WO2023159576A1

WO2023159576A1 - Method and apparatus of scheduling applications

Info

Publication number: WO2023159576A1
Application number: PCT/CN2022/078314
Authority: WO
Inventors: Xu Zhao; Wenfeng Liu; Zijian Wang; Shunjie Fan
Original assignee: Siemens Aktiengesellschaft; Siemens Ltd., China
Priority date: 2022-02-28
Filing date: 2022-02-28
Publication date: 2023-08-31

Abstract

A method of scheduling Apps, which is used for scheduling the Apps among a plurality of edge devices, includes: classifying the Apps into static Apps (SAs) and dynamic Apps (DAs) (110), where the SAs are associated with the edge devices, and the DAs are not associated with the edge devices; deploying a plurality of Apps to the plurality of edge devices (120); and calculating a load of each of the edge devices when the Apps thereon are running, comparing the load with a threshold, and in a case that the load is greater than the threshold, scheduling DAs deployed on the edge device to run on another edge device, until the loads of all the edge devices are less than the threshold (130).

Description

METHOD AND APPARATUS OF SCHEDULING APPLICATIONS

TECHNICAL FIELD

The present invention generally relates to the field of industrial digitization, and in particular, to a method and apparatus of scheduling applications (Apps) .

BACKGROUND

The convergence of information technology (IT) and operational technology (OT) is a major trend in the field of industrial digitalization, and edge devices play an important role in the convergence of IT and OT. There are a plurality of applications (Apps) running on the edge devices to achieve different functions. An App orchestration system is used for deploying a plurality of Apps to run on different edge devices. Existing orchestration technologies (such as Kubernetes) usually use predefined rules to specify the packaging and configuration of Apps, resulting in excessive loads on some edge devices and affecting the performance and real-time performance of the edge devices.

SUMMARY

To resolve the above technical problems, the present invention provides a method and apparatus of scheduling applications (Apps) to balance loads of edge devices, avoid overloading of some edge devices, and improve the performance and real-time performance of the edge devices.

To achieve the above purpose, the present invention provides a method of scheduling Apps, which is used for scheduling the Apps among a plurality of edge devices. The method of scheduling the Apps includes: classifying the Apps into static Apps (SAs) and dynamic Apps (DAs) , where the SAs are associated with the edge devices, and the DAs are not associated with the edge devices; deploying a plurality of Apps to the plurality of edge devices; and calculating a load of each of the edge devices when the Apps thereon are running, comparing the load with a threshold, and in a case that the load is greater than the threshold, scheduling DAs deployed on the edge device to run on another edge device, until the loads of all the edge devices are less than the threshold. Therefore, Apps are classified into SAs and DAs, loads of edge devices when the Apps thereon are running are calculated, and DAs on edge devices with excessively large loads are scheduled to run on edge devices with free loads, so that the loads of the edge devices are balanced, overloading of some edge devices is avoided, and the performance and real-time performance of the edge devices are improved.

Optionally, the classifying the Apps into SAs and DAs includes: obtaining classification labels of the Apps during development, and classifying the Apps into the SAs and the DAs according to the classification labels. Therefore, the developer manually adds a classification label to the App. According to the classification label, it can be determined whether the type of the App is an SA or a DA. Therefore, the flexibility of App classification is improved.

Optionally, the classifying the Apps into SAs and DAs includes: obtaining an amount of data exchange of each App, comparing the amount of data exchange with a specified value, classifying the App as the SA in a case that the amount of data exchange of the App is greater than the specified value, and classifying the App as the DA in a case that the amount of data exchange of the App is less than the specified value. Therefore, whether the type of an App is an SA or a DA can be automatically determined.

Optionally, the calculating a load of each of the edge devices when the Apps thereon are running includes: summing up memory usage of the SAs, memory usage of the DAs, and bandwidth usage of the DAs on the edge devices. Therefore, loads of edge devices when the Apps thereon are running are calculated.

Optionally, the scheduling DAs deployed on the edge device to run on another edge device includes: calculating load consumption of the DAs, and scheduling the DAs to run on the another edge device according to the load consumption. Therefore, edge devices to be scheduled are determined by load consumption, which implements the scheduling of DAs.

