CN117251253A

CN117251253A - Cloud management platform design method based on multi-source heterogeneous

Info

Publication number: CN117251253A
Application number: CN202311331120.5A
Authority: CN
Inventors: 任俊武; 徐斌; 李怀旭; 杨砚世
Original assignee: Nanjing Laiwangxin Technology Research Institute Co ltd
Current assignee: Nanjing Laiwangxin Technology Research Institute Co ltd
Priority date: 2023-10-16
Filing date: 2023-10-16
Publication date: 2023-12-19

Abstract

The invention provides a cloud management platform design method based on multisource heterogeneous, which comprises the steps of establishing a cloud resource data resource catalog, setting data resource types and distinguishing data coding rules; determining an acquisition mode and data loading work, and carrying out data acquisition and data resource catalog loading operation on the received Guan Yigou cloud platform; customizing an API interface in a data mapping mode, and performing self-adaptive access management on the received Guan Yigou cloud platform interface; and establishing a sample library, carrying out modeling analysis, and carrying out data management work on the received Guan Yigou cloud platform. According to the cloud management system, the cloud server, the cloud hard disk, the VPC virtual private network, the elastic IP, the container mirror image and other data resource data provided by the cloud platform are integrated, and the problems of unified nano-tube, collaboration, monitoring and the like among the multi-source heterogeneous cloud platforms are solved through operations such as data acquisition, data mapping and data management, so that a quality-effect type cloud management service is constructed, and the utilization rate of the whole cloud resources is improved.

Description

Cloud management platform design method based on multi-source heterogeneous

Technical Field

The invention relates to a cloud management platform design method, in particular to a cloud management platform design method based on multi-source heterogeneous.

Background

The cloud computing technology is taken as one of modern IT novel infrastructures, has been developed and popularized in depth in recent years, the cloud demand of application systems of enterprises and public institutions of all levels is increased, and the cloud computing technology is an important basic support and key element for promoting industrial digitization and development of the digitization industry and plays an important role in the construction and development of digital Chinese and national economy.

With the rapid development of cloud computing technology, technical architecture evolution is not uniform, architectures are different, technical standard specifications are not uniform among different cloud platform manufacturers, interconnection and intercommunication do not exist among the cloud platforms, and management staff cannot perform uniform resource allocation, monitoring and management. Sharing and collaborative scheduling of different kinds of cloud resources and resources among cloud platforms are difficult to achieve.

In order to solve the problems of unified nano-tube, coordination, monitoring and the like among the multi-source heterogeneous cloud platforms, the cloud management platform based on multi-source heterogeneous is designed, unified nano-tube and fine operation of the multi-source heterogeneous cloud are realized, the utilization rate of the whole cloud resources is improved, and a quality-effect type cloud management service is constructed.

Disclosure of Invention

The invention aims to: the invention aims to solve the technical problem of providing a cloud management platform design method based on multi-source isomerism aiming at the defects of the prior art.

In order to solve the technical problems, the invention discloses a cloud management platform design method based on multi-source isomerism, which comprises the following steps:

step 1, establishing a cloud resource data resource catalog;

step 2, determining a data acquisition mode;

step 3, performing self-adaptive access management on the interface of the heterogeneous cloud platform of the nanotube, and accessing data;

step 4, performing data conversion processing on the accessed data;

step 5, carrying out data management analysis on the data subjected to conversion treatment;

and 6, the cloud management platform issues a management operation instruction to the heterogeneous cloud platform of the nanotube to finish cloud platform management.

Further, establishing the cloud resource data resource directory in the step 1 specifically includes:

step 1-1, defining interface parameters provided by a heterogeneous cloud platform and a cloud management platform of a nanotube according to a data resource type, and encoding the interface parameters;

step 1-2, setting a data resource type, including: infrastructure resources, virtualized platform resources, cloud service product resources, and setting parameter names and corresponding parameter descriptions of the resource opening, changing, canceling and management control interfaces provided by the heterogeneous cloud platform and the cloud management platform of the nano tube;

and step 1-3, distinguishing a data coding rule, namely coding the data resources for managing the data resources.

Further, the determining data acquisition manner in the step 2 specifically includes:

step 2-1, setting a data acquisition protocol, which specifically comprises:

the cloud management platform performs data acquisition on heterogeneous cloud platforms of the nanotubes through a set protocol;

step 2-2, when the heterogeneous cloud platform and the cloud management platform of the nanotube are not in the same network, adopting a front-end acquisition method to acquire data;

and 2-3, synchronizing to a system cache of the cloud management platform in a single or batch mode.

