Distributed adaptive computing

• US 6,480,861 B1
• Filed: 02/26/1999
• Issued: 11/12/2002
• Est. Priority Date: 02/26/1999
• Status: Expired due to Term

• Current Assignee: Bank of America Corp.
• Original Assignee: Merrill Lynch & Co. Incorporated (Bank of America Corp.)
• Inventors: Tyler, Daniel, Kanevsky, Paul, Pizi, Anthony C., Tsao, Thomas
• Primary Examiner(s): ROBINSON, GRETA LEE
• Application Number:
  US09/258,711
• Time in Patent Office: 1,355 Days
• Field of Search: 707/10, 707/104, 707/103, 705/70, 705/79, 705/80, 705/10, 705/34, 705/35, 705/400, 709/229, 709/203, 709/217, 709/202
• US Class Current: 1/1
• CPC Class Codes: G06F 2209/5015 Service provider selection
• CPC Class Codes: G06F 9/50 Allocation of resources, e.g. of the central processing unit [CPU]
• CPC Class Codes: G06Q 30/0283 Price estimation or determination
• CPC Class Codes: G06Q 50/188 Electronic negotiation
• CPC Class Codes: Y10S 707/99944 Object-oriented database structure
• CPC Class Codes: Y10S 707/99945 Object-oriented database structure processing

Abstract

A system and a method for managing, organizing, and allocating service providers in the operational environment of a distributed computer network by applying trade and price mechanisms to a plurality of resource allocation decisions. Local resource allocation rules are set forth for maintaining a near-optimal, global load distribution. The service providers are dynamically allocated based upon the supply of the providers and the demand thereof. An automated mechanism, based on service provider reputation, channels demand away from failing or broken service providers. Strategic load balancing rules cause the elimination of ineffective service providers, and also provide a dynamic replication of service providers that cannot handle the current demand. Further, a method for managing the overall system behavior utilizes administrative surcharges.

Claims

1. A resource allocation system comprising:

(a) a plurality of service providers,

(b) a plurality of service requestors, and

(c) a service broker,

wherein the service providers, service requestors, and service broker are all accessible from a computer network, the service broker using a service provider allocation directory to allocate service providers to service requestors based upon dynamically-changing pricing constraints,

wherein the service provider allocation directory associates each of a plurality of service providers with a first indicator specifying a type or class of service provider, a second indicator specifying the location of the service provider on the computer network, and a third indicator specifying a base price for the service.

2. The system of claim 1 wherein the service provider allocation directory also associates each of a plurality of service providers with at least one of:

a fourth indicator specifying attribute pricing, a fifth indicator specifying a load premium, a sixth indicator specifying a reputation premium, and a seventh indicator specifying an administrative premium.

3. The system of claim 1 wherein the prices are updated according to changes in resource usage conditions.

4. The system of claim 1 wherein a relative business priority of a request for a given information resource is expressed as a budget.

5. In a computer network, a method including the steps of:

(a) allocating a plurality of service providers among a plurality of service requestors according to the relative priorities of processes which request the use of a respective service provider, and

(b) associating each process with information setting forth the manner in which a service requestor will use that service if it is allocated to that service requestor;

wherein the plurality of service providers include any of a database, a computer program, a person providing services over a computer network, an information resource, or a hardware resource such as a fax machine, a printer, or a data storage drive;

wherein the service requestors include any of computer programs and devices accessible from the computer network for use by one or more individuals requesting one or more service providers;

wherein the service request includes the type or class of service desired, as well as a budget specifying the maximum price that the service requestor will pay for that service;

further including the step of service providers sending availability messages to the service broker indicating the availability of one or more services, each availability message identifying the type or class of service, location of the service and associating each service with a base price and attribute prices.

6. The method of claim 5 further including the step of the service broker storing the availability messages in a service provider allocation directory.

