Multi-operation and maintenance management method and system for batch management of business machine clusters based on application group dimension

Abstract

A multi-operation and maintenance management method and system for batch managing a business machine cluster in an application group dimension, the method comprising: S1: setting a bastion machine architecture; S2: setting a sentinel program under each Master machine; and setting an initialization script for accessing a new machine on the Master machine of each node; S3: once a node accesses a new business host, the Master machine of the node triggers a Salt-Minion access event: after the new machine executes the initialization script, the initialization steps including naming the new machine according to the three-stage rule of application group-intranet IP-network area, installing Salt-Minion on the machine, and configuring the Master machine to which it belongs are completed; S4: when an application group updates operation and maintenance management data, the node where the application group is located is determined, and through the Master machine of the corresponding node, according to the field of application group, all machine names of the same application group are obtained, and then the configuration files of the batch operation and maintenance management are written to the machines respectively through Salt-Minion.

Description

Technical Field

[0001] The present invention relates to an operation and maintenance management method, and in particular to a multi-operation and maintenance management method and system for batch managing business machine clusters in an application group dimension. Background Art

[0002] Alibaba Group Holding Limited disclosed a server management method, device and equipment in CN201910585545.6, including: the first bastion host obtains the operation and maintenance proxy operation, the first bastion host is used to perform operation and maintenance management on the servers in the first area; the second bastion host is determined according to the operation and maintenance proxy operation, the second bastion host is used to perform operation and maintenance management on the servers in the second area; the first bastion host is used to act as the agent for the second bastion host to perform operation and maintenance management on the servers in the second area. The operation and maintenance proxy operation is obtained by the first bastion host, the second bastion host is determined according to the operation and maintenance proxy operation, and the first bastion host is used as the agent for the second bastion host to perform operation and maintenance management on the servers in the second area, which effectively realizes the unified management of cross-regional servers by the first bastion host, and no public IP is required between the first bastion host and the second bastion host.

[0003] In an operation and maintenance management system dominated by a bastion host, super administrators, administrators, operation and maintenance personnel, and developers are often set up, and they are managed uniformly according to the characteristics of the login personnel. For example, during the first registration process, the default registration role can be the developer role, and then the system administrator can authorize a user role as an operation and maintenance personnel in the system as needed. The operation and maintenance personnel have the authority to approve permission applications. In addition to the approval authority, the system administrator has the authority to assign user roles and clean up and delete resigned personnel. However, the above operation and maintenance management method is only applicable to the operation and maintenance management system corresponding to a single cloud, and cannot adapt to a highly scalable network environment. When facing operation and maintenance management scenarios corresponding to multiple clouds, there are a series of problems such as complex operation and maintenance management structure and high operation and maintenance costs.

[0004] Our company has disclosed an operation and maintenance method in CN202111535716.8. The deployment machine is named based on a preset data structure; when the deployment machine is in an abnormal operating state, the network node where the deployment machine is located and the application business of the deployment machine are obtained according to the name, where: the data structure includes the application business of the deployment machine, the name of the network node where the deployment machine is located, and the private domain IP address of the deployment machine in the network node.

[0005] How to utilize this structure for effective operation and maintenance is exactly the problem that the present invention needs to solve. Summary of the invention

[0006] The present invention provides a multi-operation and maintenance management method for batch managing a business machine cluster in an application group dimension, so as to solve the problem that effective operation and maintenance cannot be performed through naming in the prior art.

[0007] A multi-operation and maintenance management method for batch managing a business machine cluster based on an application group dimension, further comprising:

[0008] S1: Set up the fortress architecture:

[0009] Divide the network into independent areas according to the business sites, and treat each independent VPC network environment in different regions as an independent node. Then plan a machine in each node as the node's security entrance login machine and a master machine as the node's operation and maintenance control. The master is responsible for managing all business hosts of the node.