Further, the self-adapting access management of the interface of the heterogeneous cloud platform of the nanotube in step 3 specifically includes:

step 3-1, uniformly controlling the API interface of the heterogeneous cloud platform of the nano tube, comprising the following steps: maintaining a source URL, a request protocol, a request type and a destination address URI of an API interface, starting a third-party plug-in, running according to preset plug-in parameters, and automatically accessing data of the API interface;

step 2-2, setting a label for the API interface, namely classifying the API interface, and performing unified management and control in the step 3-1 in a classified manner;

step 2-3, the cloud management platform uses an authentication back end to finish user authentication when the interface is subjected to self-adaptive access;

step 2-4, fault information is injected into the interface of the heterogeneous cloud platform of the nanotube, interface abnormality is simulated, and the test of the relevant dependent interfaces and dependent logic is completed;

step 2-5, providing access control of an API (application program interface) of the heterogeneous cloud platform of the nanotube in a module assembly mode, wherein the access control at least comprises the following steps: cross-domain control, request interception, IP black and white list and UA black and white list;

step 2-6, setting a triggering current limiting parameter, and if a request sent by an API interface of a heterogeneous cloud platform of the nano tube meets the above condition, automatically triggering current limiting operation;

and 2-7, for parameter transfer requests sent by the API interface of the heterogeneous cloud platform of the nanotube, rewriting Request messages and message headers and rewriting message headers and message bodies of Response messages are achieved in a module assembly mode.

Further, when the data is accessed in the step 3, the method further includes: setting a back-end service address and a pushing mode, recording a request log of an API interface in real time, and storing the request log for log analysis and display.

Further, the data conversion processing of the accessed data in step 4 specifically includes:

and carrying out meaning identification, data assembly, data subscription and release, alarm judgment and data quality monitoring on the received original data.

Further, the data management analysis in step 5 specifically includes:

step 5-1, establishing a data convergence task according to the ownership dimension and the time dimension and automatically executing the data convergence task;

step 5-2, a sample library is established and modeling analysis is carried out, and a unified original model is established for describing the same business objects on the heterogeneous cloud platform of the nanotube;

step 5-3, discovering potential rules and atlas knowledge through machine learning, knowledge atlas analysis and natural language analysis tools, and providing decision support for problem positioning and alarm convergence;

step 5-4, managing a topology model of the same service object in the heterogeneous cloud, wherein the topology model at least comprises association maintenance, an association type table and automatic extraction of association data;

and 5-5, analyzing the original data acquired by the cloud management platform and managing a data dictionary of the original data.

Further, the cloud management platform in step 6 issues a management operation instruction to the heterogeneous cloud platform of the nanotube, which specifically includes:

and (3) according to the requirements of the steps 1 to 5, issuing management operation instruction data to the heterogeneous cloud platform of the nanotube to work.

Further, the set protocol in step 2-1 at least includes: SNMP protocol, database platform commands, and SQL statements.

Further, the raw data in step 5-5 at least includes operation data, alarm event data, performance index data, object resource data, and log data.

The beneficial effects are that:

according to the invention, through a dynamic self-defined interface adaptation design method, unified management and refined management of resources such as cloud servers, cloud hard disks, VPC virtual private networks, elastic IP, container mirroring, load balancing, virtualization, physical servers and the like of cloud platforms with different technical architectures can be realized under a multi-cloud heterogeneous environment, the problem of resource management of heterogeneous cloud is solved, and the functions of monitoring cloud resource operation indexes, processing alarm information, backing up the platform, processing resource application work orders, charging cloud resources and the like are supported.

Drawings

The foregoing and/or other advantages of the invention will become more apparent from the following detailed description of the invention when taken in conjunction with the accompanying drawings and detailed description.

FIG. 1 is a flow chart of cloud management platform data management operations.

Fig. 2 is a diagram of a cloud management platform adapted to access an overall architecture.

Fig. 3 is a schematic diagram of data resource types and data coding rules.

Fig. 4 is a schematic deployment diagram of a front-end collector of the cloud management platform.

