WO2020211253A1

WO2020211253A1 - Elastic scaling method and apparatus for number of hosts in distributed system, and computer device

Info

Publication number: WO2020211253A1
Application number: PCT/CN2019/103546
Authority: WO
Inventors: 赵骏
Original assignee: 平安科技（深圳）有限公司
Priority date: 2019-04-16
Filing date: 2019-08-30
Publication date: 2020-10-22
Also published as: CN110175068A

Abstract

An elastic scaling method and apparatus for the number of hosts in a distributed system, and a computer device. Said method comprises: monitoring, in real time, index parameter data of a Docket container corresponding to hosts (S1); comparing the index parameter data with a preset standard index threshold to obtain a comparison result (S2); and according to the comparison result, performing elastic scaling processing corresponding to the current traffic on the number of hosts by means of a preset host scaling strategy (S3). The present invention can effectively and accurately achieve the elastic scaling for the number of hosts.

Description

Method, device and computer equipment for elastic scaling of host number in distributed system

This application claims the priority of a Chinese patent application filed with the Chinese Patent Office on April 16, 2019, the application number is 201910305187.9, and the invention title is "Methods, devices, and computer equipment for the elastic scaling of the number of hosts in a distributed system", and its entire contents Incorporated in this application by reference.

Technical field

This application relates to the field of cloud computing technology, and in particular to a method, device and computer equipment for elastic scaling of the number of hosts in a distributed system.

Background technique

Many existing distributed systems deployed with Docker containers do not have the ability to elastically scale hosts. It is often necessary to prepare enough redundant hosts in advance to meet the needs of distributed systems at the peak of business calls. When the integrated system is in the low period of business call, a lot of idle resources will be wasted. In addition, there are also some distributed systems with the ability of host elastic scaling, but the host elastic scaling of these distributed systems is based on periodic elastic scaling, that is, the elastic scaling of the host is realized through a fixed cycle, for example, distributed The business call usage of the system is very large during the working hours from 9:00 to 18:00. You can customize the host to expand a certain number of hosts before 9:00 every day, and to shrink a certain number of hosts after 18:00 every day. This kind of periodic elastic scaling cannot perform timely elastic scaling according to the actual data volume of the business call, the elastic scaling of the host is not accurate, the flexibility is low, and the resource utilization rate of the distributed system is not high.

technical problem

The main purpose of this application is to provide a method, device, computer equipment and storage medium for elastic scaling of the number of hosts in a distributed system, which aims to solve the problem that the existing distributed system does not have the ability of host elastic scaling and has periodicity-based Distributed systems with elastic scaling capabilities will have technical problems such as inaccurate elastic scaling of the host and low flexibility.

Technical solutions

In order to achieve the purpose of the above application, this application proposes a method for elastic scaling of the number of hosts in a distributed system. The method includes the steps:

Monitor the indicator parameter data of the Docker container corresponding to the host in real time, where the indicator parameter data is parameter data corresponding to the indicator input by the user, and the indicator is set by the user according to the Docker container corresponding to the host. Indicators include one or more of system response time, the number of pod functions in the Docker container, the resource consumption of the programs running in the Docker container, CPU usage, and memory usage;

Comparing the index parameter data with a preset standard index threshold to obtain a comparison result, where the standard index threshold includes at least a first preset index threshold corresponding to the expansion process;

Judging whether the comparison result is that the index parameter data is greater than the first preset index threshold;

If yes, obtain the current first business volume data;

Expanding the current number of first original hosts to the number of first hosts matching the first service volume data.

This application also provides an elastic scaling device for the number of hosts in a distributed system, including:

The monitoring module is used to monitor the indicator parameter data of the Docker container corresponding to the host in real time, wherein the indicator parameter data is parameter data corresponding to the indicator input by the user, and the indicator is determined by the user according to the Docker container corresponding to the host Setting, the indicators include one or more of system response time, the number of pod functions in the Docker container, the resource consumption of programs running in the Docker container, the CPU usage rate, and the memory usage rate;

A comparison module, configured to compare the index parameter data with a preset standard index threshold to obtain a comparison result, where the standard index threshold includes at least a first preset index threshold corresponding to the expansion process;

The first judgment module is configured to judge whether the comparison result is that the index parameter data is greater than the first preset index threshold;

The first obtaining module is used to obtain the current first business volume data if yes;

The expansion module is used to expand the current number of first original hosts to the number of first hosts matching the first service volume data.

The application also provides a computer device, including a memory and a processor, the memory stores computer readable instructions, and the processor implements the steps of the above method when the computer readable instructions are executed.

The present application also provides a computer non-volatile readable storage medium on which computer readable instructions are stored, and when the computer readable instructions are executed by a processor, the steps of the above method are realized.

Beneficial effect

The method, device, computer equipment and storage medium for elastic scaling of the number of hosts in the distributed system provided in this application monitor the indicator parameter data of the Docker container corresponding to the host in real time; compare the indicator parameter data with the preset standard indicator threshold to obtain Comparison result: According to the comparison result, the number of hosts is subjected to elastic scaling processing corresponding to the current business volume through a preset host scaling strategy, and the elastic scaling processing includes expansion processing or shrinking processing. This application monitors the indicator parameter data of the Docker container corresponding to each host in real time after pre-setting the indicator of the Docker container corresponding to the host and the standard indicator threshold corresponding to the indicator, and then the indicator parameter data and the standard indicator threshold The comparison results are obtained, and the above host scaling strategy is used according to the comparison results to expand or shrink the current number of hosts corresponding to the current business volume, so as to realize the control of the host based on the real-time index parameter data of each host. Effective and accurate elastic scaling of the quantity ensures the efficient and stable operation of the system and effectively improves the resource utilization rate of the system.

