Load balance control method and apparatus in data-processing system

• US 20050081211A1
• Filed: 08/11/2004
• Published: 04/14/2005
• Est. Priority Date: 08/11/2003
• Status: Active Grant

• Current Assignee: Hitachi, Ltd.
• Original Assignee: Hitachi, Ltd.
• Inventors: Inoue, Akinori, Koga, Kentaro
• Granted Patent: US 7,779,416 B2
• Time in Patent Office: Days
• US Class Current: 718/105
• CPC Class Codes: G06F 9/505 considering the load
• CPC Class Codes: G06F 9/5055 considering software capabilities, i.e. software resources associated or available to the machine

Abstract

A load balancing method for processing a plurality of job requests from a plurality of clients includes receiving a first request from one of the clients at a load control module. The first request relates to performing a first function by one or more of a plurality of application servers that are coupled to the plurality of clients. The first request received is associated to one of a plurality of job categories that each of the servers is configured to perform, the one job category to which the first request is associated being a first job category. Whether or not a first condition associated with the first job category is satisfied is determined. The first request is sent to a load balancer if the first condition associated with the first job category is determined to be satisfied, so that the load balancer can assign the first request to one of the application server in order to perform the requested first function, wherein the load balancer is coupled to the servers and configured to balance loads of the servers by distributing requests to the servers. At the load control module first information from the load balancer, the first information relating to the time expended in processing the first request is received. The load control module uses second information relating to the first information to determine whether to send a second request received from one of the clients to the load balancer or to a queue, the second request relating to the same job category as the first request.

Claims

1. A load balancing method for processing a plurality of job requests from a plurality of clients, the method comprising:

receiving a first request from one of the clients at a load control module, the first request relating to peforming a first function by one or more of a plurality of application servers that are coupled to the plurality of clients;

associating the first request received to one of a plurality of job categories that each of the servers is configured to perform, the one job category to which the first request is associated being a first job category;

determining whether or not a first condition associated with the first job category is satisfied;

sending the first request to a load balancer if the first condition associated with the first job category is determined to be satisfied, so that the load balancer can assign the first request to one of the application server in order to perform the requested first function, wherein the load balancer is coupled to the servers and configured to balance loads of the servers by distributing requests to the servers; and

receiving at the load control module first information from the load balancer, the first information relating to the time expended in processing the first request, wherein the load control module uses second information relating to the first information to determine whether to send a second request received from one of the clients to the load balancer or to a queue, the second request relating to the same job category as the first request.

2. The load balancing method of claim 1, wherein the first information relates to the time expended in executing the first request and the second information relates to an average time expended in executing requests relating to the first job category, the second information being derived using the first information.

3. The method of claim 1, wherein the first request is associated to the first job category based on an address of the first function indicated on the first request.

4. The method of claim 3, wherein the address of the first function is a Universal Resource Locator (URL).

5. The method of claim 1, wherein the first category includes first and second job types, the first job type relating to a batch process, the second job type relating to an on-line process.

6. The method of claim 1, wherein the first condition is determined to be not satisfied if an average time for processing requests relating to the first job category is determined to be longer than a predefined time period.

7. The method of claim 1, further comprising:

determining whether or not a second condition associated with the first job category is satisfied, wherein the first request is sent to a load balancer if both of the first and second conditions associated with the first job category are determined to be satisfied.

8. The method of claim 7, wherein the load control module updates a request information table upon receiving the first request, the request information table includes information on a number of requests being processed that are related to the first job category and an average time needed to complete the execution of requests relating to the first job type of the first job category.

9. The method of claim 8, wherein the load control module accesses a restriction information table to determine whether or not the first and second conditions associated with the first request are satisfied, the restriction information table includes a maximum number of requests that may be executed concurrently for requests relating to the first job category and a maximum delay time allowed for executing a request relating to the first category,

wherein the first condition is satisfied if the maximum number is not exceeded and the second condition is satisfied if the maximum delay time is not exceeded.

10. The method of claim 1, wherein the load control module is a load control apparatus that is remotely located from the load balancer.

11. The method of claim 1, wherein the load control module and the load balancer are provided within the same housing.

12. The method of claim 1, wherein the second request relates to performing a second function by one or more of the plurality of application servers, wherein the first function requested by the first request relates to a batch process and the second function requested by the second request relates to an on-line process.

13. A load control apparatus for processing a plurality of job requests from a plurality of clients, the load conrol apparatus comprising:

means for receiving a first request from one of the clients at a load control apparatus, the first request relating to peforming a first function by one or more of a plurality of application servers that are coupled to the plurality of clients;

means for associating the first request received to one of a plurality of job categories that each of the servers is configured to perform, the one job category to which the first request is associated being a first job category;

means for determining whether or not a first condition associated with the first job category is satisfied;

means for sending the first request to a load balancer if the first condition associated with the first job category is determined to be satisfied, so that the load balancer can assign the first request to one of the application server in order to perform the requested first function, wherein the load balancer is coupled to the servers and configured to balance loads of the servers by distributing requests to the servers; and

means for receiving at the load control module first information from the load balancer, the first information relating to the time expended in processing the first request, wherein the load control module uses second information relating to the first information to determine whether to send a second request received at the load control module to the load balancer or to a queue, the second request relating to the same job category as the first request.

14. A load controling device provided in a data processing system, the data processing system including a plurality of clients and a pluarlity of servers, the load controling device comprising:

a request registration component to receive a first request from one of the clients, the first request relating to peforming a first function by one or more of a plurality of application servers that are coupled to the plurality of clients, the request registration part being configured to associate the first request received to one of a plurality of job categories that each of the servers is configured to perform, the one job category to which the first request is associated being a first job category;

a load restriction component to determine whether or not a first condition associated with the first job category is satisfied, the load restriction component being configured to send the first request to a load balancer if the first condition associated with the first job category is determined to be satisfied, so that the load balancer can assign the first request to one of the application server in order to perform the requested first function, wherein the load balancer is coupled to the servers and configured to balance loads of the servers by distributing requests to the servers; and

a process result registration component to receive first information from the load balancer, the first information relating to the time expended in executing the first request,

wherein the load control device uses second information relating to the first information to determine whether to send a second request received from one of the clients to the load balancer or to a queue, the second request relating to the same job category as the first request.

15. The load controling device of claim 14, wherein the load balancer couples the load control device to the plurality of servers.

16. The load controlling device of claim 14, wherein the load controling device and th load balancer are integrated as a single network device.

17. The load controling device of claim 14, wherein the first information relates to the time expended in executing the first request and the second information relates to an average time expended in executing requests relating to the first job category, the second information being derived using the first information.

18. The load controling device of claim 14, wherein the first request is associated to the first job category based on an address of the first function indicated on the first request.

19. The load controling device of claim 18, wherein the address of the first function is a Universal Resource Locator (URL).

20. The load controling device of claim 14, wherein the first category includes first and second job types, the first job type relating to a batch process, the second job type relating to an on-line process.