WO2024120205A1

WO2024120205A1 - Method and apparatus for optimizing application performance, electronic device, and storage medium

Info

Publication number: WO2024120205A1
Application number: PCT/CN2023/133455
Authority: WO
Inventors: 叶可江; 张永贺; 须成忠
Original assignee: 中国科学院深圳先进技术研究院
Priority date: 2022-12-05
Filing date: 2023-11-22
Publication date: 2024-06-13

Abstract

The present invention relates to the technical field of cloud computing. Disclosed are a method and apparatus for optimizing application performance, an electronic device, and a storage medium. The method for optimizing application performance is applied to a mixed cluster. The method comprises: obtaining indicator data; on the basis of first indicator data of applications, detecting whether the applications are interfered with; if there is an abnormal application that is interfered with, on the basis of second indicator data, allocating a CPU core for the abnormal application from a CPU shared pool; and updating a control group Cgroup file of the abnormal application according to the CPU core allocated to the abnormal application. According to the present invention, interference received by applications is monitored and solved in real time by means of indicator data of the applications and a system core, and CPU cores allocated for the applications are dynamically adjusted, so that the application stability of a mixed cluster is greatly guaranteed, application performance is improved, the overall utilization rate of the entire machine is improved, and the problem of interference to application performance caused by CPU core preemption in the mixed cluster is solved.

Description

Application performance optimization method, device, electronic device and storage medium

Technical Field

The present invention belongs to the technical field of cloud computing, and more specifically, relates to an application performance optimization method, device, electronic device and storage medium.

Background technique

Currently, applications in colocation clusters are mainly deployed in containers in the form of microservices. Containerized applications are deployed on servers in two different CPU usage modes: one is the CPU set mode that binds the application to a fixed CPU core. In this mode, the CPU cores that the application can use are pre-allocated and exclusively used by the application and cannot be preempted by other applications; the other mode is the CPU share mode that shares all CPU cores in the CPU sharing pool. In this mode, all applications share the same batch of CPU cores that are not exclusively used by applications in the CPU set mode.

With the rapid development of cloud computing, more and more applications are switching from CPU set mode to CPU share mode. However, in CPU share mode, applications share CPU cores in the same CPU sharing pool, which will inevitably lead to the problem of applications preempting CPU cores. This will inevitably cause serious interference during application operation, which in turn causes a sharp increase in scheduling overhead. The operating system spends a lot of time swapping in and out threads on the CPU core, but not much CPU time slice is actually used by the application, which seriously affects application performance.

It can be seen that in the prior art, there is a problem in which application performance is disturbed due to CPU core preemption in a colocation cluster.

technical problem

In view of the defects of the related art, the present invention provides an application performance optimization method, device, electronic device and storage medium, aiming to solve the problem of application performance being disturbed due to CPU core preemption in a mixed cluster existing in the related art.

Technical Solutions

The technical solution is as follows:

According to one aspect of the present application, an application performance optimization method is applied to a colocation cluster, the method comprising: obtaining indicator data, the indicator data comprising first indicator data of each application during its operation in a current time period, and second indicator data related to a system kernel; based on the first indicator data of each application, detecting whether each application is interfered with; if there is an abnormal application that is interfered with, based on the second indicator data, allocating a CPU core to the abnormal application from a CPU shared pool; and updating a control group Cgroup file of the abnormal application according to the CPU core allocated to the abnormal application.

According to one aspect of the present application, an application performance optimization device is deployed in a colocation cluster, and the device includes: an acquisition module, used to obtain indicator data, the indicator data including first indicator data of each application during its operation in a current time period, and second indicator data related to the system kernel; an interference detection module, used to detect whether each application is interfered with based on the first indicator data of each application; a resource allocation module, used to allocate a CPU core to the abnormal application from a CPU sharing pool based on the second indicator data if there is an abnormal application that is interfered with; and a file update module, used to update the control group Cgroup file of the abnormal application according to the CPU core allocated to the abnormal application.

According to one aspect of the present application, an electronic device includes: at least one processor, at least one memory, and at least one communication bus, wherein a computer program is stored in the memory, and the processor reads the computer program in the memory through the communication bus; when the computer program is executed by the processor, the application performance optimization method as described above is implemented.

According to one aspect of the present application, a storage medium stores a computer program thereon, and when the computer program is executed by a processor, the application performance optimization method as described above is implemented.

