WO2014173339A1

WO2014173339A1 - Task scheduling service system and method

Info

Publication number: WO2014173339A1
Application number: PCT/CN2014/077702
Authority: WO
Inventors: 刘海波; 赖天建; 漆动波; 王祺
Original assignee: 中兴通讯股份有限公司
Priority date: 2013-08-07
Filing date: 2014-05-16
Publication date: 2014-10-30
Also published as: CN104346215A

Abstract

A task scheduling service system and method, said task scheduling service system comprising: a task calling end module, configured to initiate task scheduling requests; a service interface component module, configured to create a task in accordance with the task scheduling request, and to create a corresponding scheduling task record in a database; a task scanning component module, configured to scan the scheduling request record, to calculate task priority in accordance with a task scheduling priority algorithm, and to then enter the task into a corresponding priority queue; a task queue component module, configured to select, from the priority queue and in accordance with a priority queue element queue algorithm, a current task to be executed; a task execution module, configured to execute the current task to be executed. The task scheduling service system and method provided in the embodiment of the present invention cause task scheduling to distinguish between priority levels, thus increasing task scheduling service execution efficiency and simultaneously improving the user's differentiated experience in regard to task scheduling services having different priority levels.

Description

Task scheduling service system and method

Technical field

The present invention relates to the field of communications, and more particularly to a task scheduling service system and method. Background technique

In recent years, electronic communication has spread our daily lives. The application of voice mail is also very extensive. At present, the voice mailbox includes new mail phone notification, guest timed message, outbound call fetching, schedule reminder and other timed tasks. Some of these timed tasks are business operations specified by the user at the date and time, and some are caused by the user's specified date and time business operations. Subsequent processing, some of which are implemented by internal task mechanisms for implementing a function.

The invocation of these timing tasks is based on the internal processing steps of the business application layer, which increases the complexity of the business application module, resulting in strong coupling between the business application layer and the data structure, data distribution and module relationships within the system, and Continuous expansion and change will result in repeated development tests of the same function, reducing R&D efficiency, and making it difficult to ensure product quality and performance. Summary of the invention

The technical problem to be solved by the present invention is to provide a task scheduling service system and method. The processing logic of the business application layer can be simplified, and the execution efficiency and user experience of the task scheduling service can be improved.

To solve the above technical problem, an embodiment of the present invention provides a task scheduling service system and method, where the task scheduling service system includes:

The task invoking module is set to: initiate a task scheduling request;

The service interface component module is configured to: create a task according to the task scheduling request, and create a corresponding scheduling task record in the database;

The task scanning component module is configured to: scan the scheduled task record, calculate the task priority according to the task scheduling priority algorithm, and then put the task into the corresponding priority queue;

The task queue component module is configured to: select a task to be executed in the priority queue according to a priority queue element dequeue algorithm; The task execution module is set to: execute the current task to be executed.

The task scheduling service system further includes:

The task statistics module is configured to: obtain and count the total number of task scheduling requests initiated by the task invoking module in the scheduling period, and the number of tasks successfully executed by the task execution module and the number of tasks that fail to execute.

The WEB management module is configured to: configure the scheduling rules of the task, system parameters, and query the processing progress of the current task.

The service interface component module is configured to: when creating a task, calculate a startup time and a task number for the task according to the task scheduling algorithm.

The service interface component module is further configured to: receive a task execution result report of the task execution module, and respond to the task execution module to perform task retry or perform task return processing on the task that fails to perform.

The service interface component module is further configured to: when the task is retried, calculate a restart time of the task according to the task scheduling rule calculation algorithm.

The task scheduling rule calculation algorithm includes: a current system time when the task is created

+Maximum startup delay duration * (The sum of the proportions of all scheduling priorities greater than the priority of the new task scheduling priority).

The service interface component module is configured to: when performing task receipt processing, return the task receipt information to the task invoking module according to the information of the task.

The service interface component module is further configured to: move the task that ends the execution to the history table for saving.