[0010] Set up a Central master machine as the Login machine of each independent node and the superior Master machine of the Master machine; the Central master machine is deployed with the WEB system, which has the construction files and configuration information of the global system. The Central master machine regularly synchronizes the relevant data of each node with the Master machine of each node;

[0011] S2: Set up a sentinel program under each Master machine, which is responsible for monitoring and recording events including business host access and offline of the node, and updating the configuration files used for local monitoring or batch operation and maintenance. The configuration files used for batch operation and maintenance further include configuration files of application group dimension; and set up the initialization script for accessing new machines on the Master machine of each node;

[0012] S3: Once a node is connected to a new service host, the master machine of the node triggers a Salt-Minion access event: after the initialization script is executed on the new machine, the initialization steps are completed, including naming the new machine according to the three-stage rule of application group-intranet IP-network area, installing Salt-Minion on the machine, and configuring the master machine to which it belongs;

[0013] S4: When an application group updates its operation and maintenance management data, the node where the application group is located is determined, and through the Master machine of the corresponding node, all machine names of the same application group are obtained according to the application group field, and then the configuration files of the batch operation and maintenance management are written to the machines respectively through Salt-Minion.

[0014] Compared with the existing technology, the system standardizes the identification of machine ID information into a three-segment structure, application group-private network IP-regional identifier, such as the machine ID information of bigdata-172.16.4.48-Hangzhou. By parsing the first field, all application group information can be obtained, and then the machines in the application group can be obtained, and in which network areas they are distributed, and then batch management capabilities based on the application group dimension and real-time monitoring of a certain application cluster can be viewed based on the application group dimension. Daily communication is also very convenient. The machine ID is pasted over without much communication. The whole system will understand which business machine it is, what the private network IP is, and where the machine is located. It can also be considered to be standardized with the machine name. With this standardization of machine ID information, the system can very elegantly generate relevant management configuration files and monitoring configuration files, which is extremely convenient for management and has high parsing efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is an example diagram of the principle of a multi-operation and maintenance management system that manages business machine clusters in batches based on the application group dimension;

[0016] Figure 2 The schematic diagram of the multi-operation and maintenance management method for batch managing business machine clusters based on the application group dimension. DETAILED DESCRIPTION

[0017] There are many solutions for the safe and efficient management of massive hosts in the industry, such as Ansible, Puppet, and SaltStack, which can solve the problem of efficient and safe management of massive hosts in enterprises to a certain extent. Judging from the time when these tool systems appeared, Puppet (full-featured and complex) came first, followed by Ansible (simple and convenient), and then SaltStack (efficient and simple). It can be understood that it has balanced the two and introduced many new ideas. SaltStack is a centralized management tool for server infrastructure based on C / S architecture. The management end is called Master and the client is called Minion. SaltStack has functions such as configuration management, remote execution, and monitoring. The open source version of SaltStack only provides basic solutions, but in actual work it still requires a lot of maintenance skills, such as grouping and monitoring configuration, as well as challenges such as operation and maintenance management of multiple independent network nodes in the enterprise. We hope to be able to unify the management and automatic monitoring of multiple network environments (without relying on cross-node network connectivity. Network connectivity can certainly solve some problems, but considering the more scientific real-time collection of massive business host monitoring data through their own intranets and the expansion of future ultra-large-scale IT systems, we should consider the divide-and-conquer node internal self-governance solution in advance).

[0018] See also Figure 1, which is a schematic diagram of a multi-operation and maintenance management system for batch management of business machine clusters in the application group dimension. The applicant will divide the independent network areas of our respective business sites, regard each independent VPC network environment in different regions as an independent node, and then plan two machines in this node, one as a secure entrance Login machine (similar to our current jumpserver), and one as an operation and maintenance control Master machine. The Login and Master will be provided with relevant basic services, and the Master will be responsible for managing all business hosts of the node. We can plan another central Central master control machine as the superior Master machine of the Login and Master of each independent node, so that the Central master control machine can manage the Login and Master machines of each independent node as a scheduling center, and then indirectly manage the hosts of each node in a unified manner. Our Central will deploy a WEB system, merge various daily necessary functions, and use the unified collected data as a data overview to show the IT resource status of all our nodes, as well as display real-time monitoring related data according to the application grouping dimension, so that we can have a clear view of various IT resources.

