Distributed computing system having hierachical organization

• US 20060173895A1
• Filed: 01/31/2005
• Published: 08/03/2006
• Est. Priority Date: 01/31/2005
• Status: Active Grant

• Current Assignee: Computer Associates Think Inc. (Broadcom, Inc.)
• Original Assignee: Computer Associates Think Inc. (Broadcom, Inc.)
• Inventors: Engquist, James D., Jackson, Jerry R., Lindley, Craig A., Resch, David L., Nordby, Jonathan D., Church, Dann M., Vosburgh, Craig A., Berliner, Brian, Collins, Doreen E.
• Granted Patent: US 7,685,148 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 2209/505 Clust
• CPC Class Codes: G06F 9/5061 Partitioning or combining of resources

Abstract

A distributed computing system conforms to a multi-level, hierarchical organizational model. One or more control nodes provide for the efficient and automated allocation and management of computing functions and resources within the distributed computing system in accordance with the organization model. The model includes four distinct levels: fabric, domains, tiers and nodes that provide for the logical abstraction and containment of the physical components as well as system and service application software of the enterprise. A user, such as a system administrator, interacts with the control nodes to logically define the hierarchical organization of distributed computing system. The control nodes are responsible for all levels of management in accordance with the model, including fabric management, domain creation, tier creation and node allocation and deployment.

Claims

1. A method comprising:

receiving input that defines a model for a hierarchical organization of a distributed computing system having a plurality of computing nodes, wherein the model specifies a fabric having a one or more domains, and wherein each domain has at least one tier that includes at least one node slot; and

automatically configuring the distributed computing system in accordance with the hierarchical organization defined by the model.

2. The method of claim 1,

wherein the model defines a different computing function for each of the tiers, and

wherein automatically configuring the distributed computing system comprises:

mapping software images associated with the computing functions to the node slots in accordance with model;

assigning the computing nodes to the node slots of the tiers; and

deploying the software images on assigned computing nodes.

3. The method of claim 1, further comprising:

receiving input that defines specific node requirements of each of the tiers of the fabric; and

assigning the computing nodes to the node slots in accordance with the node requirements.

4. The method of claim 3, wherein the node requirements include at least one of a central processing unit (CPU) count, a CPU speed, an amount of memory or local disk characteristics.

5. The method of claim 3, further comprising assigning the computing nodes having attributes that most closely match the node requirements defined for the tiers.

6. The method of claim 3, wherein assigning the computing nodes comprises:

selecting one of the tiers;

identifying the node requirements of the selected tier;

retrieving from a database computing node attributes associated with an unallocated set of the computing nodes; and

comparing the node attributes of the unallocated set of computing nodes to the node requirements of the selected tier.

7. The method of claim 6, further comprising:

determining whether the node attributes for each of the unallocated computing nodes meet a set of minimum node requirements of the selected tier; and

removing the unallocated computing nodes from a list of candidate computing nodes for the selected tier when the node attributes do not meet the minimum node requirements of the selected tier.

8. The method of claim 6, further comprising assigning the unallocated computing nodes to the node slots of the tier when the node attributes of the unallocated computing nodes and the node requirements of the selected tier are an exact match.

9. The method of claim 6, further comprising:

calculating a processing energy for each of the unallocated computing nodes based on differences between the node attributes of the unallocated computing nodes and the node requirements of the selected tier; and

recording the processing energy of the unallocated computing node.

10. The method of claim 9, further comprising:

calculating the processing energy of other unallocated computing nodes;

selecting the unallocated computing node having a lowest positive processing energy; and

assigning the selected unallocated computing node to a node slot of the selected tier.

11. The method of claim 1, further comprising:

identifying a need for additional processing capacity on one of the tiers; and

assigning an additional computing node to the tier.

12. The method of claim 11, wherein assigning an additional computing node comprises:

receiving input that defines specific node requirements of each of the tiers of the fabric;

selecting an unallocated computing node from a free pool based the node requirements associated with the tier in need of the additional processing capacity; and

assigning the selected unallocated computing node to the tier in need of processing capacity.

13. The method of claim 1, further comprising:

receiving input that defines a re-provisioning policy for one or more of the tiers of the fabric;

identifying a need for additional processing capacity on one of the tiers; and

transferring a node from one of the other tiers to the tier in need of additional processing capacity according to the re-provisioning policy.

14. The method of claim 13, wherein the tiers are assigned priority values, and wherein transferring a node comprises:

identifying the tiers with a priority lower than the tier needing the additional processing capacity;

determining which of the computing nodes associated with the lower priority tiers satisfy requirements of the tier in need of processing capacity;

calculating a processing energy of each of the nodes of the lower priority tiers that satisfy the requirements; and

assigning the computing node with the highest processing energy to the tier in need of processing capacity

15. The method of claim 1, wherein each of the domains corresponds to one of an enterprise application or an enterprise service.

16. The method of claim 1, wherein the node slots represent a data set that describes specific information for a corresponding node.

17. A method comprising:

receiving input defining an organization model for a distributed computing system, wherein the model includes a plurality of tiers that represent different computing functions, and wherein the model defines a priority and computing node requirements for each of the tiers; and

assigning unallocated computing nodes to the tiers from a free pool based on the priority and computing requirements for each of the tiers.

18. The method of claim 17, wherein assigning unallocated computing nodes further comprises:

selecting one of the tiers based on the priorities of the tiers;

identifying the node requirements associated with the selected tier;

selecting one of the unallocated computing nodes from the free pool;

retrieving computing node attributes associated with the selected unallocated computing node;

comparing the node attributes of the selected unallocated computing node to the node requirements associated with the selected tier; and

assigning the selecting unallocated computing node to the selected tier based on the comparison.

19. The method of claim 18, further comprising:

determining whether the node attributes of the selected unallocated computing node meets a minimum node requirements of the selected tier; and

removing the selected unallocated node from a list of candidate nodes when the node attributes do not meet the minimum node requirements of the selected tier.

20. The method of claim 18, further comprising assigning the selected unallocated computing node to a node slot of the tier when the node attributes of the selected unallocated computing node match the node requirements of the selected tier.

21. The method of claim 17, wherein assigning unallocated computing nodes:

calculating a processing energy for each of the unallocated computing node based on differences between node attributes of the unallocated computing nodes and the node requirements of the selected tier; and

assigning the unallocated computing nodes to tier based on the calculated processing energy for each of the unallocated computing nodes.

22. A distributed computing system comprising:

a plurality of application nodes interconnected via a communications network;

a database that stores a hierarchical model for the distributed computing system, wherein the model specifies a fabric having one or more domains, and wherein each domain has at least one tier that includes one or more node slots; and

a control node that automatically configures the application nodes in accordance with the hierarchical organization defined by the model.

23. The distributed computing system of claim 22, wherein the model defines policies that are used by the control node to assign the application nodes to the node slots of the tiers and to re-provision the application nodes between the tiers.

24. The distributed computing system of claim 22, further comprising an image repository storing master software images and software image instances generated from the master software images,

wherein the control node deploys the software images instances to the application nodes by assigning the software images to the node slots in accordance with the model.

25. The distributed computing system of claim 24,

wherein each tier represents a different computing function and is associated with a different master software image, and

wherein each node slot identifies one of the software image instance associated and a network address to be assigned to the respective application node.