CN113132456B - Edge cloud cooperative task scheduling method and system based on deadline perception - Google Patents
Edge cloud cooperative task scheduling method and system based on deadline perception Download PDFInfo
- Publication number
- CN113132456B CN113132456B CN202110229562.3A CN202110229562A CN113132456B CN 113132456 B CN113132456 B CN 113132456B CN 202110229562 A CN202110229562 A CN 202110229562A CN 113132456 B CN113132456 B CN 113132456B
- Authority
- CN
- China
- Prior art keywords
- task
- tasks
- server
- deadline
- time
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
- H04L67/1004—Server selection for load balancing
- H04L67/101—Server selection for load balancing based on network conditions
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
- H04L67/1004—Server selection for load balancing
- H04L67/1008—Server selection for load balancing based on parameters of servers, e.g. available memory or workload
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Computer Hardware Design (AREA)
- General Engineering & Computer Science (AREA)
- Computer And Data Communications (AREA)
Abstract
The invention belongs to the technical field of computer application, and discloses a method and a system for scheduling a side cloud cooperative task based on deadline sensing, wherein the method for scheduling the side cloud cooperative task based on deadline sensing comprises the following steps: the data of the terminal equipment are sent to the unified access module; pre-classifying the tasks, and respectively sending the tasks to different destinations; the task scheduling center extracts the task type and the task deadline and inquires a server set Stargetl with corresponding service; the task is planned to be distributed to the servers in the start 1, whether the time delay meets the deadline or not is calculated, and the task is distributed to the servers meeting the condition; and when the task reaches the target server, generating a new task execution queue. According to the invention, the unified access module receives the task data generated by the terminal equipment, presorts the tasks and then sends the tasks to the remote cloud center or the task scheduling center, so that the load balance of the edge cloud system is ensured, and the resource utilization rate of the whole edge cloud architecture is improved.
Description
Technical Field
The invention belongs to the technical field of computer application, and particularly relates to a method and a system for scheduling edge cloud cooperative tasks based on deadline awareness.
Background
At present, with the rapid development of technologies such as internet of things and mobile internet, the data volume of terminal equipment is increasing in a blowout manner. The continuous and rapid increase of the data volume promotes the evolution of the whole computing mode, and simultaneously, higher requirements are put on data storage and processing technologies, especially real-time processing and intelligent analysis of network edge side services. Thanks to cloud computing technology, more and more mobile applications and internet of things devices can handle local compute-intensive tasks (such as image processing tasks) using rich computing resources of remote cloud centers. Theoretically, such task offloading can significantly extend the capabilities of the mobile device, but the high communication latency of the device to the remote cloud can severely degrade the quality of service. Due to inherent disadvantages of traditional cloud computing, such as insufficient real-time performance, insufficient bandwidth, large energy consumption, lack of mobility and the like, the key technology of the traditional cloud computing cannot efficiently process massive data and computing tasks generated by edge devices. Therefore, edge computing is produced as a new computing mode, and by extending computing capacity from a cloud data center to a network edge, internet data volume transmission is reduced, so that time delay is reduced, bandwidth is saved, and usability and expandability of the whole system are improved.
Many computing jobs are shifted from the remote cloud to the edge server processing, especially for latency sensitive tasks, which typically require a fast response speed. Due to the limited storage resources and computing power of edge servers, efficient task scheduling algorithms are needed to ensure that tasks generated by edge devices are executed in a reasonable order where appropriate. The current common task scheduling algorithm, such as a first-come first-serve scheduling algorithm, executes tasks according to the arrival sequence of the tasks, and does not consider the priority level of the tasks so that the delay-sensitive tasks cannot be processed in time; the shortest job priority algorithm prioritizes the task with the shortest completion time, and the algorithm performance starts to degrade when the number of large tasks exceeds that of small tasks. Moreover, these basic task scheduling algorithms are no longer applicable in complex "end-edge-cloud" computing models. Some task scheduling algorithms based on end edge cloud cooperation unload terminal tasks to edge computing nodes through cooperation, allow an edge cloud with limited resources to unload part or all tasks to a remote cloud to achieve resource cooperative sharing, and the technology mainly focuses on an unloading strategy from the edge cloud to the remote cloud, but neglects resource utilization rate and load balance of a plurality of servers in an edge cloud system. Related research also suggests a method for scheduling tasks of the internet of things using data mining, using a modified Apriori association rule algorithm to classify tasks, and then applying a new algorithm named TSFC (task scheduling in fog computing) to perform scheduling. This approach is purely time-based, but ignores the availability of network and fog node resources.