Detailed Description

The principle of the invention is as follows: an adapter is created between the cloud management platform and the accessed cloud platform, different types of interfaces can be converted into object interfaces, so that the interfaces can interact with cloud platforms of different technical architectures, and modeling analysis is carried out by gathering data resources of the received Guan Yigou cloud platform. By encapsulating objects, the complex conversion process is hidden behind the scenes, and the adapter not only can convert data in different formats, but also can facilitate cooperation between objects employing different interfaces. The specific technical scheme is as follows: a cloud management platform design method based on multi-source heterogeneous comprises the following steps:

step 1: establishing a cloud resource data resource catalog, and determining an acquisition mode and data loading work;

step 2: performing self-adaptive access management on the received Guan Yigou cloud platform interface;

step 3: carrying out data management work on the received Guan Yigou cloud platform;

step 4: and the cloud management platform manages the issuing work of the operation instruction data.

The step 1 comprises the following steps:

step 1-1, configuring a resource data resource catalog of the received Guan Yigou cloud platform in the platform according to the data acquisition standard of the cloud management platform;

and step 1-2, setting a data resource type distinguishing data coding rule, coding forcing or not and the like. The list describes the encoding rules and may be associated with detailed dictionary data for various resource types. The list display content mainly comprises resource type codes, resource type names, coding rules, whether codes are forced, resource coding attribution, dictionary data and the like.

And step 1-3, the cloud management platform performs data acquisition on the received Guan Yigou cloud platform through a certain protocol. The method mainly supports the standard MIB data acquisition of the SNMP protocol and the engine acquisition mode of database platform commands and SQL sentences, wherein database connection information, separators and the like are configured and realized in the engine commands.

And step 1-4, aiming at the nanotube cloud platform which is not in the same network, adopting a front-end acquisition mode to realize, and transmitting data on the front-end processor to the network where the nanotube cloud platform is located in a file mode after the acquisition is completed. And after the buffer memory is out of date, the data file on the front-end processor executes automatic deletion operation, so that the situation that the acquisition program cannot run due to the occupied space is avoided.

And step 1-5, according to the data acquisition standard, the configured data resource catalogue is synchronized into a system cache in a single or batch mode, so that data connection is saved, data loading efficiency is improved, and meanwhile, an updating function is supported. The data loading service supports different database platforms such as a relational database, a time sequence database and the like.

As shown in fig. 2, step 2 includes:

and 2-1, uniformly controlling all southbound API interfaces, mainly managing interfaces of mainstream manufacturers (including but not limited to Huacheng, langchao, ali, tech, and the like) of the accommodated Guan Yigou cloud platform, customizing the API interfaces in a data mapping mode, and recording service scenes such as calling times of the interfaces, interface return states, interface failure times and the like.

And 2-2, classifying the specific API interfaces in a manner of labeling the API interfaces so as to uniformly set the API interface sets corresponding to the labels. And setting a plug-in which needs to be executed for a certain API interface or a certain API interface, and carrying out specific parameter configuration on the corresponding plug-in. When the API gateway is matched with the corresponding request message, message processing is sequentially carried out according to the plug-in set started by configuration, and the processed message is forwarded to the configured destination address.

And 2-3, supporting integration BaseAuth, JWT, OIDC, LDAP and other various authentication back ends to realize a user authentication function, selecting the integrated authentication back end after a user authentication integration interface is selected and started for one or a class of API interfaces, and setting parameters corresponding to different authentication back ends.

And 2-4, selecting and starting a user authentication integrated interface for one or a class of API interfaces, wherein SSL certificate information can be set, and the API gateway automatically carries out HTTPS security protection on the destination address interface.

And 2-5, starting fault detection setting, configuring message matching conditions, returning fault response codes and message body information, and directly returning the set fault information after the matched request message accords with the matching conditions.

Step 2-6, setting different information aiming at different control types, wherein the cross-domain control can realize accurate cross-domain interception by setting an address, a domain name, a request type, a request head and the like which allow cross-domain access; the request interception can be performed by setting certain special request types and request heads in the non-cross-domain request; the IP black-and-white list and the UA black-and-white list can realize the request interception of special IP addresses and user-agents by setting the black-and-white lists of the corresponding IP and UA.

And 2-7, after one or a class of API interfaces select to start flow management, processing logic such as current limiting, degradation, fusing and the like can be set. Setting parameters such as the concurrency number and the request number of the trigger current limit, and automatically triggering the current limit operation if the request meets the setting conditions.