Optionally, after the calculating load consumption of the DAs, the method further includes: sorting the DAs in descending order according to the load consumption, and preferentially scheduling top-ranked DAs to run on the another edge device. Therefore, the scheduling of DAs consuming more loads is prioritized, which can reduce the overall scheduling complexity and improve the processing efficiency of scheduling.

The present invention further provides an apparatus of scheduling Apps, which is used for scheduling the Apps among a plurality of edge devices. The apparatus of scheduling the Apps includes: a classification module, configured to classify the Apps into SAs and DAs, where the SAs are associated with the edge devices, and the DAs are not associated with the edge devices; a deployment module, configured to deploy a plurality of Apps to the plurality of edge devices; and a scheduling module, configured to calculate a load of each of the edge devices when the Apps thereon are running, compare the load with a threshold, and in a case that the load is greater than the threshold, schedule DAs deployed on the edge device to run on another edge device, until the loads of all the edge devices are less than the threshold.

Optionally, the classifying, by the classification module, the Apps into SAs and DAs includes: obtaining classification labels of the Apps during development, and classifying the Apps into the SAs and the DAs according to the classification labels.

Optionally, the classifying, by the classification module, the Apps into SAs and DAs includes: obtaining an amount of data exchange of each App, comparing the amount of data exchange with a specified value, classifying the App as the SA in a case that the amount of data exchange of the App is greater than the specified value, and classifying the App as the DA in a case that the amount of data exchange of the App is less than the specified value.

Optionally, the calculating, by the scheduling module, a load of each of the edge devices when the Apps thereon are running includes: summing up memory usage of the SAs, memory usage of the DAs, and bandwidth usage of the DAs on the edge devices.

Optionally, the scheduling, by the scheduling module, DAs deployed on the edge device to run on another edge device includes: calculating load consumption of the DAs, and scheduling the DAs to run on the another edge device according to the load consumption.

Optionally, after the calculating, by the scheduling module, load consumption of the DAs, the method further includes: sorting the DAs in descending order according to the load consumption, and preferentially scheduling top-ranked DAs to run on the another edge device.

The present invention further provides an electronic device, including a processor, a memory and instructions stored in the memory, where the instructions, when executed by the processor, implement the method as described above.

The present invention further provides a computer-readable storage medium, storing computer instructions thereon, where the computer instructions, when executed, implement the method as described above.

BRIEF DESCRIPTION OF THE DRAWINGS

The accompanying drawings below are merely intended to provide exemplary descriptions and explanations for the present invention, but are not intended to limit the scope of the present invention. In the figures:

FIG. 1 is a flowchart of a method of scheduling according to an embodiment of the present invention;

FIG. 2 is a schematic diagram of an industrial automation system according to an embodiment of the present invention;

FIG. 3 to FIG. 5 are schematic diagrams of scheduling applications (Apps) in an industrial automation system according to an embodiment of the present invention;

FIG. 6 is a schematic diagram of an apparatus of scheduling according to an embodiment of the present invention; and

FIG. 7 is a schematic diagram of an electronic device according to an embodiment of the present invention.

DESCRIPTION OF REFERENCE NUMERALS

100 Method of scheduling

110-130 Steps

200 Industrial automation system

210 Edge management device

220-1, 220-2, 220-3, and 220-N Edge devices

230-1, 230-2, 230-3, and 230-N Field devices

300 Industrial automation system

310 Edge management device

321, 322, and 323 Edge devices

331, 332, and 333 Field devices

600 Apparatus of scheduling

610 Classification module

620 Deployment module

630 Scheduling module

700 Electronic device

710 Processor

720 Memory

DETAILED DESCRIPTION

To provide a clearer understanding of the technical features, objectives, and effects of the present invention, specific implementations of the present invention are described with reference to the accompanying drawings.

In the following description, many specific details are provided to give a full understanding of the present invention. However, the present invention may also be implemented in other manners different from those described herein. Therefore, the present invention is not limited to the specific embodiments disclosed below.

As shown in the present application and the claims, terms such as "a/an" , "one" , "one kind" , and/or "the" do not refer specifically to singular forms and may also include plural forms, unless the context expressly indicates an exception. In general, terms "comprise" and "include" merely indicate including clearly identified steps and elements. The steps and elements do not constitute an exclusive list. A method or a device may also include other steps or elements.

