So I woke up this morning to an enormous amount of websockets errors in my log. They all look like this:
2020-04-20 11:59:52 ERROR (MainThread) [homeassistant.components.websocket_api.http.connection.2755931280] Client exceeded max pending messages [2]: 512
2020-04-20 11:59:52 ERROR (MainThread) [homeassistant.components.websocket_api.http.connection.2755931280] Client exceeded max pending messages [2]: 512
2020-04-20 11:59:52 ERROR (MainThread) [homeassistant.components.websocket_api.http.connection.2755931280] Client exceeded max pending messages [2]: 512
You can see with the timestamps how fast they were coming and this had been happening for hours, it started in the middle of the night. That’s literaly all the info I have on them, it didn’t say anything else in the log other then that there was an error.
My first thought was Node RED since I know that add-on opens a websockets connection to HA. But that appeared to be working fine. Everything still working and no unusual messages in the Node RED logs at all. So pretty confident that isn’t it.
When I opened up Portainer to see the logs there i found an enormous amount of messages that looked like this:
/usr/local/lib/python3.7/site-packages/zeroconf/__init__.py:1580: DeprecationWarning: ServiceInfo.address is deprecated, use addresses instead
warnings.warn("ServiceInfo.address is deprecated, use addresses instead", DeprecationWarning)
/usr/local/lib/python3.7/site-packages/zeroconf/__init__.py:1580: DeprecationWarning: ServiceInfo.address is deprecated, use addresses instead
warnings.warn("ServiceInfo.address is deprecated, use addresses instead", DeprecationWarning)
/usr/local/lib/python3.7/site-packages/zeroconf/__init__.py:1580: DeprecationWarning: ServiceInfo.address is deprecated, use addresses instead
warnings.warn("ServiceInfo.address is deprecated, use addresses instead", DeprecationWarning)
There’s no timestamps on these but you’ll have to trust me that they were firing right alongside the first set. Intermixed with those sporadically were some very odd logs that looked like these:
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0
100 155 100 155 0 0 25833 0 --:--:-- --:--:-- --:--:-- 31000
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0
100 67 100 67 0 0 13400 0 --:--:-- --:--:-- --:--:-- 13400
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0
100 351 100 351 0 0 50142 0 --:--:-- --:--:-- --:--:-- 50142
I didn’t really know what to make of these.
I tried to restart and turn on debug logging for homeassistant.components.websocket_api.http.connection
but that didn’t work. HA just immediately starts pumping out a massive stream of logs for all the websockets messages being flung around. And on top of that after I restarted the error seems to have disappeared entirely.
Does anyone know what this might be? Is this zeroconf or is zeroconf a red herring here? And does anyone know what I can do to debug this in the future if it happens again?
I didn’t report a bug on this since I don’t really have enough info to say much about what’s going on and it appears I can’t reproduce it. Really just hoping this jumps out at someone or someone has some future debugging strategies I can use next time.