System and method for clustering servers for performance and load balancing

• US 6,965,938 B1
• Filed: 09/07/2000
• Issued: 11/15/2005
• Est. Priority Date: 09/07/2000
• Status: Active Grant

• Current Assignee: Kyndryl Incorporated, ServiceNow Incorporated
• Original Assignee: International Business Machines Corporation
• Inventors: D'Esposito, John J., Beasley, Robert C., Neuberger, Wendy D., Cirulli, Susan B.
• Primary Examiner(s): Thompson, Marc D.
• Application Number:
  US09/657,215
• Time in Patent Office: 1,895 Days
• Field of Search: 709/105, 709/227, 709/202, 709/250, 709/201, 709/205, 709/248, 709/229, 235/379, 707/100, 707/104, 707/102, 707/2, 707/4, 718/101, 718/105, 718/106
• US Class Current: 709/229
• 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 69/329 in the application layer [OSI layer 7]

Abstract

A hybrid Notes/DB2 environment provides a requisition catalog on the Web. Client browsers are connected to a GWA infrastructure including a first network dispatcher and a virtual cluster of Domino.Go servers. The network dispatcher sprays out browser requests among configured .nsf servers in virtual server cluster. Communications from this virtual server cluster are, in turn, dispatched by a second network dispatcher servers in a Domino cluster. External objects, primarily for a GUI, are served in a .dfs and include graphic files, Java files, HTML images and net.data macros. The catalog is built from supplier provided flat files. A front end is provided for business logic and validation, as also is a relation database backend. HTML forms are populated using relational database agents. A role table is used for controlling access both to Notes code and DB2 data. Large amounts of data are quickly transferred using an intermediate agent and window.

Claims

1. A scalable system for providing a web processing tool, comprising:

a browser;

a first server cluster including a plurality of first servers, each said first server running identical first system and first application code for routing client requests among a plurality of enterprise applications;

a storage and execution unit, including a file server, a second server cluster, and an application server, for storing and executing an enterprise application comprising a plurality of web application code tables and data;

said file server including configuration file proxy statements for mapping user requests directed to said web application to said second server cluster;

said second server cluster including a plurality of second servers, each said second server running identical second system and identical selected components of said web application code, said second servers within said second server cluster periodically replicating with each other by moving identical data and code tables into identical data structures in each said second server so as to maintain data and code table consistency between them;

a relational database for storing data tables including a user profile specifying user roles;

said web application code tables including a first web application code table for authorizing access to other web application code tables based upon said user roles;

a database server for interfacing said relational database to said second server cluster and said application server;

a first network dispatcher for dynamically balancing client workload by redirecting client requests to one of said first servers based on current workload of first servers within said first server cluster;

a second network dispatcher responsive said first clustered servers for dynamically balancing client workload by redirecting clients to one of said second clustered servers based on current workload of servers within said plurality of second clustered servers;

said application server asynchronously responsive to said second servers for running agents to process application data requests and bridge said application data with respect to said database server and other back end servers.

2. The scalable system of claim 1, said first server being operable for presenting a graphical user interface to the said browser and for caching data on behalf of an end user.

3. The scalable system of claim 2, further comprising an external objects dynamic file for storing external objects in one place for dynamic access by said first servers, and for generating said gui.

4. The scalable system of claim 1, each said first server being a domino.go server operable for presenting a graphical user interface to said browser and redirecting said client via said second network dispatcher to a second server.

5. The scalable system of claim 4, further comprising a configuration file of proxy statements for mapping user requests to said second server cluster.

6. The scalable system of claim 1, said web processing tool being a web requisition catalog application.

7. The scalable system of claim 1, said second servers being operable for performing workflow, providing security, and serving as a document repository.

8. The scalable system of claim 7, said second servers being domino network servers.

9. The scalable system of claim 8, said document repository being requisitions stored in domino nsf files.

10. The scalable system of claim 1, said database server being a relational database server.

11. The scalable system of claim 1, said other back end server comprising an enterprise resource planning system, including an accounting application having an accounts payable function.

