Management of software images for computing nodes of a distributed computing system

• US 7,516,206 B2
• Filed: 01/28/2005
• Issued: 04/07/2009
• Est. Priority Date: 01/28/2005
• Status: Active Grant

• Current Assignee: Computer Associates Think Inc. (Broadcom, Inc.)
• Original Assignee: Cassatt (Broadcom, Inc.)
• Inventors: Cahir, Margaret, Gilbertson, Jeanne C., Henseler, David A.
• Primary Examiner(s): Barqadle; Yasin M
• Application Number:
  US11/046,133
• Publication Number:
  US 20060173993A1
• Time in Patent Office: 1,530 Days
• Field of Search: 709/201, 709/203, 709216-224, 709246-248, 717168-172
• US Class Current: 709/223
• CPC Class Codes: G06F 8/63 Image based installation; Cloning; Build to order

Abstract

In general, the invention is directed to software image management within a distributed computing system. The distributed computing system includes a plurality of application nodes interconnected via a communications network, and an image repository storing a plurality of pre-generated software image instances. A control node may pre-assign a network address to each of the image instances stored within the image repository, and assigns the network addresses of the image instances to the application nodes upon subsequent deployment of the software images to the application nodes.

Claims

1. A method comprising:

storing a master software image for a type of computing function to be deployed within a distributed computing system;

generating K software image instances from the master software image prior to selecting any application nodes of the distributed computing system upon which to deploy any of the K software image instances, wherein the distributed computing system is configured to deploy a maximum of K of the software image instances for the type of computing function, and wherein the K software image instances are generated such that, when all or a subset of the K software image instances are deployed on a set of application nodes, the set of applications nodes combine to provide a computing application for the type of computing function;

when generating the K software image instances, pre-assigning a unique network address to each of the K software image instances prior to selecting any of the application nodes upon which to deploy any of the K software image instances;

after generating the K software image instances and after pre-assigning the unique network address to each of the K software image instances, selecting M application nodes upon which to deploy a first subset of the K software image instances, where M is less than K;

deploying the first subset of the K software image instances on the M application nodes within the distributed computing system;

after deploying the first subset of the K software image instances, detecting a need to add one or more additional application nodes to the distributed computing system for the type of computing function;

selecting a second subset of the K software image instances that are not currently deployed; and

deploying the second subset of the K software image instances on the one or more additional application nodes.

2. The method of claim 1, further comprising storing the image instances within a software image repository.

3. The method of claim 2, further comprising:

designating the image instances deployed on the application nodes as active; and

designating the image instances that have not been deployed as inactive.

4. The method of claim 1, further comprising:

when deploying the image instances, assigning the network addresses of the image instances to the application nodes.

5. The method of claim 4, wherein pre-assigning a network address to each of the image instances comprises assigning an Internet Protocol (IP) address and a host name to each of the image instances.

6. The method of claim 1, further comprising generating the master software image by temporarily installing an operating system and one or more software applications on a computing node and copying the installed software applications.

7. The method of claim 1, further comprising maintaining a data structure that stores data defining each of the image instances including:

an instance identification that identifies an instance number and a version of the master software image from which the image instance was created,

a network address associated with the image instance, and

a status indicating whether the image instance is currently deployed on one of the application nodes.

8. The method of claim 1, wherein detecting the need to add one or more additional application nodes comprises automatically detecting a high processing load on the M application nodes.

9. The method of claim 1, wherein detecting the need to add one or more additional application nodes comprises receiving input from a system administrator requesting deployment of the additional image instances.

18. The method of claim 1, further comprising:

storing a second master software image for a second type of computing function to be deployed within a distributed computing system, wherein the second type of computing function is different from the type of computing function used for the first master software image;

generating L software image instances from the second master software image prior to selecting any application nodes upon which to deploy any of the L software image instances, wherein the distributed computing system is configured to deploy a maximum of L of the software image instances for the second type of computing function;

selecting N application nodes upon which to deploy a subset of the L software image instances; and

deploying the subset of the L software image instances on the N application nodes within the distributed computing system, where N is less than L.

10. A distributed computing system comprising:

a plurality of application nodes interconnected via a communications network; and

an image repository storing a plurality of K pre-generated software image instances, wherein K represents a maximum number of the software image instances to be deployed within the distributed computing system for a type of computing function, wherein the K pre-generated software image instances are generated and are each assigned a unique network address prior to selecting any application nodes upon which to deploy any of the K pre-generated software image instances, and wherein the K pre-generated software image instances are generated such that, when all or a subset of the K pre-generated software image instances are deployed on a set of application nodes, the set of applications nodes combine to provide a computing application for the type of computing function;

a control node coupled to a communications network, wherein the control node selects M application nodes upon which to deploy a first subset of the K pre-generated software image instances, deploys the first subset of the K pre-generated software image instances on the M application nodes within the distributed computing network, where M is less than K, detects a need to add one or more additional application nodes to the distributed computing system for the type of computing function after deploying the first subset of the K software image instances, selects a second subset of the K software image instances that are not currently deployed, and deploys the second subset of the K software image instances on the one or more additional application nodes.

11. The distributed computing system of claim 10, wherein the control node assigns the network addresses of the image instances to the application nodes when deploying the image instances.

12. The distributed computing system of claim 10,

wherein the image repository stores a master software image for a type computing function to be deployed within a distributed computing system, and

wherein the control node generates the software image instances from the master software image.

13. The distributed computing system of claim 12, wherein each of the image instances includes:

(1) a copy of a first portion of the master image software that will be writable once deployed, and (2) a read-only reference to a remaining portion of the master image.

14. The distributed computing system of claim 10, further comprising:

wherein the control node presents a user interface to receiving a request to deploy an additional image instance.

15. The distributed computing system of claim 10, further comprising a database storing data that designates the image instances associated with the deployed application nodes as active, and designates the image instances that have not been associated with the deployed application as inactive.

16. The distributed computing system of claim 15, wherein the data includes:

an instance identification that identifies an instance number and a version of the master software image from which the image instance was created,

a network address associated with the image instance, and

a status indicating whether the image instance is currently deployed.

17. A computer-readable medium comprising instructions that cause a programmable processor to:

present a first user interface to receive input identifying a master software image for a type of computing function to be deployed within a distributed computing system;

present a second user interface to receive input identifying a maximum number of software image instances that can be deployed within the distributed computing system for the type of computing function;

generate a software image instance from the master software image for each of the image instances that can be deployed prior to selecting any application nodes upon which to deploy any of the software image instances, and wherein the software image instances are generated such that, when all or a subset of the software image instances are deployed on a set of application nodes, the set of applications nodes combine to provide a computing application for the type of computing function;

when generating the software image instances, pre-assign a unique network address to each of the software image instances prior to selecting any of the application nodes upon which to deploy any of the software image instances;

after generating the software image instances and after pre-assigning the unique network address to each of the software image instances, select M application nodes upon which to deploy a first subset of the software image instances, where M is less than the maximum number of software instances that can be deployed for the type of computing function;

deploy the first subset of the software image instances on the M application nodes within the distributed computing system;

after deploying the first subset of the software image instances, detect a need to add one or more additional application nodes to the distributed computing system for the type of computing function;

select a second subset of the software image instances that are not currently deployed; and

deploy the second subset of the software image instances on the one or more additional application nodes.