7. The method of claim 5 wherein, in response to the receipt of a service request, the service broker uses a first indicator in the allocation directory to generate a trial candidate list of service providers, the trial candidate list only including service providers of the type or class desired by the service requestor.

8. The method of claim 7 further including the step of the service broker calculating a levied price for each of a plurality of service providers on the trial candidate list.

9. The method of claim 8 wherein the levied price is the base price.

10. The method of claim 8 wherein the levied price is a summation of the base price and at least one of:

an attribute price, a load premium, a reputation premium, an administrative premium, and a delivery premium.

11. The method of claim 10 further including the step of the service broker eliminating any service providers on the trial candidate list that do not have a levied price below the budget specified in the service request, thus providing a final candidate list.

12. The method of claim 11 further including the step of the service broker generating a draft contract for one or more service providers on the final candidate list, the draft contract specifying at least one of:

an identity for the service provider, a location of this service provider on a computer network, an identity for a service requestor, a location of the service requestor on the computer network, the levied price, the base price, the attribute price, the load premium, the reputation premium, the administrative premium, the delivery premium, a type or class of service, a schedule for performance of the service, and a contract expiration date and time.

13. The method of claim 12 wherein, in response to the receipt of one or more draft contracts, the service requestor redeems any of these draft contract through the use of an execution manager software component accessible from the computer network.

14. The method of claim 13 wherein the execution manager schedules performance of services specified in draft contracts to be executed.

15. The method of claim 14 wherein the execution manager causes the services provider specified in the redeemed contract to perform the service as requested by the service requestor.

16. The method of claim 15 wherein the execution manager causes the service requestor to pay the levied price specified in the redeemed contract by instructing the accounting manager to debit the service requestor account.

17. The method of claim 15 wherein the execution manager causes the service provider account to be credited for completed service by instructing the accounting manager to credit the service provider account in the amount of the sum of the base price, the attribute price and the load premium.

18. The method of claim 17 wherein an accounting manager software component, coupled to the computer network, records all credits and debits of service requestor and service provider accounts.

19. The method of claim 8 wherein at least one of:

a base price component, an attribute price component, a load premium component, a reputation premium component, an administrative premium component, and a delivery premium component of the levied price of each service provider are dynamically adjusted over time in accordance with a series of adjustment rules.

20. The method of claim 19 wherein a first adjustment rule specifies that, as the demand for a particular service provider increases, its load premium also increases.

21. The method of claim 20 wherein a second adjustment rule specifies that the load premium of a service provider that is idle for too long will be decreased until sufficient demand for the resource is generated.

22. The method of claim 21 wherein a third adjustment rule specifies that a given service provider generates sufficient demand or be removed from the allocation directory.

23. The method of claim 22 wherein a fourth adjustment rule specifies that, as a service provider'"'"'s costs increase, its base price also increases.

24. The method of claim 23 wherein a fifth adjustment rule specifies that, as a service provider'"'"'s level of service or prioritization changes, its attribute prices also change.

25. The method of claim 24 wherein a sixth adjustment rule specifies that as a service provider'"'"'s performance changes the service provider'"'"'s reputation premium also changes.

26. The method of claim 25 wherein, as the service provider'"'"'s failure rate increases, its reputation premium also increases.

27. The method of claim 26 wherein a seventh adjustment rule specifies that a system administrator can influence the use of a service provider by applying an administrative premium to the service provider allocation directory.

28. The method of claim 27 wherein an eighth adjustment rule specifies that the service broker will determine the delivery premium at the time of creating the draft contract.

29. The method of claim 28 wherein a service broker will issue a service request for delivery based on a location of the service provider and a location of the service requestor.

30. The method of claim 19 further including the step of applying strategic load balancing rules to each service provider to dynamically replicate service providers that cannot handle all of the current demand from service requestors.

31. The method of claim 30 wherein the adjustment rules, as well as any rules used to determine levied prices, utilize information that is locally available on the computer network, thereby producing a globally adaptive, self-configurable behavior for a plurality of service providers.