The task scanning component module is further configured to: when scanning a timeout and expired task, report the task result to the service interface component module, so that the service interface component module performs a task retry or performs a task for timeout and expired tasks. Receipt processing of tasks.

The task queue component module is further configured to: when the startup task is abnormal, report the task result to the service interface component module, so that the service interface component module performs task retry or task receipt processing on the task that initiates the abnormality. .

The task scheduling priority algorithm includes: a system level priority corresponding to a task type of the new task * 2 + a task priority flag + 1. The priority queue element outbound algorithm includes: The task queue component module extracts the task processing from each task queue one by one according to the scheduling priority of each queue, first processes the tasks in the high priority queue, and then processes the low priority. The tasks in the level queue, then loop.

An embodiment of the present invention further provides a task scheduling method, including:

Initiating a task scheduling request;

Creating a task according to the task scheduling request, and creating a corresponding scheduling task record in the database;

Scanning the scheduled task record, calculating the task priority according to the task scheduling priority algorithm, and then placing the task into the corresponding priority queue;

Selecting a task to be executed in the priority queue according to a priority queue element dequeue algorithm;

Perform the task you want to perform now.

The task scheduling method further includes: calculating, when the task is retried, a restart time of the task according to the task rule algorithm.

The beneficial effects of the above technical solutions of the embodiments of the present invention are as follows:

The foregoing embodiment of the present invention initiates a task scheduling request by using a task invoking module; the service interface component module creates a task according to the task scheduling request, and creates a corresponding scheduling task record in the database; the task scanning component module processes the scheduling task The record is scanned, the task priority is calculated according to the task scheduling priority algorithm, and the task is placed in the corresponding priority queue; the task queue component module selects the current execution in the priority queue according to the priority queue element dequeue algorithm. The task execution module executes the task to be executed currently; simplifies the processing logic of the business application layer, improves the execution efficiency of the task scheduling service, and enhances the differentiated experience of external users for different priority task scheduling services. . BRIEF abstract

1 is a structural diagram of a task scheduling service system according to an embodiment of the present invention;

FIG. 2 is a flowchart of a method for implementing task scheduling according to an embodiment of the present invention; FIG. 5 is a flowchart of a method for scheduling a task according to an embodiment of the present invention. Preferred embodiment of the invention

The detailed description will be made below in conjunction with the accompanying drawings and specific embodiments.

The call of the present invention for the existing timing tasks is based on the problem that the business application layer performs internal processing and is inefficient, and provides a task scheduling service system and method.

As shown in FIG. 1 , an embodiment of the present invention provides a task scheduling service system, including: a task invoking end module 101, configured to initiate a task scheduling request;

The service interface component module 102 is configured to create a task according to the task scheduling request, and create a corresponding scheduling task record in the database;

The task scanning component module 103 is configured to scan the scheduled task record, calculate a task priority according to a task scheduling priority algorithm, and then put the task into a corresponding priority queue;

The task queue component module 104 is configured to select a task to be executed in the priority queue according to a priority queue element dequeue algorithm;

The task execution module 105 is configured to execute the task to be executed currently.

The foregoing embodiment of the present invention initiates a task scheduling request by using a task invoking module; the service interface component module creates a task according to the task scheduling request, and creates a corresponding scheduling task record in the database; the task scanning component module processes the scheduling task The record is scanned, the task priority is calculated according to the task scheduling priority algorithm, and the task is placed in the corresponding priority queue; the task queue component module selects the current execution in the priority queue according to the priority queue element dequeue algorithm. The task execution module executes the current task to be executed; not only simplifies the processing logic of the business application layer, but also improves the execution efficiency of the task scheduling service, and also enhances the differentiated experience of external users for different priority task scheduling services. .

As shown in FIG. 1 , the embodiment of the task scheduling service system of the present invention may further include:

The task statistics module 106 is configured to obtain a total number of task scheduling requests initiated by the task invoking module in the scheduling period, a number of tasks successfully executed by the task execution module, and a number of tasks that fail to execute.

