A unified construction method for heterogeneous cloud platforms

Abstract

The present invention relates to the technical field of cloud platform construction, and in particular to a method for uniformly constructing a heterogeneous cloud platform, comprising: monitoring whether the operation of the heterogeneous cloud platform is qualified based on the average response time of each task dispatched and the variance of each response time; when determining that the operation of the cloud platform is abnormal, classifying each task abnormality category based on log information, so as to determine the processing method in a targeted manner according to the specific abnormal situation; adjusting the operation parameters of the cloud platform based on the proportion of each task abnormality category, thereby improving the resource management efficiency of the heterogeneous cluster.

Description

Technical Field

[0001] The present invention relates to the technical field of cloud platform construction, and in particular to a method for uniformly constructing a heterogeneous cloud platform. Background Art

[0002] Chinese patent publication number: CN110545260A, discloses a cloud management platform construction method based on mimetic construction, including: a distributor receives user request information from a user end or a host machine, pre-processes it and sends it to multiple active heterogeneous cloud management platform executors in a heterogeneous executor group at the same time; the heterogeneous cloud management platform executors receive the user request information forwarded by the distributor, perform corresponding request processing operations, and send the processing results to the arbitrator; the arbitrator receives the processing result information sent by each heterogeneous cloud management platform executor, and classifies the result information according to the request source, performs consistency voting on multiple processing result information from the same request source, and sends the voted result information to the user end and the host machine after appropriate processing, and the arbitrator reports its own status and ruling information to the negative feedback scheduling unit as needed; after the user end receives the result information sent by the arbitrator, it displays the processing result to the user through an interactive interface; it can be seen that the above technical solution has the following problems: for the heterogeneous cloud platforms under various operating conditions, the same operating parameters are used to complete the processing of tasks, which affects the resource management efficiency of the heterogeneous cluster. Summary of the invention

[0003] To this end, the present invention provides a unified construction method for heterogeneous cloud platforms to overcome the problem in the prior art that heterogeneous cloud platforms in various operating conditions all use the same operating parameters to complete task processing, which affects the resource management efficiency of the heterogeneous cluster.

[0004] To achieve the above object, the present invention provides a method for uniformly constructing a heterogeneous cloud platform, comprising:

[0005] Complete the construction of the cloud platform;

[0006] Dispatching several tasks and determining the response time of each task, determining whether the operation of the cloud platform is qualified based on the average response time and the variance of each response time, and when determining the operation abnormality, classifying the abnormality category of each task into the scheduling abnormality category and the transmission abnormality category based on the log information, and obtaining the proportion of the scheduling abnormality category and the proportion of the transmission abnormality category respectively;

[0007] Determine the processing of the service process based on the comparison result of the proportion of the scheduling abnormality category and the proportion of the transmission abnormality category, or process the service process;

[0008] Determine the handling method for facility deployment based on the proportion of scheduling exception categories, including adjusting the number of servers to a corresponding value or determining changes in heterogeneous support methods;

[0009] Determine the handling method for the service process based on the proportion of transmission exception categories, including adjusting the flow limit standard to the corresponding value, or determining to update the interface;

[0010] When the cloud platform is deemed to be running properly, it will be deployed online and continuously monitored after it is launched.

[0011] Furthermore, the process of determining whether the operation of the cloud platform is qualified based on the average response time and the variance of each response time includes:

[0012] Get the average response time, and determine whether the cloud platform is running properly based on the average response time;

[0013] If the average response time is less than or equal to the first preset time, it is determined that the operation of the cloud platform is qualified;

[0014] If the average response time is less than or equal to the second preset time and greater than the first preset time, determining whether the operation of the cloud platform is qualified based on the variance of each response time;

[0015] If the average response time is longer than the second preset time, the operation of the cloud platform is determined to be abnormal, and each task is classified into abnormal categories based on the log information, and the operation parameters of the cloud platform are adjusted based on the proportion of each task abnormal category.

