FAQ¶
Note
Many questions asked in websockets’ issue tracker are actually
about asyncio. Python’s documentation about developing with
asyncio is a good complement.
Server side¶
Why does the server close the connection after processing one message?¶
Your connection handler exits after processing one message. Write a loop to process multiple messages.
For example, if your handler looks like this:
async def handler(websocket, path):
print(websocket.recv())
change it like this:
async def handler(websocket, path):
async for message in websocket:
print(message)
Don’t feel bad if this happens to you — it’s the most common question in websockets’ issue tracker :-)
Why can only one client connect at a time?¶
Your connection handler blocks the event loop. Look for blocking calls. Any call that may take some time must be asynchronous.
For example, if you have:
async def handler(websocket, path):
time.sleep(1)
change it to:
async def handler(websocket, path):
await asyncio.sleep(1)
This is part of learning asyncio. It isn’t specific to websockets.
See also Python’s documentation about running blocking code.
How can I pass additional arguments to the connection handler?¶
You can bind additional arguments to the connection handler with
functools.partial():
import asyncio
import functools
import websockets
async def handler(websocket, path, extra_argument):
...
bound_handler = functools.partial(handler, extra_argument='spam')
start_server = websockets.serve(bound_handler, ...)
Another way to achieve this result is to define the handler coroutine in
a scope where the extra_argument variable exists instead of injecting it
through an argument.
How do I get the IP address of the client connecting to my server?¶
It’s available in remote_address:
async def handler(websocket, path):
remote_ip = websocket.remote_address[0]
How do I set which IP addresses my server listens to?¶
Look at the host argument of create_server().
serve() accepts the same arguments as
create_server().
How do I close a connection properly?¶
websockets takes care of closing the connection when the handler exits.
How do I run a HTTP server and WebSocket server on the same port?¶
This isn’t supported.
Providing a HTTP server is out of scope for websockets. It only aims at providing a WebSocket server.
There’s limited support for returning HTTP responses with the
process_request hook.
If you need more, pick a HTTP server and run it separately.
Client side¶
How do I close a connection properly?¶
The easiest is to use connect() as a context manager:
async with connect(...) as websocket:
...
How do I reconnect automatically when the connection drops?¶
See issue 414.
How do I stop a client that is continuously processing messages?¶
You can close the connection.
Here’s an example that terminates cleanly when it receives SIGTERM on Unix:
#!/usr/bin/env python
import asyncio
import signal
import websockets
async def client():
uri = "ws://localhost:8765"
async with websockets.connect(uri) as websocket:
# Close the connection when receiving SIGTERM.
loop = asyncio.get_event_loop()
loop.add_signal_handler(
signal.SIGTERM, loop.create_task, websocket.close())
# Process messages received on the connection.
async for message in websocket:
...
asyncio.get_event_loop().run_until_complete(client())
How do I disable TLS/SSL certificate verification?¶
Look at the ssl argument of create_connection().
connect() accepts the same arguments as
create_connection().
Both sides¶
How do I do two things in parallel? How do I integrate with another coroutine?¶
You must start two tasks, which the event loop will run concurrently. You can
achieve this with asyncio.gather() or asyncio.wait().
This is also part of learning asyncio and not specific to websockets.
Keep track of the tasks and make sure they terminate or you cancel them when the connection terminates.
How do I create channels or topics?¶
websockets doesn’t have built-in publish / subscribe for these use cases.
Depending on the scale of your service, a simple in-memory implementation may do the job or you may need an external publish / subscribe component.
What does ConnectionClosedError: code = 1006 mean?¶
If you’re seeing this traceback in the logs of a server:
Error in connection handler
Traceback (most recent call last):
...
asyncio.streams.IncompleteReadError: 0 bytes read on a total of 2 expected bytes
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
...
websockets.exceptions.ConnectionClosedError: code = 1006 (connection closed abnormally [internal]), no reason
or if a client crashes with this traceback:
Traceback (most recent call last):
...
ConnectionResetError: [Errno 54] Connection reset by peer
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
...
websockets.exceptions.ConnectionClosedError: code = 1006 (connection closed abnormally [internal]), no reason
it means that the TCP connection was lost. As a consequence, the WebSocket connection was closed without receiving a close frame, which is abnormal.
You can catch and handle ConnectionClosed to prevent it
from being logged.
There are several reasons why long-lived connections may be lost:
End-user devices tend to lose network connectivity often and unpredictably because they can move out of wireless network coverage, get unplugged from a wired network, enter airplane mode, be put to sleep, etc.
HTTP load balancers or proxies that aren’t configured for long-lived connections may terminate connections after a short amount of time, usually 30 seconds.
If you’re facing a reproducible issue, enable debug logs to see when and how connections are closed.
How can I pass additional arguments to a custom protocol subclass?¶
You can bind additional arguments to the protocol factory with
functools.partial():
import asyncio
import functools
import websockets
class MyServerProtocol(websockets.WebSocketServerProtocol):
def __init__(self, extra_argument, *args, **kwargs):
super().__init__(*args, **kwargs)
# do something with extra_argument
create_protocol = functools.partial(MyServerProtocol, extra_argument='spam')
start_server = websockets.serve(..., create_protocol=create_protocol)
This example was for a server. The same pattern applies on a client.
Why do I get the error: module 'websockets' has no attribute '...'?¶
Often, this is because you created a script called websockets.py in your
current working directory. Then import websockets imports this module
instead of the websockets library.
Are there onopen, onmessage, onerror, and onclose callbacks?¶
No, there aren’t.
websockets provides high-level, coroutine-based APIs. Compared to callbacks, coroutines make it easier to manage control flow in concurrent code.
If you prefer callback-based APIs, you should use another library.
Can I use websockets synchronously, without async / await?¶
You can convert every asynchronous call to a synchronous call by wrapping it
in asyncio.get_event_loop().run_until_complete(...).
If this turns out to be impractical, you should use another library.
Miscellaneous¶
How do I set a timeout on recv()?¶
Use wait_for():
await asyncio.wait_for(websocket.recv(), timeout=10)
This technique works for most APIs, except for asynchronous context managers. See issue 574.
How do I keep idle connections open?¶
websockets sends pings at 20 seconds intervals to keep the connection open.
In closes the connection if it doesn’t get a pong within 20 seconds.
You can adjust this behavior with ping_interval and ping_timeout.
How do I respond to pings?¶
websockets takes care of responding to pings with pongs.
Is there a Python 2 version?¶
No, there isn’t.
websockets builds upon asyncio which requires Python 3.