Step 2-8, after the start Request is rewritten according to a certain or a certain API interface, setting a Request rule, a Response rule and a transmission information injection mode which need to be rewritten, and realizing the rewriting of a message header of a Request message and the rewriting of a message header and a message body of a Response message by a plug-in mode so as to meet the transmission requirement of special parameters of the interface.

And 2-9, data conversion mainly realizes the work from original data to meaning identification, data assembly, data subscription and release, alarm judgment, data quality monitoring and the like. The data conversion processing object is a file or an API interface, and is mainly subjected to buffer processing through a message queue.

And 2-10, setting a pushed back-end service address and a pushing mode, recording an API interface request log in real time, and storing the request log in a local system or pushing the request log to a three-party log system in an interface mode for log analysis and display.

The step 3 comprises the following steps:

and 3-1, the cloud management platform collects data from each nanotube cloud platform, forms tasks according to two paths of a weight dimension and a time dimension and automatically executes data collection. The raw data after data classification is stored in a time sequence database, and the summarized data is stored in a relational database.

And 3-2, establishing a sample library, carrying out modeling analysis, carrying out data prediction tasks to predict the future change trend of the data index to be predicted through the linear, exponential, logarithmic and ARMA algorithms required by data prediction, carrying out analysis of actual values, predicted values and deviation values on the result, improving the data prediction precision, and judging whether the predicted result is effective or not.

And 3-3, based on the platform data, realizing potential knowledge discovery of rules, patterns and the like through tools such as machine learning, knowledge pattern analysis, natural language analysis and the like, and providing decision support for problem positioning and alarm convergence.

And 3-4, carrying out topology model management on the same business object in the heterogeneous cloud, wherein the topology model management comprises association relation maintenance, an association relation type table, automatic extraction of association data and the like.

And 3-5, providing maintenance service for the butt joint relations of different database tables, forming association information and then performing automatic task execution by defining the association relation types among the tables, and completing the relation maintenance of the access data.

And 3-6, establishing a storage table of the association information of the three types including, bearing and connecting, and automatically and periodically executing and updating the information table according to task information maintained by the association relation.

And 3-7, forming periodic execution tasks based on the information in the relation maintenance table. When the task is maintained, a mark whether the task is valid or not is set, and the task is only started effectively and is executed periodically.

And 3-8, performing operation and analysis service on various original data of the cloud management platform, wherein the operation data, the alarm event data, the performance index data, the object resource data and the log data are mainly managed and analyzed, and providing service for managing a data dictionary of the original data.

Step 4 comprises:

and (3) according to the operations of the steps 1-4, carrying out cloud management platform management operation instruction data issuing work, wherein the operation steps are similar.

Examples:

as shown in fig. 1, the present invention provides a cloud management platform design method based on multi-source heterogeneous, which includes the following steps:

1. establishing a cloud resource data resource catalog, and determining an acquisition mode and data loading work;

1-1, establishing a cloud resource data resource catalog

According to the requirements of functions and performance indexes of the cloud management platform, management, control and monitoring type interface parameters are defined and encoded, a cloud management platform data acquisition standard is established, the Guan Yigou cloud platform provides resource data interface parameter information outwards according to the standard requirements, and according to the data interface types and the parameter types, the cloud resource data catalog is carded and imported into the cloud management platform for configuration.

1-1-1. Setting data resource types

Meanwhile, the configuration of the data types required by the cloud management platform is synchronously completed, and a complete cloud resource data resource directory system is established. Aiming at resources such as infrastructure, virtualization platforms, cloud service products and the like, the cloud management platform and the received Guan Yigou cloud platform can provide interface parameter names and corresponding parameter descriptions such as opening, changing, canceling, management control and the like of cloud resources.

1-1-2. Differentiate data encoding rules

And encoding interfaces provided by each nano-tube cloud platform according to the management and control interface types, and performing core metadata management on the shared resources, wherein the core metadata management comprises resource type encoding, resource type names, encoding rules, whether encoding is mandatory, resource encoding attribution, dictionary data and the like. See figure 3 for details.

1-2, determining data acquisition mode and data loading work

The cloud management platform is realized through the engine type acquisition mode of database platform commands and SQL sentences through the standard MIB data acquisition of the SNMP protocol, wherein the database connection information, separators and the like are configured in the engine commands. Support pass GET (161 port), POST (162 port).

