Este conteúdo foi traduzido por máquina para sua conveniência e a Huawei Cloud não pode garantir que o conteúdo foi traduzido com precisão. Para exibir o conteúdo original, use o link no canto superior direito para mudar para a página em inglês.

Nesta página

Mostrar todos

Central de ajuda/ MapReduce Service/ Visão geral de serviço/ Componentes/ IoTDB/ IoTDB Basic Principles

IoTDB Basic Principles

Atualizado em 2023-08-14 GMT+08:00

Ver PDF

Database for Internet of Things (IoTDB) is a software system that collects, stores, manages, and analyzes IoT time series data. Apache IoTDB uses a lightweight architecture and features high performance and rich functions.

IoTDB sorts time series and stores indexes and chunks, greatly improving the query performance of time series data. IoTDB uses the Raft protocol to ensure data consistency. In time series scenarios, IoTDB pre-computes and stores data to improve analysis performance. Based on the characteristics of time series data, IoTDB provides powerful data encoding and compression capabilities. In addition, its replica mechanism ensures data security. IoTDB is deeply integrated with Apache Hadoop and Flink to meet the requirements of massive data storage, high-speed data reading, and complex data analysis in the industrial IoT field.

This section applies to MRS 3.1.5 or later.

IoTDB Architecture

The IoTDB suite consists of multiple components to provide a series of functions such as data collection, data writing, data storage, data query, data visualization, and data analysis.

Figura 1 shows the overall application architecture after all components of the IoTDB suite are used. IoTDB refers to the time series database component in the suite.

Figura 1 IoTDB architecture
Clique para ampliar

Users can use Java Database Connectivity (JDBC) to import the time series data and system status data (such as server load, CPU usage and memory usage) collected from device sensors, as well as time series data in message queues, applications, or other databases, to the local or remote IoTDB. Users can also directly write the preceding data into a local TsFile file or a TsFile file in the HDFS.
Users can write TsFile files to the HDFS to implement data processing tasks such as exception detection and machine learning on the Hadoop or Flink data processing platform.
The TsFile-Hadoop or TsFile-Flink connector can be used to allow Hadoop or Flink to process the TsFile files written to the HDFS or local host.
The analysis result can be written back to a TsFile in the same way.
IoTDB and TsFile also provide client tools to meet users' requirements for viewing and writing data in SQL, script, and graphical formats.

IoTDB Principles

Based on the attribute hierarchy, attribute coverage, and subordinate relationships between data, the IoTDB data model can be represented as the attribute hierarchy, as shown in Figura 2. The hierarchy is as follows: power group layer - power plant layer - device layer - sensor layer. ROOT is a root node, and each node at the sensor layer is a leaf node. According to the IoTDB syntax, the path from ROOT to a leaf node is separated by a dot (.). The complete path is used to name a time series in the IoTDB. For example, the time series name corresponding to the path on the left in the following figure is ROOT.ln.wf01.wt01.status.

Figura 2 IoTDB data model
Clique para ampliar