Concept

The TAG table is a virtual table that can be created by only one user, and is responsible for data storage and related additional information management for sensor data processing.

The TAG table provides three conceptual data processing spaces as described below.

Sensor Partition

This is an internal sensor data table based on the schema defined by the user when TAG table is created.

This data can be extracted through a SELECT query on the TAG table.

This table has very strong data management capabilities as listed below.

1. Tens of thousands to Hundreds of thousands of sensor data can be loaded at high speed.
2. Tens of thousands of sensor data can be retrieved at high speed given the time range condition.
3. Real-time compression allows long-term storage of sensor data.
4. Chronological deletion of sensor data beginning with the oldest is possible.

The user sensor data to be stored is, in basic, a time series data and is a specific data type with the corresponding tag of the name, time, and 64-bit real value.

ROLLUP Partition

This is an internal table that automatically generates statistical data based on the sensor data stored in the sensor storage.

This was developed in order to obtain statistical data over a long period of time within a few seconds in real time analysis.

ROLLUP table supports three types of tables: hour, minute, and second; and five types of statistics functions: MIN, MAX, AVG, SUM, COUNT.

To get this ROLLUP result value, you must execute a SELECT query with HINT in the TAG table.

META Table

This is another table that stores the name and additional meta information for TAG data.

Users can generate this meta information with the INSERT clause and manipulate it in several ways through SELECT, UPDATE, and DELETE.

How to use Tag table

• Creation and Dropping of Tag table
• Managing tag meta (tag name)
• Manipulating tag data
• Select in Rollup Table
• An example of tag table