Description of the drawings

FIG. 1 is a schematic flowchart of a method for elastic scaling of the number of hosts in a distributed system according to an embodiment of the present application;

2 is a schematic flowchart of a method for elastic scaling of the number of hosts in a distributed system according to another embodiment of the present application;

FIG. 3 is a schematic structural diagram of an elastic scaling device for the number of hosts in a distributed system according to an embodiment of the present application;

Fig. 4 is a schematic structural diagram of a computer device according to an embodiment of the present application.

The best mode of the invention

It should be understood that the specific embodiments described here are only used to explain the application, and are not used to limit the application.

1, the method for elastically scaling the number of hosts in a distributed system according to an embodiment of the present application includes:

S1: Monitor the indicator parameter data of the Docker container corresponding to the host in real time, where the indicator parameter data is parameter data corresponding to an indicator input by the user, and the indicator is set by the user according to the Docker container corresponding to the host, The indicators include one or more of system response time, the number of pod functions in the Docker container, the resource consumption of programs running in the Docker container, the CPU usage rate, and the memory usage rate;

S2: Compare the index parameter data with a preset standard index threshold to obtain a comparison result;

S3: According to the comparison result, the number of hosts is subjected to an elastic scaling process corresponding to the current business volume through a preset host scaling strategy, and the elastic scaling process includes capacity expansion processing or shrinking processing.

As described in step S1 above, the designated subject of the method in this embodiment is an elastic scaling device for the number of hosts in a distributed system, which can be applied in a distributed system. The aforementioned distributed system is a software system built on the network, the aforementioned host is a Node host installed in the distributed system, the aforementioned Docker container runs on the Node host, and one Node host may include multiple Docker containers, where Docker container is an open source application container engine that allows developers to package their applications and dependent packages into a portable container, and then publish to any popular Linux machine, which can also be virtualized. The above indicator parameter data is the parameter data corresponding to the indicator input by the user. The above indicator is set by the user according to the Docker container corresponding to the host. The above indicator corresponding to the Docker container of the host includes the system response time and the pod function in the Docker container. One or more of the number, the resource consumption of the program running in the Docker container, CPU usage, memory usage, etc. The above host scaling strategy is set with a corresponding standard index threshold for each index, and each standard index threshold is correspondingly set with a first preset standard threshold corresponding to the expansion process and a second preset standard corresponding to the scaling process The threshold, that is, the first preset standard threshold and the second preset standard threshold are respectively used to determine whether to perform expansion processing or shrinking processing on the number of hosts. The above host scaling strategy is an evaluation condition set based on the above indicators. After obtaining the above indicator parameter data, the number of hosts can be expanded or contracted according to the host scaling strategy. For example, the host scaling strategy It can include the first host scaling strategy and the second host scaling strategy. If the currently detected indicator is the CPU usage rate, the standard indicator value corresponding to the indicator is the CPU usage rate: the first preset standard threshold is 70%, and the second preset If the standard threshold is 30%, the first host scaling strategy is: when the CPU usage is greater than 70%, the number of hosts will be expanded. The number of expansions can be determined according to the current actual business volume; the second host will scale The strategy is: when the CPU usage rate is less than 30%, the number of hosts is scaled down, and the number of downscaling can also be determined according to the current actual business volume. In this embodiment, the indicator parameter data of the Docker container corresponding to each host is monitored in real time, and then the indicator parameter data is compared with the standard indicator value to obtain a comparison result, and the above host scaling strategy is used according to the comparison result to compare the current The number of hosts performs elastic scaling processing corresponding to the current business volume, that is, expansion processing or shrinking processing, which realizes the effective and accurate elastic scaling of the number of hosts according to the real-time index parameter data of each host, ensuring the efficient and stable operation of the system , Effectively improve the resource utilization rate of the system.

2, further, in an embodiment of the present application, the standard index threshold includes a first preset index threshold corresponding to the expansion process, and the above step S3 includes:

S300: Determine whether the comparison result is that the index parameter data is greater than the first preset index threshold;

S301: If yes, obtain the current first business volume data;

S302: Expand the current number of first original hosts to the number of first hosts matching the first service volume data.

