IT resource real-time state monitoring index automatic recommendation method and system

Abstract

The invention provides an IT resource real-time state monitoring index automatic recommendation method and system. The method comprises the following steps: carrying out abstract modeling on monitoring resources and monitoring state information thereof in a cloud computing platform; the alarm state of the monitoring index is quantified; quantifying the relationship between the monitoring resource entities and the relationship between the monitoring indexes; the query operation of the monitoring indexes is quantified; and integrating and calculating various quantitative factors to obtain a recommendation value of each index, ranking the recommendation values according to different scene rules, and finally obtaining an index recommendation result of a certain configuration item. According to the method, the monitoring indexes of all dimensions related to real-time state monitoring of various IT type resources are quantified, different weights are given, the heat calculation of the specified monitoring resource indexes is automatically realized, the automatic recommendation of the monitoring resource indexes in different operation and maintenance scenes is realized, and the operation and maintenance efficiency is improved. Operation and maintenance personnel are helped to quickly complete state inspection of monitoring resources and analysis and positioning of monitoring fault reasons.

Description

Technical Field

[0001] The present invention relates to the technical field of real-time monitoring status inspection and monitoring fault diagnosis, analysis and positioning of IT resources, and in particular to an automatic recommendation method and system for real-time status monitoring indicators of IT resources. Background Art

[0002] With the continuous development of IT systems, the monitoring platform has taken over a huge amount of resources. The need for status monitoring at all levels, including computing, storage, network, virtualization, containers, application software, service interfaces, and resource status analysis in various dimensions has accumulated a huge amount of monitoring indicators. These indicators are built by different personnel over many years of continuous construction. The operation and maintenance personnel change frequently, especially those who lack experience. Faced with a large number of monitoring indicators, it is difficult to quickly determine which indicators need to be focused on, which greatly affects the efficiency of monitoring status operation and maintenance and fault diagnosis.

[0003] Patent document CN116048814A discloses a method for automatic optimization of application resource specifications based on monitoring utility data, and its technical solution is: the first step: collecting load data of application nodes from the nodes; the second step: connecting each node to an external real-time monitoring system, and obtaining business monitoring indicators through the real-time monitoring system; the third step: setting parameters for changing configurations according to needs; the fourth step: calculating recommended specifications and offsets based on the parameters of the changed configuration, and choosing to manually or automatically implement the changed new configurations in batches during business off-peak periods.

[0004] The existing solutions have the following problems: there are many types of monitoring indicators and a large number of them. It is difficult for operation and maintenance personnel, especially those who lack experience, to quickly determine which indicators need to be focused on, which indicators can help quickly determine whether the status of monitoring resources is normal when a fault occurs, and which indicators can assist in locating the cause of the fault more quickly. Summary of the invention

[0005] In view of the defects in the prior art, the object of the present invention is to provide a method and system for automatically recommending indicators for real-time status monitoring of IT resources.

[0006] The method for automatically recommending real-time status monitoring indicators of IT resources provided by the present invention includes:

[0007] Abstract modeling step: abstract modeling of monitoring resources and their monitoring status information in the cloud computing platform;

[0008] Status quantification step: Quantify the alarm status of the monitoring indicators. The indicators that are currently in alarm are recommended first. When multiple indicators are in alarm at the same time, the recommendation priority is determined by the recommendation index of the indicator type and alarm level.

[0009] Relationship quantification step: quantify the relationship between monitoring resource entities and the relationship between monitoring indicators;

[0010] Behavior quantification step: quantify the query operations of users on monitoring indicators through visual web pages or API interface calls by call tracking and log tracking;

[0011] Recommendation calculation steps: Integrate the above-mentioned quantitative factors to calculate the recommended value of each indicator, and then rank the recommended values ​​according to different scenario rules to finally obtain the indicator recommendation result of a certain configuration item.