According to one aspect of the present application, a computer program product includes a computer program, the computer program is stored in a storage medium, a processor of a computer device reads the computer program from the storage medium, and the processor executes the computer program, so that the computer device implements the application performance optimization method as described above when executing the computer program.

The beneficial effects of the technical solution provided by this application are:

In the above technical solution, various indicator data about each application and the system kernel on the mixed cluster are obtained, and based on the indicator data of each application, it is detected whether each application is disturbed; when there is an abnormal application that is disturbed, the CPU core allocated to the application is dynamically adjusted based on the indicator data of the system kernel in the indicator data to ensure application performance. The present invention monitors and solves the interference to the application in real time based on the indicator data of each application, and dynamically adjusts the CPU resources based on the indicator data related to the system kernel, which greatly ensures the stability of the mixed cluster application, improves the application performance and the utilization rate of the whole machine, and solves the problem of application performance being disturbed due to CPU core preemption in the mixed cluster.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in describing the embodiments of the present application are briefly introduced below.

FIG1 is a schematic diagram of an implementation environment of an application performance optimization method provided in an embodiment of the present application;

FIG2 is a flow chart of an application performance optimization method provided in an embodiment of the present application;

FIG3 is a flow chart of step 240 in one embodiment of the embodiment corresponding to FIG2 ;

FIG4 is a flow chart of step 240 in the embodiment corresponding to FIG2 in another embodiment;

FIG5 is a schematic diagram of a specific implementation of an application performance optimization method in an application scenario;

FIG6 is a block diagram of an application performance optimization device according to an exemplary embodiment;

FIG7 is a hardware structure diagram of a server according to an exemplary embodiment;

Fig. 8 is a block diagram of an electronic device according to an exemplary embodiment.

Embodiments of the present invention

The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and cannot be interpreted as limiting the present application.

It will be understood by those skilled in the art that, unless expressly stated, the singular forms "one", "said", and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present application refers to the presence of the features, integers, steps, operations, elements and/or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and/or groups thereof. It should be understood that when we refer to an element as being "connected" or "coupled" to another element, it may be directly connected or coupled to the other element, or there may be an intermediate element. In addition, the "connection" or "coupling" used herein may include wireless connection or wireless coupling. The term "and/or" used herein includes all or any unit and all combinations of one or more associated listed items.

The following is an introduction and explanation of several terms involved in this application:

A CPU core is usually considered to be a logical core allocated to applications. Different logical cores can come from the same physical core or from different physical cores.

Socket refers to the CPU slot, which is the socket for installing the CPU. Often CPU resources include CPU cores from multiple sockets. The CPU cores bound to the application should belong to the same socket as much as possible. Cross-socket will waste performance resources.

Cgroup, or control groups, is a function of the Linux kernel that is used to limit, control, and separate the resources of a process group (such as CPU, memory, disk input and output, etc.). CPU share pool, also known as CPU sharing pool, all CPU cores in it can be called by all processes.

The CPU share mode means that each application can share all CPU cores in the CPU sharing pool.

CPU set mode means that the CPU cores that can be used by an application are pre-allocated and exclusively used by the application and cannot be preempted by other applications. Most applications in this mode are online services with relatively high priority.

Colocation clusters are a technology that mixes clusters and schedules different types of tasks to the same physical resources. Through scheduling, resource isolation and other control measures, it improves resource utilization and greatly reduces costs while ensuring SLO. Colocation means that loads with different business characteristics, priorities, and resource usage models are mixed and run on the same machine, which is inevitably accompanied by problems such as resource preemption.

As mentioned above, in the related art, there are often interference problems caused by resource preemption between applications in a colocation cluster.

Usually, in order to solve the interference between applications in a colocation cluster, we start from the perspective of optimizing application deployment, and continuously improve the scheduling and deployment algorithm to save deployment costs and reduce application interference. Although these methods have improved the problems of difficult to control application deployment costs and prevent some performance interference to a certain extent, as the scale of applications increases sharply, the application density on servers is getting higher and higher. It is difficult to prevent interference between applications by simply improving the scheduling and deployment methods. In addition, applications themselves are "different for different people" and the application characteristics vary greatly. It is difficult to cover all situations with one mechanism or algorithm.

As can be seen from the above, the related technology still has the limitation that the application preempts the CPU core, causing the performance to be disturbed.