The WEB management module 107 is configured to configure scheduling rules of the task, system parameters, and query processing progress of the current task. When the task is created, the service interface component module 102 is specifically configured to: calculate a startup time and a task number for the task according to the task scheduling algorithm.

The service interface component module 102 is further configured to: receive a task execution result report of the task execution module 105, respond to the task execution module 105, perform task retry on the task that fails to perform, or perform a receipt process of the task.

The service interface component module 102 is further configured to: when the task is retried, calculate a restart time of the task according to the task scheduling rule calculation algorithm.

The task scheduling rule calculation algorithm specifically refers to: the current system time when the task is created + the maximum startup delay duration * (the sum of the task counts of all the scheduling priorities greater than the new task scheduling priority).

When the service interface component module 102 performs the task receipt processing, it is specifically configured to: return the task receipt information to the task invoking end module 101 according to the information of the task.

The service interface component module 102 is further configured to: move the task that ends the execution to a history table to save.

The task scanning component module 103 is further configured to: when scanning a timeout and expired task, report the task result to the service interface component module 102, so that the service interface component module 102 performs a task retry on the timeout and expired tasks or Carry out the receipt processing of the task.

The task queue component module 104 is further configured to: when the startup task has an abnormality, report the task result to the service interface component module 102, so that the service interface component module 102 performs a task retry or a task receipt for the task that initiates the abnormality. deal with.

The task scheduling priority algorithm specifically refers to: a system level priority corresponding to a task type of a new task * 2 + task priority flag + 1.

The priority queue element dequeue algorithm specifically refers to: the task queue component module 104 extracts the task processing from each task queue one by one according to the scheduling priority of each queue, and first processes the tasks in the high priority queue, and the processing is low. The tasks in the priority queue are then looped.

The following describes how the above system is implemented in conjunction with FIG. 2 to FIG.

As shown in FIG. 2, a specific flowchart of the method for implementing the task scheduling service system according to the present invention mainly includes the following steps:

Step 201: The WEB management module 107 is mainly responsible for configuring task scheduling rules, including tasks. The system-level priority, the forbidden scheduling time, the retry interval, and the maximum number of retries, and the like; Step 202, the task invoking module 101 sends a task scheduling request to the service interface component module 102, and the type of the current scheduling task needs to be specified in the request message. Whether the field of the scheduling result receipt is required, and the scheduling task startup time field is optional;

Step 203: After receiving the task scheduling request, the service interface component module 102 calculates the startup time of the task according to the task scheduling rule calculation algorithm, and creates a corresponding task scheduling record in the database; wherein, each scheduling priority in the statistical period is required. The task counter calculates the ratio of the number of tasks for each scheduling priority to the total number of tasks.

Step 204: The task scanning component 103 queries the task that meets the task start condition according to the task scheduling priority algorithm, scans the task scheduling record, and puts the task scheduling record into the corresponding task priority queue; wherein, for example, system level priority The level is 3 (0, 1, 2). Each task type is divided into two types: normal and emergency (0, 1). Therefore, the maximum scheduling priority level of the whole system is 6, but considering the possible extension. The system-level priority number, so the task scheduling service can configure the maximum number of scheduling system-level priorities it supports. The default is 6.

Step 205: The task queue component module 104 calls out the task that meets the requirements in the priority queue according to the queue function of the priority queue element, and sends the task to the task execution module 105. The task priority queue element dequeue algorithm includes: The task queue component takes the task processing from each task queue one by one according to the scheduling priority of each queue, processes the high priority queue first, then processes the low priority queue, and then loops. Each round of the loop takes the elements from the corresponding queue according to the priority values 6, 5, 4, 3, 2, 1;

Step 206, after receiving the task scheduling execution request, the task execution module 105 triggers the intelligent network platform to perform task scheduling, and feeds the execution result to the service interface component module 102;