[0012] Preferably, the abstract modeling step includes:

[0013] Monitoring resource modeling, abstracting various monitoring domain objects into configuration items. Each configuration item has a configuration item code as a globally unique identifier;

[0014] Monitoring status modeling abstracts the status information of various dimensions during the operation of monitoring resources into indicators. Each indicator has a unique identifier, and its name consists of the indicator code and the indicator name. Each indicator includes one or more attributes to express the specific monitoring status connotation represented by the indicator. The indicator is finally mounted on the configuration item instance.

[0015] Preferably, the relationship quantification step includes: when there is no direct relationship but only an indirect relationship between resource entities, the relationship quantification weight is hierarchically attenuated.

[0016] Preferably, the recommended value of the monitoring indicator = initial recommended value + state increment + relationship increment + behavior increment;

[0017] Initial recommended value: determined by the type of monitoring indicator and the enablement rate of the indicator on the monitoring resource;

[0018] Status increment: The change in recommended popularity based on the current alarm status is determined by the alarm status generated by the monitoring resource, and the indicator that is currently alerting is given priority recommendation;

[0019] Relationship increment: the change in recommendation popularity based on the strength of the correlation, including the correlation between monitoring resources and the correlation between indicators;

[0020] Behavior increment: The change in recommendation popularity based on user operations, generated through various user operations on monitoring indicators.

[0021] Preferably, when the user performs status inspection or fault analysis and location of a specified monitoring resource object, a weighted calculation is performed based on the quantitative values ​​of the recommended impact factors of the monitoring indicators based on daily periodic statistical updates of offline calculations and the real-time status of the resource object to obtain a recommended list of monitoring indicators.

[0022] The IT resource real-time status monitoring indicator automatic recommendation system provided by the present invention comprises: an abstract modeling module, a status quantification module, a relationship quantification module, a behavior quantification module and a recommendation degree calculation module;

[0023] The abstract modeling module is responsible for abstract modeling of monitoring resources and monitoring status information in the cloud computing platform;

[0024] The state quantification module quantifies the alarm status of the monitoring indicators, and recommends the indicators that are currently in alarm first. When multiple indicators are in alarm at the same time, the recommendation priority is determined by the recommendation index of the indicator type and the alarm level.

[0025] The relationship quantification module quantifies the relationship between monitoring resource entities and the relationship between monitoring indicators;

[0026] The behavior quantification module quantifies the query operation of the user on the monitoring indicator through the visual web page or API interface call by calling tracking and log embedding;

[0027] The recommendation degree calculation module integrates the above-mentioned various quantitative factors to calculate the recommended value of each indicator, and then ranks the recommended values ​​according to different scenario rules to finally obtain the indicator recommendation result of a certain configuration item.

[0028] Preferably, the abstract modeling module includes:

[0029] Monitoring resource modeling, abstracting various monitoring domain objects into configuration items. Each configuration item has a configuration item code as a globally unique identifier;

[0030] Monitoring status modeling abstracts the status information of various dimensions during the operation of monitoring resources into indicators. Each indicator has a unique identifier, and its name consists of the indicator code and the indicator name. Each indicator includes one or more attributes to express the specific monitoring status connotation represented by the indicator. The indicator is finally mounted on the configuration item instance.

[0031] Preferably, the relationship quantification module includes: when there is no direct relationship but only an indirect relationship between resource entities, the relationship quantification weight is hierarchically attenuated.

[0032] Preferably, the recommended value of the monitoring indicator = initial recommended value + state increment + relationship increment + behavior increment;

[0033] Initial recommended value: determined by the type of monitoring indicator and the enablement rate of the indicator on the monitoring resource;

[0034] Status increment: The change in recommended popularity based on the current alarm status is determined by the alarm status generated by the monitoring resource, and the indicator that is currently alerting is given priority recommendation;

[0035] Relationship increment: the change in recommendation popularity based on the strength of the correlation, including the correlation between monitoring resources and the correlation between indicators;

[0036] Behavior increment: The change in recommendation popularity based on user operations, generated through various user operations on monitoring indicators.

[0037] Preferably, when the user performs status inspection or fault analysis and location of a specified monitoring resource object, a weighted calculation is performed based on the quantitative values ​​of the recommended impact factors of the monitoring indicators based on daily periodic statistical updates of offline calculations and the real-time status of the resource object to obtain a recommended list of monitoring indicators.