[0016] Furthermore, the process of determining whether the operation of the cloud platform is qualified based on the variance of each response time includes:

[0017] If the variance of each response time is less than or equal to the preset time variance, the cloud platform is judged to be operating abnormally, and the abnormal categories of each task are divided based on the log information, and the operating parameters of the cloud platform are adjusted based on the proportion of each task abnormal category;

[0018] If the variance of each response time is greater than the preset time variance, the first preset time is adjusted based on the response floating parameter, and whether the operation of the cloud platform is qualified is re-determined based on the adjusted first preset time.

[0019] Further, the process of adjusting the first preset duration based on the response floating parameter includes:

[0020] Determine the response floating parameter, calculate the difference between the variance of each response time and the preset time variance, and obtain the response floating parameter;

[0021] The first preset time length determined based on the response floating parameter is positively correlated with the response floating parameter.

[0022] Furthermore, the process of classifying each task exception category based on log information includes:

[0023] Filter logs related to task scheduling and traverse the logs to identify abnormal prompt entries;

[0024] Determine the task with the task retry prompt entry as a scheduling exception category;

[0025] Tasks with packet loss prompt entries are identified as transmission anomalies.

[0026] Furthermore, the process of adjusting the operating parameters of the cloud platform based on the abnormality category ratio of each task includes:

[0027] Determine the proportion of scheduling exception categories, calculate the proportion of the number of scheduling exception category tasks to the total number of tasks, and obtain the proportion of scheduling exception categories;

[0028] Determine the proportion of transmission anomaly categories, calculate the proportion of the number of transmission anomaly category tasks to the total number of tasks, and obtain the proportion of transmission anomaly categories;

[0029] The proportion of scheduling anomaly categories is compared with the proportion of transmission anomaly categories, where:

[0030] If the proportion of the scheduling anomaly category is greater than the proportion of the transmission anomaly category, a handling method for facility deployment is determined based on the proportion of the scheduling anomaly category;

[0031] If the proportion of the scheduling exception category is less than or equal to the proportion of the transmission exception category, a processing method for the service process is determined based on the proportion of the transmission exception category.

[0032] Furthermore, based on the proportion of the scheduling exception categories, a processing method for facility deployment is determined, including:

[0033] If the ratio of the scheduling abnormality category is less than or equal to the preset abnormality ratio, the number of servers is adjusted to a corresponding value based on the difference between the preset abnormality ratio and the ratio of the scheduling abnormality category;

[0034] If the proportion of scheduling exception categories is greater than the preset exception proportion, it is determined that the heterogeneous support mode is changed.

[0035] Furthermore, based on the proportion of transmission anomaly categories, the processing method for the service process is determined, including:

[0036] If the ratio of transmission abnormal categories is less than or equal to the preset abnormal ratio, the current limiting standard is adjusted to the corresponding value based on the CPU utilization;

[0037] If the ratio of the transmission abnormality category is greater than the preset abnormality ratio, it is determined to update the interface.

[0038] Furthermore, the number of servers is adjusted to a corresponding value based on the difference between the preset abnormal ratio and the ratio of the scheduling abnormality category, wherein:

[0039] The difference between the preset abnormality ratio and the scheduling abnormality category is recorded as the scheduling ratio difference;

[0040] The adjustment range of the number of servers determined based on the scheduling ratio difference is inversely proportional to the scheduling ratio difference.

[0041] Furthermore, the current limiting standard is adjusted to a corresponding value based on the CPU utilization, wherein:

[0042] The current limiting standard determined based on CPU utilization is inversely proportional to the CPU utilization.

[0043] Compared with the prior art, the beneficial effect of the present invention lies in that the operation of the heterogeneous cloud platform is monitored based on the average response time of each task dispatched and the variance of each response time. When the operation abnormality of the cloud platform is determined, the task abnormality category is divided based on the log information, so as to determine the processing method according to the specific abnormal situation, and the operation parameters of the cloud platform are adjusted based on the proportion of each task abnormality category, thereby improving the resource management efficiency of the heterogeneous cluster.

