Connecting to Rithmic

Basic Connection

To connect to Rithmic, instantiate a RithmicClient with your credentials and call await client.connect():

import asyncio
from async_rithmic import RithmicClient

async def main():
    client = RithmicClient(
        user="your_username",
        password="your_password",
        system_name="Rithmic Test",
        app_name="my_test_app",
        app_version="1.0",
        url="rituz00100.rithmic.com:443"  # Example: test gateway only
    )
    await client.connect()
    await client.disconnect()

asyncio.run(main())

Conformance

Important

To obtain the list of production URLs, you must first pass the conformance test.

First, contact Rithmic.
You will be asked to connect to the order plant and leave the app running.
You can use the conformance.py script to accomplish this.
See this GitHub issue comment for one user’s notes on the conformance process.

Custom Reconnection Settings

async_rithmic will automatically attempt to reconnect if the connection to Rithmic is lost (no matter the reason for the disconnection). This includes network drops, internet hiccups, server restarts, or unexpected protocol errors.

You can customize the reconnection behavior using ReconnectionSettings. This allows you to define backoff strategies like exponential delay with jitter:

from async_rithmic import ReconnectionSettings

reconnection = ReconnectionSettings(
    max_retries=None,  # retry forever
    backoff_type="exponential",
    interval=2,
    max_delay=60,
    jitter_range=(0.5, 2.0)
)

client = RithmicClient(
    ...
    reconnection_settings=reconnection
)

Custom Retry Settings

If a request takes longer than the configured timeout (in seconds), it will be retried up to max_retries times. This helps handle situations where a response might be unexpectedly slow due to network delays, backend congestion, or temporary service issues.

You can customize the retry behavior using RetrySettings.

from async_rithmic import RetrySettings

retry = RetrySettings(
    max_retries=3,           # Maximum number of attempts
    timeout=30.0,            # Timeout in seconds for each attempt
    jitter_range=(0.5, 2.0)  # Random delay (in seconds) between retries
)

client = RithmicClient(
    ...
    retry_settings=retry
)

Note

Only timeouts are retried; other errors (like bad requests) will not trigger retries.
Each retry uses a new request ID, so every attempt is a fresh request to Rithmic.
If all attempts time out, a TimeoutError is raised.
This mechanism makes your client more robust to occasional slowdowns in the Rithmic infrastructure, network hiccups, or other unpredictable delays.

Event Handlers

You can register callbacks to respond to connection lifecycle events such as successful plant connection or disconnection.

async def on_connected(plant_type: str):
    print(f"Connected to plant: {plant_type}")

async def on_disconnected(plant_type: str):
    print(f"Disconnected from plant: {plant_type}")

client.on_connected += on_connected
client.on_disconnected += on_disconnected

Debugging & Logging

async_rithmic uses Python’s standard logging module to emit log messages. You can easily control the verbosity of logging to help with debugging or to better understand what your client is doing.

To see detailed logs from the package (such as connection events, requests, and errors), set the logger named “rithmic” to DEBUG level:

import logging

logging.getLogger("rithmic").setLevel(logging.DEBUG)

Setting the level to DEBUG will print more detailed information to the console, which can be very helpful when troubleshooting.