To this end, the application performance optimization method provided in the present application can dynamically adjust CPU resources to effectively improve application performance. Accordingly, the application performance optimization method is suitable for an application performance optimization device, which can be deployed in an electronic device. The electronic device can be a computer device configured with a von Neumann architecture, for example, the computer device can be a desktop computer, a laptop computer, a server, etc.

Example

In order to make the objectives, technical solutions and advantages of the present application more clear, the implementation methods of the present application will be further described in detail below with reference to the accompanying drawings.

Please refer to FIG1 , which shows a schematic diagram of an implementation environment involved in an application performance optimization method provided by the present application. The implementation environment includes a monitoring component 101 , a trigger component 102 , a CPU scheduling and allocation component 103 , a tuning component 104 , and a management and control component 105 .

The monitoring component 101 collects indicator data from applications and the system kernel, and sends the indicator data to the trigger component 102 and the CPU scheduling and allocation component 103. The indicator data includes first indicator data of each application running in the current time period, and second indicator data related to the system kernel.

The trigger component 102 detects whether each application is interfered with based on the first indicator data of each application, and sends the detection result to the tuning component 104 .

Among them, the first indicator data refers to the indicator data related to the operation of the application in the current time period, such as the response delay of the request, the end-to-end delay, the task completion time, etc.

The CPU scheduling and allocation component 103 calculates the CPU account book according to the second indicator data related to the system kernel collected by the monitoring component 101.

Among them, the second indicator data may include the idle status of the CPU core, the location of the CPU core in each socket, which CPU cores are in the same physical core, etc. The CPU account book refers to the CPU cores divided into different levels. The division rule may be whether the CPU cores are in the same socket, whether the CPU cores are in the same physical core, whether there is an exclusive CPU core in the physical core where the CPU core is located, or one or more of the idleness of each CPU core, which is not limited here.

If the detection result sent by the trigger component 102 is that there is an abnormal application that is disturbed, the tuning component 104 allocates CPU cores to the abnormal application from the CPU shared pool according to the CPU ledger obtained from the CPU scheduling and allocation component 103, obtains the CPU allocation strategy, and sends the CPU allocation strategy to the management and control component 105.

After receiving the CPU allocation policy sent by the tuning component 104, the management and control component 105 detects whether the CPU allocation policy is correct. If correct, it traverses each abnormal application that needs to adjust the CPU resources, finds the Cgroup file of the abnormal application, and modifies the CPU core bound to the abnormal application in the Cgroup file to the target value to achieve the binding of the abnormal application with the CPU core indicated by the CPU allocation policy, that is, the CPU resources of the abnormal application are adjusted, and then the performance of the abnormal application is optimized, so that the abnormal application returns to normal.

Please refer to FIG. 2 . An embodiment of the present application provides an application performance optimization method, which is applied to a colocation cluster.

In the following method embodiments, for ease of description, the execution subject of each step of the method is taken as an example of a server in a colocation cluster, but this does not constitute a specific limitation.

As shown in FIG. 2 , the method may include the following steps:

Step 200, obtaining indicator data.

The indicator data includes first indicator data of each application during operation in the current time period, and second indicator data related to the system kernel.

The second indicator data refers to indicator data related to the system kernel, such as the scheduling delay of the application on the CPU core, the CPI (average number of cycles required for instruction execution) of the application, the utilization rate of each CPU core, the position of the CPU core in each socket, which CPU cores are on the same physical core, etc.

Regarding the acquisition of indicator data, in one possible implementation, the indicator data is obtained by real-time monitoring and collection of the application running process and/or the system kernel by a monitoring component deployed in the colocation cluster.

Step 220: Based on the first indicator data of each application, detect whether each application is interfered.

In a possible implementation, whether the application is interfered with is determined by detecting whether the application has performance fluctuations, that is, if the application has performance fluctuations, it is determined that the application is interfered with.

Specifically, step 220 may include the following steps: obtaining historical indicator data of each application during its operation in a historical time period, obtaining performance fluctuation data of each application by comparing and analyzing the first indicator data of each application with the historical indicator data of each application during its operation in the historical time period, and then determining whether each application is interfered with based on the performance fluctuation data, thereby completing the detection of whether each application is interfered with.

The historical indicator data refers to the indicator data related to the operation of each application in the historical time period, such as the response delay of the request, the end-to-end delay, the task completion time, etc.