[0019] The multi-operation and maintenance management system of the present invention further comprises:

[0020] Multiple independent nodes: Divide the network areas of each business site into independent networks, and treat each independent VPC network (Virtual Private Cloud) environment in different regions as an independent node. Each node includes at least one machine as the node's security entrance login machine and one machine as the node's operation and maintenance control Master machine. Among them:

[0021] A sentinel program is set up under each Master machine to monitor and record events including business host access and offline of the node, and to update the configuration files used for local monitoring or batch operation and maintenance in a linked manner. The configuration files used for batch operation and maintenance further include configuration files of application group dimensions. An initialization script for accessing new machines is set up on the Master machine of each node: once a node accesses a new business host, the Master machine of the node triggers a Salt-Minion access event. The Salt-Minion access event further includes: after the new machine executes the initialization script, the new machine is named according to the three-stage rule of application group-intranet IP-network area, and Salt-Minion is installed on the machine and the corresponding Master is configured.

[0022] A Central master control machine, serving as the Login machine and the superior Master machine of the Master machine of each independent node; the Central master control machine is deployed with the WEB system, which has the construction files and configuration information of the global system. The Central master control machine regularly synchronizes the relevant data of each node with the Master machine of each node, so that when an application group updates the operation and maintenance management data, it can determine the node where the application group is located, and obtain the names of all machines in the same application group through the Master machine of the corresponding node according to the application group field, and then write the batch operation and maintenance management configuration files to the machines respectively through Salt-Minion.

[0023] The core of the present invention lies in the separation of three powers. The Login machine is only responsible for secure login entry; the Master machine is only responsible for executing scheduling tasks; the Central machine is only responsible for dispatching tasks; in this way, the system is highly secure and reliable. For example, if there is a problem with the Central service, it will not affect the assigned permissions, and each node is independent of each other and does not affect each other; and Login and Master are in a sense a kind of "manager" role. Once an unrecoverable exception occurs in the Login or Master machine of a certain node, our master control Central machine can quickly regenerate a new Login or Master in the node, because the master control machine has the configuration data of each node Login and Master and the full amount of system permission allocation history data, which can be quickly generated, and the data of our master control machine is backed up every day and synchronized to the Master machine of each node every day, so that our master control data is very safe. This system expands and manages new nodes according to the actual project and enterprise development needs. That is, unify the configuration management of each node host: adopt the master control + distributed bastion machine architecture, the corresponding working mechanism of the supporting architecture, unify the configuration management of each node host, and realize multi-node operation and maintenance.

[0024] The specific instructions are as follows.

[0025] See also Figure 2 , which is a flowchart of a multi-operation and maintenance management method for batch management of business machine clusters based on application group dimensions. It further includes:

[0026] S110: Set up the fortress structure:

[0027] Divide the network into independent areas according to the business sites, and treat each independent VPC network environment in different regions as an independent node. Then plan a machine in each node as the node's security entrance login machine and a master machine as the node's operation and maintenance control. The master is responsible for managing all business hosts of the node.

[0028] Set up a Central master machine as the Login machine of each independent node and the superior Master machine of the Master machine; the Central master machine is deployed with the WEB system, which has the construction files and configuration information of the global system. The Central master machine regularly synchronizes the relevant data of each node with the Master machine of each node;

[0029] S120: Set a sentinel program under each Master machine, which is responsible for monitoring and recording events including business host access and offline of the node, and linking to update the configuration files used for local monitoring or batch operation and maintenance, wherein the configuration files used for batch operation and maintenance further include configuration files of application group dimension; and set the initialization script for accessing new machines on the Master machine of each node;

[0030] S130: Once a node is connected to a new service host, the master machine of the node triggers a Salt-Minion access event: after the new machine executes the initialization script, it completes the initialization steps including naming the new machine according to the three-stage rule of application group-intranet IP-network area, installing Salt-Minion on the machine, and configuring the master machine to which it belongs;

[0031] S140: When an application group updates its operation and maintenance management data, the node where the application group is located is determined, and through the Master machine of the corresponding node, all machine names of the same application group are obtained according to the application group field, and then the configuration files of the batch operation and maintenance management are written to the machines respectively through Salt-Minion.

