Broken Pipe errors / Bad write retry errors

[peterbuga]

Broken Pipe errors / Bad write retry errors

[peterbuga]

this is a weird low level python or ssl error, a bit out of the scope of this project. i'll try to fix it by catching it and internally retrying the request a few times, more than this i have no idea how to reproduce it :/

if possible try to look out for more variables that might trigger this error, like: are you switching on and off fast-ish the switches (one or more of them) ? does HA has a constant and good connection to the internet ?

[peterbuga]

just triggered something similar on my side too. it's possible that websocket token (the one that updates the switch states) might expire after 3-6 hours, don't know for sure yet.

it'll help if you could test it on your side too: restart HA to obtain new login token for sonoff component then try to flip any switch every 1 hour

[2016for]

Haven't restarted HA for a while (maybe a day or so), still working properly.

I think that error was triggered with fast switching of states / or by turning off a light group with multiple switches

[2016for]

Don't think I am getting these any more with the new commits.