[0044] Furthermore, when the average response time is less than or equal to the second preset time and greater than the first preset time, the variance of each response time is comprehensively considered to determine whether the operation of the heterogeneous cloud platform is qualified; when the variance of each response time is greater than the preset time variance, there are some tasks whose response time is too long, lowering the average response time. In this case, considering that the monitoring deviation caused by network anomalies in some areas affects the judgment accuracy, the evaluation standard of the average response time is adjusted, and the qualified operation of the cloud platform is re-determined based on the adjusted evaluation standard, thereby improving the resource management efficiency of the heterogeneous cluster.

[0045] Furthermore, when determining that the cloud platform is operating abnormally, specific solutions are determined, and each task abnormality category is divided according to the log. The scheduling abnormality category represents that the response time is too long due to task scheduling abnormality, and the transmission abnormality category represents that the response time is long due to data transmission abnormality. The proportion of each category is counted. When the proportion of the scheduling abnormality category is greater than the proportion of the transmission abnormality category, there are a large number of situations where the response time is abnormal due to task scheduling abnormality. The scheduling abnormality is processed. At this time, the processing facility is deployed based on the proportion of the scheduling abnormality category. When the proportion of the scheduling abnormality category is greater than the preset abnormality proportion, due to the policy abnormalities in the resource allocation and management of the heterogeneous architecture, the compatibility abnormalities between the heterogeneous mode and the components or systems on which the task depends, there are a large number of tasks with too long response time. At this time, the heterogeneous support mode is determined to be changed; when the proportion of the scheduling abnormality category is less than or equal to the preset abnormality proportion, there are a small number of tasks with too long response time due to scheduling problems. At this time, the number of servers is increased to improve data processing efficiency, improve the operating stability of the heterogeneous cloud platform, and further improve the utilization of servers.

[0046] Furthermore, when the proportion of scheduling exception categories is less than or equal to the proportion of transmission exception categories, data transmission exceptions cause the task response time to be too long. In this case, the service process is processed. When the proportion of transmission exception categories is greater than the preset exception proportion, there are a large number of tasks with data transmission exceptions. In this case, due to compatibility issues with the interface, improper data packet processing leads to loss, and the interface is updated; when the proportion of transmission exception categories is less than or equal to the preset exception proportion, there are only a small number of cases where the task response time is too long due to unreasonable task allocation caused by too low a current limiting standard. At this time, the current limiting standard is adjusted, which effectively improves the response speed of the task while improving the cloud platform's processing efficiency for the task, further improving the resource management efficiency of the heterogeneous cluster. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 A flowchart of the steps of the method for uniformly building a heterogeneous cloud platform according to an embodiment of the present invention;

[0048] Figure 2 A logic decision diagram for determining whether a construction is qualified based on the average response time in an embodiment of the present invention;

[0049] Figure 3 A logic decision diagram is constructed for determining whether a qualified logic decision diagram is constructed based on the variance of each response time in an embodiment of the present invention;

[0050] Figure 4 This is a logic decision diagram for determining a processing method for facility deployment based on the proportion of scheduling exception categories according to an embodiment of the present invention. DETAILED DESCRIPTION

[0051] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0052] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0053] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the drawings. This is merely for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.

[0054] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0055] See also Figure 1 , Figure 2 , Figure 3 as well as Figure 4 As shown, they are respectively a flowchart of a method for uniformly building a heterogeneous cloud platform according to an embodiment of the present invention, a logic decision diagram for determining whether a construction is qualified based on an average response time, a logic decision diagram for determining whether a construction is qualified based on a variance of each response time, and a logic decision diagram for determining a processing method for facility deployment based on a proportion of scheduling abnormality categories; a method for uniformly building a heterogeneous cloud platform according to an embodiment of the present invention, comprising:

[0056] Upgrade and adjust hardware, install virtualization software and operating systems on servers, set up and manage network infrastructure, configure virtualization environments, build containerized environments, and build storage resource pools;

[0057] Integrate automated deployment tools and build monitoring and alarm systems;

[0058] Dispatching several tasks and determining the response time of each task, determining whether the cloud platform is running properly based on the average response time and the variance of each response time, and when determining that the cloud platform is running abnormally, classifying each task abnormality category based on log information, and adjusting the cloud platform's operating parameters based on the proportion of each task abnormality category, including:

[0059] Changes to heterogeneous support methods, redetermination of server quantity, interface updates, or redetermination of current limiting standards;

[0060] Deploy and launch the system, and then continuously monitor it after it goes online.