[0032] When the Central control machine parses the name of the business host, it finds the Master to which it belongs based on the network area field and executes the corresponding task;

[0033] After the initialization script is executed on the new machine, the new machine is named according to the three-part rule of application group-intranet IP-network area. Each network area has an independently named intranet IP, and the intranet IPs in different network areas can be the same.

[0034] The default machine naming rule of the system is "application group-intranet IP-network area". Application group-intranet IP-network area describes the application group to which the machine belongs, the network area to which the machine belongs (i.e. node information) and the intranet IP information of the machine in the network area. The new machine is located by the network area to which the machine belongs (i.e. node information) and the intranet IP information of the machine in the network area. For example:

[0035] gateway-192.168.104.214-Hangzhou

[0036] This machine name can be resolved in the entire multi-operation and maintenance management system as a gateway machine, whose intranet IP is 192.168.104.214. This machine is in the Hangzhou network area.

[0037] When the user submits the permission application for the above machines, the Central control machine can find the scheduling Master to which it belongs to execute the task based on the network area field, so the local IP in each network area can be repeated without affecting the global management of the system.

[0038] For the "sentinel" program of each node, when a new machine is connected to the node, it can naturally split by field, such as according to the application group field, get the names of all machines in the same application group, and then write them into the configuration file of batch operation and maintenance management.

[0039] Such as / etc / salt / master.d / nodegroup.conf and / etc / ansible / hosts

[0040] At the same time, we can also automatically generate prometheus monitoring configuration files, for example:

[0041] ```

[0042] - targets:

[0043] - 192.168.104.214:9100

[0044] labels:

[0045] group: gateway

[0046] ip: 192.168.104.214

[0047] region: Hangzhou

[0048] instance: gateway-192.168.104.214-Hangzhou

[0049] ```

[0050] In this way, once an alarm is triggered, the system can clearly see which network area the alarm comes from and what business machine it belongs to. These capabilities are derived from the system's very concise three-part naming structure.

[0051] In addition, this method also needs to explain the benefits of the management method of separation of powers, including Login machines and Master machines. The Login machine serves as the secure entry machine of the node, exposing the standard sshd port to the outside. Users log in to the corresponding Login machine by being assigned login_id_rsa to enter the intranet of the node. Deploy basic components on the Master machine of the node, which further includes: deploying Salt-Minion to receive scheduling tasks from the Central master control machine; deploying Redis to record relevant data corresponding to user permissions-machines in this node, and deploying Salt-Master to receive and manage the business host with Salt-Minion installed in this node. The Master machine serves as the operation and maintenance control machine of the network area where it is located. It does not expose any ports to the outside and trusts the network access of the Central master control machine internally. After the initial installation of Login and Master of each node is completed, the Central master control machine synchronizes the basic template management of each of them:

[0052] # salt -N login state.apply group.login completes the template synchronization for the Login machine.

[0053] # salt -N master state.apply group.master Complete the template synchronization of the Master machine.

[0054] In fact, for the Central machine, one command salt '*' state.apply can complete the template synchronization management of two groups of machines on a node, including the synchronization of the basic configurations of the two groups of machines, the distribution of the "sentinel" program on each node, and so on.

[0055] Specifically, the initialization of a new machine means that the new machine is accessed and managed by the Salt-Master of the node where it is located. The Master of each node has an access machine initialization script, which is executed immediately after the new machine is initialized. The core function of the script is to normalize the naming of the new machine, install Salt-Minion, and configure the Master to which it belongs. After execution, it is immediately accessed and managed by the Master machine of the node where it is located, and the access event is triggered at the same time, recording the Redis in this area and the Redis of the remote master control.

[0056] The Central master control machine contains the build files and configurations of the global system, including the Login machine and Master machine of each node, and regularly synchronizes and backs up the core data to the Master machine of each node. Generally, all files of the master control global system are also synchronized to the enterprise Git repository to ensure that they are not lost. This setting prevents the loss of the core configuration files of the system; and if the Central master control machine has a problem, the system can download the build files of the global system from any Master or git repository to quickly build a new Central master control machine. In addition, if a certain Login machine or a certain Master machine has a problem, the Central master control machine can also generate a corresponding new Login machine or Master machine through templates and core data.

