Can you delete data from influxdb?

sql

data-retention

influxdb

database-management

byAlex Kataev·Jan 28, 2025

Yes, you can delete data from InfluxDB using the DELETE influx query language (InfluxQL):

-- SQL: old but reliable, like your grandma's apple pie recipe
DELETE FROM "your_measurement" WHERE "your_condition"

To prune data points over a certain threshold:

-- Eliminate all those "hot" data points... they're not so "cool" anymore
DELETE FROM "temperature" WHERE "value" > 100

Heads up: Remember, deletion is permanent. Confirm your conditions carefully, you don't want to lose your favorite data point!

Check your InfluxDB version first

Before you whip out the deletion commands, confirm your InfluxDB version. Different versions might require different syntaxes. For instance, InfluxDB 2.0 uses CLI and its REST API endpoints, unlike InfluxQL DELETE and DROP SERIES commands familiar to the 1.x users.

Let's talk data retention policies

One smart way to manage data life cycle in InfluxDB is through data retention policies (DRPs). DRPs automatically age out data after a specified duration, tidying up your dbase just like a virtual housekeeper!

-- Even your data needs good hygiene habits!
CREATE RETENTION POLICY "one_week" ON "database_name" DURATION 1w REPLICATION 1 DEFAULT

InfluxDB 2.0 has a different deletion style

In InfluxDB 2.0, you will have to trade the simple InfluxQL with a more syntax-heavy deletion command which requires the bucket, start-time, stop-time and predicate:

# Become a time-traveling data assassin!
influx delete --bucket your_bucket --start '1970-01-01T00:00:00Z' --stop '2021-01-01T00:00:00Z' --predicate '_measurement="your_measurement"'

The softer side of deletion: soft delete method

Sometimes, you might want to retain data points for audit trails or condition-based deletions. The soft delete method adds a boolean field (like "ForUse") to mark records inactive, keeping the data intact for any rainy day!

-- Like a relationship status, it's complicated...
UPDATE "your_measurement" SET "ForUse" = FALSE WHERE "condition"

Before you delete...

Trial run in a controlled environment

Firstly, you want to rehearse your deletion act in a controlled environment (development or staging) before stepping onto the production stage. Always backup data before a grand deletion performance!

Fresh from the known bugs office

For any known bugs related to deletion, always check in on the InfluxDB GitHub Issues. It's best not to repeat mistakes someone else has already made!