In a possible implementation, the performance fluctuation data may be obtained by calculating the difference between the first indicator data and certain historical indicator data.

In a possible implementation, the performance fluctuation data may be obtained by comparing the first indicator data with an average value of all historical indicator data within a certain historical time period.

Step 240: If there is an abnormal application that is disturbed, a CPU core is allocated to the abnormal application from the CPU shared pool based on the second indicator data.

Among them, the application that is disturbed is regarded as an abnormal application. Since the CPU cores in the CPU shared pool are shared by various applications, in order to avoid CPU core preemption, before allocating CPU cores to abnormal applications, it is necessary to select a specific number and position of CPU cores from the CPU shared pool based on the second indicator data, and then allocate them to the abnormal applications, thereby dynamically adjusting the CPU resources of the abnormal application.

For example, based on the second indicator data, the CPU core with the highest idleness in the CPU shared pool is allocated to the abnormal application.

Step 260: Update the control group Cgroup file of the abnormal application according to the CPU core allocated to the abnormal application.

Cgroup (control groups) is a function of the Linux kernel that is used to limit, control and separate the resources of a process group (such as CPU, memory, disk input and output, etc.). By updating the Cgroup file, the CPU core bound to the application can be updated.

In an exemplary embodiment, after step 260, the method may further include the following steps:

Step 261 : After the Cgroup file of the abnormal application is updated, based on the first indicator data of the abnormal application during operation in the current time period, it is detected whether the abnormal application has returned to normal.

That is to say, continue to obtain the first indicator data of the abnormal application during the operation process after completing the Cgroup file update, and obtain the performance fluctuation data of the abnormal application after completing the Cgroup file update by comparing and analyzing the first indicator data of the abnormal application with the historical indicator data of the abnormal application during the operation process in the historical time period. Then, based on the performance fluctuation of the abnormal application after completing the Cgroup file update indicated by the performance fluctuation data, determine whether the abnormal application has returned to normal after completing the Cgroup file update.

If the abnormal application returns to normal after completing the Cgroup file update, execute step 262; otherwise, if the abnormal application is still abnormal after completing the Cgroup file update, return to execute step 240 and continue to adjust the CPU resources for the abnormal application until the abnormal application returns to normal.

Step 262: If it is detected that the abnormal application has returned to normal, the CPU core allocated to the abnormal application is restored to the CPU shared pool.

By modifying the Cgroup file of the abnormal application again and setting the number and position of the CPU cores bound to the abnormal application, the CPU cores allocated to the abnormal application when it is abnormal can be restored to the CPU sharing pool.

In the above process, the application interference is monitored and resolved in real time through the indicator data of each application, and the CPU resources are dynamically adjusted based on the indicator data related to the system kernel. This greatly ensures the application stability of the colocation cluster, improves application performance and the utilization rate of the entire machine, and solves the problem of application performance interference caused by CPU core preemption in the colocation cluster.

Referring to FIG. 3 , in an exemplary embodiment, step 240 may include the following steps:

Step 241 : if the abnormal application supports the CPU share mode, the CPU cores in the CPU sharing pool are divided into a plurality of idle levels according to the second indicator data of the system.

The CPU cores in the same idle level have the same allocation priority. It should be noted that the higher the idle level, the lower the allocation priority, which means that the CPU cores in the idle level are more difficult to be allocated.

In one possible implementation, the CPU cores in the same socket are allocated to the same idle level. To prevent cross-socket CPU allocation to applications, the CPU cores in the same socket are allocated to the same idle level as much as possible. This method can effectively reduce the redundant performance consumed by running CPUs across sockets, which is further beneficial to improving application performance.

In a possible implementation, CPU cores in the same physical core are divided into the same idle level.

In one possible implementation, based on the second indicator data of the system, the idleness of the CPU core is determined, and the CPU cores with idleness within the same set range are divided into the same idle level. The idleness can be obtained based on the utilization rate of each CPU core, and the utilization rate of each CPU core is obtained from the second indicator data related to the system core. For example, CPU cores with idleness between 10% and 20% are divided into one idle level, and CPU cores with idleness between 20% and 30% are divided into another idle level.

In one possible implementation, if there is an exclusive CPU resource in the physical core where the CPU core is located, the idle level of the CPU core is higher than the idle levels of other CPU cores. The other CPU core refers to a CPU core that does not have an exclusive CPU resource in the physical core. In this way, interference with applications that support the CPU set mode of the exclusive logical core can be effectively avoided, which is further beneficial to improving application performance.

