Migrating from InfluxDB

This page describes the steps to importing data from InfluxDB OSS or InfluxDB Cloud to QuestDB.

Overview

Data stored in InfluxDB needs to be exported in order to import it into QuestDB. There are two ways of exporting data from InfluxDB:

• Run a SQL query using InfluxDB API and get results in JSON
• Run the inspect command and get results in ILP

The first approach is simpler and might suffice if you are migrating a small to moderate dataset. For larger datasets it is advised to use the second option.

SQL query for importing small tables

note

This approach is recommended only for small to moderate datasets.

When using InfluxDB API to run a SQL query, results will be in JSON. Since we cannot import JSON files directly into QuestDB, we will need to convert the results into CSV. There are many ways to do so, and one of those is using the jq JSON processor.

To run the SQL query you will need to have an API token.

The below is an example to query a table using the SQL API endpoint and convert the results to CSV:

curl --get http://localhost:8086/query --header "Authorization: Token zuotzwwBxbXIor4_D-t-BMEpJj1nAj2DGqNSshTUyHUcX0DMjI6fiBv_pgeW-xxpnAwgEVG0uJAucEaJKtvpJA==" --data-urlencode "db=bench" --data-urlencode "q=SELECT * from readings LIMIT 1000;" | jq '.results[].series[].values[] | @csv'

The resulting CSV can be then imported into QuestDB.

The inspect command and ILP for importing datasets at scale

To move data from InfluxDB into QuestDB at scale, it is best to use the influxd inspect command to export the data, as the export-lp subcommand allows exporting all time-structured merge tree (TSM) data in a bucket as ILP messages in a big text file.

The text file can then be inserted into QuestDB. This assumes you are migrating from self-managed InfluxDB and have access to execute the inspect command.

For InfluxDB Cloud users, the first step should be exporting the data from cloud to InfluxDB OSS before following the instructions.

Instructions

Generate admin token

Make sure you have an admin token generated and set the env variable export INFLUX_TOKEN.

For example:

export INFLUX_TOKEN=xyoczzmBxbXIor4_D-t-BMEpJj1nAj2DGqNSshTUyHUcX0DMjI6fiBv_pgeW-xxpnAwgEVG0uJAucEaJKtvpJA==

Find out your org_id and bucket_id

You need to know your org_id and bucket_id you will export. If you don’t know them you can issue influx org list and influx bucket list --org-id YOUR_ID to find those values.

Export the bucket contents

Now you can just export the bucket contents by using inspect export-lp command and by defining a destination folder:

influxd inspect export-lp --bucket-id YOUR_BUCKET_ID --output-path /var/tmp/influx/outputfolder

Please note the step above can take a while. As an example, it took almost an hour for a 160G bucket on a mid-AWS EC2 instance.

Connect to QuestDB

Connect to your QuestDB instance and issue a CREATE TABLE statement. This is not technically needed as once you start streaming data, your table will be automatically created. However, this step is recommended because this allows fine tuning some parameters such as column types or partition.

Since the data is already in ILP format, there is no need to use the official QuestDB client libraries for ingestion.

You only need to connect via a socket to your instance and stream row by row.

The below is an example Python code streaming the instance:

import socket

import sys

sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)

def send_utf8(msg):

print(msg)

sock.sendall(msg.encode())

if __name__ == '__main__':

try:

sock.connect(('localhost', 9009))

with open("ilp_influx_export.ilp") as infile:

for line in infile:

print(line)

send_utf8(line)

except socket.error as e:

sys.stderr.write(f'Got error: {e}')

sock.close()

Transform data in QuestDB

Since InfluxDB exports only one metric for each ILP line, this means that if you are storing more than one metric for the same series, one row will create multiple ILP lines with one valid metric value and the other metrics shown as NULL. Therefore, we recommend transforming your data in QuestDB.

For example, if you query a table with several metrics:

SELECT * FROM diagnostics WHERE timestamp = '2016-01-01T00:00:00.000000Z' AND driver='Andy' AND name='truck_150')

Your result may be something like this:

A way to solve this is to execute a SQL query grouping data by all the dimensions and selecting the maximum values for all the metrics:

SELECT

timestamp,

device_version,

driver,

fleet,

model,

name,

max(current_load) AS current_load,

max(fuel_capacity) AS fuel_capacity,

max(fuel_state) AS fuel_state,

max(load_capacity) AS load_capacity,

max(nominal_fuel_consumption) AS nominal_fuel_consumption,

max(status) AS status

FROM

diagnostics;

This produces aggregated rows containing all the metrics for each dimension group:

You can use the INSERT keyword to output the processed result into a new table.

See also

• Comparing InfluxDB, TimescaleDB, and QuestDB time series databases
• Comparing InfluxDB and QuestDB databases