[0057] The management of user machine rights is to dispatch the Central master control machine to the adaptation Master to issue a task to the user to create his account under the corresponding machine; the machine monitoring operation is to dispatch the Central master control machine to all business machines in the node to deploy the monitoring agent program; the daily operation and maintenance operation is to execute a task on the adaptation group machine or the global machine through the Central master control machine in the adaptation Master to achieve unified management. Therefore, whether it is the allocation of user rights, the daily deployment of machine monitoring and network monitoring, or the batch management of daily operation and maintenance, this system can be abstracted as the Central master control machine dispatches the Master of a certain regional node to issue a task to execute a certain task, which naturally achieves unified management; the subsequent invention can continue to follow this design concept, integrate the management of container clusters, and continuously expand the management objects to make its functionality more powerful and efficient. To some extent, there is no difference between the system managing dozens of network areas and managing one network area; whether it is a few machines or dozens or thousands of machines, the system management cost is similar.

[0058] According to the above, through the Master machine of the corresponding node, according to the field of application group, all machine names of the same application group are obtained, and then the configuration files of the batch operation and maintenance management are written to the machines through Salt-Minion, which further includes: through the Central master control machine scheduling adaptation Master, the corresponding configuration files are executed for the machines that meet the requirements, and the said meeting the requirements further includes that the application group name meets the requirements. The application group can be divided into application groups of different machines according to the functions implemented in the management system, or it can be divided into application groups of different machines according to the applications implemented in the management system. This three-segment naming method and system management method, the present invention can realize batch management of business machine clusters based on the application group dimension.

[0059] In addition, in this method, when a new business host is connected, a prometheus monitoring configuration file can be automatically generated; the local Master machine detects a certain alarm information, and the alarm information including the host name where the alarm is located is synchronized to the Central master control machine; the Central master control machine parses the application group and network area where the alarm information host is located through the host name where the alarm information occurs; if the corresponding business host of this network area and / or this application group needs to be processed according to the alarm information, the corresponding execution task is sent to the Master machine of the adapter node.

[0060] The monitoring configuration file of Prometheus automatically generated on the business host further includes:

[0061] The core configuration is uniformly distributed from the Central Master to each node Master, specifically to the monitoring object of each node: when a new service host is connected to the node, the configuration is automatically generated on the host;

[0062] Through the cloud service provider's API, Aliyun and AWS cloud monitoring data collection and alarm configuration are configured, and the prometheus network monitoring component blackbox_exporter is deployed on the Master machine as the network monitoring detection service of the node;

[0063] When the business host accesses the node, it obtains the prometheus monitoring configuration file through the master machine of this node;

[0064] Set three alarm levels of info / warning / critical: this node collects monitoring data -> this node matches the warning rules -> this node issues alarm self-management including telephone alarms, and also initiates monitoring detection of the Central master control machine itself, and timely issues warning notifications. Of course, the Central master control machine itself also initiates relevant service detection of the Master of each node to complete the self-monitoring of the system itself. In this way, no matter whether there is a problem with the basic service of the Central master control machine or the basic service managed by each Master, an alarm will be received and handled in a timely manner.

[0065] After the access script is executed, the management of the Master machine in the area is completed. According to the default SaltStack execution template rules set in the pre-dock, a series of template actions will be performed on the new machine, which can mainly include but not limited to 7 general templates;

[0066] 1. Synchronize a series of tool scripts in the / opt / sys / directory of the new machine, including the init.sh initialization script, the name.sh renaming script, and the src.sh basic software installation script we wrote. These scripts can be continuously expanded according to our daily work. That is to say, after the new machine is taken over, we agree that some of the tool scripts required by default are in the / opt / sys / directory, which is convenient for every technician (including operation and maintenance personnel) to obtain;

[0067] 2. Write the default crontab tasks, including disk cleaning tasks and salt-minion automatic pull-up tasks (placed in salt-minion due to various abnormal disconnection issues). Of course, these basic tasks can be continuously expanded according to the needs of daily work;

