Apparatus and method for processing task

• US 9,009,713 B2
• Filed: 03/11/2013
• Issued: 04/14/2015
• Est. Priority Date: 10/05/2012
• Status: Expired due to Fees

• Current Assignee: Electronics and Telecommunications Research Institute
• Original Assignee: Electronics and Telecommunications Research Institute
• Inventors: Lee, Soo Jeon, Jung, Won Chan, Kim, Jae Hoon
• Primary Examiner(s): Puente, Emerson
• Assistant Examiner(s): KIM, SISLEY NAHYUN
• Application Number:
  US13/793,108
• Publication Number:
  US 20140101669A1
• Time in Patent Office: 764 Days
• Field of Search: None
• US Class Current: 718/102
• CPC Class Codes: G06F 2209/506 Constraint
• CPC Class Codes: G06F 9/5038 considering the execution order of a plurality of tasks, e.g. taking priority or time dependency constraints into consideration scheduling strategies G06F9/4881 and subgroups
• CPC Class Codes: G06F 9/54 Interprogram communication

Abstract

Provided is a task processing apparatus and method that may select a task corresponding to predetermined task selection information when a task execution is completed and thus, an idle server occurs in at least one server, may separate the selected task into a first task and a second task, and may control the first task and the second task to be allocated to an existing allocation server of the selected task and an idle server, respectively.

Claims

1. a task processing apparatus, comprising:

an allocator to allocate a plurality of tasks to a plurality of servers, each task being allocated to a different server;

an executor to execute the plurality of tasks through the plurality of servers;

a task selector to select a task, which is allocated to an existing allocation server from among the plurality of servers, to be parallel processed by multiple servers from among the plurality of servers, the task being selected from among the plurality of tasks based on predetermined task selection information when a task execution is completed and an idle server occurs in the plurality of servers, wherein the predetermined task selection information to be parallelized comprises rule information used to select a task having a relatively large number of child nodes from the plurality of tasks and a relatively small prediction interval up to a child node from the plurality of tasks;

a separator to separate the selected task into a first task and a second task applying a parallelization level of the first task and second task;

a controller to control the first task to be allocated to the existing allocation server and the second task to be allocated to the idle server in order to parallel process the first and second tasks; and

a combiner to combine the first task and the second task into a single task, wherein the allocator allocates the combined single task to the existing allocation server.

2. The task processing apparatus of claim 1, wherein the controller decreases a parallelization level of the combined task compared to a parallelization level of the first task and the second task.

3. The task processing apparatus of claim 1, wherein the executor parallel-processes the first task and the second task through the existing allocation server and the idle server, respectively.

4. The task processing apparatus of claim 1, wherein the task selection information comprises rule information used to minimize the average execution time of entire nodes.

5. The task processing apparatus of claim 1, wherein the task selection information comprises rule information used to minimize an execution time of a node having the longest execution time.

6. a task processing method, comprising:

allocating a plurality of tasks to a plurality of servers, each task being allocated to a different server;

executing the plurality of tasks through the plurality of servers;

selecting a task, which is allocated to an existing allocation server from among the plurality of servers, to be parallel processed by multiple servers from among the plurality of servers, the task being selected from among the plurality of tasks based on predetermined task selection information when a task execution is completed and an idle server occurs in the plurality of servers, wherein the predetermined task selection information to be parallelized comprises rule information used to select a task having a relatively large number of child nodes from the plurality of tasks and a relatively small prediction interval up to a child node from the plurality of tasks;

separating the selected task into a first task and a second task applying a parallelization level of the first task and second task;

controlling the first task to be allocated to the existing allocation server and the second task to be allocated to the idle server in order to parallel process the first and second tasks; and

combining the first task and the second task into a single task, wherein the allocator allocates the combined single task to the existing allocation server.

7. The method of claim 6, further comprising:

decreasing a parallelization level of the combined task compared to a parallelization level of the first task and the second task.

8. The method of claim 6, further comprising:

parallel-processing the first task and the second task through the existing allocation server and the idle server, respectively.

9. The method of claim 6, wherein the task selection information comprises rule information used to minimize the average execution time of entire nodes.

10. The method of claim 6, wherein the task selection information comprises rule information used to minimize an execution time of a node having the longest execution time.