1
votes

I've been trying trying with no luck to create a endless chain of client instances.

I'm developing an asyncio app, this app among many others things, like running a server with loop.create_server(), needs every 10 seconds to connect to a list of servers, send some data and then disconnect.

I keep getting 2 errors: "runtimeError: Event loop is running." or "asyncio > Task was destroyed but it is pending!"

The code down below works.

import asyncio
from tcp.client import Client


def send_to_peers(data):
    for index in range(1, 3): #this for-loop just for simulating a list of peers
    try:
        loop = asyncio.get_event_loop()
        coro = loop.create_connection(lambda: Client(), '127.0.0.1', 10000 + index)
        _, proto = loop.run_until_complete(coro)
        msg = data + "-" + str(index) + "\n"
        proto.transport.write(str.encode(msg))
        proto.transport.close()
    except ConnectionRefusedError as exc:
        print(exc)

def infinite():
    for index in range(5): #again this should be a While True:
        #there should be here an asyncio.sleep(10)
        send_to_peers(str(index))

infinite()

But when I call it from the main_loop things start to break.

async def infinite_loop():
    for index in range(5):
        print("loop n " + str(index))
        task = asyncio.Task(send_to_peers(str(index)))
        await asyncio.sleep(10)
        task.cancel()
        with suppress(asyncio.CancelledError):
            await task

main_loop = asyncio.get_event_loop()
main_loop.run_until_complete(infinite_loop())
main_loop.run_forever()

I've tried to giving the main_loop to send_to_peers , giving it to the Client(loop) class,I tried to stoop & close the loop, delete the task, use weird combination of ensure_future but nothing works.

I googled as much as I could, I read it is not good to nest infinite loops but i didn't find any other way.

My last hope is to use threading but even if I think it will work, it won't be an elegant solution, nor the right one.

I'm used to work with Node so please excuse me if i made a silly mistake, I thought that after 2 weeks i could make it but here I am.

I would really appreciate any help. I'm stuck. Thanks!

PS: The Client() class is very basic:

import asyncio
import logging
import sys

logging.basicConfig(
    level=logging.DEBUG,
    format='%(name)s > %(message)s',
    stream=sys.stderr
)

class Client(asyncio.Protocol):

    def __init__(self):
        self.log = logging.getLogger('client')
        self.address = None
        self.transport = None

    def connection_made(self, transport):
        self.transport = transport
        self.address = transport.get_extra_info('peername')
        self.log.debug('{}:{} connected'.format(*self.address))

    def data_received(self, data):
            self.log.debug('{}:{} just sent {!r}'.format(*self.address, data))

    def eof_received(self):
        self.log.debug('{}:{} sent EOF'.format(*self.address))

    def connection_lost(self, error=""):
        self.log.debug('{}:{} disconnected'.format(*self.address))
        self.transport.close()
1

1 Answers

1
votes

I keep getting 2 errors: "runtimeError: Event loop is running." or "asyncio > Task was destroyed but it is pending!"

As you discovered, asyncio event loops do not nest.

To remove the nesting, you should define send_to_peers as a coroutine using async def. Inside it loop.run_until_complete(coro) should be changed to await coro. Once send_to_peers is a coroutine, you can call it:

  • from blocking code such as infinite using loop.run_until_complete(send_to_peers(...))

  • from async code such as infinite_loop using await send_to_peers(...).

In case of infinite_loop, you can implement the timeout using asyncio.wait_for:

try:
    await asyncio.wait_for(send_to_peers(str(index)), 10)
except asyncio.TimeoutError:
    # ... timeout ...