1-2-1. Data acquisition is carried out on the received Guan Yigou cloud platform

The acquisition processing is required to be carried out on the premise that the service operation is not affected, and the overall performance occupation is not more than 3%. Aiming at a nano-tube cloud platform which is not in the same network, a mode of deploying a front-end collector in a service area of the cloud platform is adopted to collect relevant interface data of the platform (through a gateway), after the collection is completed, the data on the front-end collector is transmitted to a service area network agent front-end processor of the cloud tube platform in a file mode through a virtual private network, and the cloud tube platform recalls corresponding interface data (through the gateway) on the agent front-end processor. And after the buffer memory is out of date, the data file on the front-end processor executes automatic deletion operation, so that the situation that the acquisition program cannot run due to the occupied space is avoided. See fig. 4 for details.

1-2-2 data resource directory load operations

Synchronizing the configured cloud resource data resource catalog into a system cache according to a single or batch, calculating according to 15 ten thousand management objects with design capacity, wherein the support bottom limit is 1000 pieces/second (assuming a period of 5 minutes, each management object corresponds to two rows of data); the upper limit may support 5000 bars/second (assuming a 1 minute period, two rows of data for each management object).

2. Performing self-adaptive access management on the received Guan Yigou cloud platform interface;

API interface control

And uniformly controlling the API interface related to the cloud platform, and customizing the API interface in a data mapping mode. Maintaining API interface information to be managed and controlled, namely connecting an API interface data field from a source system (a nano-tube cloud platform database) to a target system (a cloud-tube platform database), wherein main maintenance contents comprise a source URL, a request protocol, a request type, a target address URI and the like; the method comprises the steps of adding a label to an API interface (an API interface for completing data mapping) which is created (manually input, semi-automatic or fully automatic input), selecting a plug-in set to be started (a third party tool can replace manual part or complete automatic matching of data fields), setting corresponding parameters of the plug-in to be started, namely interface parameters, according to the data interface parameter input rule, setting (resource catalog template requirement) so as to realize plug-in processing of a certain or a certain type of API interface message (the plug-in is automatically realized according to the set rule).

2-2 tag management

And classifying the specific API interfaces in a manner of labeling the API interfaces so as to uniformly set the API interface sets corresponding to the labels. And the new page is added to the API interface, so that the labels can be managed, and the labels can be flexibly added and deleted. The interface is interface parameter information of the nano-tube cloud platform which is combed according to the resource catalog template; the labels refer to different nanotube cloud platforms, and the background may have different labels for the same field;

for example: by setting a load balancing policy (one of functions of a nanotube cloud platform product, which may be configuring a cloud server, a cloud hard disk, an elastic IP, etc.) on a certain tag, a unified load balancing policy setting may be performed on a class of API interfaces matched to the tag;

if the API interface is provided with a load balancing strategy, processing according to the load balancing strategy which is provided by the API interface; otherwise, processing according to a load balancing strategy set by the label to which the label belongs;

if the labels are provided with different load balancing strategies (such as specific parameter configurations of the number of inflow data packets, the number of outflow data packets, the network inflow flow rate, the network outflow flow rate and the like, the security protection strategy can be analogically configured), the labels are processed according to a default polling mode.

2-3 user authentication management

After a user authentication integration interface is selected and started for one or a class of API interfaces, an integrated authentication back end (namely system user information management) is required to be selected, and interface parameters corresponding to different authentication back ends are set. And a plurality of authentication back ends such as integration BaseAuth, JWT, OIDC, LDAP are supported to realize the user authentication function.

After the setting is completed, when the gateway is matched with the corresponding API request, the cloud management platform calls the authenticated corresponding interface of the nano-tube platform and automatically calls the set authentication back end to carry out user authentication: if the authentication is not passed, directly returning a failure result; otherwise, the next plug-in processing (interface adaptation by the third party tool software) is continued.

2-4 fault monitoring management

The fault monitoring management simulates abnormal return of the interface by injecting fault information into the resource monitoring interface of the nano-tube cloud platform, so that the test (simulated fault alarm return value test) of the relevant dependent interfaces and dependent logic after the interface is abnormal is realized.

When one or a class of API interfaces select to start fault injection, the type of the injected fault and relevant parameters needed by the type of the fault need to be set. After fault injection is set, when the interface is accessed, the interface automatically feeds back corresponding error information according to the fault injection setting.

After the fault injection plug-in is started, setting a message matching condition, returning a fault response code and message body information, and directly returning the set fault information after the matched request message accords with the matching condition.

