Dynamic software composition in a component-based software system

• US 7,448,022 B1
• Filed: 02/10/2004
• Issued: 11/04/2008
• Est. Priority Date: 02/10/2004
• Status: Expired due to Fees

• Current Assignee: Prasad Ram
• Original Assignee: Prasad Ram
• Inventors: Ramanujapuram, Arun, Ram, Prasad, Schuler, Robert
• Primary Examiner(s): CHAVIS, JOHN Q
• Application Number:
  US10/775,383
• Time in Patent Office: 1,729 Days
• Field of Search: 717/120
• US Class Current: 717/120
• CPC Class Codes: G06F 8/00 Arrangements for software engineering testing or debugging G06F11/36; administrative, planning or organisation aspects of software project management G06Q10/06
• CPC Class Codes: G06F 9/45504 Abstract machines for programme code execution, e.g. Java virtual machine [JVM], interpreters, emulators

Abstract

A resource catalog system comprises a catalog organized as a set of metadata structures representing distributed system resources, language for resource invocation, and a query interpreter and optimizer. Query optimization includes dynamic functional composition by modeling resources according constraints and determining constraint-satisfaction to compose an execution plan. Query optimization may also include dynamic selection and scheduling of resources by combining a resource scheduling with adaptive feedback. The resources may comprise distributed software components, application programs, shell scripts, executables, and services. The resource catalog system accepts queries for operations on distributed resources through a declarative language or a procedural language. A query optimizer derives catalog operators from a query or elsewhere and accesses the system resources according to interpretation of the catalog operators. The optimizer also retrieves information concerning the location and/or physical interfaces of the system resources to fulfill the commands specified by the catalog operators.

Claims

1. A system for adaptive functional composition of software components, to fulfill requests for higher-order functions, comprising a processor for utilizing:

a catalog of system resources including tools, applications, web services, and data repositories each implementing one or more functions and constraint information relating to the system resources;

a constraint solver configured to select a subset of the system resources based upon the request and the constraint information in said catalog; and

a composer configured to optimally sequence the functions and cause execution of the sequence of functions based upon the selection of system resources by said constraint solver;

wherein said catalog comprises an abstract model of functions, their inter-relationships, and their constraints;

wherein said constraint solver is configured to determine one or more sequences of functions that satisfy all constraints of the selected system resources, and to generate a function graph based thereon; and

wherein said composer is configured to generate a procedural process script from the function graph generated by said constraint solver.

2. The system of claim 1, wherein said constraint information comprises:

(i) function hierarchies of abstract function comprised of primitive functions, and (ii) ordering relations between functions.

3. The system of claim 1, wherein said catalog comprises metadata relating to the functions included therein, and wherein said composer utilizes said metadata in generating said procedural process script.

4. The system of claim 1, wherein said procedural process script is provided to a process engine to cause execution of the selected sequence of functions.

5. The system of claim 1, wherein said constraint solver is provided with a set of decision variables that correspond to the system resources, and wherein it is configured to determine an assignment of values for the decision variables that satisfies all constraints of the selected system resources.

6. The system of claim 5, wherein said constraint solver employs constraint propagation to attempt to solve for the decision variables, and a search technique to find values for any unsolved decision variables.

7. The system of claim 6, wherein said constraint solver is configured to determine a set of valid decision variables that correspond to primitive functions eligible for inclusion within the subset of system resources, and to generate a path that traverses all of the eligible primitive functions to generate said function graph.

8. The system of claim 1, wherein said constraints are processed by said constraint solver according to specific domains.

9. The system of claim 1, wherein said constraint solver utilizes a constraint satisfaction parameter to select said subset of system resources, and wherein constraint satisfaction may be relaxed or restricted to allow alternative selections of system resources.