Step 207, the service interface component module 102 determines whether the scheduling is successful, if successful, proceeds to step 208; if it fails, proceeds to step 209;

Step 208, the service interface component module 102 determines whether the task scheduling request requires a receipt. If the receipt is required, the process proceeds to step 210; if not, the process proceeds to step 214, and the current task scheduling ends, and the service interface component module 102 imports the scheduled task. In the task history table;

Step 209, the service interface component module 102 determines whether the number of retries of the scheduled task has exceeded the maximum number of retries, if yes, returns to step 208; if not, proceeds to step 212, by the service The interface component module 102 calculates the task retry start time for the task scan component module 103 to scan; Step 210, the service interface component module 102 determines whether the receipt execution result is successful to the task call end 101, and if successful, proceeds to step 214, the task Scheduling end, service interface component module

102, the scheduling task is imported into the task history table; if not, the process proceeds to step 213, where the task retry receipt time is calculated by the service interface component module 102 for scanning by the task scanning component module 103.

As shown in FIG. 3, tasks that request scheduling from the service interface component module 102 can be divided into two phases: task execution and task receipt during the creation and execution process:

The first is that the service interface component module 102 creates a task in the task table, and needs to wait for execution after the task is created;

There are three situations that occur while waiting for execution:

When the scheduling is started successfully, the task execution module 105 executes the task;

The task call end 101 actively cancels the task;

Waiting for execution and timeout;

There are usually three outcomes during the execution of the task phase:

If the execution is successful, the task execution module 105 returns that the task execution is successful;

The execution fails, and can be divided into: the task execution module 105 returns the task execution failure; the task queue component module 104 returns the task startup failure; the task in the waiting state times out;

Executing timeout;

Tasks in the task table require a receipt. There are usually four situations in the task receipt phase: Waiting for a receipt, the receipt fails but can be retried;

Successful receipt

The receipt failed, the receipt failed and cannot be retried, or the receipt failed and the maximum number of retries has been exceeded; the receipt timed out, waiting for the receipt to time out;

As shown in FIG. 4, the task statistics module 106 obtains the task counters of the respective scheduling priorities in the scheduling period, and calculates the proportion of the tasks of the respective scheduling priorities to the total number of tasks. In order to avoid the delay of a certain scheduling priority task, the algorithm uses the proportion of the total number of tasks according to each scheduling priority to the total number of tasks of all scheduling priorities to calculate the task start of each scheduling priority. time.

As shown in FIG. 5, an embodiment of the present invention further provides a task scheduling method, including: Step 501: Initiate a task scheduling request.

Step 502: Create a task according to the task scheduling request, and create a corresponding scheduling task record in the database.

Step 503: Scan the scheduled task record, calculate the task priority according to the task scheduling priority algorithm, and then put the task into the corresponding priority queue.

Step 504 selects a task to be executed in the priority queue according to a priority queue element out queue algorithm;

Step 505 performs the task currently to be executed.

Further, the foregoing embodiment may further include:

After the task 506 is successfully executed, the execution result is returned to the task service interface component module 102. Step 507 The service interface component module 102 moves the completed task to the task history table for saving. Further, the task scheduling method further includes: when the task is retried, the service interface component should calculate the restart time of the task according to the task rule algorithm.

Further, the task scheduling method further includes: for the task that ends and requests the receipt, the service interface component should return the task receipt information to the task invoking end according to the task information.

The embodiment of the method is an embodiment corresponding to the above system. All the implementation processes in the above system embodiment are applicable to the embodiment of the method, and the same technical effects can be achieved.

The above is a preferred embodiment of the present invention, and it should be noted that those skilled in the art can also make several improvements and retouchings without departing from the principles of the present invention. It should be considered as the scope of protection of the present invention.