Through the above analysis, the problems and defects of the prior art are as follows:
(1) in the prior art, load balance of servers in the edge cloud system is neglected, and the utilization rate of the whole resources is not high.
(2) In the task scheduling algorithm, the network resources, the computing resources, the storage resources, the task completion deadline, the overall service quality and other factors of the server are not comprehensively considered, and most of the technologies only consider part of the influencing factors, so that the optimal scheduling result cannot be achieved.
The difficulty in solving the above problems and defects is: in a system architecture of cooperative operation of an edge cloud and a remote cloud center, a plurality of influence factors such as computing resources, storage resources, network conditions and the like of terminal computing task scheduling are considered comprehensively and used as important parameters of a scheduling method, and then a reasonable scheduling model is established to achieve an optimal scheduling effect, so that the computing tasks are completed efficiently, the edge cloud load is balanced, and great difficulty is achieved.
The significance of solving the problems and the defects is as follows: with the increasing demand of the delay sensitive service, the number of nodes and resources on the edge cloud will increase continuously, and it is very important to reasonably distribute the tasks generated by the terminal device to the edge server or the cloud center, so that the utilization rate of the resources of the cloud center and the edge cloud can be improved, and the improvement of the service quality is greatly facilitated. The invention innovatively provides a side cloud cooperative task scheduling method, which realizes the on-line scheduling of real-time tasks, ensures the load balance of an edge cloud server, and ensures the on-time processing completion of delay sensitive tasks.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a method and a system for scheduling a side cloud cooperative task based on deadline sensing. The invention provides a novel edge cloud cooperative computing model architecture and an edge cloud cooperative task scheduling method based on the edge cloud cooperative computing model architecture, the algorithm can efficiently schedule terminal tasks on line, the load balance of an edge cloud system is ensured, and the resource utilization rate of the whole edge cloud architecture is improved.
The invention is realized in such a way that a method for scheduling edge cloud cooperative tasks based on deadline awareness comprises the following steps:
firstly, terminal equipment data are sent to a uniform access module, the access module performs uniform processing on heterogeneous data, the difference of the data is eliminated, task data are identified, and scheduling of subsequent calculation tasks is facilitated;
the tasks are pre-classified and are respectively sent to different destinations, and the tasks are pre-classified according to prior experience, so that the task processing efficiency is improved, and the pressure of a task scheduling center is reduced;
step three, the task scheduling center extracts the task type and the task deadline, inquires a server set Stargetl with corresponding service, and ensures that the task is accurately scheduled to a service node capable of executing processing;
step four, the task is planned to be distributed to a server in the Start 1, whether the time delay meets the deadline time is calculated, the task is distributed to the server meeting the condition, and the task is guaranteed to be dispatched to the server capable of meeting the deadline time through the pre-estimation of the transmission time delay, the waiting time delay and the calculation time;
and step five, the task reaches the target server, a new task execution queue is generated, and the task is waited to be executed in the queue and can be finished at the fastest speed under the condition of meeting the requirement of the deadline.
Further, the second step specifically comprises the following steps: according to the data type and the service characteristics of the terminal, tasks generated by the terminal equipment are pre-classified, the tasks with high calculation cost and insensitive time delay are sent to a remote cloud center for processing in a unified access module, and other tasks are sent to a task scheduling center.
