Streams in Vineyard#

Stream is an abstraction upon the immutable data sharing storage that allows convenient pipelining between computing engines. Like pipe, stream in vineyard enable efficiently inter-engine communication without introducing the overhead of data serialization/deserialization and copying.

A common use case of stream in vineyard is one process continues to producing data chunks (e.g., an IO reader) and another process can do some scan computing over the data (e.g., filtering and aggregation operators). A stream is consist of a sequence of immutable data chunks that produced by the former engine and consumed by the later engine.

This section will cover the usage of streams in vineyard.

Using streams#

We first import required packages:

import threading
import time
from typing import List

import numpy as np
import pandas as pd

import vineyard
from import RecordBatchStream


Vineyard has defined some built-in stream types, e.g., For other stream types, you could refer to Streams.

Then we define a producer which generate some random dataframe chunks and put into the stream:

A producer of RecordBatchStream#
 def generate_random_dataframe(dtypes, size):
     columns = dict()
     for k, v in dtypes.items():
         columns[k] = np.random.random(size).astype(v)
     return pd.DataFrame(columns)

 def producer(stream: RecordBatchStream, total_chunks, dtypes, produced: List):
     writer = stream.writer
     for idx in range(total_chunks):
         chunk = generate_random_dataframe(dtypes, 2)  # np.random.randint(10, 100))
         chunk_id = vineyard_client.put(chunk)
         produced.append((chunk_id, chunk))

And a consumer which takes the chunks from the stream in a loop until receive a StopIteration exception:

A consumer of RecordBatchStream#
 def consumer(stream: RecordBatchStream, total_chunks, produced: List):
     reader = stream.reader
     index = 0
     while True:
             chunk =
             print('reader receive chunk:', type(chunk), chunk)
             pd.testing.assert_frame_equal(produced[index][1], chunk)
         except StopIteration:
         index += 1

Finally, we can test the producer and consumer using two thread:

Connect the producer and consumer threads using vineyard stream#
 def test_recordbatch_stream(vineyard_client, total_chunks):
     stream =
     dtypes = {
         'a': np.dtype('int'),
         'b': np.dtype('float'),
         'c': np.dtype('bool'),

     client1 = vineyard_client.fork()
     client2 = vineyard_client.fork()
     stream1 = client1.get(
     stream2 = client2.get(

     produced = []

     thread1 = threading.Thread(target=consumer, args=(stream1, total_chunks, produced))

     thread2 = threading.Thread(target=producer, args=(stream2, total_chunks, dtypes, produced))


 if __name__ == '__main__':
     vineyard_client = vineyard.connect("/tmp/vineyard.sock")
     test_recordbatch_stream(vineyard_client, total_chunks=10)

For more detailed API about the streams, please refer to Streams.