The present invention provides a method of scheduling applications (Apps) , which is used for scheduling the Apps among a plurality of edge devices. FIG. 1 is a flowchart of a method 100 of scheduling according to an embodiment of the present invention. As shown in FIG. 1, the method 100 of scheduling Apps includes:

Step 110: Classify the Apps into static Apps (SAs) and dynamic Apps (DAs) , where the SAs are associated with the edge devices, and the DAs are not associated with the edge devices.

Apps can be classified into SAs and DAs during development. The SAs are associated with the edge devices and are bound to field devices controlled by the edge devices. The SAs exchange a large amount of data with the field devices, and only run on the deployed edge devices. Once being deployed, the SAs are not scheduled to run on other edge devices. For example, the SAs may be data monitoring apps. The DAs are not associated with the edge devices, and are not bound to the field devices controlled by the edge devices. The DAs exchange a small amount of data with the field devices, and can be scheduled to run on different edge devices. For example, the DAs may be alarm response apps.

In some embodiments, the classifying the Apps into SAs and DAs may include: obtaining classification labels of the Apps during development, and classifying the Apps into the SAs and the DAs according to the classification labels. Specifically, when developing an App, the developer manually adds a classification label to the App. According to the classification label, it can be determined whether the type of the App is an SA or a DA. Therefore, the flexibility of App classification is improved.

In some embodiments, the classifying the Apps into SAs and DAs may include: obtaining an amount of data exchange of each App, comparing the amount of data exchange with a specified value, classifying the App as the SA in a case that the amount of data exchange of the App is greater than the specified value, and classifying the App as the DA in a case that the amount of data exchange of the App is less than the specified value. Specifically, an amount of data exchange is designed during the development of an App. If the amount of data exchange is too large, it indicates that the App is closely associated with a field device, otherwise, the correlation is not closely associated with a field device. The specified value can be used as a criterion to measure the amount of data exchange. Therefore, whether the type of an App is an SA or a DA can be automatically determined.

Step 120: Deploy a plurality of Apps to the plurality of edge devices.

After the development is completed, a plurality of Apps are deployed to a plurality of edge devices. Because SAs are associated with field devices, and cannot be scheduled after being deployed, the SAs are deployed to corresponding edge devices. For example, data monitoring software of a sensor is deployed to an edge device corresponding to the sensor. Then, DAs can be deployed to a plurality of edge devices randomly, or the DAs can be deployed to the plurality of edge devices according to other rules.

FIG. 2 is a schematic diagram of an industrial automation system 200 according to an embodiment of the present invention. As shown in FIG. 2, the industrial automation system 200 includes an edge management device 210, a plurality of edge devices 220-1, 220-2, 220-3, …, and 220-N, and a plurality of field devices 230-1, 230-2, 230-3, …, and 230-N. The edge devices control the corresponding field devices. That is, the edge device 220-1 controls the field device 230-1, the edge device 220-2 controls the field device 230-2, the edge device 220-3 controls the field device 230-3, and the edge management device 210 deploys Apps to the plurality of edge devices. As shown in FIG. 2, an SA 1.1, an SA 1.2, a DA 1.1, and a DA 1.2 are deployed on the edge device 220-1, an SA 2.1 and a DA 2.1 are deployed on the edge device 220-2, an SA 3.1 and a DA 3.1 are deployed on the edge device 220-3, and an SA N. 1 and a DA N. 1 are deployed on the edge device 220-N.

Step 130: Calculate a load of each of the edge devices when the Apps thereon are running, compare the load with a threshold, and in a case that the load is greater than the threshold, schedule DAs deployed on the edge device to run on another edge device, until the loads of all the edge devices are less than the threshold.

After the Apps are deployed to the edge devices, Apps that are run by the edge devices can be determined. Then, the loads of the Apps are calculated. The threshold is deployed to the edge devices with the Apps at the same time. The loads calculated on the edge devices are compared with the threshold. In a case that the loads of the edge devices are less than the threshold, it indicates that the edge devices are not overloaded. In a case that the loads of the edge devices are greater than the threshold, it indicates that the edge devices are overloaded. At this time, the DAs on the edge devices are scheduled to run on other edge devices until the loads of all the edge devices are less than the threshold.