[0061] Specifically, the facility deployment process includes upgrading and adjusting hardware, installing virtualization software and operating systems on servers, setting up and managing network infrastructure, including physical networks and virtual networks; installing KVM on servers and configuring virtualization environments, installing Docker and Kubernetes to build containerized environments, and using distributed storage services or GlusterFS to build storage resource pools. These will not be elaborated here.

[0062] Specifically, the process of integrating automation tools and determining service processes includes integrating Ansible and Terraform automated deployment tools, building Prometheus and Grafana monitoring and alarm systems; analyzing business processes, determining service process requirements, designing service processes, and integrating Kubernetes orchestration tools to optimize service processes. This will not be repeated here.

[0063] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0064] It should be pointed out that the data in this embodiment are obtained by comprehensive analysis and evaluation of the historical detection data and corresponding historical detection results of the three months before this detection by the method described in the present invention. Before this detection, the method described in the present invention comprehensively determines the values ​​of various preset parameter standards for this detection based on the task dispatching of 25,863 batches detected cumulatively in the first three months and the response time of each task of each batch of dispatched tasks, the detection data information of each batch of tasks and the judgment results. It can be understood by those skilled in the art that the method described in the present invention can be used to determine the single parameter mentioned above by selecting the value with the highest proportion as the preset standard parameter according to the data distribution, using weighted summation to use the obtained value as the preset standard parameter or other selection methods, as long as the method described in the present invention can clearly define different specific situations in the single judgment process through the obtained value.

[0065] Specifically, the process of determining whether the operation of the cloud platform is qualified based on the average response time and the variance of each response time includes:

[0066] Determine whether the cloud platform is running properly based on the average response time;

[0067] If the average response time is less than or equal to the first preset time, the cloud platform is judged to be running properly and deployed online;

[0068] If the average response time is less than or equal to the second preset time and greater than the first preset time, determining whether the operation of the cloud platform is qualified based on the variance of each response time;

[0069] If the average response time is longer than the second preset time, the operation of the cloud platform is determined to be abnormal, and each task is classified into abnormal categories based on the log information, and the operation parameters of the cloud platform are adjusted based on the proportion of each task abnormal category.

[0070] Specifically, the first preset time length T1 is selected within the interval [0.5, 2], and the second preset time length T2 is selected within the interval [5, 30], wherein the unit of the preset time length is seconds.

[0071] Specifically, the operation of the heterogeneous cloud platform is monitored based on the average response time of each task dispatched and the variance of each response time. When the operation abnormality of the cloud platform is determined, the task abnormality category is divided based on the log information, so as to determine the processing method according to the specific abnormal situation. The operation parameters of the cloud platform are adjusted based on the proportion of each task abnormality category, which improves the resource management efficiency of the heterogeneous cluster.

[0072] Specifically, the process of determining whether the operation of the cloud platform is qualified based on the variance of each response time includes:

[0073] If the variance of each response time is less than or equal to the preset time variance, the cloud platform is judged to be operating abnormally, and the abnormal categories of each task are divided based on the log information, and the operating parameters of the cloud platform are adjusted based on the proportion of each task abnormal category;

[0074] If the variance of each response time is greater than the preset time variance, the first preset time is adjusted based on the response floating parameter, and whether the operation of the cloud platform is qualified is re-determined based on the adjusted first preset time.

[0075] Specifically, the preset time variance C0 is selected within the interval [0.05T, 0.09T], and T is the average response time.

