to_kafka throughput

未关闭
#398 13 条评论 1 个 reaction 已指派 0 人 在 GitHub 查看

还没有人认领这个 Issue。

评估

难度
4/5
预计耗时
3-5 天
新手友好度
35/100
Issue 类型
缺陷
描述清晰度
基本清楚
活跃度
停滞
技术栈
kafka, python
领域
stream-processing

调研方向

Start at the to_kafka sink and its self.producer.poll(0) call, then reproduce the behavior with the proposed test_to_kafka_throughput test. Done means 100 emitted items reach sink_to_list() within the test timeout without callbacks being limited by the 0.2-second poll interval.

由索引模型根据 Issue 内容生成。

描述

I'm testing to_kafka sink and its throughput is limited by polltime (0.2 sec). Looks like self.producer.poll(0) only polls for one message at a time and so only one callback is called every 0.2 seconds.

This fails:

def test_to_kafka_throughput():
    ARGS = {'bootstrap.servers': 'localhost:9092'}
    with kafka_service() as kafka:
        _, TOPIC = kafka
        source = Stream.from_iterable(range(100)).map(lambda x: str(x).encode())
        kafka = source.to_kafka(TOPIC, ARGS)
        out = kafka.sink_to_list()

        source.start()
        wait_for(
            lambda: len(out) == 100,
            5,
            period=0.1,
            fail_func=lambda: print("len(out) ==", len(out))
        )

The existing test_to_kafka test doesn't catch this, because it starts waiting on the result only after all the items are emitted.

I spent some time tinkering with the code, but can't figure out what's wrong and how to fix this, so any ideas are appreciated.

主要语言
Python
星标
1.3k
派生
149
30 天内没有已合并 PR

贡献指南

打开贡献指南

从这里开始

  1. 先读完整个 Issue，再读项目的贡献指南。
  2. 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
  3. Fork 仓库，在一个分支上完成修改。
  4. 提交 Pull Request，并在描述里引用这个 Issue 编号。

python-streamz/streamz 的其他 Issue

查看 python-streamz/streamz 的全部 Issue

相似的 Issue

更多 Python Issue

把新 issue 发到你的邮箱

精选适合新手参与的 GitHub issue 摘要。