Further, in the third step, the task scheduling center processes the received task data, extracts key information such as task type and task completion deadline, judges a service required by processing the task according to the task type, and then queries an edge server installed with the service from a service registry to obtain a server set S meeting conditions target1 。
Further, the task is assigned to the obtained S target1 Each server in the system calculates task transmission delay, calculates delay and allocates the task to cause the increased waiting time of other tasks, and the set of servers meeting the task cut-off time is recorded as S target2 ；
The transmission delay is calculated according to the size of the task data volume and the real-time network condition from the task scheduling center to the target server, and the calculated delay comprises waiting delay and execution time; the sum of the transmission delay and the calculation delay is smaller than the task deadline so as to meet the requirement of the task deadline.
Further, the obtained S target2 If the task is empty, namely no edge server capable of meeting the task deadline exists, discarding the task and sending a notification;
if S is target2 If only one server meets the condition, the task is distributed to the server for processing;
if S is target2 If there are multiple servers that meet the condition, then the server that has the shortest latency for other tasks due to the task being assigned is selected.
Further, the task arrives at the obtained target server s target Executing a task scheduling algorithm in the single server:
if s is target If no other task waits for processing, the task is directly processed; if s is target The method comprises the steps that a task waiting queue is arranged, the task is firstly planned to be placed at the tail of the queue, the execution time of all the preposed tasks is the waiting time delay of the current position of the task, whether the sum of the waiting time delay and the execution time of the task meets the task deadline time or not is judged, and if yes, the task waits to be executed at the tail of the queue;
if the time exceeds the deadline time, the task is moved forward in sequence, the influence of the insertion of the task on the waiting time delay of other tasks is calculated, and if the other tasks are overtime, the task is discarded, namely the server cannot complete the task within the specified deadline time;
if no other task is overtime, calculating the task waiting time at the current position, judging whether the task waiting time meets the deadline, and if so, inserting the task into the current position to generate a new task execution queue; the insertion of the newly arrived task causes the delay of the execution sequence of other tasks and the increase of the waiting time.
Further, the task execution time is calculated as follows:
the server records the data size and the execution time of executing a certain type of task each time within the latest time t, and the unit execution time of processing the task most recently is calculated by dividing the data size by the execution time;
and dividing the size of the task data quantity by the latest execution time of the task unit to calculate the task execution time.
Further, the steps of creating and updating the service registry are as follows:
the remote cloud center creates and maintains a service registry, the registry mainly records a currently installed service list on each edge server, and the remote cloud is identified with the edge servers of all services;
whenever the edge server has an operation of installing or uninstalling the service, the operation is updated to a service registry of the remote cloud center;
and after the service registry of the remote cloud center is updated, the service registry of the task scheduling center can be automatically synchronized.
Another object of the present invention is to provide a side cloud cooperative task scheduling system for implementing the side cloud cooperative task scheduling method, where the side cloud cooperative task scheduling system includes:
the unified access module is used for uniformly receiving and pre-classifying the calculation task data generated by the terminal equipment;
the task scheduling center receives and extracts the task key information and stores relevant state information of the server, including real-time network channel conditions, services installed by the server and the like;
and the edge server cluster and the remote cloud determine and process the task execution sequence.
By combining all the technical schemes, the invention has the advantages and positive effects that: a traditional end-edge-cloud computing model is optimized, the coordination capability of a server in an edge cloud is expanded, a method for edge-cloud coordination task scheduling based on deadline sensing is provided, task data generated by terminal equipment is received by a unified access module, the tasks are pre-classified and then sent to a remote cloud center or a task scheduling center, the task scheduling center is used for scheduling the tasks on line by comprehensively considering factors such as server network conditions and computing resources based on the task deadline, the load balance of an edge cloud system is guaranteed, and the resource utilization rate of the whole edge cloud architecture is improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments of the present application will be briefly described below, and it is obvious that the drawings described below are only some embodiments of the present application, and it is obvious for those skilled in the art that other drawings can be obtained from the drawings without creative efforts.