[0076] Specifically, when the average response time is less than or equal to the second preset time and greater than the first preset time, the variance of each response time is comprehensively considered to determine whether the operation of the heterogeneous cloud platform is qualified; when the variance of each response time is greater than the preset time variance, there are some tasks with excessively long response times, which lowers the average response time. In this case, considering that the monitoring deviation caused by network anomalies in some areas affects the judgment accuracy, the evaluation standard of the average response time is adjusted, and the qualified operation of the cloud platform is re-determined based on the adjusted evaluation standard, thereby improving the resource management efficiency of the heterogeneous cluster.

[0077] Specifically, the process of adjusting the first preset duration based on the response floating parameter includes:

[0078] Determine the response floating parameter, calculate the difference between the variance of each response time and the preset time variance, and obtain the response floating parameter;

[0079] The first preset time length determined based on the response floating parameter is positively correlated with the response floating parameter.

[0080] In this embodiment, optionally,

[0081] Compare the response floating parameter with the first preset floating comparison threshold and the second preset floating comparison threshold,

[0082] If the response floating parameter is less than or equal to the first preset floating comparison threshold, the first preset duration is adjusted to the corresponding value using the first preset evaluation adjustment coefficient, and the first preset duration adjusted using the first preset evaluation adjustment coefficient is 1.11 times the initial first preset duration;

[0083] If the response floating parameter is less than or equal to the second preset floating comparison threshold and greater than the first preset floating comparison threshold, the first preset duration is adjusted to the corresponding value using the second preset evaluation adjustment coefficient, and the first preset duration adjusted using the second preset evaluation adjustment coefficient is 1.17 times the initial first preset duration;

[0084] If the response floating parameter is greater than the second preset floating comparison threshold, the first preset duration is adjusted to the corresponding value using the third preset evaluation adjustment coefficient, and the first preset duration adjusted using the third preset evaluation adjustment coefficient is 1.23 times the initial first preset duration;

[0085] Among them, the first preset floating comparison threshold is 2.1C0, and the second preset floating comparison threshold is 5.3C0.

[0086] Specifically, the process of re-determining whether the operation of the cloud platform is qualified based on the adjusted first preset time includes:

[0087] If the average response time is less than or equal to the adjusted first preset time, the cloud platform is judged to be running properly and deployed online;

[0088] If the average response time is greater than the adjusted first preset time, the operation of the cloud platform is determined to be abnormal, and each task is classified into abnormal categories based on log information, and the operation parameters of the cloud platform are adjusted based on the proportion of each task abnormal category.

[0089] Specifically, the process of classifying each task exception category based on log information includes:

[0090] Filter logs related to task scheduling and traverse the logs to identify abnormal prompt entries;

[0091] Determine the task with the task retry prompt entry as a scheduling exception category;

[0092] Tasks with packet loss prompt entries are identified as transmission anomalies.

[0093] Specifically, the process of adjusting the operating parameters of the cloud platform based on the abnormality category ratio of each task includes:

[0094] Determine the proportion of scheduling exception categories, calculate the proportion of the number of scheduling exception category tasks to the total number of tasks, and obtain the proportion of scheduling exception categories;

[0095] Determine the proportion of transmission anomaly categories, calculate the proportion of the number of transmission anomaly category tasks to the total number of tasks, and obtain the proportion of transmission anomaly categories;

[0096] Compare the proportion of the scheduling anomaly category with the proportion of the transmission anomaly category;

[0097] If the proportion of the scheduling anomaly category is greater than the proportion of the transmission anomaly category, a handling method for facility deployment is determined based on the proportion of the scheduling anomaly category;

[0098] If the proportion of the scheduling exception category is less than or equal to the proportion of the transmission exception category, a processing method for the service process is determined based on the proportion of the transmission exception category.

[0099] Specifically, the handling method for facility deployment is determined based on the proportion of scheduling exception categories, including:

[0100] If the ratio of the scheduling abnormality category is less than or equal to the preset abnormality ratio, the number of servers is adjusted to a corresponding value based on the difference between the preset abnormality ratio and the ratio of the scheduling abnormality category;

[0101] If the proportion of scheduling exception categories is greater than the preset exception proportion, it is determined that the heterogeneous support mode is changed.