[0038] Compared with the prior art, the present invention has the following beneficial effects:

[0039] (1) The present invention realizes full-dimensional quantification of monitoring indicators through abstract modeling, state quantification, relationship quantification, behavior quantification, etc. When operation and maintenance personnel judge the state of monitoring resources or perform fault analysis and location, relevant monitoring indicator data information can be automatically recommended to the operation and maintenance personnel, thereby improving the efficiency of monitoring state inspection and fault analysis and location;

[0040] (2) The present invention realizes the full-dimensional quantification and recommendation popularity quantitative ranking of IT resource monitoring indicators. When operation and maintenance personnel check the status of monitoring resources or perform fault analysis, the monitoring indicator data information that needs the most attention can be automatically recommended to the operation and maintenance personnel, which significantly improves the efficiency of operation and maintenance personnel, especially inexperienced operation and maintenance personnel, in implementing monitoring status inspections and fault analysis and positioning, and greatly reduces the operation and maintenance threshold. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0042] Figure 1 The entity relationship topology of the resource;

[0043] Figure 2 Flowchart of a method for automatically recommending indicators for real-time status monitoring of IT resources. DETAILED DESCRIPTION

[0044] The present invention is described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those of ordinary skill in the art, several changes and improvements can also be made without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0045] Example 1

[0046] like Figure 2 The present invention provides an automatic recommendation method for real-time status monitoring indicators of IT resources, comprising:

[0047] Abstract modeling step: abstract modeling of monitoring resources and their monitoring status information in the cloud computing platform;

[0048] Status quantification step: Quantify the alarm status of the monitoring indicators. The indicators that are currently in alarm are recommended first. When multiple indicators are in alarm at the same time, the recommendation priority is determined by the recommendation index of the indicator type and alarm level.

[0049] Relationship quantification step: quantify the relationship between monitoring resource entities and the relationship between monitoring indicators;

[0050] Behavior quantification step: quantify the query operations of users on monitoring indicators through visual web pages or API interface calls by call tracking and log tracking;

[0051] Recommendation calculation steps: Integrate the above-mentioned quantitative factors to calculate the recommended value of each indicator, and then rank the recommended values ​​according to different scenario rules to finally obtain the indicator recommendation result of a certain configuration item.

[0052] The abstract modeling step includes:

[0053] Monitoring resource modeling, abstracting various monitoring domain objects into configuration items. Each configuration item has a configuration item code as a globally unique identifier;

[0054] Monitoring status modeling abstracts the status information of various dimensions during the operation of monitoring resources into indicators. Each indicator has a unique identifier, and its name consists of the indicator code and the indicator name. Each indicator includes one or more attributes to express the specific monitoring status connotation represented by the indicator. The indicator is finally mounted on the configuration item instance.

[0055] The relationship quantification step includes: when there is no direct relationship between resource entities but only an indirect relationship, the relationship quantification weight is hierarchically attenuated.

[0056] Recommended value of monitoring indicator = initial recommended value + state increment + relationship increment + behavior increment;

[0057] Initial recommended value: determined by the type of monitoring indicator and the enablement rate of the indicator on the monitoring resource;

[0058] Status increment: The change in recommended popularity based on the current alarm status is determined by the alarm status generated by the monitoring resource, and the indicator that is currently alerting is given priority recommendation;

[0059] Relationship increment: the change in recommendation popularity based on the strength of the correlation, including the correlation between monitoring resources and the correlation between indicators;

[0060] Behavior increment: The change in recommendation popularity caused by user operations is generated by recording various user operations on monitoring indicators through tracking points.

[0061] When users conduct status inspection or fault analysis and location of specified monitoring resource objects, a weighted calculation is performed based on the quantitative values ​​of the recommended impact factors of the monitoring indicators based on daily periodic statistical updates of offline calculations and the real-time status of the monitoring objects to obtain a recommended list of monitoring indicators.

