AUTOMATED MECHANISMS FOR ENSURING CORRECTNESS OF EVOLVING DATACENTER CONFIGURATIONS

• US 20200133688A1
• Filed: 10/30/2018
• Published: 04/30/2020
• Est. Priority Date: 10/30/2018
• Status: Active Grant

• Current Assignee: Oracle International Corporation (Oracle Corporation)
• Original Assignee: Oracle International Corporation (Oracle Corporation)
• Inventors: SHINDE, PRAVIN, SCHMIDT, FELIX, SCHELP, CRAIG
• Granted Patent: US 10,795,690 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 16/9024 Graphs; Linked lists G06F16/9027 takes precedence
• CPC Class Codes: G06F 2009/4557 Distribution of virtual machine instances; Migration and load balancing
• CPC Class Codes: G06F 9/44505 Configuring for program initiating, e.g. using registry, configuration files
• CPC Class Codes: G06F 9/45558 Hypervisor-specific management and integration aspects

Abstract

Herein are computerized techniques for generation, costing/scoring, optimal selection, and reporting of intermediate configurations for a datacenter change plan. In an embodiment, a computer receives a current configuration of a datacenter and a target configuration. New configurations are generated based on the current configuration. A cost function is applied to calculate a cost of each new configuration based on measuring a logical difference between the new configuration and the target configuration. A particular new configuration is selected that has a least cost. When the particular configuration satisfies the target configuration, the datacenter is reconfigured based on the particular configuration. Otherwise, this process is (e.g. iteratively) repeated with the particular configuration instead used as the current configuration. In embodiments, new configurations are randomly, greedily, and/or manually generated. In an embodiment, new configurations obey design invariants that constrain which changes and/or configurations are attainable.

Claims

1. A method comprising:

receiving a current configuration of a datacenter and a target configuration of said datacenter;

generating a plurality of new configurations of said datacenter that are based on said current configuration;

applying a cost function to calculate a cost of each configuration of said plurality of new configurations based on measuring a logical difference between said each configuration and said target configuration;

selecting a particular configuration of said plurality of new configurations that has a least cost;

when the particular configuration satisfies said target configuration, reconfiguring said datacenter based on said particular configuration;

when the particular configuration does not satisfy said target configuration, repeating said method with said particular configuration as said current configuration and with said target configuration;

wherein the method is performed by one or more computers.

2. The method of claim 1 wherein:

said method further comprises receiving design invariants to impose upon said plurality of new configurations;

said plurality of new configurations comprises a first configuration that satisfies said design invariants and a second configuration that does not satisfy said design invariants;

said calculate said cost of said each configuration comprises increasing said cost of said second configuration based on detecting that said second configuration that does not satisfy said design invariants;

selecting said particular configuration comprises selecting said first configuration based on said increasing said cost of said second configuration.

3. The method of claim 2 wherein:

the method further comprises generating a detailed change plan for achieving said reconfiguring;

said detailed change plan comprises a sequence of intermediate configurations;

each configuration of said sequence of intermediate configurations satisfies said design invariants.

4. The method of claim 2 wherein said design invariants constrain at least one of:

power, or space.

5. The method of claim 1 wherein:

said plurality of new configurations comprises a first configuration that passes a simulation test and a second configuration fails said simulation test;

said calculate said cost of said each configuration comprises increasing said cost of said second configuration based on detecting that said second configuration fails said simulation test;

selecting said particular configuration comprises selecting said first configuration based on said increasing said cost of said second configuration.

6. The method of claim 1 wherein said plurality of new configurations are further based on said target configuration.

7. The method of claim 1 wherein said cost of said each configuration is based on at least one of:

a count of hardware elements that are required by said each configuration,

a capital cost of hardware elements that are required by said each configuration,

a count of types of hardware elements that are required by said each configuration,

a count of stock keeping units (SKUs) of hardware elements that are required by said each configuration,

a count of redundant communication paths between hardware elements that are achieved by said each configuration, or

an amount that is required by said each configuration of at least one of;

power, or space.

8. The method of claim 1 wherein:

vertices of a logical graph are associated with respective hardware elements of said each configuration;

said cost of said each configuration is increased based on a count of communication links in a minimum cut that divides said logical graph into two partitions of similar size.

9. The method of claim 8 further comprising storing said logical graph into a graph database.

10. The method of claim 9 wherein applying said cost function comprises applying a user defined function (UDF) that is defined in said graph database.

11. The method of claim 1 wherein:

said plurality of new configurations comprises a first configuration and a second configuration that is more similar to said current configuration than is said first configuration to said current configuration;

a cost of said first configuration exceeds a cost of said second configuration.

12. The method of claim 1 wherein:

said applying said cost function comprises applying a plurality of cost functions that includes said cost function;

said calculate said cost of said each configuration comprises calculating a weighted cost for each cost function of said plurality of cost functions.

13. The method of claim 1 wherein said generating said plurality of new configurations comprises at least two of:

manually generating a configuration,

randomly generating a configuration, or

generating a configuration based on said current configuration and said target configuration.

14. The method of claim 1 wherein said selecting said particular configuration comprises:

identifying a fixed size subset of said plurality of new configurations that have less costs than other configurations of said plurality of new configurations;

adjusting said costs of said fixed size subset of configurations of said fixed size subset based on performance measurements of simulated communication traffic in said fixed size subset.

15. The method of claim 14 wherein said simulated communication traffic comprises at least one of:

discrete event network simulation, or

dynamic routing by one or more hardware elements of a configuration of said fixed size subset.

16. The method of claim 14 wherein said simulated communication traffic occurs for a particular real duration.

17. The method of claim 1 wherein:

a compute cloud contains contains said datacenter and a second datacenter;

generating said plurality of new configurations of said datacenter comprises generating at least one of;

a second plurality of new configurations of new configurations of said second datacenter, or said plurality of new configurations that span both of said datacenter and said second datacenter.

18. The method of claim 1 wherein at least one of:

the method further comprises generating a detailed change plan for achieving said reconfiguring;

a topology of said particular configuration is none of;

a tree, or a star;

or

said current configuration contains at least one of;

a floor plan, or a wiring plan.

19. The method of claim 18 wherein said detailed change plan comprises a sequence of intermediate configurations.

20. One or more non-transient computer-readable media storing instructions that, when executed by one or more processors, cause:

receiving a current configuration of a datacenter and a target configuration of said datacenter;

generating a plurality of new configurations of said datacenter that are based on said current configuration;

applying a cost function to calculate a cost of each configuration of said plurality of new configurations based on measuring a logical difference between said each configuration and said target configuration;

selecting a particular configuration of said plurality of new configurations that has a least cost;

when the particular configuration satisfies said target configuration, reconfiguring said datacenter based on said particular configuration;

when the particular configuration does not satisfy said target configuration, repeating said method with said particular configuration as said current configuration and with said target configuration.