As described in the above steps S300 to S302, the above indicators may include system response time, the number of pod functions in the Docker container, the resource consumption of programs running in the Docker container, CPU usage, memory usage, etc. In this embodiment The standard index threshold includes the first preset index threshold corresponding to the expansion process. After comparing the index parameter data with the preset standard index threshold to obtain the comparison result, it is further judged whether the comparison result is an index parameter The data is greater than the first preset index threshold. When the comparison result is that the index parameter data is greater than the first preset index threshold, that is, when the index parameter data obtained by real-time detection is greater than the first preset standard threshold, the expansion process will be performed . Wherein, the aforementioned first preset standard threshold corresponds to a plurality of preset thresholds corresponding to each different index. The above-mentioned host expansion processing may specifically include the following: when the above-mentioned indicator is the system response time, the first preset index threshold may include a first preset threshold, and when the system response time is greater than the first preset threshold, Then the number of hosts is expanded. The above system response time can be obtained indirectly according to the number of calls of the process of the Docker container opened on the host. The greater the number of calls of the process of the opened Docker container, the system response time The larger, or the above system response time can be obtained indirectly according to whether there is a backlog of message intermediates in the system. The message intermediate is a queue. Common message middlewares include kafka, RocketMq, etc. If there is a backlog in the message intermediate, It means that many tasks are inserted in the queue during this period, the processing speed of the system cannot keep up, and the system response time is large; when the above indicator is the number of pod functions in the Docker container, the first preset indicator threshold can include the second preset Set a threshold. When the number of pod functions in the Docker container is greater than the second preset threshold, the number of hosts is expanded. The above pod function is the process of the Docker container. When the business volume is large, the Docker container service will first Increase the number of function pods, so when the number of pods increases, it can indicate that the current business volume has increased, and the processed business volume is more; when the above indicator is the resource consumption of the program running in the Docker container, the first preset indicator threshold can include The third preset threshold, when the resource consumption of the program running in the Docker container is greater than the third preset threshold, the number of hosts is expanded; when the above indicator is the CPU usage rate, the first preset indicator threshold may include the first Four preset thresholds. When the CPU usage rate is greater than the fourth preset threshold, the number of hosts is expanded; when the above indicator is the memory usage rate, the first preset indicator threshold may include the fifth preset threshold. When the rate is greater than the fifth preset threshold, the number of hosts is expanded. In this embodiment, the above steps of performing expansion processing on the host need to perform expansion processing on the number of hosts according to the current first service volume data of the system. Specifically, when the comparison result is that the index parameter data is greater than the first preset index threshold When, first obtain the current first service volume data, then expand the current first original host number to the first host number matching the first service volume data, and compare the first host number with the The difference in the number of the first original host is recorded as the number of newly added hosts. Finally, it is necessary to create a specified number of Docker containers corresponding to the number of newly added hosts to complete the final expansion process, where the number of the first original host is Refers to the number of hosts before expansion. The above specified number can be set according to the actual situation. For example, the ratio of a new host during expansion to the number of Docker containers in the new host that needs to be created can be set in advance. Then, the specified number can be determined based on the number of newly added hosts. In this embodiment, by expanding the current number of first original hosts to the number of first hosts matching the first service volume data, it is beneficial to realize effective flexibility in the number of hosts according to the current first service volume data. Scaling ensures the efficient and stable operation of the system and effectively improves the resource utilization rate of the system. Further, the host quantity value corresponding to the service volume can be preset, so that multiple hosts corresponding to the host quantity value can effectively process the service volume and ensure the processing efficiency of the system. Then, each business volume and each host quantity value are stored in the database in a one-to-one correspondence. When the current business volume data is known, the designated host quantity value corresponding to the business volume data can be queried from the database, and the current host quantity can be expanded or contracted according to the designated host quantity value. .

Further, in an embodiment of the present application, the standard index threshold includes a second preset index threshold corresponding to the shrinking process, and after the above step S300, the method includes:

S303: Determine whether the comparison result is that the index parameter data is less than the second preset index threshold, where the second preset threshold is less than the first preset threshold;

S304: If yes, obtain the current second business volume data;

S305: Reduce the current number of second original hosts to the number of second hosts matching the second service volume data.

As described in the above steps S303 to S305, the standard index threshold also includes a second preset index threshold corresponding to the above-mentioned shrinking process, and the second preset index threshold is less than the first preset index threshold, and it is determined that After the comparison result is not that the index parameter data is greater than the first preset index threshold, it is further necessary to determine whether the comparison result is that the index parameter data is less than the second preset index threshold. When the comparison result is that the index parameter data is greater than The first preset index threshold, that is, when the index parameter data obtained by real-time detection is less than the second preset index threshold, the scaling process will be performed. Wherein, the second preset index threshold corresponds to a plurality of preset thresholds corresponding to each index. Specifically, when the index is the system response time, the second preset index threshold may include the sixth preset threshold. When the system response time is less than the sixth preset threshold, the number of hosts will be scaled down; when the above indicator is the number of pod functions in the Docker container, the second preset indicator threshold may include the seventh preset threshold. When the number of pod functions in the container is less than the seventh preset threshold, the number of hosts is scaled down; when the above indicator is the resource consumption of the program running in the Docker container, the first preset indicator threshold may include the eighth preset threshold. Set a threshold. When the resource consumption of the program running in the Docker container is less than the eighth preset threshold, the number of hosts is scaled down; when the above indicator is the CPU usage rate, the first preset indicator threshold may include the ninth preset threshold. Set a threshold. When the CPU usage rate is less than the ninth preset threshold, the number of hosts will be scaled down; when the above indicator is the memory usage rate, the first preset indicator threshold may include the tenth preset threshold. When the memory usage rate is When it is less than the tenth preset threshold, the number of hosts is reduced. In addition, the above-mentioned steps of shrinking processing need to shrink the number of hosts according to the current second business volume data of the system. Specifically, first obtain the current second business volume data, and then change the current second original host The number is scaled down to the number of second hosts matching the second service volume data to complete the scaling process of the host, where the second original host number refers to the number of hosts before the scaling process. In this embodiment, when the detected index parameter data is less than the aforementioned second preset index threshold, the current number of second original hosts is reduced to the number of first hosts matching the second service volume data, It is beneficial to realize the effective elastic scaling of the number of hosts according to the current second business volume data, ensure the efficient and stable operation of the system, effectively improve the resource utilization rate of the system, and reduce the waste of system resources.

