InfluxDB data schema

This page documents an earlier version of InfluxDB. InfluxDB v2.7 is the latest stable version. View this page in the v2.7 documentation.

InfluxDB data elements are stored in time-structured merge tree (TSM) and time series index (TSI) files to efficiently compact stored data.

InfluxDB also provides a tabular data schema that includes the following:

• Annotation rows
• Header row
• Data rows
• Other columns
• Group keys

The tabular data schema is used for the following:

• To view raw data when exploring metrics with InfluxDB
• To return query results in annotated CSV syntax

Annotation rows

Annotation rows describe column properties, for example:

• #group
• #datatype
• #default

Header row

The header row defines column labels that describe data in each column, for example:

• table
• _time
• _value
• _field
• _measurement
• tag key names (without underscore prefix): tag-1, tag-2

Data rows

Each data row contains the data specified in the header row for one point.

Other columns

In addition to the columns in each data row (specified in the header row), the following columns are optional:

• annotation
• result
• table

Group keys

Determine the contents of output tables in Flux by grouping records that share common values in specified columns. Learn more about grouping your data with Flux.

key concepts