InfluxDB 3 Python Client

Warnings

• gotcha The `org` parameter, which was mandatory in InfluxDB 2.x and earlier `influxdb-client` versions, is no longer required or supported by `influxdb3-python`. Including it can lead to unexpected behavior or errors.
  Fix: Remove the `org` parameter from client initialization or write calls.
• gotcha Using `write_dataframe()` or `query_dataframe()` methods requires either `pandas` or `polars` to be installed as optional dependencies. If these libraries are not installed, attempting to use these methods will raise an `ImportError`.
  Fix: Install the desired DataFrame library: `pip install 'influxdb3-python[pandas]'` or `pip install 'influxdb3-python[polars]'`.
• gotcha The `query_async()` method (introduced in v0.12.0) is an `awaitable` coroutine. It must be called with `await` within an `async` function. Calling it directly without `await` will not execute the query and might lead to unexpected behavior or a `RuntimeWarning` if not awaited.
  Fix: Ensure `query_async()` is called as `await client.query_async(query)` within an `async def` function and an active `asyncio` event loop.
• gotcha The `WriteOptions.no_sync` flag (introduced in v0.14.0) allows for faster writes by not waiting for Write-Ahead Log (WAL) persistence confirmation. While faster, data is not guaranteed to be durable if the server crashes immediately after the write. Default is `False`.
  Fix: Only set `WriteOptions.no_sync=True` when write performance is critical and some data loss upon server failure is acceptable. Otherwise, rely on the default `False` for durability.
• gotcha Prior to version 0.18.0, `InfluxDBClient3.write_file()` and `InfluxDBClient3.write_dataframe()` could fail when batching mode was enabled due to a bug. This was fixed in v0.18.0.
  Fix: Upgrade to `influxdb3-python` v0.18.0 or newer to ensure correct behavior with `write_file()` and `write_dataframe()` in batching mode.

Install

• pip install influxdb3-python Install core client

Imports

• InfluxDBClient3

  from influxdb_client_3 import InfluxDBClient3

  This client is specifically for InfluxDB 3.0 and uses 'influxdb_client_3' for its package name, distinct from the InfluxDB 2.x client ('influxdb_client').
• Point

  from influxdb_client_3 import Point

  For InfluxDB 3.0, the `Point` class for data construction is directly available from the top-level `influxdb_client_3` package.

Quickstart

Initializes the client, writes a single data point, and then queries data from InfluxDB 3.0. Credentials are sourced from environment variables for secure and flexible deployment. Remember to replace 'YOUR_API_TOKEN' and 'YOUR_DATABASE' or set corresponding environment variables.

import os
from influxdb_client_3 import InfluxDBClient3, Point

# Ensure these environment variables are set or replace with actual values
# Example: export INFLUXDB_V3_HOST="us-east-1-1.aws.cloud2.influxdata.com"
HOST = os.environ.get("INFLUXDB_V3_HOST", "us-east-1-1.aws.cloud2.influxdata.com")
TOKEN = os.environ.get("INFLUXDB_V3_TOKEN", "YOUR_API_TOKEN")
DATABASE = os.environ.get("INFLUXDB_V3_DATABASE", "YOUR_DATABASE")

client = InfluxDBClient3(host=HOST, token=TOKEN, database=DATABASE)

try:
    # Write data
    point = Point("measurement1").tag("tag1", "value1").field("field1", 1.0)
    client.write(point)
    print("Data written successfully.")

    # Query data
    query = f"SELECT * FROM measurement1 WHERE time > now() - interval '1 hour'"
    table = client.query(query=query, database=DATABASE)

    print("\nQuery results:")
    for row in table:
        print(row)

except Exception as e:
    print(f"An error occurred: {e}")
finally:
    client.close()