Apparatus, systems and methods for deployment and management of distributed computing systems and applications

• US 8,862,933 B2
• Filed: 02/09/2011
• Issued: 10/14/2014
• Est. Priority Date: 02/09/2011
• Status: Active Grant

• Current Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Original Assignee: CliQr Technologies, Inc. (Cisco Systems, Inc.)
• Inventors: Manglik, Gaurav, Fu, Tianying
• Primary Examiner(s): Ng, Fan
• Application Number:
  US13/024,302
• Publication Number:
  US 20120203823A1
• Time in Patent Office: 1,343 Days
• Field of Search: 709/203, 714/11
• US Class Current: 714/11
• CPC Class Codes: G06F 9/5072 Grid computing
• 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]
• CPC Class Codes: H04L 67/565 Conversion or adaptation of application format or content adding application control or application functional data H04L67/561

Abstract

Embodiments of apparatus, systems and methods facilitate the running a distributed computing application on at least one of a plurality of cloud infrastructures. In some embodiments, the processor implemented method comprises: deriving an infrastructure independent representation of the distributed computing application; and running the distributed computing application on one of a plurality of cloud infrastructures by utilizing a cloud-specific implementation of the infrastructure independent representation of the distributed computing application. The cloud-specific implementation of the infrastructure independent representation corresponds to the cloud infrastructure on which the distributed application is run.

Claims

1. A processor-implemented method for running a distributed computing application on at least one of a plurality of cloud infrastructures, the processor implemented method comprising:

deriving a cloud infrastructure independent representation of the distributed computing application, wherein the cloud infrastructure independent representation comprises a representation of an underlying pattern of resource utilization of the distributed computing application; and

running the distributed computing application on the at least one cloud infrastructure by utilizing at least one cloud-specific implementation of the cloud infrastructure independent representation of the distributed computing application, and wherein the at least one cloud-specific implementation corresponds to the at least one cloud infrastructure.

2. The processor-implemented method of claim 1, wherein the cloud infrastructure independent representation comprises pattern primitives and system primitives.

3. The processor-implemented method of claim 2, wherein the at least one cloud-specific implementation corresponding to the at least one cloud infrastructure is realized by utilizing a library of program routines that implement the system and pattern primitives on the at least one cloud infrastructure.

4. The processor-implemented method of claim 1, wherein the cloud infrastructure independent representation is specified in terms of composite interfaces and wherein the composite interfaces are built using pattern primitives and system primitives as building blocks.

5. The processor-implemented method of claim 2, wherein the cloud infrastructure independent representation further includes cluster synchronization operators, wherein the cluster synchronization operators are used to specify a workflow during runtime on the at least one cloud infrastructure.

6. The processor-implemented method of claim 1, wherein the plurality of cloud infrastructures includes one or more of Amazon EC2, Terremark vCloud, Rackspace Cloudserver, Microsoft Azure, Savvis, or private clusters.

7. The processor-implemented method of claim 2, wherein the pattern primitives comprise parallel execution pattern primitives, batch execution pattern primitives and three-tier web primitives.

8. The processor-implemented method of claim 5, wherein the cloud infrastructure independent representation uses primitives and synchronization operators to specify templates for MPI and map-reduce.

9. The processor implemented method of claim 2, wherein the system primitives comprise launching a cluster with a specified number of compute nodes and launching shared storage.

10. The processor implemented method of claim 5, wherein the cluster synchronization operators comprise wait-on-cluster and wait-on-node.

11. A system comprising:

at least one system manager, wherein the system manager derives a cloud infrastructure independent representation of the distributed computing application comprising a representation of an underlying pattern of resource utilization of the distributed computing application;

at least one coordinator, wherein the coordinator maintains a library of program routines that are used to implement the cloud infrastructure independent representation of the distributed computing application on at least one of a plurality of cloud infrastructures; and

wherein the at least one system manager sends the cloud infrastructure independent representation of the distributed computing application to the at least one coordinator and the coordinator executes routines from the library of program routines to implement the infrastructure independent representation of the distributed computing application on the at least one cloud infrastructure.

12. The system of claim 11, wherein the cloud infrastructure independent representation comprises pattern primitives and system primitives.

13. The system of claim 12, wherein the library of program routines maintained by the coordinator implement the system and pattern primitives on the at least one cloud infrastructure.

14. The system of claim 11, wherein the cloud infrastructure independent representation is specified in terms of composite interfaces and wherein the composite interfaces are built using pattern primitives and system primitives as building blocks.

15. The system of claim 12, wherein the cloud infrastructure independent representation further includes cluster synchronization operators, wherein the cluster synchronization operators are used to specify a workflow during the running of the distributed computing application on the at least one cloud infrastructure.

16. The system of claim 11, wherein system manager and the coordinator communicate over secure communication links using RESTful web service calls.

17. The system of claim 15, wherein the coordinator further launches agents to monitor and report on workflow execution state for the distributed computing application.

18. The system of claim 11, wherein the plurality of cloud infrastructures includes one or more of Amazon EC2, Terremark vCloud, Rackspace Cloudserver, Microsoft Azure, Savvis, or private clusters.

19. A non-transitory computer-readable medium that stores instructions, which when executed by a processor perform steps in a method for running a distributed computing application on at least one of a plurality of cloud infrastructures, the processor implemented method comprising:

deriving a cloud infrastructure independent representation of the distributed computing application, wherein the cloud infrastructure independent representation comprises a representation of an underlying pattern of resource utilization of the distributed computing application; and

running the distributed computing application on the at least one cloud infrastructure by utilizing at least one cloud-specific implementation of the cloud infrastructure independent representation of the distributed computing application, and wherein the at least one cloud-specific implementation corresponds to the at least one cloud infrastructure.

20. The computer-readable medium of claim 19, wherein the cloud infrastructure independent representation comprises pattern primitives and system primitives.