Streaming live trades

You will record live BTC/USDT trades from Binance to Parquet for a few seconds, then stop the stream cleanly. This builds on Your first backfill.

Stream for a while, then stop

stream runs until the stop_event is set. Here we stop after ten seconds; in a daemon you would set it on shutdown.

import asyncio
from dccd import Client

async def main():
    async with Client() as c:
        stop = asyncio.Event()
        loop = asyncio.get_running_loop()
        loop.call_later(10, stop.set)          # stop after 10 s
        await c.stream("binance", "BTC/USDT", "trades", stop_event=stop)

asyncio.run(main())

Read what you recorded

async def main():
    async with Client() as c:
        df = c.read("binance", "BTC/USDT", "trades")
        print(len(df), "trades")
        print(df.select(["TS", "price", "amount", "side"]).head(3))

asyncio.run(main())

Each row is one trade, deduplicated on its tid, with a nanosecond TS. Trades land in a per-day file under …/binance/trades/BTC-USDT/.

Run it as a service instead

Instead of a script, declare the stream as a job in your config and let the daemon supervise it (auto-reconnect, restart on failure):

jobs:
  - exchange: binance
    pairs: [BTC/USDT]
    data_type: trades
    operation: stream
    trigger_kind: supervised

dccd start        # scheduler + streams + web UI

See Web UI to watch it live and Configuration Reference for the full job schema.