DataView standardizes indicator management to help enterprises quickly formulate SLOs

In the wave of digital transformation of enterprises, more and more enterprises are moving towards comprehensive and refined operations. Among them, the operation and maintenance department often evaluates the health status of services by formulating service level indicators (SLIs), and ensures the stability of enterprise services by achieving service level objectives (SLOs), providing strong operational support for enterprises. At present, when formulating service level indicators (SLIs), enterprises often face the problem of scattered data and the inability to associate definitions. When a service fails, they are faced with logging into various monitoring platforms such as IT infrastructure monitoring, APM application performance monitoring, middleware monitoring, and log monitoring to check monitoring data. Since the data has no correlation, it is difficult to locate the fault, which affects the achievement of service level objectives (SLOs).

Bonree DataView—Building an integrated, standardized operation and maintenance data platform <br /> To solve the problem of enterprise operation and maintenance data analysis difficulties, Bonree DataView provides an integrated, standardized, and visualized operation and maintenance data platform construction solution. Through four steps, it helps enterprises quickly build a comprehensive data analysis platform.

DataView provides a rich and fast data integration method to help enterprises centralize operation and maintenance data and lay a solid foundation for subsequent data analysis.

Data integration methods include:

1. Monitoring data collection: Through SmartAgent, collect monitoring data of various service components such as IT infrastructure, containers, databases, middleware, etc.

2. Monitoring platform connection: Supports quick connection with all Bori products, Zabbix, Promethus, Alibaba Cloud monitoring and other monitoring platforms, and obtains data collected by each monitoring platform

3. Database connection: connect to Mysql, Oracle and other databases to obtain data, and connect to Kafka message middleware to consume data

4. Custom reporting: reporting data via HTTP

After accessing data from different sources, the first problem is that the inconsistency of data format and data language leads to difficulty in using the data and the inability to associate data. To this end, DataView provides standardized indicators and entity management functions to standardize the accessed data from two aspects. The functions include:

1. Standardized indicator definition : supports converting data in different formats into standardized indicator formats: indicator name, indicator dimension; uniformly obtains data in the form of indicators, and supports multi-dimensional indicator analysis.

2. Standardized entity definition: supports defining the same analysis objects in different data sources as standardized entity types; thus supporting the analysis of indicators in multiple data sources.

Usage Examples

Table 1: For a certain application, the user request data collected by the front end is as follows:

Table 2: For a certain application, the response time data collected by the server is as follows:

After DataView standardizes indicators and entity definitions, you can obtain:

• 2 indicators

Among them, through the definition of dimensions, it supports obtaining indicator statistical data of different dimensions, such as obtaining the number of application front-end user requests for a certain operating system.

• 1 entity type

Among them, by associating the application with the indicators in Table 1 and Table 2 through data association, the application's request count statistics and response time statistics for a certain time period can be obtained at the same time.

After data standardization, DataView supports the definition of health scoring models for entity types. It can evaluate the health status of entity types based on multiple indicators of entity types, helping enterprises quickly define service level indicators (SLIs).

Example:

The expected configuration service health score is calculated as follows: average response time score * 0.2 + error rate score * 0.6 + Apdex index score * 0.2, where:

• Average Response Time (T) Score:
• T ≥ 2000, score = 0
• 1000 < T < 2000, score = [1 - (T - 1000) / 1000] * 100
• T ≤ 1000, score = 100
• Error rate (e) score = (1 - e) * 100
• Apdex score = Apdex * 100

To meet the data analysis needs of different enterprise scenarios, DataView also provides the ability to customize visualization applications. Users can quickly configure visualization dashboards by dragging and dropping, and quickly configure scenario-based data analysis applications through functions such as custom component interactions and custom application menus.

Click to read the original text , come and apply for the DataView trial version to experience the full range of data analysis platform~