KafkaIQ

from kafka import KafkaProducer import argparse, time def produce(bootstrap, topic, n, linger_ms, batch): p = KafkaProducer( bootstrap_servers=bootstrap, acks="all", linger_ms=linger_ms, batch_size=batch, value_serializer=lambda v: v.encode("utf-8"), ) t0 = time.time() for i in range(n): p.send(topic, f"msg-{i}") # optional tiny sleep to avoid overwhelming small dev setups # time.sleep(0.0005) p.flush(30) dt = time.time() - t0 print(f"Produced {n} messages to '{topic}' in {dt:.2f}s (~{int(n/max(dt,1)):d}/s)") if __name__ == "__main__": ap = argparse.ArgumentParser() ap.add_argument("--bootstrap", default="localhost:9092") ap.add_argument("--topic", default="x-topic") ap.add_argument("--num", type=int, default=50000) ap.add_argument("--linger-ms", type=int, default=5) ap.add_argument("--batch", type=int, default=32768) args = ap.parse_args() produce(args.bootstrap, args.topic, args.num, args.linger_ms, args.batch) # python .\producer.py --topic my-topic --num 50000