Further, in an embodiment of the present application, the above step S305 includes:

S3050: Obtain the aging value of each of the hosts and the number of the second hosts that match the second business volume data, and record the difference between the number of the second original host and the number of the second host To reduce the number of hosts;

S3051: Sort the aging value of each host in descending order of numerical value to obtain the aging value ranking;

S3052: According to the aging value ranking, sequentially select multiple designated aging values corresponding to the reduced number of hosts from the aging value ranking;

S3053: Reclaim each designated host corresponding to each designated aging value.

As described in the above steps S3050 to S3053, the above step of reducing the current number of second original hosts to the number of second hosts matching the second business volume data is to filter out old hosts from all current hosts. The number of designated hosts with the highest value and corresponding to the aforementioned reduction of the number of hosts is realized by recycling, where the number of the second original host refers to the number of hosts before the shrinking process, and the host with the highest aging value represents slow processing speed , Inefficient host. Specifically, first obtain the aging value of each of the hosts and the number of second hosts that match the above-mentioned second business volume data, and record the difference between the number of second original hosts and the number of second hosts as the reduced number of hosts, The aging value of the host can be obtained by calculating the sum of the aging value of the host and the container aging value of each Docker container in each host. After calculating the aging value of each host, the aging value of each host can be sorted in descending order of numerical value to obtain the corresponding aging value ranking, and sorting from the aging value according to the aging value ranking Multiple designated aging values corresponding to the above-mentioned reduced number of hosts are filtered out, and finally multiple designated hosts corresponding to the multiple designated aging values can be obtained, and the shrinking process of the hosts is completed by recycling the above-mentioned multiple designated hosts . In this embodiment, when the shrinking process is required, after calculating the current aging value of each host and obtaining the aging value ranking corresponding to each aging value, the multiple aging values can be filtered directly from the aging value ranking The larger designated host is recycled. By eliminating some hosts with a large aging value and a long time existence, it is helpful to reduce the resource consumption of the system and also improve the processing efficiency of the system.

Further, in an embodiment of the present application, before the above step S3050, the method includes:

S30500: Obtain the first life cycle of each host, and input the first life cycle into a preset first formula to obtain a host aging value corresponding to each host, where the first formula is : S1=50 points*D/120, S1 is the host aging value, D is the first life cycle, and D is over 120 days as 120 days; and,

S30501: Obtain the second life cycle of each Docker container in each host, and input the second life cycle into a preset second formula to find the container of each Docker container in each host Aging value, where the second formula is: S2=50 points/number of Docker containers*H1/240+...+50 points/number of Docker containers*Hn/240, Hn is the second life cycle, and Hn More than 240 hours are counted as 240 hours, and less than one hour is counted as one hour;

S30502: Calculate the sum of the host aging value of each host and the container aging value of each Docker container in each host to obtain the aging value of each host.

As described in the above steps S30500 to S30502, before the step of obtaining the aging value of each of the above hosts, it also includes a process of calculating the aging value of each of the above hosts, and the aging value of each host can be calculated by calculating the host aging value of the host and The sum of the container aging value of each Docker container in each host is calculated. Specifically, first obtain the first life cycle of each host, and input the first life cycle into a preset first formula to obtain the host aging value corresponding to each host, where the first life cycle is Refers to the creation time of the host. The first preset formula mentioned above is: host aging value=50 minutes*D/120 (D: is the number of days the host has been created, that is, the number of days that the host has been alive, and D exceeds 120 days as 120 days); Then obtain the second life cycle of each Docker container in each host, and input the second life cycle into a preset second formula to obtain the container aging value of each Docker container in each host, The second life cycle mentioned above refers to the time when the Docker container has been created. The second preset formula mentioned above is the aging value of the container=50 points/number of Docker containers*H1/240+...+50 points/Docker container Number* Hn/240 (Hn: Docker container has been created hours, that is, the number of hours alive. Hn over 240 hours is calculated as 240 hours, less than one hour is calculated as one hour); finally calculate the host aging value of each host and The aging value of each Docker container in each host is summed to obtain the aging value of each host.

Further, in an embodiment of the present application, before the foregoing step S30500, the method includes:

S305000: Obtain the first creation time corresponding to each host and the second creation time corresponding to each Docker container in each host;

S305001: Calculate each first difference between the current time and the first creation time of each of the hosts to obtain each of the first life cycles corresponding to each of the hosts; and

S305002: Calculate each second difference between the current time and the second creation time of each Docker container in each host, respectively, to obtain each second life cycle corresponding to each Docker container.

As described in the above steps S305000 to S305002, since the aging value of the host can be obtained by calculating the sum of the aging value of the host and the container aging value of each Docker container in each host, the aging value corresponding to each host is calculated Before, you need to calculate the host aging value of each host and the container aging value of the Docker container in each host. Before calculating the aging value of each host and the aging value of each container, it is necessary to calculate the first life cycle of each host and the second life cycle of the Docker container in each host. Specifically, the first creation time of each host and the second creation time of the Docker container in each host must be obtained first, and then according to the current time, the first creation time of the current time and the first creation time of each host The difference value can be used to obtain the first life cycle of each host, and the second difference between the current time and the second creation time of each Docker container can be calculated to obtain each Docker container in each host. The corresponding second life cycle.

Further, in an embodiment of the present application, before the above step S3053, the method includes:

S30530: Extract the designated data in each of the designated hosts;

S30531: Store the specified data in a preset database.

