Coordination engine for cloud selection

• US 9,781,205 B2
• Filed: 09/12/2011
• Issued: 10/03/2017
• Est. Priority Date: 09/12/2011
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Inventors: Batrouni, Marwan, Ashkar, Shady N.
• Primary Examiner(s): RUBIN, BLAKE J
• Application Number:
  US13/230,487
• Publication Number:
  US 20130067090A1
• Time in Patent Office: 2,213 Days
• Field of Search: 709226
• CPC Class Codes: H04L 41/022 Multivendor or multi-standard integration
• CPC Class Codes: H04L 41/0226 Mapping or translating multiple network management protocols
• 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 67/1012 based on compliance of requirements or conditions with available server resources
• CPC Class Codes: H04L 67/1014 based on the content of a request
• CPC Class Codes: H04L 67/1097 for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS]

Abstract

Methods, systems, and computer-readable media for selecting and managing a public cloud-computing network to host a client'"'"'s account information are provided. Initially, the client issues a request to a coordination engine, which understands rules language of various public clouds, to update the account information residing on a target cloud. The target cloud was previously selected from the various public clouds as a function of desirable criteria specified by the client in light of properties (e.g., pricing, security, and reliability) dynamically abstracted from the public clouds. When addressing the request, the coordination engine extracts command(s) from the request and automatically translates the command(s) consistent with the rules language of the target cloud. Upon delivery to the target cloud, the translated command(s) affect reading of or writing to the account information. Accordingly, the client is absolved from converting instructions into a format that is expected by the target cloud.

Claims

1. One or more computer memory devices having computer-executable instructions embodied thereon that, when executed, perform a method for assigning workload to candidate computer networks based on criteria provided from a client, the method comprising:

receiving, a request for computing resources from the client, the request is received via an abstraction layer that comprises one or more interfaces that serve as an intermediary for the client to interact with a coordination engine;

receiving at the coordination engine the criteria associated with the request, wherein the criteria specify client-preferred properties for candidate computer networks;

employing the coordination engine to perform an analysis of the criteria with respect to metrics of abstracted properties corresponding to a plurality of candidate computer networks, wherein the coordination engine employs a rules language for defining and evaluating criteria with respect to the metrics, wherein the rules languages supports client-defined weighting, ranking, absolute and optional designations for the criteria;

the coordination engine dynamically updates target computer networks with candidate computer networks based on performing the analysis, wherein performing the analysis of the criteria comprises;

(a) accessing the metrics at a metrics database, wherein the metrics are mined from the plurality of candidate computer networks, the metrics of abstracted properties are identified using agents associated with the coordination engine, the agents dynamically collect the metrics of the plurality of computer networks; and

(b) comparing the criteria of client-preferred properties against the metrics of abstracted properties of the plurality of candidate computer networks, comparing is based at least in part on a manifest comprising metrics of abstracted properties for the plurality of candidate computer networks;

based on the comparison targeting at least one computer network, from the plurality of candidate computer networks, which exhibits metrics that satisfy the criteria and designations; and

initiating interaction with the at least one targeted computer network.

2. The computer memory devices of claim 1, wherein the request comprises instructions to run an application on virtual machines available at the plurality of candidate computer networks, and wherein the application is associated with the client'"'"'s account.

3. The computer memory devices of claim 1, wherein the request comprises instructions to maintain data on a storage location available at the plurality of candidate computer networks, and wherein the data is associated with the client'"'"'s account.

4. The computer memory devices of claim 1, wherein the criteria define particular attributes of the plurality of candidate computer networks that pertain to at least one of security, availability, cost, scalability, or geo-redundancy.

5. The computer memory devices of claim 1, further comprising:

accessing updated metrics at the metrics database, the updated metrics are identified using agents of the coordination engine that dynamically collect information from the one or more computer network to update the metrics;

comparing the criteria of client-preferred properties against the updated metrics of the plurality of candidate computer networks; and

based on the comparison, dynamically updating the at least one targeted network with at least one second targeted network, from the plurality of candidate computer networks, which exhibits updated metrics that satisfy the criteria.

6. The computer memory devices of claim 1, wherein the plurality of candidate computer networks comprise a private enterprise network and at least one public cloud-computing network, and wherein the method further comprises employing the coordination engine to manage usage of the client'"'"'s account across the private enterprise network and the at least one targeted computer network.

7. The computer memory devices of claim 6, wherein the employing the coordination engine to manage usage of the client'"'"'s account across the private enterprise network and the at least one targeted computer network comprises overseeing an application running on virtual machines provisioned on the at least one targeted network.

