Real-time dataset overview

更新时间: 2026-06-17 09:53:25

A real-time dataset is a virtual table structure that serves as the foundation for creating real-time tags through metric mapping. You can define real-time datasets by using different methods based on your data source.

Feature overview

image

Real-time datasets support different definition methods based on the data source.

  • Data from Events: Define event properties or their statistical results as dataset metrics. For example, define an "Order event", create a real-time dataset based on the event, and create a real-time tag "Total consumption amount in the last 1 day".

  • Data from Tables: Parse and process fields from data source tables such as HBase and Hologres to define dataset metrics. For example, create a real-time dataset by querying transaction data in HBase and create a real-time tag "Number of orders in the last 7 days".

  • Data from API Requests: Parse and process request parameters to define dataset metrics. For example, acquire data from third-party open platforms to define real-time dataset metrics and create real-time tags.

The following table describes the available creation methods:

Creation Method

Description

Data from Events

Create a dataset for real-time analysis by preprocessing events

Preprocess events and use the results as dataset metrics.

Data from Tables

Create a real-time dataset using HBase

Parse fields from HBase data source tables by using calculation scripts to define dataset metrics.

Creating a Real-Time Dataset via MySQL

Process MySQL data sources by using SQL to define dataset metrics.

Create a dataset in real-time with Hologres

Process Hologres data sources by using SQL to define dataset metrics.

Creating a Real-Time Dataset with PostgreSQL

Process PostgreSQL data sources by using SQL to define dataset metrics.

上一篇: Real-time dataset 下一篇: Create a real-time dataset through event preprocessing
阿里云首页 智能数据建设与治理 Dataphin 相关技术圈