Fig. 1 is a flowchart of a method for scheduling edge cloud cooperative tasks according to an embodiment of the present invention.
Fig. 2 is an architecture diagram of an edge cloud collaborative computing model according to an embodiment of the present invention.
Fig. 3 is an architecture diagram of an "end-edge-cloud" computing model provided in an embodiment of the present invention.
Fig. 4 is a flowchart of a task scheduling method in a single server node according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Aiming at the problems in the prior art, the invention provides a method and a system for scheduling a side cloud cooperative task based on deadline sensing, and the invention is described in detail below with reference to the accompanying drawings.
As shown in fig. 1, a method for scheduling a cloud-edge cooperative task based on deadline awareness according to an embodiment of the present invention includes:
s101: and the terminal equipment data is sent to the unified access module.
S102: and pre-classifying the tasks, and respectively sending the tasks to different destinations.
S103: and the task scheduling center extracts the task type and the task deadline and inquires a server set Stargetl provided with the corresponding service.
S104: and (3) scheduling the tasks to be distributed to the servers in the start 1, calculating whether the time delay meets the deadline, and distributing the tasks to the servers meeting the condition.
S105: and when the task reaches the target server, generating a new task execution queue.
Persons of ordinary skill in the art may also use other steps to implement the deadline-aware-based edge cloud collaborative task scheduling method provided by the present invention, and the deadline-aware-based edge cloud collaborative task scheduling method provided by the present invention in fig. 1 is only a specific embodiment.
In S102 provided by the embodiment of the present invention, the specific process is: according to the data type and the service characteristics of the terminal, tasks generated by the terminal equipment are pre-classified, the tasks with high calculation cost and insensitive time delay are sent to a remote cloud center for processing in a unified access module, and other tasks are sent to a task scheduling center.
In step S103 provided by the embodiment of the present invention, the task scheduling center processes the received task data, extracts key information such as a task type and a task completion deadline, determines a service required for processing the task according to the task type, and then queries the edge server in which the service is installed from the service registry to obtain a server set S that satisfies a condition target1 。
Fitting tasks to the resulting S target1 Each server in the system calculates task transmission delay, calculates delay and allocates the task to cause the increased waiting time of other tasks, and the set of servers meeting the task cut-off time is recorded as S target2 . The transmission delay is calculated according to the size of the task data volume and the real-time network condition from the task scheduling center to the target server, and the calculation delay comprises waiting delay and executionAnd (3) removing the solvent. The requirement that the transmission delay plus the calculation delay is smaller than the task deadline is met.
If S is obtained target2 If the task is empty, namely no edge server which can meet the task deadline exists, the task is discarded and a notification is sent out; if S is target2 If only one server meets the condition, the task is distributed to the server for processing; if S is target2 If there are multiple servers that meet the condition, then the server that has the shortest latency for other tasks due to the task being assigned is selected.
Target server S obtained by task arrival target Executing a task scheduling algorithm in the single server: if S is target If no other task waits for processing, directly processing the task; if S is target The method comprises the steps that a task waiting queue is arranged, the task is firstly planned to be placed at the tail of the queue, the execution time of all the preposed tasks is the waiting time delay of the current position of the task, whether the sum of the waiting time delay and the execution time of the task meets the task deadline time or not is judged, and if yes, the task waits to be executed at the tail of the queue; if the time exceeds the deadline time, the task is moved forward in sequence, the influence of the insertion of the task on the waiting time delay of other tasks is calculated, and if the other tasks are overtime, the task is discarded, namely the server cannot complete the task within the specified deadline time; if no other task is overtime, the task waiting time at the current position is calculated, whether the deadline time is met is judged, and if the deadline time is met, the task is inserted into the current position to generate a new task execution queue. The insertion of the newly arrived task causes the delay of the execution sequence of other tasks and the increase of the waiting time.