[0062] The recommended heat calculation process of monitoring indicators is as follows:

[0063] 1. Recommend monitoring indicators based on the real-time status of the monitored object. The indicators that the monitored object is currently alerting are added to the recommended results in the order of alarm levels from high to low (Fatal, Error, Warning, Info); if the alarm levels are the same, they are ranked and recommended according to the popularity of the monitoring indicators.

[0064] 2. Recommend monitoring indicators based on the real-time status of the monitoring resources related to the monitoring object. The indicators that the related monitoring resources are currently warning about are included in the recommendation results. The rules for the recommendation order are the same as in step 1.

[0065] 3. According to the monitoring indicators enabled by the monitoring object and its related monitoring resources, combined with the offline recommendation calculation results, the recommendation results are completed. Offline recommendation achieves the recommended ranking of monitoring indicators by calculating the recommended heat value of the monitoring resource monitoring indicators.

[0066] The calculation formula for offline calculation of the recommended heat of the monitoring indicator of the monitoring object is: monitoring resource heat * monitoring indicator heat * monitoring resource relationship attenuation coefficient.

[0067] The calculation expressions of the factors involved in the formula are shown in the following table:

[0068]

[0069]

[0070]

[0071] Example 2

[0072] The present invention also provides an automatic recommendation system for IT resource real-time status monitoring indicators. The automatic recommendation system for IT resource real-time status monitoring indicators can be implemented by executing the process steps of the automatic recommendation method for IT resource real-time status monitoring indicators, that is, those skilled in the art can understand the automatic recommendation method for IT resource real-time status monitoring indicators as a preferred implementation of the automatic recommendation system for IT resource real-time status monitoring indicators.

[0073] The IT resource real-time status monitoring indicator automatic recommendation system includes the following modules:

[0074] Abstract modeling module, state quantification module, relationship quantification module, behavior quantification module, and recommendation degree calculation module.

[0075] The abstract modeling module is responsible for abstract modeling of monitoring resources and monitoring status information in the cloud computing platform.

[0076] Monitoring resource modeling is to abstract each type of monitoring domain object into a configuration item. Each configuration item has a configuration item code as a globally unique identifier.

[0077] The modeling objects of monitoring domain configuration items include but are not limited to:

[0078] 1. Hardware resource objects such as CPU, memory, disk array, router, switch, firewall, cabinet, computer room, data center, etc.

[0079] 2. Database, middleware, Web server, cache and other software resource objects

[0080] 3. Technical management objects such as API interfaces, Web services, logical network interfaces, clusters, subsystems, operation nodes, operation centers, projects, etc.

[0081] 4. Operation and maintenance A / B corners, service managers and other personnel

[0082] There are various types of relationships between different objects, including but not limited to:

[0083] 1. Operational relationship hierarchy: configuration item -> technical service -> business service -> application system -> tenant, etc.

[0084] 2. Resource relationship hierarchy: service->application software->operating system->virtualization->host->cabinet->cabinet block->computer room->data center, etc.

[0085] 3. Network device hierarchy: network segment->logical interface->physical interface->board->device, etc.

[0086] 4. Personnel relationship level: operation and maintenance personnel -> operation and maintenance personnel in the operation center -> projects, etc.

[0087] 5. Service call level: API interface->API interface service->service, etc.

[0088] Take the operating system (PCENDC0007) configuration item as an example:

[0089]

[0090] Based on the configuration item attributes and relationships described above, any type of resource will automatically generate a hierarchical entity relationship topology, such as Figure 1 .

[0091] Monitoring status modeling is to abstract the status information of various dimensions when monitoring resources are running into indicators, including the usage of various infrastructures such as computing, storage, and network by conventional monitoring resources, key log information, service call tracking information, application runtime analysis information, security risk information, and other information of various dimensions. Each indicator has a unique identifier. The name consists of the indicator code and the indicator name. Each indicator includes one or more attributes, which express the specific monitoring status connotation represented by the indicator. The indicator is finally mounted on the configuration item instance, indicating which dimensions of real-time status monitoring are enabled on the configuration item instance.

[0092] Take the indicator A00001-CPU instantaneous usage as an example:

[0093]

[0094] From the perspective of the importance of indicators, indicators are mainly divided into three types: life and death indicators, key indicators, and ordinary indicators.