Step 242 , according to the number of CPU cores required by the abnormal application and the idle level of the CPU cores, select CPU cores of the same idle level from the CPU sharing pool to obtain a CPU allocation strategy.

The CPU allocation policy is used to indicate the CPU cores that can be allocated to the abnormal application.

Continuing to refer to FIG. 4 , in an exemplary embodiment, step 240 may further include the following steps:

Step 243: Based on the CPU cores that can be allocated to the abnormal application as indicated by the CPU allocation policy, it is detected whether the CPU allocation policy is correct.

If it is detected that the CPU allocation strategy is correct, step 260 is executed.

On the contrary, if it is detected that the CPU allocation policy is incorrect, for example, the CPU cores that can be allocated to the abnormal application have been exclusively occupied by other applications supporting the CPU set mode, step 244 is executed.

Step 244: if it is detected that the CPU allocation policy is wrong, then a CPU core in the CPU shared pool is re-allocated to the abnormal application.

Through the cooperation of the above embodiments, the CPU cores in the CPU sharing pool are divided into several idle levels to facilitate allocation to abnormal applications. In the process of division, the CPU cores in the same socket are divided into the same idle level, which can effectively reduce the redundant performance consumed by running CPUs across sockets, and further help improve application performance. Check whether the allocation strategy is correct to prevent some CPU cores from being monopolized by CPU set type applications at this moment, resulting in allocation failure.

Figure 5 is a schematic diagram of a specific implementation of an application performance optimization method in an application scenario. In this application scenario, the CPU cores allocated to the application in the initial state are the CPU shared pool, and the servers in the colocation cluster collect the first indicator data of the application during the current time period, and then determine whether the first indicator data of the application is abnormal. If it is abnormal, the CPU cores allocated to the application are adjusted; otherwise, the data indicators of the application continue to be collected and determined whether they are abnormal.

After adjusting the CPU cores allocated to the application, continue to collect the first indicator data of the application and determine whether it is normal. If it is normal, the application returns to the initial state, that is, the CPU cores allocated to the application are restored to the CPU shared pool; otherwise, continue to adjust the CPU cores allocated to the application.

In this application scenario, the interference to the application is monitored and resolved in real time through the indicator data of each application, and the CPU core allocated to the application is adjusted dynamically, which greatly ensures the stability of the colocation cluster application, improves application performance and the utilization rate of the entire machine, and solves the problem of application performance being disturbed due to CPU core preemption in the colocation cluster.

The following is an embodiment of the device of the present application, which can be used to execute the application performance optimization method involved in the present application. For details not disclosed in the embodiment of the device of the present application, please refer to the method embodiment of the application performance optimization method involved in the present application.

Please refer to FIG. 6 . An embodiment of the present application provides an application performance optimization device 900 deployed in a colocation cluster. The device 900 includes but is not limited to: an acquisition module 910 , an interference detection module 930 , a resource allocation module 950 , and a file update module 970 .

The acquisition module 910 is used to acquire indicator data, which includes first indicator data of each application during operation in the current time period and second indicator data related to the system kernel.

The interference detection module 930 is used to detect whether each application is interfered based on the first indicator data of each application.

The resource allocation module 950 is configured to allocate a CPU core from a CPU sharing pool to an abnormal application that is disturbed based on the second indicator data.

The file updating module 970 is used to update the control group Cgroup file of the abnormal application according to the CPU core allocated to the abnormal application.

It should be noted that the application performance optimization device provided in the above embodiment only uses the division of the above-mentioned functional modules as an example when performing application performance optimization. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the application performance optimization device will be divided into different functional modules to complete all or part of the functions described above.

In addition, the application performance optimization device and the application performance optimization method provided in the above embodiments belong to the same concept, and the specific manner in which each module performs operations has been described in detail in the method embodiments and will not be repeated here.

Please refer to FIG. 7 , which shows a schematic diagram of the structure of a server according to an exemplary embodiment.

It should be noted that the server is only an example adapted to the present application and cannot be considered to provide any limitation on the scope of use of the present application. The server cannot be interpreted as needing to rely on or having to have one or more components in the exemplary server 2000 shown in FIG. 7 .