As described in the above steps S30530 and S30531, when the current host needs to be scaled down, multiple designated aging values corresponding to the reduced number of hosts are selected from the foregoing aging value sorting, and the multiple specified aging values corresponding to the number of When specifying multiple designated hosts corresponding to the aging value, because the multiple designated hosts also store designated data, the above-mentioned designated data can be the data retained when the designated host is working, and the designated data may contain some important data , And if the designated host is directly recycled at this time, the system will directly delete the designated data, resulting in the loss of important data. In this embodiment, before performing the shrinking operation of recovering each designated host corresponding to each designated aging value, the designated data in the designated host may be backed up. The designated data may be important data. The specific content of important data is not specifically limited. The important data can be screened automatically by the system or according to the wishes of the business personnel. In addition, a database can be set up in advance, and then the specified data can be extracted and transferred to the database for storage . This embodiment backs up the designated data in the designated host before recycling the designated host, which helps to avoid the situation that the designated data will be deleted and cannot be recovered after the designated host is recovered, and reduces the amount of data in the designated host. Loss caused by loss of important data.

The method for elastic scaling of the number of hosts in the distributed system of the embodiment of the present application monitors the indicator parameters of the Docker container corresponding to each host in real time after pre-setting the indicator of the Docker container corresponding to the host and the standard indicator threshold corresponding to the indicator Data, and then compare the index parameter data with the standard index threshold to obtain a comparison result, and according to the comparison result, the above host scaling strategy is used to expand or shrink the current number of hosts corresponding to the current business volume to achieve According to the real-time index parameter data of each host, the number of hosts can be effectively and accurately elastically scaled to ensure the efficient and stable operation of the system and effectively improve the resource utilization rate of the system.

3, an embodiment of the present application also provides an elastic scaling device for the number of hosts in a distributed system, including:

The monitoring module 1 is used to monitor the indicator parameter data of the Docker container corresponding to the host in real time, wherein the indicator parameter data is the parameter data corresponding to the indicator input by the user, and the indicator is determined by the user according to the Docker corresponding to the host. The container is set, and the indicators include one or more of system response time, the number of pod functions in the Docker container, the resource consumption of programs running in the Docker container, the CPU usage rate, and the memory usage rate;

The comparison module 2 is used to compare the index parameter data with a preset standard index threshold to obtain a comparison result;

The processing module 3 is configured to perform an elastic scaling process corresponding to the current business volume on the number of hosts through a preset host scaling strategy according to the comparison result, and the elastic scaling process includes capacity expansion processing or capacity reduction processing.

In this embodiment, the implementation process of the functions and roles of the monitoring module, the comparison module, and the processing module are detailed in the implementation process of the corresponding steps S1-S3 in the method for elastic scaling of the number of hosts in the distributed system, which will not be repeated here.

Further, in an embodiment of the present application, the standard index threshold includes a first preset index threshold corresponding to the expansion process, and the above processing module includes:

The first judgment unit is configured to judge whether the comparison result is that the index parameter data is greater than the first preset index threshold;

The first obtaining unit is configured to obtain current first business volume data if yes;

The expansion unit is configured to expand the current number of first original hosts to the number of first hosts matching the first service volume data.

In this embodiment, the implementation process of the functions and roles of the first judging unit, the first acquiring unit, and the capacity expansion unit are detailed in the implementation process corresponding to steps S300-S302 in the method for elastic scaling of the number of hosts in the distributed system, here No longer.

Further, in an embodiment of the present application, the standard index threshold includes a second preset index threshold corresponding to the shrinking process, and the above processing module includes:

A second judgment unit, configured to judge whether the comparison result is that the index parameter data is less than the second preset index threshold, wherein the second preset threshold is less than the first preset threshold;

The second acquiring unit is configured to, if yes, acquire the current second business volume data;

The shrinking unit is configured to shrink the current number of second original hosts to the number of second hosts matching the second service volume data.

In this embodiment, the implementation process of the functions and effects of the second judgment unit, the second acquisition unit, and the scaling unit are detailed in the implementation process corresponding to steps S303-S305 in the method for elastic scaling of the number of hosts in the distributed system. This will not be repeated here.

Further, in an embodiment of the present application, the aforementioned shrinking unit includes:

The first obtaining subunit is used to obtain the aging value of each of the hosts, and the number of the second hosts matching the second service volume data, and compare the number of the second original host with the second The difference in the number of hosts is recorded as reducing the number of hosts; the sorting subunit is used to sort the aging value of each of the hosts in numerical order from large to small to obtain the aging value ranking;

The screening subunit is used for sorting according to the aging value, and sequentially selecting a plurality of designated aging values corresponding to the reduced number of hosts from the aging value sorting;

The recovery subunit is used to recover each designated host corresponding to each designated aging value.

In this embodiment, the implementation process of the functions and roles of the first acquisition sub-unit, the screening sub-unit, and the recycling sub-unit is detailed in the implementation process corresponding to steps S3050-S3053 in the method for elastic scaling of the number of hosts in the distributed system. This will not be repeated here.

The second acquiring subunit is used to acquire the first life cycle of each of the hosts, and input the first life cycle into a preset first formula to obtain the host aging value corresponding to each of the hosts, where The first formula is: S1=50 points*D/120, S1 is the host aging value, D is the first life cycle, and D exceeds 120 days as 120 days; and,

The third acquisition subunit is used to acquire the second life cycle of each Docker container in each host, and input the second life cycle into a preset second formula to obtain each The container aging value of the Docker container, wherein the second formula is: S2=50 points/number of Docker containers*H1/240+...+50 points/number of Docker containers*Hn/240, where Hn is the first 2. Life cycle, and Hn exceeds 240 hours as 240 hours, and less than one hour as one hour;

The first calculation subunit is used to calculate the sum of the host aging value of each host and the container aging value of each Docker container in each host to obtain the aging value of each host.