2-5 Access control

After a certain API interface or a class of API interfaces select to start access control, a corresponding control type can be selected. For example: and providing cross-domain control, request interception, IP black-and-white list, UA black-and-white list and the like aiming at the API interface in a module component mode.

Setting different information for different control types, and realizing accurate cross-domain interception for cross-domain control by setting an address, a domain name, a request type, a request head and the like which allow cross-domain access; the request interception can be performed by setting certain special request types and request heads in the non-cross-domain request; the IP black-and-white list and the UA black-and-white list can realize the request interception of special IP addresses and user-agents by setting the black-and-white lists of the corresponding IP and UA.

After the corresponding access control plug-in is started, setting information such as an IP black-and-white list, a UA black-and-white list and the like, releasing the traffic conforming to the white list, and intercepting the traffic conforming to the black list.

2-6 flow threshold management

When one or a class of API interfaces select to start flow management, processing logic such as current limiting, degradation, fusing and the like can be set.

Setting parameters such as the concurrency number, the request speed and the like for triggering the current limiting, and automatically triggering the current limiting operation after the request meets the setting conditions;

aiming at the condition that the request amount is suddenly increased within a period of time, the automatic degradation of the service can be realized by setting degradation operation, so that more server resources are released for ensuring the normal operation of the core service;

aiming at the condition that the request failure rate is increased suddenly within a period of time, the automatic fusing of the service and the automatic switching of the half-open and full-open states of the interface of the service can be realized by setting fusing operation.

2-7 request for adjustment operations

And (3) rewriting the Request message Request, rewriting the back-end service return message Response, and injecting tracking information for calling chain tracking.

After opening the corresponding Request to rewrite plug-in, setting the information such as protocol, host name, URI and the like of the message to be rewritten and the information of the message header to be rewritten, the plug-in screens the flow meeting the conditions according to the set conditions and rewrites the message header of the Request message according to the configuration.

After the corresponding Response request is started to rewrite the plug-in, parameter conditions of the message to be rewritten are set, and the rewritten message header and message body information are set, the plug-in can screen the flow meeting the conditions according to the set conditions, and rewrite the Response message according to the configuration.

2-8 data conversion operation

The data processing is divided into three stages of acquisition, conversion and loading, the data identification processing is not carried out on the acquisition and the loading, and the data mapping operation is completed in the conversion stage. The data conversion realizes the work from original data to meaning identification, data assembly, data subscription and release, alarm judgment, data quality monitoring and the like according to the cloud resource data resource catalog.

For file format conversion, the corresponding process from the original variable to the model attribute is used for calling the variable of different data sources according to a preset formula, mapping, functions and the like, calculating the variable as a data value required by the model attribute, and assembling the data value into a loadable and warehouse-in format.

The corresponding relation of the normalized data models of different data sources is maintained in the data mapping, and the normalized data model can be used in data needing automatic batch warehousing such as alarms, performances, resources and the like. The data formulas are obtained and used in format conversion, and four operations of the formulas, variable mapping, derived formulas and the like are executed according to word recognition and corresponding grammar calling.

In the fields of the normalization model, the required values come from a formula combination of the raw data. The derived formula provides the processing of the original data variables and mainly realizes the splicing, partial value taking and case-to-case conversion of the character strings; four kinds of numerical values are mixed and calculated, and the value with judging conditions is obtained. The value of the judgment condition is calculated according to the value of one variable as the judgment condition, if the denominator is 0, the four operations cannot be calculated, and if the calculated value is the success rate, 100% is returned, and if the calculated value is the failure rate, 0% is returned.

2-9 Log management operations

After the corresponding request log pushing plug-in is started, an address to be pushed is set, and the plug-in background can automatically send the request log information to the corresponding address.

3. And carrying out data management work on the received Guan Yigou cloud platform.

3-1. Data summarization

And summarizing the collected data to a corresponding summary table according to time granularity according to summary indexes corresponding to each data type, wherein the summary of the time granularity comprises the following steps: hours granularity summary (day data), days granularity summary (week data), weeks granularity summary (month data), months granularity summary (year data), years granularity summary. The summary of time granularity follows the summary of small granularity to large granularity.

3-2 data prediction

And establishing a sample library, carrying out modeling analysis, carrying out data prediction tasks to predict the future change trend of the data index to be predicted through the linear, exponential, logarithmic and ARMA algorithms required by data prediction, carrying out analysis of actual values, predicted values and deviation values on the results, improving the data prediction precision, and judging whether the predicted results are effective or not.