[0095] a. Life and death indicators are monitoring indicators used to measure the life and death of the monitoring resource entity state, and have the highest quantitative weight. They are usually expressed as yes or no, such as normal / abnormal, yes / no, OK / ERROR, correct / wrong, YES / NO, etc. The default recommended weight quantitative value is 20.

[0096] b. Key indicators are mandatory indicators on monitoring resource entities. They have a higher weight and are usually configured with analysis linkage strategies to generate alarms and events. The default recommended weight quantization value is 10.

[0097] c. Ordinary indicators usually play an auxiliary analysis role and have a lower weight. They are generally not directly used to generate alarms or events. The default recommended weight quantization value is 3.

[0098] From the perspective of indicator monitoring dimensions, it is mainly divided into availability, reliability, maintainability, security, performance, etc. Each monitoring indicator has at least one monitoring dimension identifier.

[0099] Here are a few examples of metric dimension identifiers:

[0100]

[0101]

[0102] The state quantification module quantifies the alarm state of the monitoring indicator and recommends the indicator with the current alarm first. When multiple indicators are alarmed at the same time, the recommendation priority is determined by the recommendation index of the indicator type and alarm level mentioned above.

[0103] Alarm levels include Fatal, Error, Warning, Info, Normal, etc. Alarms of different levels have different default quantization weights.

[0104] Alarm level Quantization weights illustrate Fatal 100 The indicator is currently in the Fatal level alarm state Error 50 The indicator is currently in the Error level alarm state Warning 10 The indicator is currently in the Warning level alarm state Info 5 The indicator is currently in the Info level alarm state Normal 1 The indicator is not currently in a level warning state

[0105] State quantification is mainly affected by the current state of the monitored resources and the comprehensive state in the past period (such as the past natural month).

[0106] The relationship quantification module quantifies the relationship between monitoring resource entities and the relationship between monitoring indicators, and each relationship has a different default quantization weight.

[0107] The relationships between monitoring resource entities include but are not limited to the following categories:

[0108] Relationship Type Quantization weights Example Host 10 Application servers are hosted on top of operating systems use 5 Virtual machines use virtual hard disks combination 1 Two core switches are combined to form a stack Cluster 1 Two databases are a cluster High frequency calls 20 Application server high-frequency call cache Intermediate frequency call 15 Application servers call databases infrequently … … …

[0109] When there is no direct relationship between resource entities but only an indirect relationship, the relationship quantification weight will decay hierarchically.

[0110] The relationships between monitoring indicators include but are not limited to the following categories:

[0111] a. Master-slave relationship: The slave indicator is the detailed drill-down indicator of the master indicator and is the auxiliary analysis indicator of the master indicator.

[0112] Primary indicator relationship type Secondary indicator

[0113] A00001-CPU instantaneous usage master-slave A00008-10 processes using the most CPU

[0114] A00012- Physical memory instantaneous usage master-slave A00009- Use physical memory up to 10 processes

[0115] …Master-Slave…

[0116] b. Alarm attribution relationship: The relationship between the monitoring indicators of the monitoring resource and the monitoring indicators of other monitoring resources obtained based on historical alarm correlation analysis. Different quantitative weights are calculated based on the correlation analysis within a specified time period in the past.

[0117] Configuration item alarm indicator attribution Configuration item attribution indicator quantitative weight

[0118] Operating system A00001-CPU Web application server A00034-Java thread 10

[0119] PCENDC0007 Instantaneous usage rate PTOMDC0017 Pool usage

[0120] Operating system A00001-CPU Web application server A00034-Java thread 6

[0121] PCENDC0007 Instantaneous usage rate PTOMDC0018 Pool usage

[0122] …………

[0123] The user behavior quantification module quantifies the query operations of users (including but not limited to operation and maintenance personnel) on monitoring indicators through visual web pages or API interface calls by calling tracking and log embedding, and accurately describes the following information:

[0124] [What user] accessed the data of [what configuration item], [what indicator], [what time period] at [what point in time] and [what method].

