Forming neighborhood groups from disperse cloud providers

• US 10,334,029 B2
• Filed: 01/10/2017
• Issued: 06/25/2019
• Est. Priority Date: 01/10/2017
• Status: Active Grant

• Current Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Original Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Inventors: Agrawal, Mohit
• Primary Examiner(s): Donabed, Ninos
• Application Number:
  US15/403,074
• Publication Number:
  US 20180198852A1
• Time in Patent Office: 896 Days
• CPC Class Codes: G06F 9/5072 Grid computing
• CPC Class Codes: H04L 67/10 in which an application is distributed across nodes in the network software deployment G06F8/60; multiprogramming arrangements G06F9/46
• CPC Class Codes: H04L 67/1001 for accessing one among a plurality of replicated servers
• CPC Class Codes: H04W 4/08 User group management

Abstract

A cloud manager, for forming neighborhood groups from disperse cloud providers, receives provider data relating to a plurality of disperse cloud providers for a plurality of data subcategories classified under N-number main categories. The cloud manager generates a respective vector to represent each of the plurality of disperse cloud providers based on the provider data. The cloud manager generates an N-number axis space comprising the vectors. The cloud manager groups the plurality of disperse cloud providers in the N-number axis space into at least one cloud provider group, based on the vectors and a clustering algorithm.

Claims

1. A method for forming neighborhood groups from disperse cloud providers, by a cloud manager, the method comprising:

receiving provider data relating to a plurality of disperse cloud providers for a plurality of data subcategories classified under N-number main categories including at least;

a feasibility category with corresponding data subcategories of ISP provider, cloud provider location, network type and computing type;

an offering similarity category with corresponding data subcategories of service level agreement (“

SLA”

), cost model, allocation model, and addon resource; and

provider credibility with corresponding data subcategories of percentage of SLAs met, user ratings, resources used, and cloud size;

scoring the feasibility category, offering similarity category and provider credibility category to produce a feasibility score, offering similarity score and provider credibility score;

generating a respective vector to represent each of the plurality of disperse cloud providers based on the provider data to yield vectors, each respective vector including as an element of the vector the feasibility score, the offering similarity sore and the provider credibility score of the respective cloud provider;

generating an N-number axis space comprising the vectors; and

grouping the plurality of disperse cloud providers in the N-number axis space into at least one cloud provider group, based on the vectors and a clustering algorithm wherein the clustering algorithm is a suggested and supervised KMeans (SS-KMeans) clustering algorithm;

providing access to the one cloud provider group such that resources within the one cloud provider group are collectively available for use.

2. A method of claim 1, wherein each respective element of the respective vector is based on the provider data for one of the N-number main categories.

3. A method of claim 1, wherein grouping the plurality of disperse cloud providers comprises grouping each respective vector in the N-number axis space into one of the at least one cloud provider group, based on the clustering algorithm.

4. The method of claim 1, further comprising monitoring for a change in the provider data for a disperse cloud provider.

5. The method of claim 4, further comprising in response to the change in the provider data for a disperse cloud provider, generating an updated vector for the disperse cloud provider, and regrouping the disperse cloud provider.

6. The method of claim 1, wherein each grouping of one of the plurality of disperse cloud providers reinforces the SS-KMeans clustering algorithm.

7. The method of claim 1, wherein grouping first separates the data points on a single plane of the N-number axis space representing a corresponding main category, before applying the clustering algorithm.

8. A cloud manager device for forming neighborhood groups from disperse cloud providers, the cloud manager device comprising:

at least one processor; and

a memory coupled to the at least one processor, the memory storing instructions which, when executed by the at least one processor, cause the at least one processor to perform operations comprising;

receiving provider data relating to a plurality of disperse cloud providers for a plurality of data subcategories classified under N-number main categories including at least;

a feasibility category with corresponding data subcategories of ISP provider, cloud provider location, network type and computing type;

an offering similarity category with corresponding data subcategories of service level agreement (“

SLA”

), cost model, allocation model, and addon resource; and

provider credibility with corresponding data subcategories of percentage of SLAs met, user ratings, resources used, and cloud size;

scoring the feasibility category, offering similarity category and provider credibility category to produce a feasibility score, offering similarity score and provider credibility score;

generating a respective vector to represent each of the plurality of disperse cloud providers based on the provider data to yield vectors, each respective vector including as an element of the vector the feasibility score, the offering similarity sore and the provider credibility score of the respective cloud provider;

generating an N-number axis space comprising the vectors; and

grouping the plurality of disperse cloud providers in the N-number axis space into at least one cloud provider group, based on the vectors and a clustering algorithm wherein the clustering algorithm is a suggested and supervised KMeans (SS-KMeans) clustering algorithm;

providing access to the one cloud provider group such that resources within the one cloud provider group are collectively available for use.

9. A cloud manager device of claim 8, wherein each respective element of the respective vector is based on the provider data for one of the N-number main categories.

10. A claim device of claim 8, wherein grouping the plurality of disperse cloud providers comprises grouping each respective vector in the N-number axis space into one of the at least one cloud provider group, based on the clustering algorithm.

11. A cloud manager device of claim 8, wherein the KMeans clustering algorithm groups the data points in the N-number axis space into k-number clusters.

12. A cloud manager device of claim 8, wherein grouping first separates the data points on a single place of the N-number axis space representing a corresponding main category before applying the clustering algorithm.

13. A non-transient computer readable storage device storing program instructions which, when executed by a processor, cause the processor to perform operations comprising:

receiving provider data relating to a plurality of disperse cloud providers for a plurality of data subcategories classified under N-number main categories including at least;

a feasibility category with corresponding data subcategories of ISP provider, cloud provider location, network type and computing type;

an offering similarity category with corresponding data subcategories of service level agreement (“

SLA”

), cost model, allocation model, and addon resource; and

provider credibility with corresponding data subcategories of percentage of SLAs met, user ratings, resources used, and cloud size;

scoring the feasibility category, offering similarity category and provider credibility category to produce a feasibility score, offering similarity score and provider credibility score;

generating a respective vector to represent each of the plurality of disperse cloud providers based on the provider data to yield vectors, each respective vector including as an element of the vector the feasibility score, the offering similarity sore and the provider credibility score of the respective cloud provider;

generating an N-number axis space comprising the vectors; and

grouping the plurality of disperse cloud providers in the N-number axis space into at least one cloud provider group, based on the vectors and a clustering algorithm wherein the clustering algorithm is a suggested and supervised KMeans (SS-KMeans) clustering algorithm;

providing access to the one cloud provider group such that resources within the one cloud provider group are collectively available for use.