# US 2010/0146516 A1 · Distributed Task System and Distributed Task Management Method

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2010/0146516 A1
- Document ID: US20100146516A1
- Status: Published application
- Kind: A1
- Application number: 11/997,351
- Filed: 2008-01-29
- Published: 2010-06-10
- Earliest claimed: 2007-01-30
- Assignees: Alibaba Group Holding Limited
- Inventors: Jiandong Yao
- Web page: https://patentgrep.com/patent/US20100146516A1

## CPC

- G06Q 10/06: Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06F 9/505: {considering the load}
- G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

A distributed task system has a task transaction server and at least one task server. Instead of being merely passively called by the task transaction server for executing a task, the task server performs self-balancing according to task execution conditions and operation conditions of the task server. The task transaction server receives task requests from the task server, records the execution conditions, and provides feedback to the task server, and the task server executes the task according to the received feedback and the operation conditions of the task server. The task transaction server may determine if the task server can execute the task according to the execution conditions of the task, and feedback to the task server. A self-balancing unit of the task server may further determine whether the task server is busy, and if not busy, trigger a task execution unit of the task server to execute the task.

## Claims

### Claim 1

A distributed task system, comprising:
a task transaction server; and
at least one task server, wherein the task transaction server is adapted to receive from the task server a task request requesting a task, provide a feedback to the task server, and record an execution condition of the task requested, and wherein the task server is adapted to send the task request to the task transaction server, and execute the task according to the feedback received from the task transaction server.

### Claim 2 (depends on 1)

The distributed task system as recited in claim 1, wherein the at least one task server further comprises:
a self-balancing unit; and
a task execution unit, wherein the self-balancing unit is adapted to send the task request to the task transaction server, receive the feedback from the task transaction server, and trigger the task execution unit to execute the task according to the feedback, and wherein the task execution unit is to connected with the self-balancing unit and is adapted to execute the task.

### Claim 3 (depends on 2)

The distributed task system as recited in claim 2, wherein the task execution unit is adapted to execute the task under control of the self-balancing unit.

### Claim 4 (depends on 2)

The distributed task system as recited in claim 2, wherein the at least one task server further comprises:
a task strategy unit connected with the self-balancing unit and used to trigger the self-balancing unit to send the task request.

### Claim 5 (depends on 2)

The distributed task system as recited in claim 2, wherein the self-balancing unit is configured to check whether the task server is busy when the feedback indicates that the task server can execute the task, and to further send a task check message to the task transaction server if the task server is busy, and wherein the task transaction server is configured to subsequently check if another task server is requesting for the task, and to return result of checking to the self-balancing unit.

### Claim 6 (depends on 5)

The distributed task system as recited in claim 5, wherein the system is adopted to determine a current number of task threads and a current number of long-duration task threads, and wherein the task server is considered to be busy when at least one the following conditions is met:
the current number of threads is greater than or equal to a preset bound for number of threads; and
the current number of long-duration threads is greater than or equal to a preset bound for number of long-duration threads.

### Claim 7 (depends on 5)

The distributed task system as recited in claim 5, wherein the system is adopted to determine an execution encumbrance of the task server at least partially based on current number of task threads and current number of long-duration task threads, and wherein the task server is considered to be busy when the execution encumbrance of the task server has a value greater than or equal to a preset threshold.

### Claim 8 (depends on 1)

The distributed task system as recited in claim 1, wherein the system is adapted to obtain execution data of the task, group the execution data into a plurality of data groups according to the execution data and the task's execution times, and process each data group.

### Claim 9 (depends on 1)

The distributed task system as recited in claim 1, wherein the task transaction server and the at least one task server of are connected through a network.

### Claim 10 (depends on 1)

The distributed task system as recited in claim 1, wherein the task transaction server and the at least one task server are part of an e-commerce system used for processing online commercial transactions.

### Claim 11

A distributed task system, comprising:
a task transaction server; and
at least one task server having a self-balancing unit and a task execution unit, wherein the task transaction server is adapted to receive from the task server a task request, to record an execution condition of the task requested, and to provide a feedback to the task server, and the at least one task server is adapted to send the task request to the task transaction server and to receive the feedback from the task transaction server, and wherein the self-balancing unit of the task server is adapted to process the feedback from the task transaction server and determine whether an execution criterion is met for the task server to execute the task requested, and if the execution criterion is met, trigger the task execution unit to execute the task requested.

### Claim 12 (depends on 11)

The distributed task system as recited in claim 11, wherein the self-balancing unit determines whether the execution criterion is met by analyzing the feedback from the task transaction server and examining an operation condition of the task server.

### Claim 13

A distributed task management method, comprising:
sending a task request from a task server to a task transaction server;
determining by the task transaction server whether the task server can execute the requested task based on an execution condition of the requested task, and sending a result thus determined to the task server; and
executing the task by the task server if it is determined by the task transaction server that the task server can execute the requested task.

### Claim 14 (depends on 13)

The distributed task management method as recited in claim 13, wherein the task server executes the task if a feedback is received from the task transaction server within a first time interval after sending the task request to task transaction server by task server, or otherwise sends another task request to the task transaction server.

### Claim 15 (depends on 14)

The distributed task management method as recited in claim 14, further comprising:
abandoning the task request if no feedback is received from the task transaction server within a second time interval.

### Claim 16 (depends on 13)

The distributed task management method as recited in claim 13, wherein determining by the task transaction server whether the task server can execute the requested task comprises:
checking by the task transaction server the execution condition of requested task; and
concluding that the task server cannot execute the task if the requested task is being presently executed by another task server, or otherwise concluding that the task server can execute the task.

### Claim 17 (depends on 13)

The distributed task management method as recited in claim 13, wherein the method further comprising:
checking whether the task server is busy;
sending a task check message to the task transaction server if the task server is busy;
checking by the task transaction server whether another task server is requesting for the task, and sending check result to the task server; and
abandoning the task request by the task server if the check result is affirmative, or otherwise executing the task.

### Claim 18 (depends on 17)

The distributed task management method as recited in claim 17, wherein checking whether the task server is busy comprises:
determining whether current number of threads is greater than or equal to a preset bound, and if affirmative, concluding that the task server is busy.

### Claim 19 (depends on 17)

The distributed task management method as recited in claim 17, wherein checking whether the task server is busy comprises:
determining whether current number of long-duration threads is greater than or equal to a preset bound, and if affirmative, concluding that the task server is busy.

### Claim 20 (depends on 17)

The distributed task management method as recited in claim 17, wherein checking whether the task server is busy comprises:
computing an execution encumbrance value of the task server based on CPU utilization rate, current number of threads and current number of long-duration threads; and
concluding that the task server is busy if the execution encumbrance value is at or above a preset threshold.

### Claim 21 (depends on 13)

The distributed task management method as recited in claim 13, wherein executing the task by the task server comprises:
obtaining data of the task;
grouping the data into a plurality of data groups according to amount of data that needs to be processed and data processing time; and
processing each data group.