[0125] The recommendation degree calculation module integrates the above-mentioned various quantitative factors to calculate the recommended value of each indicator, and then ranks the recommended values ​​according to different scenario rules to finally obtain the indicator recommendation result of a certain configuration item.

[0126] Simplified expression of recommended values ​​for monitoring indicators:

[0127] Recommended value of monitoring indicator = initial recommended value + state increment + relationship increment + behavior increment

[0128] The instructions are as follows:

[0129] 1. Initial recommended value: mainly determined by the type of monitoring indicator and the activation rate of the indicator on the monitoring resources.

[0130] 2. Status increment: The change in recommended heat based on the current alarm status, which is determined by the alarm status generated by the monitoring resource. Alarms are alarms based on monitoring indicators, and the indicators that are currently alarming should be recommended first.

[0131] According to the alarm rules of different resources, select the status data of the specified time range (such as the past 7 days / the past month / the past quarter, etc.) as the calculation and statistical period.

[0132] 3. Relationship increment: The change in recommendation popularity based on the strength of the correlation relationship, including the correlation between monitoring resources (resource correlation, call correlation, alarm attribution correlation, etc.) and the correlation between indicators.

[0133] 4. Behavior increment: The change in recommendation popularity based on user operations is mainly generated by various user operations on monitoring indicators. According to the user behavior rules, user behavior data in a specified time range (such as the past 7 days / the past month / the past quarter, etc.) is selected as the calculation statistical period.

[0134] When users perform status inspection or fault analysis and location of specified monitoring resource objects, a recommended list of monitoring indicators is obtained by weighted calculation based on the quantitative values ​​of the recommended impact factors of the monitoring indicators updated daily based on offline calculations and the real-time status of the resource objects.

[0135] Example 3

[0136] Embodiment 3 is a preferred example of Embodiment 1.

[0137] A certain XX cloud computing platform uses this method to achieve the quantification and recommendation popularity ranking of IT resource monitoring indicators. The following uses the real-time monitoring indicator data of the operating system with the configuration item code PCENDC0007 as an example to illustrate the specific process of indicator recommendation:

[0138] 1. The recommended indicators include all monitoring indicators of the configuration item specified by the user and its related configuration items.

[0139] 2. The indicator heat recommendation is to arrange the heat values ​​of the monitoring indicators of the current configuration item and the heat values ​​of the monitoring indicators of all related configuration items from high to low, and take the TOP-N with the highest heat values.

[0140] 3. The recommended heat value of the current alarm indicator is higher than the non-alarm indicator participating in the recommendation ranking, regardless of whether the mounted configuration item instance of the indicator is the user-specified configuration item itself or a related configuration item of the user-specified configuration item.

[0141] 4. If the user-specified configuration item itself or related configuration items have monitoring indicators that are currently alerting:

[0142] 4.1 If there are multiple alarm level indicators, it is recommended to sort them by alarm level (Warn / Error / Fatal).

[0143] 4.2 If there are multiple alarm indicators of the same level, they will be ranked and recommended according to the number of alarms in the past X days.

[0144] 4.3 If the alarm indicator has auxiliary analysis indicators, they are recommended together with the alarm indicator.

[0145] 4.4 If there is an attribution indicator for the alarm indicator, it is recommended together with the alarm indicator.

[0146] 5. If the user-specified configuration item itself and its related configuration items do not have the monitoring indicator that is currently alerting: calculate the heat value according to the monitoring indicator heat value calculation formula.

[0147] 6. The monitoring indicator heat value of the related configuration item must be multiplied by a relationship level attenuation coefficient. The indicator heat value of the current configuration item is equivalent to a relationship level attenuation coefficient of 1, that is, no attenuation. The relationship level attenuation coefficient of the indicator heat value of the related configuration item ranges from (0, 1). The more levels of relationship between the related configuration item and the current configuration item, the greater the relationship attenuation, and the smaller the attenuation coefficient value.

[0148] 7. If there are multiple isomorphic related configuration items with the same relationship type / level (for example, two equivalent application server instances are deployed on the same operating system), sort them by the number of alarms in the past X days.

