Author malinoff
Recipients malinoff, yselivanov
Date 2017-08-01.07:41:05
SpamBayes Score -1.0
Marked as misclassified Yes
Message-id <>
When there is a huge amount of `writer.write` calls followed by `await writer.drain()` on a non-paused channel, and there are no other coroutine switches, `await writer.drain()` immediately returns without a switch. This is because `` do not `yield` or `yield from` on a non-paused stream.

Use-case: AMQP basic.publish method, for which the broker (rabbitmq) do not send any replies back. Trying to publish 4k messages results in the following warnings (PYTHONASYNCIODEBUG env variable is set):
`Executing <Handle <TaskWakeupMethWrapper object at 0x1106fde28>(<Future>) created at /Users/malinoff/Projects/ideas/amqproto/amqproto/> took 2.371 seconds`
2.371 seconds is the time spent on 4k `basic_publish` calls.

You can find the test itself on github:

An easy fix would be to replace return ( with yield (and but the code below under the else clause; I'm willing to prepare a pull request), but maybe I'm missing something and such behavior is intentional?
Date User Action Args
2017-08-01 07:41:06malinoffsetrecipients: + malinoff, yselivanov
2017-08-01 07:41:06malinoffsetmessageid: <>
2017-08-01 07:41:06malinofflinkissue31096 messages
2017-08-01 07:41:05malinoffcreate