Industrial applicability

The task scheduling service system and method according to the embodiments of the present invention are based on the idea of system layered construction, and the intermediate service layer is responsible for interacting with internal modules to implement scheduling of scheduled tasks. Specifically, first, for external users, the scheduling of tasks is more intelligent and real-time, and supports multiple deployments of the task scheduling system module; functions such as forbidden call period and start-up delay can be set; The difference in priority task scheduling services. Second, for vendor and internal engineering deployment and maintenance, the system configuration is simple and effective, and the operation is convenient; the business application layer processing logic is simplified, and the management is improved. The efficiency of the dispatching service is realized; the WEB interface of parameter configuration is realized, which is convenient for administrators to manage and maintain; the data statistics platform is adopted, and the statistics of system user quantity, performance and stability are more clear and evidence-based. The above embodiments of the present invention not only simplify the processing logic of the service application layer, but also improve the execution efficiency of the task scheduling service, and also enhance the differentiation of external user scheduling services for different priority tasks.

Claims

Claim

1. A task scheduling service system, comprising:

The task invoking module is set to: initiate a task scheduling request;

The task scanning component module is configured to: scan the scheduled task record, calculate the task priority according to the task scheduling priority algorithm, and then put the task into the corresponding priority queue; the task queue component module is set to: according to the priority The level queue element dequeue algorithm selects a task to be executed in the priority queue;

The task execution module is set to: execute the current task to be executed.

The task scheduling service system according to claim 1, further comprising: a task statistics module, configured to: obtain a total number of task scheduling requests initiated by the task invoking module in the scheduling period, and execute the task execution module The number of successful tasks and the number of tasks that failed to execute.

3. The task scheduling service system according to claim 1, further comprising:

The task scheduling service system according to claim 1, wherein the service interface component module is configured to: when creating a task, calculate a startup time and an assignment task number for the task according to the task scheduling algorithm.

The task scheduling service system according to claim 1, wherein the service interface component module is further configured to: receive a task execution result report of the task execution module, and respond to the task execution module to perform a task that fails to execute The task is retried or the receipt processing of the task is performed.

The task scheduling service system according to claim 5, wherein the service interface component module is further configured to: when the task is retried, calculate a restart time of the task according to the task scheduling rule calculation algorithm.

The task scheduling service system according to claim 6, wherein the task scheduling rule calculation algorithm comprises: Current system time + maximum startup delay duration when creating a task * (sum of the number of tasks greater than all scheduling priorities for the new task scheduling priority).

The task scheduling service system according to claim 5, wherein the service interface component module is configured to: when performing task receipt processing, return task receipt information to the task invoking module according to the information of the task.

9. The task scheduling service system according to claim 1, wherein the service interface component module is further configured to: move the task that ends execution to a history table for saving.

10. The task scheduling service system according to claim 1, wherein the task scanning component module is further configured to: report a task result to a service interface component module when scanning a timeout and expired task, to enable the service interface The component module performs task retry or timeout processing of tasks for timed out and expired tasks.

The task scheduling service module according to claim 1, wherein the task queue component module is further configured to: when an abnormality occurs in the startup task, report the task result to the service interface component module, so that the service interface component module Start an abnormal task to perform a task retry or to perform a receipt processing of the task.

The task scheduling service system according to claim 1, wherein the task scheduling priority algorithm comprises:

The system level priority corresponding to the task type of the new task * 2 + task priority flag + 1.

The task scheduling service system according to claim 1, wherein the priority queue element dequeue algorithm comprises:

The task queue component module takes the task processing from each task queue one by one according to the scheduling priority of each queue, processes the tasks in the high priority queue first, processes the tasks in the low priority queue, and then loops.

14. A task scheduling method, comprising:

Initiating a task scheduling request;

Scanning the scheduled task record, calculating the task priority according to the task scheduling priority algorithm, and then placing the task into the corresponding priority queue; Selecting a task to be executed in the priority queue according to a priority queue element dequeue algorithm;

Perform the task you want to perform now.

The task scheduling method according to claim 14, further comprising:

When the task is retried, the task restart time is calculated according to the task rule algorithm.