12. A method for generating on-line procurement requisitions, comprising the steps of:

receiving a client request;

dynamically balancing client workload among a plurality of first logically defined servers in a first server cluster by directing said request to one of said first servers based on current server workload, each said first server running identical first system and first application code for distributing said client request and a required function among a plurality of enterprise applications including a database access function;

distributing said database access function an application server and in a plurality of code tables and data tables on a plurality of second logically defined servers in a second server cluster;

storing data tables including a user profile specifying user roles in a relational database;

responsive to a client request for said database access function, dynamically balancing client workload among said plurality of second servers in said second server cluster based on current server workload;

operating one of said code tables to authorize access to other of said code tables based on said user profile;

periodically replicating said second servers with each other by moving identical data and code tables into identical data structures in each said second server so as to maintain data and code table consistency between them; and

operating said second server to direct said client request to an application server where all data is replicated and where bridges and agents execute with respect to data in said database.

13. The method of claim 12, further comprising the steps of:

replicating application data to a back-end relational database server; and

replicating application data to a back-end enterprise resource planning system including an accounting application having an accounts payable function.

14. A program storage device readable by a machine, tangibly embodying a program of instructions executable by a machine to perform method steps for processing a client request with respect to a database, said method steps comprising:

receiving a client request;

dynamically balancing server workload by directing said request to a first server within a first cluster of logically defined virtual servers based on current server usage, each server within said first cluster executing first code for determining the mapping of said client request and the function required among a plurality of enterprise applications including a database access function;

responsive to a client request for said database access function, dynamically balancing server workload by directing said client request to a second server within a second cluster of logically defined virtual servers based on current server usage, each server within said second cluster executing identical second application code including a plurality of web application workflow processes;

operating a first of said workflow processes for authorizing access to and execution of other of said workflow processes based on client user role;

periodically replicating said second servers within said second cluster of virtual servers by moving identical data and code tables into identical data structures in each said second server so as to maintain data and application code consistency between them; and

operating said second server to direct said client request to an application server where all data is replicated and where bridges and agents execute with respect to data in said database.

15. The program storage device of claim 14, said method further comprising the step of operating said first servers for presenting a graphical user interface to the said browser and for caching data on behalf of an end user.

16. The program storage device of claim 14, said method further comprising:

synchronizing all virtual servers within said first server cluster; and

synchronizing all virtual servers within said second server cluster.

17. The program storage device of claim 16, said method further comprising:

replicating application data to a back-end relational database server; and

replicating application data to a back-end enterprise resource planning system including an accounting application having an accounts payable function.

18. The program storage device of claim 14, said method further comprising storing external objects in an external objects dynamic file for dynamic access by said first cluster of servers, and for generating said gui.

19. A computer program product for generating on-line procurement requisitions, said computer program product comprising:

a computer readable medium;

first program instructions for receiving a client request;

second program instructions for dynamically balancing server workload by directing said request to a first server within a first cluster of logically defined virtual servers based on current server usage, each server within said first cluster of servers executing identical first code;

third program instructions for operating said first server to determine the mapping of said client request and the function required;

fourth program instructions, responsive to a database access function, for dynamically balancing server workload by directing said client request to a second server within a second cluster of logically defined second virtual servers based on current server usage, each said second server within said second cluster of logically defined second virtual servers executing identical second code tables, said second servers within said second cluster of second virtual servers periodically replicating with each other by moving identical data and code tables into identical data structures in each said second server so as to maintain data consistency between them; and

fifth program instructions for operating a code table in said second server to authorize said client request to access and execute selected other code tables based on client user role, and for operating said other code tables to direct said client request to an application server where all data is replicated and where bridges and agents execute with respect to data in said database; and

wherein

said first, second, third, fourth, and fifth program instructions are recorded on said medium.