[0068] 3. Execute the default basic configuration, such as the default public key information to be entered. This information can be customized and extended in / srv / pillar / master.sls of our respective nodes. For example, once a new machine is connected, the system's publishing machine public key is immediately entered, thereby completing the channel problem of the publishing machine, and automatically synchronizing some basic environment variables, and adjusting the machine connection number configuration;

[0069] 4. Install the defined basic tools, such as vim, wget, etc.;

[0070] 5. Automatically configure the defined secure sshd configuration file. By default, password authentication, full private key authentication, and the node's trusted intranet IP and whitelist users are prohibited;

[0071] 6. Automatically synchronized salt-minion related configurations;

[0072] 7. Automatically set the network parameters of the new machine;

[0073] 8. A custom template. Each node can customize related management tasks according to the needs of the node. Because the template configuration files of each Master machine are sent and synchronized by the master control machine, the extension changes of these template files are also updated from the master control and then synchronized to the Master of each node. The master control will be responsible for updating and synchronizing the above 7 common initialization templates, and the master control of 1 node’s customized template will not be updated.

[0074] To summarize, the system standard identifies the machine ID information as a three-segment structure, application group-private network IP-regional identifier. For example, the machine ID information of bigdata-172.16.4.48-Hangzhou can be parsed out of the first field to obtain all application group information, and then to obtain which machines are in the application group and which network areas they are distributed in. Then, batch management capabilities based on the application group dimension and real-time monitoring of a certain application cluster based on the application group dimension can be performed. Daily communication is also very convenient. The machine ID is pasted over without much communication. The entire system will understand which business machine it is, what the private network IP is, and where the machine is located. It can also be considered to be standardized with the machine name. With this standardization of machine ID information, the system can very elegantly generate relevant management configuration files and monitoring configuration files, which is extremely convenient for management.

[0075] The following is a further detailed description of an anchor and an anchor system for fixing soft tissue proposed by the present invention in conjunction with the accompanying drawings and specific embodiments. The advantages and features of the present invention will become clearer based on the following description. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0076] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0077] It should be noted that similar reference numerals represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined or explained in the subsequent drawings. The embodiments of the present invention are described in detail above in conjunction with the drawings, but the present invention is not limited to the above embodiments. Even if various changes are made to the present invention, if these changes fall within the scope of the claims of the present invention and their equivalents, they still fall within the protection scope of the present invention.

Claims

1. A multi-operation and maintenance management method for batch managing business machine clusters based on application group dimensions, characterized in that: include: S1: Set up the fortress architecture: Divide the network into independent areas according to the business sites, and treat each independent VPC network environment in different regions as an independent node. Then plan a machine in each node as the node's security entrance login machine and a machine as the node's operation and maintenance control master machine. The master machine is responsible for managing all business hosts of the node. Set up a Central master machine as the Login machine of each independent node and the superior Master machine of the Master machine; the Web system is deployed on the Central master machine, which has the construction files and configuration information of the global system. The Central master machine regularly synchronizes the relevant data of each node with the Master machine of each node; S2: Set up a sentinel program under each Master machine, which is responsible for monitoring and recording events including business host access and offline of the node, and updating the configuration files used for local monitoring or batch operation and maintenance. The configuration files used for batch operation and maintenance further include configuration files of application group dimension; and set up the initialization script for accessing new machines on the Master machine of each node; S3: Once a node is connected to a new service host, the master machine of the node triggers a Salt-Minion access event: after the initialization script is executed on the new machine, the initialization steps are completed, including naming the new machine according to the three-stage rule of application group-intranet IP-network area, installing Salt-Minion on the new machine, and configuring the master machine to which it belongs; S4: When an application group updates its operation and maintenance management data, the node where the application group is located is determined, and the names of all machines in the same application group are obtained through the Master machine of the corresponding node according to the application group field, and then the configuration files used for batch operation and maintenance are written to all the machines through Salt-Minion.