In some embodiments, the calculating a load of each of the edge devices when the Apps thereon are running may include: summing up memory usage of the SAs, memory usage of the DAs, and bandwidth usage of the DAs on the edge devices. For example, in a case that an SA or a DA needs to occupy 20%of memory resources, then memory usage of the App is 20%, and in a case that the DA needs to occupy 10%of bandwidth resources, then bandwidth usage of the DA is 10%. The loads of the edge devices when the Apps thereon are running can be calculated by summing up the memory usage and the bandwidth usage. The loads of the edge devices when the Apps thereon are running can be calculated by using the following formula:

where W _i is a load of an edge device i, W _s is the memory usage of the SA, W _d is the memory usage of the DA, N _d is the bandwidth consumption of the DA, D _d is the amount of data of the DA, k is a conversion coefficient, and

is the bandwidth usage.

In some embodiments, the scheduling DAs deployed on the edge device to run on another edge device may include: calculating load consumption of the DAs, and scheduling the DAs to run on the another edge device according to the load consumption. Specifically, the load consumption of the DAs is calculated. The load consumption indicates the amount of load that the DAs need to consume. Other edge devices are searched. If remaining resources of the edge devices can accommodate the load consumption of the DAs, the DAs are scheduled to run on the edge devices with the remaining resources. Preferably, the DAs are scheduled to run on the nearest edge devices. The load consumption of the DAs can be calculated according to the following formula:

C _D=W _d-k·N _d·D _d,

where C _D is the load consumption of the DA, W _d is the memory usage of the DA, N _d is the bandwidth consumption of the DA, D _d is the amount of data of the DA, and k is a conversion coefficient.

In some embodiments, after the calculating load consumption of the DAs, the method may further include: sorting the DAs in descending order according to the load consumption, and preferentially scheduling top-ranked DAs to run on the another edge device. The DA with the larger load consumption has a greater priority. That is, the scheduling of the DA consuming more load is prioritized, which can reduce the overall scheduling complexity and improve the processing efficiency of scheduling.

FIG. 3 to FIG. 5 are schematic diagrams of scheduling Apps in an industrial automation system according to an embodiment of the present invention. The industrial automation system 300 includes an edge management device 310, three

edge devices

321, 322, and 323, and three

corresponding field devices

331, 332, and 333. The edge device 321 is deployed with an SA 1.1 (memory usage 0.2) , an SA 1.2 (memory usage 0.2) , a DA 1.1 (memory usage 0.2, and bandwidth usage 0.01) , and a DA 1.2 (memory usage 0.2, and bandwidth usage 0.01) . The edge device 322 is deployed with an SA 2.1 (memory usage 0.2) , and a DA 2.1 (memory usage 0.1, and bandwidth usage 0.01) . The edge device 323 is deployed with an SA 3.1 (memory usage 0.65) , and a DA 3.1 (memory usage 0.1, and bandwidth usage 0.01) . The threshold is 0.7. The load of the edge device 321 is 0.82, and the load of the edge device 323 is 0.76, which are greater than the threshold. The load of the edge device 322 is 0.31, which is less than the threshold. If run according to the deployment mode, the

edge devices

321 and 323 are overloaded. According to the method for scheduling provided by the embodiments of the present invention, the load consumption required by the DA 1.1 on the edge device 321 is 0.19 (memory usage 0.2-bandwidth usage 0.01) . The edge device 322 has enough space to run the DA 1.1, and the DA 1.1 can be scheduled to the edge device 322, as shown in FIG. 3 and FIG. 4. For the edge device 323, the memory usage required by the DA 3.1 on the edge device 323 is 0.99 (memory usage 0.1-bandwidth usage 0.01) . The edge device 322 still has enough space to run the DA 3.1, and the DA 3.1 can be scheduled to the edge device 322, as shown in FIG. 4 and FIG. 5. Therefore, the loads of the

edge devices

321, 322, and 323 are all less than the threshold, and the

edge devices

321, 322, and 323 are not overloaded.

The embodiments of the present invention provide a method for scheduling Apps. The Apps are classified into SAs and DAs, loads of edge devices when the Apps thereon are running are calculated, and DAs on edge devices with excessively large loads are scheduled to run on edge devices with free loads, so that loads of edge devices are balanced, overloading of some edge devices is avoided, and the performance and real-time performance of the edge devices are improved.

