Leveraging remote server pools for client applications

• US 8,549,106 B2
• Filed: 06/15/2009
• Issued: 10/01/2013
• Est. Priority Date: 06/15/2009
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Howell, Jonathan R., Elson, Jeremy E., Douceur, John R., Lorch, Jacob R.
• Primary Examiner(s): SALL, EL HADJI MALICK
• Application Number:
  US12/484,356
• Publication Number:
  US 20100318630A1
• Time in Patent Office: 1,569 Days
• Field of Search: 709/203, 709/218, 709/24
• US Class Current: 709/218
• CPC Class Codes: G06F 2209/509 Offload
• CPC Class Codes: G06F 9/5055 considering software capabilities, i.e. software resources associated or available to the machine

Abstract

Techniques for enabling client computing devices to leverage remote server pools for increasing the effectiveness of applications stored on the client computing device are described herein. In some instances, the server pools comprise a “cloud”, “cluster” or “data center” that comprises hundreds or thousands of servers connected together by a network that has an extremely low latency and high bandwidth relative to the network through which the client computing device connects to the server pool. The client computing device may request that the server pool perform a certain task for an application whose canonical state resides on the client. After computation of a result of the task, a server of the server pool then provides the result to the client. By doing so, the techniques dramatically increase the amount of resources working on the request of the client and, hence, dramatically increase the speed and effectiveness of the client-side application.

Claims

1. One or more computer devices comprising:

one or more processors;

one or more computer-readable media storing computer-executable instructions that, when executed, cause one or more processors to perform acts, comprising

receiving, from a client computing device and at a pool of servers, a request to execute a task for an application whose canonical state resides on the client computing device, wherein the request is received over a network that has a substantially higher latency or a substantially lower bandwidth than a network connecting servers of the pool of servers;

computing, by the one or more processors, a result of the task by the pool of servers, the canonical state of the application residing on the client computing device throughout the computing by the one or more processors; and

providing the result to the client computing device for output by the application whose canonical state resides on the client computing device.

2. One or more computer devices as recited in claim 1, wherein the network over which the request is received comprises a Wide Area Network (WAN) and the network connecting the servers of the pool of servers comprises a Local Area Network (LAN).

3. One or more computer devices as recited in claim 1, wherein the computing of the result by the pool of servers comprises:

receiving the request from the client computing device at a first server of the pool of servers;

distributing one of multiple sub-tasks of the task to a second server of the pool of servers and another of the multiple sub-tasks of the task to a third server of the pool of servers for computing by the second and third servers, respectively; and

receiving, at the first server, respective results of the sub-tasks distributed to the second and third servers.

4. One or more computer devices as recited in claim 1, wherein the computing of the result by the pool of servers comprises:

sending a request to the client computing device for a file required for the computing of the result or for an identification of the file;

receiving a hash value associated with the file from the client computing device;

determining if the received hash value matches a hash value stored by the pool of servers; and

if it is determined that the received hash value matches a hash value stored by the pool of servers, using a file stored on the pool of servers and associated with the hash value stored on the pool of servers for computing the result of the task.

5. One or more computer devices as recited in claim 4, wherein the computing of the result by the pool of servers further comprises:

if it is determined that the received hash value does not match a hash value stored by the pool of servers;

sending a request to the client computing device for an identification of multiple blocks that comprise the requested file;

receiving hash values associated with respective blocks of the multiple blocks from the client computing device;

comparing the received hash values associated with the respective blocks with hash values stored on the pool of servers; and

if it is determined that the pool of servers does not store a hash value matching a received hash value for a block, sending a request for the block of the requested file and receiving the request block from the client computing device.

6. One or more computer devices as recited in claim 1, wherein the computing of the result by the pool of servers comprises:

receiving, from the client computing device, an identification of files previously requested by the pool of servers during a previous execution of a task for the application by the pool of servers.

7. One or more computer devices as recited in claim 6, wherein the computing of the result by the pool of servers further comprises:

using the received identification of the files until completion of the currently-requested task.

8. One or more computer devices as recited in claim 6, wherein the received identifications of the files comprise hash values associated with the respective files.

9. One or more computer devices as recited in claim 6, wherein the pool of servers has previously requested each of the identified files at least a predetermined number of times during the execution of a requested task.

10. One or more computer devices as recited in claim 6, wherein the computing of the result by the pool of servers further comprises:

at least in part in response to the receiving of the identification of the files previously requested by the pool of servers, identifying files that the pool of servers does not currently need for computing the result of the task and providing the identification of the unneeded files to the client computing device.

11. A method comprising:

receiving, over a first network and at a server of a pool of servers interconnected by a second network having a latency that is substantially lower than a latency of the first network or a bandwidth that is substantially higher than a bandwidth of the first network, a request from a client computing device to execute a task for an application whose canonical state resides on the client computing device;

receiving, from the client computing device over the first network and at the server, code of the application whose canonical state resides on the client computing device;

distributing, by the server and to multiple other servers of the pool of servers, portions of the task and portions of the code of the application whose canonical state resides on the client computing device for processing by the multiple other servers, the canonical state of the application residing on the client computing device throughout the processing by the multiple other servers; and

sending a result of the execution of the task from the server and to the client computing device over the first network.

12. A method as recited in claim 11, wherein the first network comprises a Wide Area Network (WAN) and the second network comprises a Local Area Network (LAN).

13. A method as recited in claim 11, further comprising:

receiving, from the client computing device, an identification of files previously requested by the pool of servers during a previous execution of a task for the application by the pool of servers.

14. A method as recited in claim 11, further comprising:

sending a request to the client computing device for a file required for the execution of the task or for an identification of the file;

receiving an identification of the file;

using the identification of the file to determine if the server has previously received the file; and

if it is determined that the server has previously received the file, using the file for the execution of the task.

15. A method as recited in claim 14, further comprising:

if it is determined that the server has not previously received the file, sending a request to the client computing device for an identification of each of multiple blocks of the file;

receiving an identification for each block of the multiple blocks of the file;

using the identifications of the multiple blocks to determine if the server has previously received the respective block;

at least in part in response to determining that the server has previously received a particular block, using the particular block for the execution of the task; and

at least in part in response to determining that the server has not previously received a particular block, sending a request for the particular block to the client computing device.