2. The multi-operation and maintenance management method according to claim 1, characterized in that: Also includes: Automatically generate a prometheus monitoring configuration file on the new business host; The local Master machine detects an alarm message, and synchronizes the host name of the alarm contained in the alarm message to the Central master control machine; The central master control machine parses the application group and network area of ​​the host where the alarm information is generated through the host name of the alarm information; If the corresponding service host of the network area and / or the application group needs to be processed according to the alarm information, the corresponding execution task is sent to the Master machine of the adapter node.

3. The multi-operation and maintenance management method according to claim 1, characterized in that: Also includes: When the Central control machine parses the name of the business host, it finds the Master machine to which it belongs based on the network area field and executes the corresponding task; Each network area independently names the intranet IP.

4. The multi-operation and maintenance management method according to claim 1, characterized in that: Also includes: The Login machine serves as the secure entry machine of the node, exposing the standard sshd port to the outside. Users log in to the corresponding Login machine by being assigned login_id_rsa to enter the intranet of the node.

5. The multi-operation and maintenance management method according to claim 1, characterized in that: Also includes: Deploy basic components on the master machine of the node, which further includes: deploying Salt-Minion to receive scheduling tasks from the central master control machine; Deploy Redis to record the relevant data corresponding to the user permissions and machines in this node, and deploy Salt-Master to receive and manage the business host with Salt-Minion installed in this node; The Master machine is the operation and maintenance control machine of the network area. It does not expose any ports to the outside world and trusts the network access of the Central master control machine internally. After the initial installation of the Login machine and the Master machine of each node is completed, the Central control machine synchronizes the basic template management of each of them: # salt -N login state.apply group.login completes the template synchronization of the Login machine, and # salt -N master state.apply group.master completes the template synchronization of the Master machine.

6. The method according to claim 5, characterized in that Also includes: The user machine authority management is to dispatch the Central Master Control Machine to the Master Machine to issue a task to the user to create his account under the corresponding machine; The machine monitoring operation is a task that dispatches the Central master control machine to adapt the Master machine to all business hosts in the node to deploy the monitoring agent program; Daily operation and maintenance is to achieve unified management by using the Central master control machine to perform a task on the adaptation group machine or global machine in the adaptation Master machine.

7. The multi-operation and maintenance management method according to claim 1, characterized in that: Through the Master machine of the corresponding node, according to the field of application group, all machine names of the same application group are obtained, and then the configuration files used for batch operation and maintenance are written to all the machines through Salt-Minion, further including: The Central master control machine schedules the adaptation Master machine and executes the corresponding configuration file on the machine that meets the requirements, wherein the meeting the requirements further includes that the application group name meets the requirements.

8. A multi-operation and maintenance management system for batch management of business machine clusters based on application group dimensions, characterized in that: include: Multiple independent nodes: Divide the network areas of each business site into independent networks, and treat each independent VPC network environment in different regions as an independent node. Each node includes at least one machine as the node's security entrance login machine and one machine as the node's operation and maintenance control Master machine. The Master machine is responsible for managing all business hosts of the node. A sentinel program is set up under each Master machine, which is responsible for monitoring and recording events including business host access and offline of the node, and updating the configuration files used for local monitoring or batch operation and maintenance. The configuration files used for batch operation and maintenance further include configuration files of application group dimension; And set up the initialization script for connecting new machines on the Master machine of each node; once a node connects to a new business host, the Master machine of the node triggers a Salt-Minion access event: after the new machine executes the initialization script, it completes the initialization steps including naming the new machine according to the three-stage rule of application group-intranet IP-network area, installing Salt-Minion on the new machine, and configuring the Master machine to which it belongs; A Central master control machine, serving as the Login machine and the superior Master machine of the Master machine of each independent node; the Central master control machine is deployed with a Web system, which has the build files and configuration information of the global system. The Central master control machine regularly synchronizes the relevant data of each node with the Master machine of each node, so that when an application group updates the operation and maintenance management data, it can determine the node where the application group is located, and obtain the names of all machines in the same application group through the Master machine of the corresponding node according to the application group field, and then write the configuration files used for the batch operation and maintenance to all the machines through Salt-Minion.

Images