The present invention further provides an apparatus of scheduling, which is used for scheduling Apps among a plurality of edge devices. FIG. 6 is a schematic diagram of an apparatus 600 of scheduling according to an embodiment of the present invention. As shown in FIG. 6, the apparatus 600 of scheduling Apps includes:

a classification module 610, configured to classify the Apps into SAs and DAs, where the SAs are associated with the edge devices, and the DAs are not associated with the edge devices;

a deployment module 620, configured to deploy a plurality of Apps to the plurality of edge devices; and

a scheduling module 630, configured to calculate a load of each of the edge devices when the Apps thereon are running, compare the load with a threshold, and in a case that the load is greater than the threshold, schedule DAs deployed on the edge device to run on another edge device, until the loads of all the edge devices are less than the threshold.

In some embodiments, the classifying, by the classification module 610, the Apps into SAs and DAs includes: obtaining classification labels of the Apps during development, and classifying the Apps into the SAs and the DAs according to the classification labels.

In some embodiments, the classifying, by the classification module 610, the Apps into SAs and DAs includes: obtaining an amount of data exchange of the App, comparing the amount of data exchange with a specified value, classifying the App as the SA in a case that the amount of data exchange of the App is greater than the specified value, and classifying the App as the DA in a case that the amount of data exchange of the App is less than the specified value.

In some embodiments, the calculating, by the scheduling module 630, a load of each of the edge devices when the Apps thereon are running includes: summing up memory usage of the SAs, memory usage of the DAs, and bandwidth usage of the DAs on the edge devices.

In some embodiments, the scheduling, by the scheduling module 630, DAs deployed on the edge device to run on another edge device includes: the scheduling module 630 calculates load consumption of the DAs, and schedules the DAs to run on the another edge device according to the load consumption.

In some embodiments, after the calculating load consumption of the DAs, the method further includes: sorting the DAs in descending order according to the load consumption, and preferentially scheduling top-ranked DAs to run on the another edge device.

The present invention further provides an electronic device 700. FIG. 7 is a schematic diagram of an electronic device 700 according to an embodiment of the present invention. As shown in FIG. 7, the electronic device 700 includes a processor 710 and a memory 720. The memory 720 stores instructions, and the instructions are executed by the processor 710 to implement the method 100 described above.

The present invention further provides a computer-readable storage medium, storing computer instructions thereon, where the computer instructions, when executed, implement the method 100 as described above.

Some aspects of the method and apparatus of the present invention may be entirely executed by hardware, may be entirely executed by software (including firmware, resident software, microcode, and the like) , or may be executed by a combination of hardware and software. The foregoing hardware or software may be referred to as "data block" , "module" , "engine" , "unit" , "component" or "system" . The processor may be one or more application specific integrated circuits (ASICs) , digital signal processors (DSPs) , digital signal processing devices (DSPDs) , programmable logic devices (PLDs) , field programmable gate arrays (FPGAs) , processors, controllers, microcontrollers, microprocessors, or a combination thereof. In addition, various aspects of the present invention may be embodied as computer products located in one or more computer-readable media, the product including computer-readable program code. For example, the computer-readable medium may include, but is not limited to, a magnetic storage device (for example, a hard disk, a floppy disk, a magnetic tape, etc. ) , an optical disk (for example, a compact disk (CD) , a digital versatile disk (DVD) , etc. ) , a smart card, and a flash memory device (for example, a card, a stick, a key driver, etc. ) .

Flow diagrams are used herein to illustrate operations performed by the method according to the embodiments of the present application. It should be understood that the foregoing operations are not necessarily performed precisely in order. On the contrary, the steps may be performed in reverse order or simultaneously. In addition, other operations may be alternatively added into the processes, or one or more steps are removed from the processes.

It should be understood that, although this specification is described according to each embodiment, each embodiment may not include only one independent technical solution. The description manner of this specification is merely for clarity. This specification should be considered as a whole by a person skilled in the art, and the technical solution in each embodiment may also be properly combined, to form other implementations that can be understood by the person skilled in the art.

The foregoing are merely specific schematic implementations of the present invention, and are not intended to limit the scope of the present invention. Any equivalent change, modification, and combination made by the person skilled in the art without departing from the conception and principles of the present invention should all fall within the protection scope of the present invention.