The hardware structure of the server 2000 may vary greatly due to different configurations or performances. As shown in FIG. 7 , the server 2000 includes: a power supply 210 , an interface 230 , at least one memory 250 , and at least one central processing unit (CPU) 270 .

Specifically, the power supply 210 is used to provide operating voltage for each hardware device on the server 2000 .

The interface 230 includes at least one wired or wireless network interface 231 for interacting with external devices.

Of course, in other examples adapted by this application, the interface 230 may further include at least one serial-to-parallel conversion interface 233, at least one input-output interface 235, and at least one USB interface 237, as shown in FIG. 7, which is not specifically limited here.

The memory 250 is a carrier for storing resources, which may be a read-only memory, a random access memory, a disk or an optical disk, etc. The resources stored thereon include an operating system 251, an application 253 and data 255, etc. The storage method may be temporary storage or permanent storage.

Among them, the operating system 251 is used to manage and control the hardware devices and application programs 253 on the server 2000 to enable the central processor 270 to calculate and process the massive data 255 in the memory 250. It can be Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, etc.

The application 253 is a computer program that performs at least one specific task based on the operating system 251, and may include at least one module (not shown in FIG. 7 ), each of which may include a computer program for the server 2000. For example, the application performance optimization device may be regarded as an application 253 deployed on the server 2000.

The data 255 may be photos, pictures, etc. stored in a disk, or may be indicator data, etc. stored in the memory 250 .

The central processor 270 may include one or more processors and is configured to communicate with the memory 250 through at least one communication bus to read the computer program stored in the memory 250, thereby realizing the operation and processing of the mass data 255 in the memory 250. For example, the application performance optimization method is completed in the form of the central processor 270 reading a series of computer programs stored in the memory 250.

In addition, the present application can also be implemented through hardware circuits or hardware circuits combined with software. Therefore, the implementation of the present application is not limited to any specific hardware circuits, software, or a combination of the two.

Please refer to FIG. 8 . An electronic device 4000 is provided in an embodiment of the present application. The electronic device 4000 may include a server in a colocation cluster.

In FIG8 , the electronic device 4000 includes at least one processor 4001, at least one communication bus 4002, and at least one memory 4003. The processor 4001 and the memory 4003 are connected, for example, via the communication bus 4002.

Optionally, the electronic device 4000 may further include a transceiver 4004, which may be used for data interaction between the electronic device and other electronic devices, such as data transmission and/or data reception. It should be noted that in actual applications, the transceiver 4004 is not limited to one, and the structure of the electronic device 4000 does not constitute a limitation on the embodiments of the present application.

Processor 4001 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. It may implement or execute various exemplary logic blocks, modules and circuits described in conjunction with the disclosure of this application. Processor 4001 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

The communication bus 4002 may include a path to transmit information between the above components. The communication bus 4002 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. The communication bus 4002 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, FIG8 only uses one thick line, but does not mean that there is only one bus or one type of bus.

The memory 4003 may be a ROM (Read Only Memory) or other types of static storage devices that can store static information and instructions, a RAM (Random Access Memory) or other types of dynamic storage devices that can store information and instructions, or an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory) or other optical disk storage, optical disk storage (including compressed optical disk, laser disk, optical disk, digital versatile disk, Blu-ray disk, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto.

The memory 4003 stores a computer program, and the processor 4001 reads the computer program stored in the memory 4003 through the communication bus 4002 .

When the computer program is executed by the processor 4001, the application performance optimization method in the above-mentioned embodiments is implemented.

In addition, a storage medium is provided in an embodiment of the present application, on which a computer program is stored. When the computer program is executed by a processor, the application performance optimization method in the above embodiments is implemented.

A computer program product is provided in an embodiment of the present application, the computer program product includes a computer program, the computer program is stored in a storage medium. A processor of a computer device reads the computer program from the storage medium, and the processor executes the computer program, so that the computer device executes the application performance optimization method in each of the above embodiments.

Compared with the related art, the present invention monitors and solves the interference to the application in real time based on the indicator data of each application, and dynamically adjusts the CPU resources based on the indicator data related to the system kernel, which greatly ensures the stability of the hybrid cluster application, improves the application performance and the utilization rate of the whole machine, eliminates the interference caused by the application preempting the CPU core in the hybrid server, and solves the problem of interference in application performance due to CPU core preemption in the hybrid cluster.

It should be understood that, although the steps in the flowchart of the accompanying drawings are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.

