Operational-task-oriented system and method for dynamically adjusting operational environment

Patent metadata

Field	Value
Publication number	US-9612875-B2
Application number	US-201414541377-A
Country	US
Kind code	B2
Filing date	Nov 14, 2014
Priority date	Oct 14, 2014
Publication date	Apr 4, 2017
Grant date	Apr 4, 2017

Abstract

The present invention provides an operational-task-oriented system and method for dynamically adjusting operational environment applicable to a computer cluster. Each operational node of the computer cluster has two or more operational systems installed. After receiving the operational task, the control node estimates the time required for completing different tasks requiring different operational systems by appropriate operational nodes and compares the estimated finish time and the assigned finish time for judging how to adjust the operating system running in the operational nodes. Thereby, the operational task can be completed in the assigned finish time. Another method is to use the control node to analyze the proportions of the tasks requiring different operational systems in an operational task and hence adjusts the operational system running in an operational node according to the proportion of requirement. Thereby, the operational task can be completed in the shortest time.

First claim

The invention claimed is:

1. A method for dynamically adjusting operational environment, applicable to a computer cluster, each of a plurality of operational nodes of said computer cluster installing a first operational system and a second operational system, and comprising steps of:
a control node receiving an operational task including at least a first-type task and at least a second-type task, said first-type task being operated in said first operational system, and said second-type task being operated in said second operational system;
when said plurality of operational nodes includes at least an idle operational node not performing operation, said control node judging whether said idle operational node operates in said first operational system or said second operational system;
said control nodes estimating an estimated finish time of first-type task for operating said first-type task by said idle operational node running said first operational system and an estimated finish time of second-type task for operating said second-type task by said idle operational node running said second operational system;
said control node comparing said estimated finish time of first-type task, said estimated finish time of second-type task, and an assigned finish time of operational task;
when said estimated finish time of first-type task is greater than said assigned finish time of operational task and said estimated finish time of second-type task is less than said assigned finish time of operational task, said control node assigns at least an idle node running said second operational system to said first-type task by restarting and altering the operational system of said idle operational node to said first operational system; and
when said estimated finish time of second-type task is greater than said assigned finish time of operational task and said estimated finish time of first-type task is less than said assigned finish time of operational task, said control node assigns at least an idle node running said first operational system to said second-type task by restarting and altering the operational system of said idle operational node to said second operational system.

2. The method for dynamically adjusting operational environment of claim 1 , and after said step of said control node comparing said estimated finish time of first-type task, said estimated finish time of second-type task, and an assigned finish time of operational task, further comprising a step of when said estimated finish time of first-type task is less than said assigned finish time of operational task and said estimated finish time of second-type task is less than said assigned finish time of operational task, said control node assigning said first-type task to said idle operational node running said first operational system and said second-type task to said idle operational node running said second operational system.

3. The method for dynamically adjusting operational environment of claim 2 , and after said step of said control node assigning said first-type task to said idle operational node running said first operational system and said second-type task to said idle operational node running said second operational system, further comprising a step of said idle operational node running said first operational system operating said first-type task and said idle operational node running said second operational system operating said second-type task for producing a plurality of operational results.

4. The method for dynamically adjusting operational environment of claim 3 , and after said step of said idle operational node running said first operational system operating said first-type task and said idle operational node running said second operational system operating said second-type task for producing a plurality of operational results, further comprising a step of a storage node storing said plurality of operational results.

5. The method for dynamically adjusting operational environment of claim 1 , and after said step of said control node comparing said estimated finish time of first-type task, said estimated finish time of second-type task, and an assigned finish time of operational task, further comprising a step of when said estimated finish time of first-type task is greater than said assigned finish time of operational task and said estimated finish time of second-type task is greater than said assigned finish time of operational task, said control node notifying that said operational task cannot be completed within said assigned finish time of operational task.

6. A method for dynamically adjusting operational environment, applicable to a computer cluster, each of a plurality of operational nodes of said computer cluster installing a first operational system and a second operational system, and comprising steps of:
a control node receiving an operational task including at least a first-type task and at least a second-type task, said first-type task being operated in said first operational system, and said second-type task being operated in said second operational system;
said control node estimating a time ratio of said first-type task in said operational task;
when said plurality of operational nodes include a plurality of idle operational nodes not performing operation, said control node estimating an operational-system ratio of at least an idle operational node running said first operational system in said plurality of idle operational nodes;
said control node comparing said time ratio and said operational-system ratio;
when said time ratio is greater than said operational-system ratio, said control node assigns at least an idle node running said second operational system to said first-type task by restarting and altering the operational system of said idle operational node to said first operational system; and
when said time ratio is less than said operational-system ratio, said control node assigns at least an idle node running said first operational system to said second-type task by restarting and altering the operational system of said idle operational node to said second operational system.

7. The method for dynamically adjusting operational environment of claim 6 , and after said step of said control node estimating said operational-system ratio of at least an idle operational node running said first operational system in said plurality of idle operational nodes, further comprising a step of when said time ratio is equal to said operational-system ratio, said control node assigning said first-type task to said idle operational node running said first operational system and said second-type task to said idle operational node running said second operational system.

8. The method for dynamically adjusting operational environment of claim 7 , and after said step of said control node assigning said first-type task to said idle operational node running said first operational system and said second-type task to said idle operational node running said second operational system, further comprising a step of said idle operational node running said first operational system operating said first-type task and said idle operational node running said second operational system operating said second-type task for producing a plurality of operational results.

9. The method for dynamically adjusting operational environment of claim 8 , and after said step of said idle operational node running said first operational system operating said first-type task and said idle operational node running said second operational system operating said second-type task for producing a plurality of operational results, further comprising a step of a storage node storing said plurality of operation

Assignees

• Nat Applied Res Laboratories

Inventors

• Wang Ming-Jen
• Chang Chih-Wen
• Lai Chuan-Lin
• Kuo Chia-Chen
• Lian Jiang-Siang

Classifications

• G06F9/5027

  the resource being a machine, e.g. CPUs, Servers, Terminals · CPC title

• G06F9/5061 Primary

  Partitioning or combining of resources · CPC title

• G06F9/5005

  to service a request · CPC title

• G06F9/4856

  resumption being on a different machine, e.g. task migration, virtual machine migration (G06F9/5088 takes precedence) · CPC title