[0102] Specifically, the preset abnormality ratio S0 is selected within the interval [0.62, 0.78].

[0103] Specifically, when determining that the cloud platform is operating abnormally, specific solutions are determined, and each task is classified into abnormal categories based on logs. The scheduling abnormality category represents that the response time is too long due to task scheduling abnormalities, and the transmission abnormality category represents that the response time is too long due to data transmission abnormalities. The proportion of each category is counted. When the proportion of the scheduling abnormality category is greater than the proportion of the transmission abnormality category, there are a large number of situations where the response time is abnormal due to task scheduling abnormalities. The scheduling abnormalities are processed. At this time, the processing facility is deployed based on the proportion of the scheduling abnormality category. When the proportion of the scheduling abnormality category is greater than the preset abnormality proportion, due to the policy abnormalities in the resource allocation and management of the heterogeneous architecture, the compatibility abnormalities between the heterogeneous mode and the components or systems on which the tasks depend, there are a large number of tasks with too long response times. At this time, the heterogeneous support mode is determined to be changed; when the proportion of the scheduling abnormality category is less than or equal to the preset abnormality proportion, there are a small number of tasks with too long response times due to scheduling problems. At this time, the number of servers is increased to improve data processing efficiency, improve the operating stability of the heterogeneous cloud platform, and further improve the utilization of servers.

[0104] Specifically, this solution meets the compatibility requirements of CPUs of different architectures. This solution adopts the CPUID instruction set and microcode update technology. Through the CPUID instruction set, the characteristics of the processor can be identified, thereby providing a corresponding operating environment for the virtual machine. The microcode update can ensure that the processor can support the latest instruction set and features. This solution achieves compatibility with multi-architecture CPUs such as x86, ARM, MIPS, etc., greatly improving the applicability and flexibility of the virtualization platform; the change in heterogeneous support mode is to determine the microcode update.

[0105] Specifically, the handling method for the service process is determined based on the proportion of transmission exception categories, including:

[0106] If the ratio of transmission abnormal categories is less than or equal to the preset abnormal ratio, the current limiting standard is adjusted to the corresponding value based on the CPU utilization;

[0107] If the ratio of the transmission abnormality category is greater than the preset abnormality ratio, it is determined to update the interface.

[0108] Specifically, when the proportion of scheduling exception categories is less than or equal to the proportion of transmission exception categories, data transmission exceptions cause the task response time to be too long. In this case, the service process is processed. When the proportion of transmission exception categories is greater than the preset exception proportion, there are a large number of tasks with data transmission exceptions. In this case, due to compatibility issues with the interface, improper data packet processing leads to loss, and the interface is updated; when the proportion of transmission exception categories is less than or equal to the preset exception proportion, there are only a small number of cases where the task response time is too long due to unreasonable task allocation caused by too low a current limiting standard. At this time, the current limiting standard is adjusted, which effectively improves the response speed of the task while improving the cloud platform's processing efficiency for the task, and further improves the resource management efficiency of the heterogeneous cluster.

[0109] Specifically, for interface updates, the cloud engine provides physical server interfaces, virtual machine interfaces, storage device interfaces, and network interfaces; supports the management of X86 physical servers and minicomputer server resource pools, and provides management functions such as adding, controlling, removing, and status querying. On the virtualization platform, the cloud engine supports access management and supports the access of PowerVM, VMWare, XEN, and KYM server virtualization technologies, and provides virtual machine creation, operation, deletion, configuration change, and status query. Due to the different open interfaces of each virtualization technology, they are handled separately when performing access management of the underlying virtualization technology. For interface updates, determine the currently used interface version, determine the available interface update version, automatically select the version to back up related configuration files and data, and install and configure the new interface. This is a prior art and will not be repeated here.

[0110] Specifically, the CPU utilization is the proportion of time that the CPU is in working state during the historical time period; there is no limitation on the selected duration of the historical time period. It can be understood that the workload and resource utilization can be characterized, which will not be elaborated here.

[0111] Specifically, the current limiting standard is the maximum amount of data transmission allowed to pass within a specific time interval, which will not be described in detail here.