[0149] Take the example of a user viewing the monitoring status of the operating system with the configuration item code PCENDC0007:

[0150] a. When the user queries, the operating system (PCENDC0007) has an indicator that is generating an alarm [A00001-CPU instantaneous usage]

[0151] b. The indicator [A00001-CPU instantaneous usage] has an auxiliary analysis indicator [A00008-10 processes using the most CPU]

[0152] c. The [A00001-CPU instantaneous usage] indicator of the operating system (PCENDC0007) has an attribution indicator based on the analysis of historical alarm events. The [A00034-Java thread pool usage] of the application server (PTOMDC0017)

[0153] d. The relevant configuration items include application server (PTOMDC0017), application server (PTOMDC0018), application server (PTOMDC0019), application server (PTOMDC0020), application server (PTOMDC0030) hosted on the operating system (PCENDC0007), and x86 virtual machine (PVVMDC0114) hosted by the operating system (PCENDC0007). Application server (PTOMDC0017) ranks high because of the number of indicators [A00036-Http Request] that are currently warning. Although application server (PTOMDC0030) has no current warnings, it has many historical warnings, so it ranks relatively high.

[0154] Indicator recommendation results

[0155] Recommended ranking Configuration items index 1 Operating System (PCENDC0007) A00001-CPU instantaneous usage 2 Operating System (PCENDC0007) A00008 - Top 10 processes using the most CPU 3 Application Server (PTOMDC0017) A00034-Usage of Java thread pool 4 Application Server (PNGXDD0001) A00328-Overview of application calls based on call chains 5 Operating System (PCENDC0007) A00002-Swap space usage 6 Application Server (PTOMDC0017) A00036-Http Request number 7 Application Server (PTOMDC0017) A00033-Application Server Status 8 Application Server (PTOMDC0030) A00033-Application Server Status 9 x86 virtual machine (PVVMDC0114) A00268-x86 virtual machine CPU main indicators 10 x86 virtual machine (PVVMDC0114) A00276-x86 virtual machine memory main indicators

[0156] Those skilled in the art know that, in addition to implementing the system, device and its various modules provided by the present invention in a purely computer-readable program code, it is entirely possible to implement the same program in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers and embedded microcontrollers by logically programming the method steps. Therefore, the system, device and its various modules provided by the present invention can be considered as a hardware component, and the modules included therein for implementing various programs can also be considered as structures within the hardware component; the modules for implementing various functions can also be considered as both software programs for implementing the method and structures within the hardware component.

[0157] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essence of the present invention. In the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. A method for automatically recommending real-time status monitoring indicators for IT resources, characterized in that: include: Abstract modeling step: abstract modeling of monitoring resources and their monitoring status information in the cloud computing platform; Status quantification step: Quantify the alarm status of the monitoring indicators. The indicators that are currently in alarm are recommended first. When multiple indicators are in alarm at the same time, the recommendation priority is determined by the recommendation index of the indicator type and alarm level. Relationship quantification step: quantify the relationship between monitoring resource entities and the relationship between monitoring indicators; Behavior quantification step: quantify the query operations of users on monitoring indicators through visual web pages or API interface calls by call tracking and log tracking; Recommendation calculation steps: Integrate the above-mentioned quantitative factors to calculate the recommended value of each indicator, and then rank the recommended values ​​according to different scenario rules to finally obtain the indicator recommendation result of a certain configuration item.

2. The method for automatically recommending real-time status monitoring indicators of IT resources according to claim 1, characterized in that: The abstract modeling step includes: Monitoring resource modeling, abstracting various monitoring domain objects into configuration items. Each configuration item has a configuration item code as a globally unique identifier; Monitoring status modeling abstracts the status information of various dimensions during the operation of monitoring resources into indicators. Each indicator has a unique identifier, and its name consists of the indicator code and the indicator name. Each indicator includes one or more attributes to express the specific monitoring status connotation represented by the indicator. The indicator is finally mounted on the configuration item instance.

3. The method for automatically recommending real-time status monitoring indicators of IT resources according to claim 1, characterized in that: The relationship quantification step includes: when there is no direct relationship between resource entities but only an indirect relationship, the relationship quantification weight is hierarchically attenuated.

