PONG response has not been received

[Champgy]

Hi,
I sent a lot of events (5000) at once from the client to the server via socketio, and then a socketio reconnection occurred. This seems to be caused by a ping timeout.

Q1: It looks like all current events and pings are in the same queue, which may prevent ping-pong from being sent in a timely manner (I'm not sure if this understanding is correct)?

Q2: I think pings might only be generated when there is no data, because if there is data being generated between the client and the server, it inherently indicates that their connection is normal, right?

Sending packet MESSAGE data 2/test,["test",{...}]
... skip
Sending packet MESSAGE data 2/test,["test",{...}]
Sending packet PING data None
Received packet NOOP data None
WebSocket connection was closed, aborting
Waiting for write loop task to end
Exiting write loop task
Waiting for ping loop task to end
PONG response has not been received, aborting
Exiting ping task
Engine.IO connection dropped
Connection failed, new attempt in 1.28 seconds
Exiting read loop task
Attempting polling connection to http://127.0.0.1:3000/socket.io/?transport=polling&EIO=3
Polling connection accepted with {'pingInterval': 40000, 'pingTimeout': 70000, 'sid': 'e665329c77f74801840b2a802179b251', 'upgrades': ['websocket']}
Engine.IO connection established
Received packet MESSAGE data 0
Namespace / is connected
Attempting WebSocket upgrade to ws://127.0.0.1:3000/socket.io/?transport=websocket&EIO=3
Sending packet MESSAGE data 0/test

[miguelgrinberg]

Pings and events use the same transport, that is correct.

The server emits pings at the configured ping interval. This is unrelated to what other traffic exists between the client and the server. The issue might be that your server gets too busy due to all the incoming traffic, causing the ping task to not receive CPU time. This is especially more likely if your server is asynchronous, but can also affect a threaded server given a sufficiently big load.