[0112] Specifically, the number of servers is adjusted to a corresponding value based on the difference between the preset abnormal ratio and the ratio of the scheduling abnormal category, where:

[0113] The difference between the preset abnormality ratio and the scheduling abnormality category is recorded as the scheduling ratio difference;

[0114] The adjustment range of the number of servers determined based on the scheduling ratio difference is inversely proportional to the scheduling ratio difference.

[0115] In this embodiment, optionally,

[0116] The scheduling ratio difference is compared with the first preset scheduling comparison threshold and the second preset scheduling comparison threshold,

[0117] If the scheduling ratio difference is less than or equal to the first preset scheduling comparison threshold, the number of servers is determined to be the first number, and the first number is set to 1.32 times the number of current servers;

[0118] If the scheduling ratio difference is less than or equal to the second preset scheduling comparison threshold and greater than the first preset scheduling comparison threshold, the number of servers is determined as the second number, and the second number is set to 1.25 times the number of current servers;

[0119] If the scheduling ratio difference is greater than the second preset scheduling comparison threshold, the number of servers is determined to be a third number, and the third number is set to 1.11 times the number of current servers;

[0120] Among them, the first preset scheduling comparison threshold is 0.25S0, and the second preset scheduling comparison threshold is 0.63S0.

[0121] Specifically, the current limiting standard is adjusted to the corresponding value based on the CPU utilization, where:

[0122] The current limiting standard determined based on CPU utilization is inversely proportional to the CPU utilization.

[0123] In this embodiment, optionally,

[0124] The CPU utilization is compared with a first preset occupancy comparison threshold and a second preset occupancy comparison threshold,

[0125] If the CPU utilization is less than or equal to the first preset occupancy comparison threshold, the current limiting standard is determined as the first current limiting standard, and the first current limiting standard is set to 1.35 times the initial current limiting standard;

[0126] If the CPU utilization is less than or equal to the second preset occupancy comparison threshold and greater than the first preset occupancy comparison threshold, the current limiting standard is determined as the second current limiting standard, and the second current limiting standard is set to 1.24 times the initial current limiting standard;

[0127] If the CPU utilization is greater than the second preset occupancy comparison threshold, the current limiting standard is determined as the third current limiting standard, and the third current limiting standard is set to 1.12 times the initial current limiting standard;

[0128] Among them, the first preset occupancy comparison threshold is 50%, and the second preset occupancy comparison threshold is 70%.

[0129] Specifically, when the adjustment of the current limiting standard is completed, the adjusted current limiting standard is compared with the preset maximum current limiting standard. If the adjusted current limiting standard is less than or equal to the preset maximum current limiting standard, it is determined to use the adjusted current limiting standard as the operating parameter of the cloud platform; if the adjusted current limiting standard is greater than the preset maximum current limiting standard, it is determined to use the preset maximum current limiting standard as the operating parameter of the cloud platform, and the number of servers is determined as the third number.