4. The method for automatically recommending real-time status monitoring indicators of IT resources according to claim 1, characterized in that: Recommended value of monitoring indicator = initial recommended value + state increment + relationship increment + behavior increment; Initial recommended value: determined by the type of monitoring indicator and the enablement rate of the indicator on the monitoring resource; Status increment: The change in recommended popularity based on the current alarm status is determined by the alarm status generated by the monitoring resource, and the indicator that is currently alerting is given priority recommendation; Relationship increment: the change in recommendation popularity based on the strength of the correlation, including the correlation between monitoring resources and the correlation between indicators; Behavior increment: The change in recommendation popularity based on user operations, generated through various user operations on monitoring indicators.

5. The method for automatically recommending real-time status monitoring indicators of IT resources according to claim 1, characterized in that: When users conduct status inspections or fault analysis and location of specified monitoring resource objects, a weighted calculation is performed based on the quantitative values ​​of the recommended impact factors of the monitoring indicators based on daily periodic statistical updates of offline calculations and the real-time status of the resource objects to obtain a recommended list of monitoring indicators.

6. An automatic recommendation system for real-time status monitoring indicators of IT resources, characterized in that: include: Abstract modeling module, state quantification module, relationship quantification module, behavior quantification module and recommendation degree calculation module; The abstract modeling module is responsible for abstract modeling of monitoring resources and monitoring status information in the cloud computing platform; The state quantification module quantifies the alarm status of the monitoring indicators, and recommends the indicators that are currently in alarm first. When multiple indicators are in alarm at the same time, the recommendation priority is determined by the recommendation index of the indicator type and the alarm level. The relationship quantification module quantifies the relationship between monitoring resource entities and the relationship between monitoring indicators; The behavior quantification module quantifies the query operation of the user on the monitoring indicator through the visual web page or API interface call by calling tracking and log embedding; The recommendation degree calculation module integrates the above-mentioned various quantitative factors to calculate the recommended value of each indicator, and then ranks the recommended values ​​according to different scenario rules to finally obtain the indicator recommendation result of a certain configuration item.

7. The IT resource real-time status monitoring indicator automatic recommendation system according to claim 6, characterized in that: The abstract modeling module includes: Monitoring resource modeling, abstracting various monitoring domain objects into configuration items. Each configuration item has a configuration item code as a globally unique identifier; Monitoring status modeling abstracts the status information of various dimensions during the operation of monitoring resources into indicators. Each indicator has a unique identifier, and its name consists of the indicator code and the indicator name. Each indicator includes one or more attributes to express the specific monitoring status connotation represented by the indicator. The indicator is finally mounted on the configuration item instance.

8. The IT resource real-time status monitoring indicator automatic recommendation system according to claim 6, characterized in that: The relationship quantification module includes: when there is no direct relationship between resource entities but only an indirect relationship, the relationship quantification weight is hierarchically attenuated.

9. The IT resource real-time status monitoring indicator automatic recommendation system according to claim 6, characterized in that: Recommended value of monitoring indicator = initial recommended value + state increment + relationship increment + behavior increment; Initial recommended value: determined by the type of monitoring indicator and the enablement rate of the indicator on the monitoring resource; Status increment: The change in recommended popularity based on the current alarm status is determined by the alarm status generated by the monitoring resource, and the indicator that is currently alerting is given priority recommendation; Relationship increment: the change in recommendation popularity based on the strength of the correlation, including the correlation between monitoring resources and the correlation between indicators; Behavior increment: The change in recommendation popularity caused by user operations is generated by recording various user operations on monitoring indicators through tracking points.

10. The IT resource real-time status monitoring indicator automatic recommendation system according to claim 6, characterized in that: When users conduct status inspections or fault analysis and location of specified monitoring resource objects, a weighted calculation is performed based on the quantitative values ​​of the recommended impact factors of the monitoring indicators based on daily periodic statistical updates of offline calculations and the real-time status of the resource objects to obtain a recommended list of monitoring indicators.

Images