In this embodiment, the implementation process of the functions and roles of the second acquisition subunit, the third acquisition subunit, and the first calculation subunit is detailed in the corresponding steps S30500-S30502 in the method for elastic scaling of the number of hosts in the distributed system. The realization process will not be repeated here.

The fourth obtaining subunit is configured to obtain the first creation time corresponding to each host and the second creation time corresponding to each Docker container in each host;

The second calculation subunit is configured to calculate each first difference between the current time and the first creation time of each of the hosts to obtain each of the first life cycles corresponding to each of the hosts; and

The third calculation subunit is used to calculate the second difference between the current time and the second creation time of each Docker container in each host to obtain each second life cycle corresponding to each Docker container .

In this embodiment, the implementation process of the functions and roles of the fourth acquisition subunit, the second calculation subunit, and the third calculation subunit are detailed in the corresponding steps S305000-S305002 in the method for elastic scaling of the number of hosts in the distributed system. The realization process will not be repeated here.

Extracting subunit for extracting designated data in each designated host;

The storage subunit is used to store the specified data in a preset database.

In this embodiment, the implementation process of the functions and roles of the extraction subunit and the storage subunit is detailed in the implementation process of the corresponding steps S30530-S30531 in the method for elastic scaling of the number of hosts in the distributed system, which will not be repeated here.

The device for elastically scaling the number of hosts in the distributed system of the embodiment of the present application monitors the indicator parameters of the Docker container corresponding to each host in real time after pre-setting the indicator of the Docker container corresponding to the host and the standard indicator threshold corresponding to the indicator Data, and then compare the index parameter data with the standard index threshold to obtain a comparison result, and according to the comparison result, the above host scaling strategy is used to expand or shrink the current number of hosts corresponding to the current business volume to achieve According to the real-time index parameter data of each host, the number of hosts can be effectively and accurately elastically scaled to ensure the efficient and stable operation of the system and effectively improve the resource utilization rate of the system.

Referring to FIG. 4, an embodiment of the present application also provides a computer device. The computer device may be a server, and its internal structure may be as shown in FIG. 4. The computer equipment includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor designed for the computer equipment is used to provide calculation and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, computer readable instructions, and a database. The internal memory provides an environment for the operation of the operating system and computer-readable instructions in the non-volatile storage medium. The database of the computer equipment is used to store data such as standard index thresholds and index parameter data. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer-readable instruction is executed, it executes the process of the above-mentioned method embodiment. Those skilled in the art can understand that the structure shown in FIG. 4 is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the devices and computer equipment to which the solution of the present application is applied.

An embodiment of the present application further provides a computer non-volatile readable storage medium, on which computer readable instructions are stored, and when the computer readable instructions are executed by a processor, the processes as in the foregoing method embodiments are implemented.

The above are only the preferred embodiments of this application, and do not limit the scope of this application. Any equivalent structure or equivalent process transformation made using the content of this application description and drawings, or directly or indirectly applied to other related The technical field is equally included in the scope of patent protection of this application.

Claims

A method for elastic scaling of the number of hosts in a distributed system is characterized in that it includes:

Monitor the indicator parameter data of the Docker container corresponding to the host in real time, where the indicator parameter data is parameter data corresponding to the indicator input by the user, and the indicator is set by the user according to the Docker container corresponding to the host. Indicators include one or more of system response time, the number of pod functions in the Docker container, the resource consumption of the programs running in the Docker container, CPU usage, and memory usage;

Comparing the index parameter data with a preset standard index threshold to obtain a comparison result, where the standard index threshold includes at least a first preset index threshold corresponding to the expansion process;

Judging whether the comparison result is that the index parameter data is greater than the first preset index threshold;

If yes, obtain the current first business volume data;

Expanding the current number of first original hosts to the number of first hosts matching the first service volume data.
The method for elastic scaling of the number of hosts in a distributed system according to claim 1, wherein the standard index threshold further comprises a second preset index threshold corresponding to the scaling process, and the determining whether the comparison result is After the step of the index parameter data being greater than the first preset index threshold, the method includes:

If not, determine whether the comparison result is that the index parameter data is less than the second preset index threshold, where the second preset threshold is less than the first preset threshold;

If yes, obtain the current second business volume data;

The current number of second original hosts is reduced to the number of second hosts matching the second service volume data.
The method for elastic scaling of the number of hosts in a distributed system according to claim 2, wherein the current number of second original hosts is reduced to less than the number of second hosts matching the second service volume data. The steps include:

Obtain the aging value of each of the hosts, and the number of the second hosts that match the second business volume data, and record the difference between the number of the second original host and the number of the second host as a decrease Number of hosts;

Sorting the aging value of each host in descending order of numerical value to obtain the aging value ranking;

According to the aging value ranking, a plurality of designated aging values corresponding to the reduced number of hosts are sequentially selected from the aging value ranking;

Each designated host corresponding to each designated aging value is recycled.
The method for elastic scaling of the number of hosts in a distributed system according to claim 3, wherein the acquiring the aging value of each of the hosts and the number of the second hosts matching the second service volume data Before the steps, include:

Obtain the first life cycle of each of the hosts, and input the first life cycle into a preset first formula to obtain the host aging value corresponding to each of the hosts, wherein the first formula is: S1 =50 points*D/120, S1 is the host aging value, D is the first life cycle, and D exceeds 120 days as 120 days; and,