Claims

A method (100) of scheduling applications (Apps) , used for scheduling Apps among a plurality of edge devices, the method (100) of scheduling Apps comprising:

classifying the Apps into static Apps (SAs) and dynamic Apps (DAs) (110) , wherein the SAs are associated with the edge devices, and the DAs are not associated with the edge devices;

deploying a plurality of Apps to the plurality of edge devices (120) ; and

calculating a load of each of the edge devices when the Apps thereon are running, comparing the load with a threshold, and in a case that the load is greater than the threshold, scheduling DAs deployed on the edge device to run on another edge device, until the loads of all the edge devices are less than the threshold (130) .
The method (100) of scheduling Apps according to claim 1, wherein the classifying the Apps into SAs and DAs comprises: obtaining classification labels of the Apps during development and classifying the Apps into the SAs and the DAs according to the classification labels.
The method (100) of scheduling Apps according to claim 1, wherein the classifying the Apps into SAs and DAs comprises: obtaining an amount of data exchange of each App, comparing the amount of data exchange with a specified value, classifying the App as the SA in a case that the amount of data exchange of the App is greater than the specified value, and classifying the App as the DA in a case that the amount of data exchange of the App is less than the specified value.
The method (100) of scheduling Apps according to claim 1, wherein the calculating a load of each of the edge devices when the Apps thereon are running comprises: summing up memory usage of the SAs, memory usage of the DAs, and bandwidth usage of the DAs on the edge devices.
The method (100) of scheduling Apps according to claim 4, wherein the scheduling DAs deployed on the edge device to run on another edge device comprises: calculating load consumption of the DAs and scheduling the DAs to run on the another edge device according to the load consumption.
The method (100) of scheduling Apps according to claim 5, wherein after the calculating load consumption of the DAs, the method further comprises: sorting the DAs in descending order according to the load consumption, and preferentially scheduling top-ranked DAs to run on the another edge device.
An apparatus (600) of scheduling applications (Apps) , configured to schedule the Apps among a plurality of edge devices, the apparatus (600) of scheduling Apps comprising:

a classification module (610) , configured to classify the Apps into static Apps (SAs) and dynamic Apps (DAs) , wherein the SAs are associated with the edge devices, and the DAs are not associated with the edge devices;

a deployment module (620) , configured to deploy a plurality of Apps to the plurality of edge devices; and

a scheduling module (630) , configured to calculate a load of each of the edge devices when the Apps thereon are running, compare the load with a threshold, and in a case that the load is greater than the threshold, schedule DAs deployed on the edge device to run on another edge device, until the loads of all the edge devices are less than the threshold.
The apparatus (600) of scheduling Apps according to claim 7, wherein the classifying, by the classification module (610) , the Apps into SAs and DAs comprises: obtaining classification labels of the Apps during development and classifying the Apps into the SAs and the DAs according to the classification labels.
The apparatus (600) of scheduling Apps according to claim 7, wherein the classifying, by the classification module (610) , the Apps into SAs and DAs comprises: obtaining an amount of data exchange of each App, comparing the amount of data exchange with a specified value, classifying the App as the SA in a case that the amount of data exchange of the App is greater than the specified value, and classifying the App as the DA in a case that the amount of data exchange of the App is less than the specified value.
The apparatus (600) of scheduling Apps according to claim 7, wherein the calculating, by the scheduling module (630) , a load of each of the edge devices when the Apps thereon are running comprises: summing up memory usage of the SAs, memory usage of the DAs, and bandwidth usage of the DAs on the edge devices.
The apparatus (600) of scheduling Apps according to claim 10, wherein the scheduling, by the scheduling module (630) , DAs deployed on the edge device to run on another edge device comprises: calculating load consumption of the DAs and scheduling the DAs to run on the another edge device according to the load consumption.
The apparatus (600) of scheduling Apps according to claim 11, wherein after the calculating, by the scheduling module (630) , load consumption of the DAs, the method further comprises: sorting the DAs in descending order according to the load consumption, and preferentially scheduling top-ranked DAs to run on the another edge device.
An electronic device (700) , comprising a processor (710) , a memory (720) , and instructions stored in the memory (720) , wherein the instructions, when executed by the processor (710) , implement the method according to any one of claims 1 to 6.
A computer-readable storage medium, storing computer instructions thereon, wherein the computer instructions, when executed, implement the method according to any one of claims 1 to 6.