3-3 thematic machine learning

Through tools such as machine learning, knowledge graph analysis, natural language analysis and the like, potential knowledge discovery such as rules, graphs and the like is realized, and decision support is provided for problem positioning and alarm convergence.

3-4 model management

And carrying out topology model management on the same business object in the heterogeneous cloud, wherein the topology model management comprises association relation maintenance, an association relation type table, automatic extraction of association data and the like.

And after the association relation types among the tables are defined, forming association information, and then executing automatic tasks to complete relation maintenance of access data.

The association model management respectively establishes association information storage tables containing, bearing and connecting three types, and automatically and periodically executes and updates the information tables according to task information maintained by association relations.

Based on the information in the relational maintenance table, periodic execution tasks are formed. When the task is maintained, a mark whether the task is valid or not is set, and the task is only started effectively and is executed periodically.

3-5 data analysis

The operation and analysis service for various original data of the cloud management platform mainly comprises management and analysis of operation data, alarm event data, performance index data, object resource data and log data, and provides a service for managing a data dictionary of the original data.

4. And according to the flow, carrying out issuing work of cloud management platform management operation instruction interface data.

In a specific implementation, the application provides a computer storage medium and a corresponding data processing unit, wherein the computer storage medium can store a computer program, and the computer program can run the invention content of the cloud management platform design method based on the multi-source heterogeneous and part or all of the steps in each embodiment when being executed by the data processing unit. The storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), a random-access memory (random access memory, RAM), or the like.

It will be apparent to those skilled in the art that the technical solutions in the embodiments of the present invention may be implemented by means of a computer program and its corresponding general hardware platform. Based on such understanding, the technical solutions in the embodiments of the present invention may be embodied essentially or in the form of a computer program, i.e. a software product, which may be stored in a storage medium, and include several instructions to cause a device (which may be a personal computer, a server, a single-chip microcomputer, MUU or a network device, etc.) including a data processing unit to perform the methods described in the embodiments or some parts of the embodiments of the present invention.

The invention provides a thought and a method for designing a cloud management platform based on multisource isomerism, and the method and the way for realizing the technical scheme are numerous, the above is only a preferred embodiment of the invention, and it should be pointed out that a plurality of improvements and modifications can be made to those skilled in the art without departing from the principle of the invention, and the improvements and modifications are also considered as the protection scope of the invention. The components not explicitly described in this embodiment can be implemented by using the prior art.

Claims

1. The cloud management platform design method based on the multisource isomerism is characterized by comprising the following steps of:

step 1, establishing a cloud resource data resource catalog;

step 2, determining a data acquisition mode;

step 4, performing data conversion processing on the accessed data;

2. The method for designing a cloud management platform based on multi-source heterogeneous according to claim 1, wherein the establishing a cloud resource data resource directory in step 1 specifically includes:

3. The method for designing a cloud management platform based on multi-source heterogeneous according to claim 2, wherein the determining data acquisition mode in step 2 specifically includes:

step 2-1, setting a data acquisition protocol, which specifically comprises:

4. The method for designing a heterogeneous cloud management platform based on multiple sources according to claim 1, wherein the self-adapting access management of the interface of the heterogeneous cloud platform of the nanotube in step 3 specifically comprises:

5. The method for designing a cloud management platform based on multi-source heterogeneous according to claim 1, wherein when the data is accessed in step 3, the method further comprises: setting a back-end service address and a pushing mode, recording a request log of an API interface in real time, and storing the request log for log analysis and display.

6. The method for designing a cloud management platform based on multi-source heterogeneous according to claim 1, wherein the data conversion processing of the accessed data in step 4 specifically includes:

7. The method for designing a cloud management platform based on multi-source heterogeneous according to claim 1, wherein the data management analysis performed in the step 5 specifically includes:

8. The method for designing a heterogeneous cloud management platform based on multiple sources according to claim 1, wherein the cloud management platform in step 6 issues a management operation instruction to the heterogeneous cloud platform of the nanotube, specifically comprising:

9. The cloud management platform design method based on multi-source heterogeneous system according to claim 3, wherein the set protocol in step 2-1 at least comprises: SNMP protocol, database platform commands, and SQL statements.

10. The method of claim 7, wherein the raw data in step 5-5 at least includes operation data, alarm event data, performance index data, object resource data, and log data.