Obtain the second life cycle of each Docker container in each host, and input the second life cycle into a preset second formula to obtain the container aging value of each Docker container in each host , Where the second formula is: S2=50 points/number of Docker containers*H1/240+...+50 points/number of Docker containers*Hn/240, Hn is the second life cycle, and Hn exceeds 240 Hour is calculated as 240 hours, and less than one hour is calculated as one hour;

The sum of the host aging value of each host and the container aging value of each Docker container in each host is calculated respectively to obtain the aging value of each host.
The method for elastic scaling of the number of hosts in a distributed system according to claim 4, wherein the first life cycle of each of the hosts is obtained, and the first life cycle is input into a preset first formula , Before the step of obtaining the host aging value corresponding to each host includes:

Acquiring the first creation time corresponding to each host and the second creation time corresponding to each Docker container in each host;

Respectively calculating each first difference between the current time and the first creation time of each of the hosts to obtain each of the first life cycles corresponding to each of the hosts; and,

Each second difference value between the current time and the second creation time of each Docker container in each host is calculated respectively to obtain each second life cycle corresponding to each Docker container.
The method for elastic scaling of the number of hosts in a distributed system according to claim 3, characterized in that, before the step of recycling the designated hosts corresponding to the designated aging values, the method comprises:

Extract the designated data in each of the designated hosts;

The specified data is stored in a preset database.
An elastic scaling device for the number of hosts in a distributed system, characterized in that it comprises:

The monitoring module is used to monitor the indicator parameter data of the Docker container corresponding to the host in real time, wherein the indicator parameter data is parameter data corresponding to the indicator input by the user, and the indicator is determined by the user according to the Docker container corresponding to the host Setting, the indicators include one or more of system response time, the number of pod functions in the Docker container, the resource consumption of programs running in the Docker container, the CPU usage rate, and the memory usage rate;

A comparison module, configured to compare the index parameter data with a preset standard index threshold to obtain a comparison result, where the standard index threshold includes at least a first preset index threshold corresponding to the expansion process;

The first judgment module is configured to judge whether the comparison result is that the index parameter data is greater than the first preset index threshold;

The first obtaining module is used to obtain the current first business volume data if yes;

The expansion module is used to expand the current number of first original hosts to the number of first hosts matching the first service volume data.
The device for elastic scaling of the number of hosts in a distributed system according to claim 7, wherein the standard index threshold further comprises a second preset index threshold corresponding to the scaling process, and the host in the distributed system A number of elastic expansion devices, including:

A second judgment module, configured to judge whether the comparison result is that the index parameter data is less than the second preset index threshold, wherein the second preset threshold is less than the first preset threshold;

The second obtaining module is used to obtain the current second business volume data if yes;

The shrinking module is used to shrink the current number of second original hosts to the number of second hosts matching the second service volume data.
The device for elastically scaling the number of hosts in a distributed system according to claim 8, wherein the shrinking module comprises:

The first obtaining unit is configured to obtain the aging value of each of the hosts, and the number of the second hosts that match the second service volume data, and compare the number of the second original host with the second host The difference in quantity is recorded as reducing the number of hosts; the sorting sub-unit is used to sort the aging value of each host in descending order of numerical value to obtain the aging value ranking;

The screening unit is configured to sort according to the aging value, and sequentially filter out multiple designated aging values corresponding to the reduced number of hosts from the aging value sorting;

The recovery unit is used to recover each designated host corresponding to each designated aging value.
The device for elastically scaling the number of hosts in a distributed system according to claim 9, wherein the shrinking module comprises:

The second acquiring unit is used to acquire the first life cycle of each of the hosts, and input the first life cycle into a preset first formula to obtain the host aging value corresponding to each of the hosts, where The first formula is: S1=50 points*D/120, S1 is the host aging value, D is the first life cycle, and D exceeds 120 days as 120 days; and,

The third acquiring unit is used to acquire the second life cycle of each Docker container in each host, and input the second life cycle into a preset second formula, so as to obtain each location in each host. The container aging value of the Docker container, where the second formula is: S2=50 points/number of Docker containers*H1/240+...+50 points/number of Docker containers*Hn/240, Hn is the second Life cycle, and Hn exceeds 240 hours as 240 hours, and less than one hour as one hour;

The first calculation unit is used to calculate the sum of the host aging value of each host and the container aging value of each Docker container in each host to obtain the aging value of each host.
The device for elastically scaling the number of hosts in a distributed system according to claim 10, wherein the shrinking module comprises:

The fourth obtaining unit is configured to obtain the first creation time corresponding to each of the hosts and the second creation time corresponding to each Docker container in each of the hosts;

The second calculation unit is configured to calculate each first difference between the current time and the first creation time of each of the hosts to obtain each of the first life cycles corresponding to each of the hosts; and

The third calculation unit is configured to calculate each second difference between the current time and the second creation time of each Docker container in each host to obtain each second life cycle corresponding to each Docker container.
The device for elastically scaling the number of hosts in a distributed system according to claim 9, wherein the shrinking module comprises:

An extraction unit for extracting designated data in each of the designated hosts;

The storage unit is used to store the specified data in a preset database.
A computer device includes a memory and a processor, and computer-readable instructions are stored in the memory. The method is characterized in that, when the processor executes the computer-readable instructions, a method for elastically scaling the number of hosts in a distributed system is implemented. The method for elastic scaling of the number of hosts in a distributed system includes:

Monitor the indicator parameter data of the Docker container corresponding to the host in real time, where the indicator parameter data is parameter data corresponding to the indicator input by the user, and the indicator is set by the user according to the Docker container corresponding to the host. Indicators include one or more of system response time, the number of pod functions in the Docker container, the resource consumption of the programs running in the Docker container, CPU usage, and memory usage;

Comparing the index parameter data with a preset standard index threshold to obtain a comparison result, where the standard index threshold includes at least a first preset index threshold corresponding to the expansion process;

Judging whether the comparison result is that the index parameter data is greater than the first preset index threshold;

If yes, obtain the current first business volume data;

Expanding the current number of first original hosts to the number of first hosts matching the first service volume data.
The computer device according to claim 13, wherein the standard index threshold value further comprises a second preset index threshold value corresponding to the shrinking process, and the determining whether the comparison result is that the index parameter data is greater than all After the step of describing the first preset index threshold, it includes:

If not, determine whether the comparison result is that the index parameter data is less than the second preset index threshold, where the second preset threshold is less than the first preset threshold;

If yes, obtain the current second business volume data;

The current number of second original hosts is reduced to the number of second hosts matching the second service volume data.
The computer device according to claim 14, wherein the step of reducing the current number of second original hosts to the number of second hosts matching the second service volume data comprises:

Obtain the aging value of each of the hosts, and the number of the second hosts that match the second business volume data, and record the difference between the number of the second original host and the number of the second host as a decrease Number of hosts;

Sorting the aging value of each host in descending order of numerical value to obtain the aging value ranking;

According to the aging value ranking, a plurality of designated aging values corresponding to the reduced number of hosts are sequentially selected from the aging value ranking;

Each designated host corresponding to each designated aging value is recycled.
15. The computer device according to claim 15, wherein before the step of acquiring the aging value of each of the hosts and the number of the second hosts matching the second service volume data, the method comprises:

Obtain the first life cycle of each of the hosts, and input the first life cycle into a preset first formula to obtain the host aging value corresponding to each of the hosts, wherein the first formula is: S1 =50 points*D/120, S1 is the host aging value, D is the first life cycle, and D exceeds 120 days as 120 days; and,

Obtain the second life cycle of each Docker container in each host, and input the second life cycle into a preset second formula to obtain the container aging value of each Docker container in each host , Where the second formula is: S2=50 points/number of Docker containers*H1/240+...+50 points/number of Docker containers*Hn/240, Hn is the second life cycle, and Hn exceeds 240 Hour is calculated as 240 hours, and less than one hour is calculated as one hour;

The sum of the host aging value of each host and the container aging value of each Docker container in each host is calculated respectively to obtain the aging value of each host.
A computer nonvolatile readable storage medium having computer readable instructions stored thereon, wherein the computer readable instructions are executed by a processor to implement a method for elastic scaling of the number of hosts in a distributed system. The methods for elastic scaling of the number of hosts in the system include:

Monitor the indicator parameter data of the Docker container corresponding to the host in real time, where the indicator parameter data is parameter data corresponding to the indicator input by the user, and the indicator is set by the user according to the Docker container corresponding to the host. Indicators include one or more of system response time, the number of pod functions in the Docker container, the resource consumption of the programs running in the Docker container, CPU usage, and memory usage;

Comparing the index parameter data with a preset standard index threshold to obtain a comparison result, where the standard index threshold includes at least a first preset index threshold corresponding to the expansion process;

Judging whether the comparison result is that the index parameter data is greater than the first preset index threshold;

If yes, obtain the current first business volume data;

Expanding the current number of first original hosts to the number of first hosts matching the first service volume data.
The computer non-volatile readable storage medium according to claim 17, wherein the standard index threshold value further comprises a second preset index threshold value corresponding to the shrinking process, and the determining whether the comparison result is After the step of the index parameter data being greater than the first preset index threshold, the method includes:

If not, determine whether the comparison result is that the index parameter data is less than the second preset index threshold, where the second preset threshold is less than the first preset threshold;

If yes, obtain the current second business volume data;

The current number of second original hosts is reduced to the number of second hosts matching the second service volume data.
The computer non-volatile readable storage medium according to claim 18, wherein the current number of the second original host is reduced to a maximum of the number of the second host matching the second service volume data The steps include:

Obtain the aging value of each of the hosts, and the number of the second hosts that match the second business volume data, and record the difference between the number of the second original host and the number of the second host as a decrease Number of hosts;

Sorting the aging value of each host in descending order of numerical value to obtain the aging value ranking;

According to the aging value ranking, a plurality of designated aging values corresponding to the reduced number of hosts are sequentially selected from the aging value ranking;

Each designated host corresponding to each designated aging value is recycled.
The computer non-volatile readable storage medium according to claim 19, wherein the acquiring the aging value of each of the hosts, and the number of the second hosts matching the second service volume data Before the steps, include:

Obtain the first life cycle of each of the hosts, and input the first life cycle into a preset first formula to obtain the host aging value corresponding to each of the hosts, wherein the first formula is: S1 =50 points*D/120, S1 is the host aging value, D is the first life cycle, and D exceeds 120 days as 120 days; and,

Obtain the second life cycle of each Docker container in each host, and input the second life cycle into a preset second formula to obtain the container aging value of each Docker container in each host , Where the second formula is: S2=50 points/number of Docker containers*H1/240+...+50 points/number of Docker containers*Hn/240, Hn is the second life cycle, and Hn exceeds 240 Hour is calculated as 240 hours, and less than one hour is calculated as one hour;

The sum of the host aging value of each host and the container aging value of each Docker container in each host is calculated respectively to obtain the aging value of each host.