8. The computer memory devices of claim 6, wherein the employing the coordination engine to manage usage of the client'"'"'s account across the private enterprise network and the at least one targeted computer network comprises tracking data maintained at a storage location provisioned on the at least one targeted network.

9. The computer memory devices of claim 6, wherein the method further comprises employing the coordination engine to provision the computer resources on the at least one targeted computer network in order to meet the request.

10. The computer memory devices of claim 6, wherein the employing the coordination engine to manage usage of the client'"'"'s account across the private enterprise network and the at least one targeted computer network comprises load-balancing usage between the at least one targeted computer network and another public cloud-computing network.

11. The computer memory devices of claim 1, wherein the process of analyzing further comprises:

accessing rules from a rules database, the rules comprising additional terms for selecting the targeted computer network; and

applying the rules to affect an outcome of the comparison of the criteria against the metrics.

12. A computerized method for distributing workload to one or more public computing networks external to a private enterprise network, the method comprising:

receiving a request, issued from a user of the private enterprise network to update account information hosted on the one or more public computing networks, the request is received at a coordination engine, the coordination engine dynamically provisions target computing networks from one or more public computing networks to distribute and load balance workload of the private enterprise network on the target computing network;

identifying using the coordination engine a target computing network from the one or more public computing networks, the target computing network is responsible for hosting the account information, wherein the coordination engine employs a rules language for defining and evaluating criteria with respect to the metrics, wherein the rules languages supports client-defined weighting, ranking, absolute and optional designations for the criteria;

extracting one or more commands from the request, wherein the one or more commands represent, in part, instructions for implementing the update;

translating the one or more commands into a format consistent with a rules language observed by the target computing network when interacting with an external source, and

initiating a distribution of the one or more translated commands to computing resources, associated with the target computing network, that are designated to implement the update to the account information.

13. The computerized method of claim 12, further comprising, upon establishing the account information on the target computing network, releasing to the administrator a token that exposes, among other things, at least one location of the account information within the one or more public computing networks, wherein the coordination engine utilizes the token as corresponding to the at least one location such that the one or more commands are translated for the at least one location, in particular.

14. The computerized method of claim 12, wherein dynamically provisioning the target computing network further comprises:

(a) accessing a rules data store to examine terms provided by an administrator associated with the private enterprise network, wherein the terms expose criteria the administrator deems valuable for an external cloud-computing network to embody;

(b) accessing a metrics data store to examine metrics that describe qualities of the one or more public computing networks designated as candidates for hosting the account information; and

(c) selecting the target computing network as a function of an analysis of the metrics in light of the terms.

15. The computerized method of claim 12, wherein evaluating the coordination engine comprises:

reviewing past decisions of the coordination engine; and

evaluating an impact of the past decisions.

16. A computer system for performing a method that monitors properties of one or more public clouds and selects an appropriate public cloud for hosting account information based on the properties, the computer system comprising a processing unit coupled to a computer memory device, the computer memory device having stored thereon a plurality of computer software components executable by the processing unit, the computer software components comprising:

a rules data store that persists terms provided by an administrator associated with a private cloud, wherein the terms expose criteria the administrator deems valuable for an external cloud-computing network to embody;

a metrics data store that accepts and maintains the properties that describe qualities of the one or more public clouds designated as candidates for hosting the account information;

one or more agents that are programmed to periodically and dynamically collect the properties by crawling the one or more public clouds and to report the collected properties to the metrics data store, wherein the one or more agents are instantiated with parameters on how to interface with the one or more public clouds; and

a coordination engine utilizes one or more interfaces of an abstraction layer to operate as an intermediary between private clouds requesting computing resources from public clouds, wherein the coordination engine employs a rules language for defining and evaluating criteria with respect to the metrics, wherein the rules languages supports client-defined weighting, ranking, absolute and optional designations for the criteria.

17. The computer system of claim 16, wherein the private cloud represents a private cloud-computing network operated by the administrator, and wherein the target cloud represents at least one public cloud-computing network operated by a cloud-service provider.

18. The computer system of claim 16, wherein the software components further comprise the feedback mechanism that evaluates the decision of the coordination engine for establishing rules or modifying rules to optimize provisioning target clouds.

19. The computer system of claim 16, wherein the one or more agents comprise a pricing agent that is programmed to retrieve expected fees for usage from the one or more candidate public clouds, and wherein the one or more agents comprise a security agent that is programmed to measure a level of security imposed by the one or more candidate public clouds.

20. The computer system of claim 16, wherein the target cloud is dynamically updated with a second target cloud based on updated properties of one or more public clouds of the public clouds, the updated properties identified based on the one or more agents dynamically collecting properties of the public clouds.