Grid service scheduling of related services using heuristics

• US 20040167980A1
• Filed: 02/20/2003
• Published: 08/26/2004
• Est. Priority Date: 02/20/2003
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Doyle, Ronald P., Kaminsky, David Louis
• Granted Patent: US 7,171,470 B2
• Time in Patent Office: Days
• US Class Current: 709/225
• CPC Class Codes: G06F 2209/5019 Workload prediction
• CPC Class Codes: G06F 2209/508 Monitor
• CPC Class Codes: G06F 9/5072 Grid computing
• CPC Class Codes: G06F 9/5083 Techniques for rebalancing the load in a distributed system

Abstract

A method and system for creating service instances in a computing grid. The method can include scheduling a service in the computing grid to process at least a portion of a requested transaction. At least one additional service related to the scheduled service can be identified, and a load condition can be assessed in the at least one additional service related to the scheduled service. A new instance of the at least one additional service can be created if the load condition exceeds a threshold load. In this way, an enhanced capacity for processing transactions can be established in the related services in advance of a predicted increase in load in the grid.

Claims

1. A method for creating service instances in a computing grid, the method comprising the steps of:

scheduling a service in the computing grid to process at least a portion of a requested transaction;

identifying at least one additional service related to said scheduled service;

assessing a load condition in said at least one additional service related to said scheduled service; and

,

creating a new instance of said at least one additional service if said load condition exceeds a threshold load.

2. The method of claim. 1, wherein said scheduling step comprises the steps of:

locating a service instance in the computing grid which has been created from a service configured to access resources required to satisfy at least a portion of said requested transaction;

assessing a load condition in said located service instance; and

,

creating a new instance of said configured service if said load condition in said located service instance exceeds a threshold load, but assigning said portion of said requested transaction to said located service instance if said load condition does not exceed a threshold load.

3. The method of claim 1, wherein said identifying step comprises the steps of:

monitoring individual resources accessed by different types of transactions processed by different services in the computing grid;

querying said different services to determine underlying resource capabilities of said different services; and

,

identifying selected ones of said different services as said related services where individual ones of said resources accessed by said selected ones of said different services, and said underlying resource capabilities are pre-determined by way of transaction type to relate to transactions processed in said scheduled service.

4. The method of claim 1, wherein said creating step comprises the step of creating a new instance of said at least one additional service if said load condition exceeds a threshold load specified according to the terms of a service level agreement (SLA).

5. The method of claim 2, wherein said step of creating a new instance of said configured service comprises the step of creating a new instance of said configured service if said load condition in said located service instance exceeds a threshold load specified according to the terms of a service level agreement (SLA), but assigning said portion of said requested transaction to said located service instance if said load condition does not exceed said threshold load.

6. A predictive grid service instance creation system comprising:

a service creation factory disposed in a grid services host and configured to instantiate grid services on demand in a corresponding computing grid;

a monitor programmed to monitor transactions scheduled in said computing grid to identify related grid services, and to assess contemporaneous load conditions in individual instantiated ones of grid services in said computing grid; and

,

a scheduler coupled both to said service creation factory and to said monitor, said scheduler (1) scheduling received transaction requests to be processed by selected instantiated ones of said grid services in the computing grid where load conditions in said instantiated ones permit, (2) requesting that said service creation factory create new instances of said grid services where load conditions in said instantiated ones require, and (3) requesting that said service creation factory create new instances of related grid services identified by said monitor where load conditions in said related grid services require.

7. The system of claim 6, further comprising at least one service level agreement (SLA) having at least one of availability and responsiveness terms based upon which it can be determined when load conditions permit scheduling of a specified grid service, and when load conditions require creating a new instance of a specified grid service.

8. The system of claim 6, further comprising a data store of transaction metrics collected by said monitor for use in identifying said related grid services.

9. The system of claim 8, wherein said transaction metrics comprise resources used by types of transactions processed by said grid services, and resource capabilities specified by said grid services in accessing said resources.

10. A machine readable storage having stored thereon a computer program for creating service instances in a computing grid, the computer program comprising a routine set of instructions for causing the machine to perform the steps of:

scheduling a service in the computing grid to process at least a portion of a requested transaction;

identifying at least one additional service related to said scheduled service;

assessing a load condition in said at least one additional service related to said scheduled service; and

,

creating a new instance of said at least one additional service if said load condition exceeds a threshold load.

11. The machine readable storage of claim 10, wherein said scheduling step comprises the steps of:

locating a service instance in the computing grid which has been created from a service configured to access resources required to satisfy at least a portion of said requested transaction;

assessing a load condition in said located service instance; and

,

creating a new instance of said configured service if said load condition in said located service instance exceeds a threshold load, but assigning said portion of said requested transaction to said located service instance if said load condition does not exceed a threshold load.

12. The machine readable storage of claim 10, wherein said identifying step comprises the steps of:

monitoring individual resources accessed by different types of transactions processed by different services in the computing grid;

querying said different services to determine underlying resource capabilities of said different services; and

,

identifying selected ones of said different services as said related services where individual ones of said resources accessed by said selected ones of said different services, and said underlying resource capabilities are pre-determined by way of transaction type to relate to transactions processed in said scheduled service.

13. The machine readable storage of claim 10, wherein said creating step comprises the step of creating a new instance of said at least one additional service if said load condition exceeds a threshold load specified according to the terms of a service level agreement (SLA).

14. The machine readable storage of claim 11, wherein said step of creating a new instance of said configured service comprises the step of creating a new instance of said configured service if said load condition in said located service instance exceeds a threshold load specified according to the terms of a service level agreement (SLA), but assigning said portion of said requested transaction to said located service instance if said load condition does not exceed said threshold load.