The steps of creating and updating the service registry are as follows:
the remote cloud center creates and maintains a service registry that primarily records a list of currently installed services on each edge server, a particular, remote cloud being identified as an edge server that has all services installed.
Whenever an edge server has an operation to install or uninstall a service, the operation is updated to the service registry of the remote cloud center.
And after the service registry of the remote cloud center is updated, the service registry of the task scheduling center can be automatically synchronized.
The task execution time calculation steps are as follows:
the server records the data size and the execution time of executing a certain type of task each time within the latest time t, and the unit execution time of processing the task most recently is calculated by dividing the data size by the execution time;
and dividing the size of the task data quantity by the latest execution time of the task unit to calculate the task execution time.
The edge cloud cooperative task scheduling system provided by the embodiment of the invention comprises:
the unified access module is used for uniformly receiving and pre-classifying the calculation task data generated by the terminal equipment;
the task scheduling center receives and extracts the task key information and stores relevant state information of the server, including real-time network channel conditions, services installed by the server and the like;
and the edge server cluster and the remote cloud determine and process the task execution sequence.
The technical solution of the present invention is further described with reference to the following specific examples.
The invention is realized in such a way, and provides a deadline awareness-based edge cloud cooperative task scheduling method, which comprises the following steps:
step one, on the basis of a traditional 'end-edge-cloud' computing model, the coordination capacity between edge servers is expanded, a unified access module and a task scheduling center are arranged in an edge cloud, and a new edge cloud coordination computing model is designed, as shown in fig. 2.
And step two, sending the data and the calculation task generated by the terminal equipment to the unified access module.
And step three, pre-classifying tasks generated by the terminal equipment according to the type and the service characteristics of the terminal data, sending the tasks with high calculation overhead and insensitive time delay to a remote cloud center for processing in a unified access module, and sending the rest tasks to a task scheduling center.
TABLE 1 task Pre-Classification
Step four, the task scheduling center processes the received task data, extracts key information such as task types and task completion deadline and the like, judges services required by processing the task according to the task types, and then inquires an edge server provided with the services from a service registry to obtain a server set S meeting conditions target1 。
Step five, the task is planned to be distributed to the S obtained in the step four target1 Each server in the system calculates task transmission delay, calculates delay and allocates the task to cause the increased waiting time of other tasks, and the set of servers meeting the task ending time is recorded as S target2 . The transmission delay is calculated according to the size of the task data volume and the real-time network condition from the task scheduling center to the target server, and the calculation delay comprises waiting delay and execution time. The requirement that the transmission delay plus the calculation delay is smaller than the task deadline is met.
Step six, if S obtained in step five target2 If the task is empty, namely no edge server which can meet the task deadline exists, the task is discarded and a notification is sent out; if S is target2 If only one server meets the condition, the task is distributed to the server for processing; if S is target2 If there are multiple servers that meet the condition, then the server that has the shortest latency for other tasks due to the task being assigned is selected.
Step seven, the task reaches the target server S obtained in the step six target Executing a task scheduling algorithm in the single server: if s is target If no other task waits for processing, the task is directly processed; if s is target The method comprises the steps that a task waiting queue is arranged, the task is firstly planned to be placed at the tail of the queue, the execution time of all the preposed tasks is the waiting time delay of the current position of the task, whether the sum of the waiting time delay and the execution time of the task meets the task deadline time or not is judged, and if yes, the task waits to be executed at the tail of the queue; if the time exceeds the deadline time, the task is moved forward in sequence, the influence of the insertion of the task on the waiting time delay of other tasks is calculated, and if the other tasks are overtime, the task is discarded, namely the server cannot complete the task within the specified deadline time; if no other task is overtime, the task waiting time at the current position is calculated, whether the deadline is met is judged, and if the deadline is met, the task is inserted into the current position to generate a new task execution queue. The insertion of the newly arrived task causes the delay of the execution sequence of other tasks and the increase of the waiting time.
