to said network address and for accepting a connection request from a router as a function of available capacity of said respective server with respect to overall available capacity of servers connected to said router and the connection value associated with said connection request.

12. The system of claim 11, wherein said connection value being within a predefined range of values.

13. The system of claim 12, wherein said module in each server connected to said router transmits the available capacity of said each server to other servers connected said router and determines the overall available capacity from the available capacity of the servers connected to said router.

14. The system of claim 13, wherein said module in said each server connected to said router determines a sub-range within said predefined range of values as a function of the available capacity of said each server with respect to the overall available capacity, each subrange not overlapping with another sub-range.

15. The system of claim 14, wherein said connection value being a pseudo-random number;

and wherein said module accepts only connection request having a connection value within the sub-range of said associated server.

16. A method for distributing load within a network comprising at least one cluster having a plurality of servers connected to each other, each server in said cluster being addressable by a common network address, comprising the steps of:

calculating a connection value for each connection request addressed to said network address by said each server; and

accepting a connection request by a server as a function of the connection value associated with said connection request and available capacity of said server with respect to overall available capacity of said cluster.

17. The method of claim 16, wherein said connection value being within a predefined range of values.

18. The method of claim 17, further comprising the steps of:

transmitting the available capacity of said server to other servers in said cluster; and

determining by each server the overall available capacity from the available capacity of said plurality of servers in said cluster.

19. The method of claim 18, further comprising the step of determining a sub-range within said predefined range of values for said each server as a function of the available capacity of said each server with respect to the overall available capacity, each sub-range not overlapping with another sub-range.

20. The method of claim 19, wherein said connection value being a pseudo-random number;

and wherein said module accepts only connection request having a connection value within the sub-range of said associated server.