It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

An application performance optimization method, characterized in that it is applied to a colocation cluster, and the method comprises:

Acquire indicator data, the indicator data including first indicator data of each application during operation in a current time period, and second indicator data related to the system kernel;

Based on the first indicator data of each application, detecting whether each application is interfered with;

If there is an abnormal application that is disturbed, allocating a CPU core from a CPU shared pool to the abnormal application based on the second indicator data;

According to the CPU core allocated to the abnormal application, a control group Cgroup file of the abnormal application is updated.
The method according to claim 1, characterized in that after updating the control group Cgroup file of the abnormal application according to the CPU core allocated to the abnormal application, the method further comprises:

After the Cgroup file of the abnormal application is updated, based on the first indicator data of the abnormal application during operation in the current time period, detecting whether the abnormal application has returned to normal;

If yes, the CPU core allocated to the abnormal application is restored to the CPU shared pool.
The method according to claim 1, wherein detecting whether each of the applications is interfered with based on the first indicator data of each application comprises:

Obtaining historical indicator data of each of the applications during operation in a historical time period;

Calculating performance fluctuation data of each application respectively according to the first indicator data and the historical indicator data of each application;

If the performance fluctuation data of the application indicates that the application has performance fluctuation, the application is detected as an abnormal application that is disturbed.
The method according to claim 1, wherein allocating a CPU core to the abnormal application from a CPU shared pool based on the second indicator data comprises:

If the abnormal application supports the CPU share mode, the CPU cores in the CPU sharing pool are divided into a plurality of idle levels according to the second indicator data of the system; the CPU cores in the same idle level have the same allocation priority;

According to the number of CPU cores required by the abnormal application and the idle level of the CPU cores, a CPU core is selected from the CPU sharing pool to obtain a CPU allocation policy; the CPU allocation policy is used to indicate the CPU cores that can be allocated to the abnormal application.
The method according to claim 4, wherein the allocating a CPU core to the abnormal application from a CPU shared pool based on the second indicator data further comprises:

Based on the CPU cores that can be allocated to the abnormal application indicated by the CPU allocation policy, detecting whether the CPU allocation policy is correct;

If the CPU core that can be allocated to the abnormal application has been exclusively occupied by other applications supporting the CPU set mode, the CPU allocation policy error is detected, and a CPU core in the CPU shared pool is re-allocated to the abnormal application.
The method according to claim 4, characterized in that the dividing the CPU cores in the CPU sharing pool into a plurality of idle levels according to the second indicator data of the system comprises:

Assign CPU cores in the same socket to the same idle level; or

Assign CPU cores in the same physical core to the same idle level; or

Based on the second indicator data of the system, determine the idleness of the CPU cores, and classify the CPU cores whose determined idleness is within the same set range into the same idle level; or

If there is an exclusively occupied CPU core in the physical core where the CPU core is located, the idle level of the CPU core is higher than the idle levels of other CPU cores; the other CPU core refers to the CPU core that is not exclusively occupied in the physical core where the CPU core is located.
The method according to any one of claims 1 to 6, characterized in that the updating of the Cgroup file of the abnormal application according to the CPU core allocated to the abnormal application comprises:

Determine a Cgroup location corresponding to the abnormal application, and find the Cgroup file of the abnormal application according to the determined Cgroup location;

In the Cgroup file of the abnormal application, the CPU core allocated to the abnormal application is bound to the abnormal application.
An application performance optimization device, characterized in that it is deployed in a colocation cluster, and comprises:

An acquisition module, used to acquire indicator data, wherein the indicator data includes first indicator data of each application during operation in a current time period, and second indicator data related to the system kernel;

An interference detection module, configured to detect whether each of the applications is interfered with based on the first indicator data of each application;

a resource allocation module, configured to allocate a CPU core from a CPU sharing pool to an abnormal application that is disturbed based on the second indicator data if there is an abnormal application that is disturbed;

The file updating module is used to update the control group Cgroup file of the abnormal application according to the CPU core allocated to the abnormal application.
An electronic device, characterized in that it comprises: at least one processor, at least one memory, and at least one communication bus, wherein:

The memory stores a computer program, and the processor reads the computer program in the memory through the communication bus;

When the computer program is executed by the processor, the application performance optimization method according to any one of claims 1 to 7 is implemented.
A storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the application performance optimization method according to any one of claims 1 to 7 is implemented.