The steps of creating and updating the service registry are as follows:
the remote cloud center creates and maintains a service registry that primarily records a list of services currently installed on each edge server, a particular, remote cloud being identified edge servers that have all services installed.
Whenever the edge server has an operation of installing or uninstalling the service, the operation is updated to a service registry of the remote cloud center;
and after the service registry of the remote cloud center is updated, the service registry of the task scheduling center can be automatically synchronized.
The task execution time calculation steps are as follows:
the server records the data size and the execution time of executing a certain type of task each time within the latest time t, and calculates the unit execution time of processing the task most recently by dividing the data size by the execution time.
And dividing the size of the task data quantity by the latest execution time of the task unit to calculate the task execution time.
The edge cloud cooperative task scheduling system provided by the embodiment of the invention comprises:
and the unified access module is used for uniformly receiving and pre-classifying the calculation task data generated by the terminal equipment.
And the task scheduling center receives and extracts the task key information and stores the relevant state information of the server, including the real-time network channel condition, the service installed by the server and the like.
And the edge server cluster and the remote cloud determine and process the task execution sequence.
It should be noted that the embodiments of the present invention can be realized by hardware, software, or a combination of software and hardware. The hardware portion may be implemented using dedicated logic; the software portions may be stored in a memory and executed by a suitable instruction execution system, such as a microprocessor or specially designed hardware. Those skilled in the art will appreciate that the apparatus and methods described above may be implemented using computer executable instructions and/or embodied in processor control code, such code being provided on a carrier medium such as a disk, CD-or DVD-ROM, programmable memory such as read only memory (firmware), or a data carrier such as an optical or electronic signal carrier, for example. The apparatus of the present invention and its modules may be implemented by hardware circuits such as very large scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field programmable gate arrays, programmable logic devices, or software executed by various types of processors, or a combination of hardware circuits and software, e.g., firmware.
The above description is only for the purpose of illustrating the embodiments of the present invention, and the scope of the present invention should not be limited thereto, and any modifications, equivalents and improvements made by those skilled in the art within the technical scope of the present invention as disclosed in the present invention should be covered by the scope of the present invention.
Claims (4)
1. A method for scheduling edge cloud cooperative tasks based on deadline awareness is characterized by comprising the following steps:
the data of the terminal equipment are sent to the unified access module;
pre-classifying the tasks, and respectively sending the tasks to different destinations;
the task scheduling center extracts the task type and the task deadline and inquires a server set S provided with corresponding services target1 ；
To assign tasks to S target1 The server calculates whether the time delay meets the deadline time, and distributes the task to the server meeting the condition;
when the task reaches the target server, generating a new task execution queue;
the specific process of pre-classifying the tasks and respectively sending the tasks to different destinations is as follows: according to the type and the service characteristics of terminal data, tasks generated by terminal equipment are pre-classified, the tasks with high calculation overhead and insensitive time delay are sent to a remote cloud center for processing in a unified access module, and other tasks are sent to a task scheduling center;
the task scheduling center processes the received task data, extracts key information of task types and task completion deadline, judges services required by processing the task according to the task types, and then queries an edge server provided with the services from a service registry to obtain a server set S meeting conditions target1 ；
The task is planned to be allocated to the obtained S target1 Each server in the system calculates task transmission delay, calculates delay and allocates the task to cause the increased waiting time of other tasks, and the set of servers meeting the task deadline is recorded as S target2 ；