[0130] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

[0131] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for uniformly constructing a heterogeneous cloud platform, characterized in that: include: Complete the construction of the cloud platform; Dispatching several tasks and determining the response time of each task, determining whether the operation of the cloud platform is qualified based on the average response time and the variance of each response time, and when determining the operation abnormality, classifying the abnormality category of each task into the scheduling abnormality category and the transmission abnormality category based on the log information, and obtaining the proportion of the scheduling abnormality category and the proportion of the transmission abnormality category respectively; Determine the processing of the service process based on the comparison result of the proportion of the scheduling abnormality category and the proportion of the transmission abnormality category, or process the service process; Determine the handling method for facility deployment based on the proportion of scheduling exception categories, including adjusting the number of servers to a corresponding value or determining changes in heterogeneous support methods; Determine the handling method for the service process based on the proportion of transmission exception categories, including adjusting the flow limit standard to the corresponding value, or determining to update the interface; When the cloud platform is deemed to be running properly, it will be deployed online and continuously monitored after it is launched. The process of determining whether the operation of the cloud platform is qualified based on the average response time and the variance of each response time includes: Get the average response time, and determine whether the cloud platform is running properly based on the average response time; If the average response time is less than or equal to the first preset time, it is determined that the operation of the cloud platform is qualified; If the average response time is less than or equal to the second preset time and greater than the first preset time, determining whether the operation of the cloud platform is qualified based on the variance of each response time; If the average response time is longer than the second preset time, the operation of the cloud platform is determined to be abnormal, and each task is classified into abnormal categories based on the log information, and the operation parameters of the cloud platform are adjusted based on the proportion of each task abnormal category; The process of determining whether the operation of the cloud platform is qualified based on the variance of each response time includes: If the variance of each response time is less than or equal to the preset time variance, the cloud platform is judged to be operating abnormally, and each task is classified into abnormal categories based on the log information, and the operating parameters of the cloud platform are adjusted based on the proportion of each task abnormal category; If the variance of each response time is greater than the preset time variance, adjusting the first preset time based on the response floating parameter, and re-determining whether the operation of the cloud platform is qualified based on the adjusted first preset time; The process of adjusting the first preset duration based on the response floating parameter includes: Determine the response floating parameter, calculate the difference between the variance of each response time and the preset time variance, and obtain the response floating parameter; The first preset time length determined based on the response floating parameter is positively correlated with the response floating parameter; The process of classifying each task exception category based on log information includes: Filter logs related to task scheduling and traverse the logs to identify abnormal prompt entries; Determine the task with the task retry prompt entry as a scheduling exception category; The tasks with the packet loss prompt entry are identified as the transmission anomaly category; The process of adjusting the operating parameters of the cloud platform based on the abnormality category ratio of each task includes: Determine the proportion of scheduling exception categories, calculate the proportion of the number of scheduling exception category tasks to the total number of tasks, and obtain the proportion of scheduling exception categories; Determine the proportion of transmission anomaly categories, calculate the proportion of the number of transmission anomaly category tasks to the total number of tasks, and obtain the proportion of transmission anomaly categories; The proportion of scheduling anomaly categories is compared with the proportion of transmission anomaly categories, where: If the proportion of the scheduling anomaly category is greater than the proportion of the transmission anomaly category, a handling method for facility deployment is determined based on the proportion of the scheduling anomaly category; If the proportion of the scheduling exception category is less than or equal to the proportion of the transmission exception category, a processing method for the service process is determined based on the proportion of the transmission exception category.

2. The method for uniformly constructing a heterogeneous cloud platform according to claim 1, characterized in that: The handling method for facility deployment is determined based on the proportion of scheduling exception categories, including: If the ratio of the scheduling abnormality category is less than or equal to the preset abnormality ratio, the number of servers is adjusted to a corresponding value based on the difference between the preset abnormality ratio and the ratio of the scheduling abnormality category; If the proportion of scheduling exception categories is greater than the preset exception proportion, it is determined that the heterogeneous support mode is changed.

3. The method for uniformly constructing a heterogeneous cloud platform according to claim 2, characterized in that: The handling method for the service process is determined based on the proportion of transmission exception categories, including: If the ratio of transmission abnormal categories is less than or equal to the preset abnormal ratio, the current limiting standard is adjusted to the corresponding value based on the CPU utilization; If the ratio of the transmission abnormality category is greater than the preset abnormality ratio, it is determined to update the interface.

4. The method for uniformly constructing a heterogeneous cloud platform according to claim 3, characterized in that: The number of servers is adjusted to a corresponding value based on the difference between the preset abnormal ratio and the ratio of the scheduling abnormal category, where: The difference between the preset abnormality ratio and the scheduling abnormality category is recorded as the scheduling ratio difference; The adjustment range of the number of servers determined based on the scheduling ratio difference is inversely proportional to the scheduling ratio difference.

5. The method for uniformly constructing a heterogeneous cloud platform according to claim 4, characterized in that: The current limiting standard is adjusted to the corresponding value based on the CPU utilization, where: The current limiting standard determined based on CPU utilization is inversely proportional to the CPU utilization.

Images