The transmission delay is calculated according to the size of the task data volume and the real-time network condition from the task scheduling center to the target server, and the calculated delay comprises waiting delay and execution time; the sum of the transmission delay and the calculated delay is smaller than the task deadline so as to meet the task deadline requirement;
if obtained S target2 If the task is empty, namely no edge server which can meet the task deadline exists, the task is discarded and a notification is sent out;
if S is target2 If only one server meets the condition, the task is distributed to the server for processing;
if S is target2 If a plurality of servers meeting the conditions exist in the system, selecting the server which causes the shortest waiting time for other tasks due to the task distribution;
target server s obtained by the task arrival target Executing a task scheduling algorithm in the single server: if s is target If no other task waits for processing, directly processing the task; if s is target The method comprises the steps that a task waiting queue is arranged, the task is firstly planned to be placed at the tail of the queue, the execution time of all the preposed tasks is the waiting time delay of the current position of the task, whether the sum of the waiting time delay and the execution time of the task meets the task deadline time or not is judged, and if yes, the task waits to be executed at the tail of the queue;
if the time exceeds the deadline time, the task is moved forward in sequence, the influence of the insertion of the task on the waiting time delay of other tasks is calculated, and if the other tasks are overtime, the task is discarded, namely the server cannot complete the task within the specified deadline time;
if no other task is overtime, calculating the task waiting time at the current position, judging whether the task waiting time meets the deadline, and if so, inserting the task into the current position to generate a new task execution queue; the insertion of the newly arrived task causes delay of the execution sequence of other tasks and further increase of waiting time;
the task execution time calculation steps are as follows:
the server records the data size and the execution time of executing a certain type of task each time within the latest time t, and the unit execution time of processing the task most recently is calculated by dividing the data size by the execution time;
and dividing the size of the task data quantity by the latest execution time of the task unit to calculate the task execution time.
2. The edge cloud collaborative task scheduling method according to claim 1, wherein the steps of creating and updating the service registry are as follows: the remote cloud center creates and maintains a service registry, the registry mainly records a currently installed service list on each edge server, and the remote cloud is identified with the edge servers of all services;
whenever the edge server has an operation of installing or uninstalling the service, the operation is updated to a service registry of the remote cloud center;
and after the service registry of the remote cloud center is updated, the service registry of the task scheduling center can be automatically synchronized.
3. A terminal device, wherein the terminal device uses computer executable instructions and/or is included in a processor control code to implement the edge cloud collaborative task scheduling method according to any one of claims 1-2.
4. A side cloud cooperative task scheduling system for implementing the side cloud cooperative task scheduling method according to any one of claims 1 to 2, the side cloud cooperative task scheduling system comprising:
the unified access module is used for uniformly receiving and pre-classifying the calculation task data generated by the terminal equipment;
the task scheduling center receives and extracts the task key information and stores the relevant state information of the server, including the real-time network channel condition and the service installed by the server;
and the edge server cluster and the remote cloud determine and process the task execution sequence.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110229562.3A CN113132456B (en) | 2021-03-02 | 2021-03-02 | Edge cloud cooperative task scheduling method and system based on deadline perception |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110229562.3A CN113132456B (en) | 2021-03-02 | 2021-03-02 | Edge cloud cooperative task scheduling method and system based on deadline perception |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113132456A CN113132456A (en) | 2021-07-16 |
| CN113132456B true CN113132456B (en) | 2022-09-30 |
Family
ID=76772416
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110229562.3A Active CN113132456B (en) | 2021-03-02 | 2021-03-02 | Edge cloud cooperative task scheduling method and system based on deadline perception |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113132456B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114500543B (en) * | 2022-01-13 | 2024-01-23 | 徐工汉云技术股份有限公司 | Distributed elastic edge acquisition system and application method thereof |
| CN114609329A (en) * | 2022-01-28 | 2022-06-10 | 西安电子科技大学 | Gas monitoring system based on sensor networking under industrial environment |
| CN115061800A (en) * | 2022-06-30 | 2022-09-16 | 中国联合网络通信集团有限公司 | Edge computing task processing method, edge server and storage medium |
| CN116362526B (en) * | 2023-06-01 | 2023-08-11 | 广州健新科技有限责任公司 | Cloud edge cooperative resource management and control method and system for digital power plant |
| CN118708325A (en) * | 2024-08-28 | 2024-09-27 | 苏州元脑智能科技有限公司 | Task scheduling method, device, equipment and computer readable storage medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110717300A (en) * | 2019-09-27 | 2020-01-21 | 云南电网有限责任公司 | Edge calculation task allocation method for real-time online monitoring service of power internet of things |
| CN111711962A (en) * | 2020-06-15 | 2020-09-25 | 重庆邮电大学 | Cooperative scheduling method for subtasks of mobile edge computing system |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108829519A (en) * | 2018-05-31 | 2018-11-16 | 深圳市零度智控科技有限公司 | Method for scheduling task, cloud platform and computer readable storage medium based on cloud platform |
| CN110647403B (en) * | 2019-10-31 | 2023-04-11 | 桂林电子科技大学 | Cloud computing resource allocation method in multi-user MEC system |
| CN111580990A (en) * | 2020-05-08 | 2020-08-25 | 中国建设银行股份有限公司 | Task scheduling method, scheduling node, centralized configuration server and system |
| CN112004239B (en) * | 2020-08-11 | 2023-11-21 | 中国科学院计算机网络信息中心 | Cloud edge collaboration-based computing and unloading method and system |
-
2021
- 2021-03-02 CN CN202110229562.3A patent/CN113132456B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110717300A (en) * | 2019-09-27 | 2020-01-21 | 云南电网有限责任公司 | Edge calculation task allocation method for real-time online monitoring service of power internet of things |
| CN111711962A (en) * | 2020-06-15 | 2020-09-25 | 重庆邮电大学 | Cooperative scheduling method for subtasks of mobile edge computing system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113132456A (en) | 2021-07-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113132456B (en) | Edge cloud cooperative task scheduling method and system based on deadline perception | |
| CN109669768B (en) | Resource allocation and task scheduling method for edge cloud combined architecture | |
| CN108509276B (en) | Video task dynamic migration method in edge computing environment | |
| US10733019B2 (en) | Apparatus and method for data processing | |
| CN105978960B (en) | A kind of cloud scheduling system and method based on massive video structuring processing | |
| Jung et al. | Synchronous parallel processing of big-data analytics services to optimize performance in federated clouds | |
| CN109697122B (en) | Task processing method, device and computer storage medium | |
| CN109564528B (en) | System and method for computing resource allocation in distributed computing | |
| US20120198462A1 (en) | Workflow control of reservations and regular jobs using a flexible job scheduler | |
| WO2024021489A1 (en) | Task scheduling method and apparatus, and kubernetes scheduler | |
| US9141436B2 (en) | Apparatus and method for partition scheduling for a processor with cores | |
| CN103927225A (en) | Multi-core framework Internet information processing and optimizing method | |
| WO2020108337A1 (en) | Cpu resource scheduling method and electronic equipment | |
| CN114168302A (en) | Task scheduling method, device, equipment and storage medium | |
| CN116010064A (en) | DAG job scheduling and cluster management method, system and device | |
| EP3662368B1 (en) | Method and apparatus for managing scheduling of services during boot-up | |
| CN116610422A (en) | Task scheduling method, device and system | |
| CN111506407B (en) | Resource management and job scheduling method and system combining Pull mode and Push mode | |
| CN117608760A (en) | Cloud application hybrid deployment method applied to Kubernetes | |
| CN117290006A (en) | VEC-oriented dependency task unloading and service cache joint optimization method | |
| CN116996941A (en) | Calculation force unloading method, device and system based on cooperation of cloud edge ends of distribution network | |
| Yang et al. | An offloading strategy based on cloud and edge computing for industrial Internet | |
| WO2020246965A1 (en) | Task distribution across multiple processing devices | |
| WO2022111466A1 (en) | Task scheduling method, control method, electronic device and computer-readable medium | |
| CN112887347B (en) | Dynamic migration method and device for edge calculation in industrial internet |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |