The only thing I’ve changed, is turning on the Camera Motion within Fully Kiosk Browser.
Now both of my ESPs keep disconnecting…
Following this thread WARNING Disconnected from API: Timeout while waiting for message response! · Issue #1194 · esphome/issues · GitHub
Feels like the similar issue - But these haven’t moved, the connection has been solid for months (one is next to the modem)
Anybody shed any light on how to debug further?
Well I would start with basics as always:
→ is it your only esphome device ? is it only one going offline ?
→ how are you running ha ? hardware platform too ?
→ your ESP and HA have both fixed IP (real fixed ip, not fixed ip in dhcp leases !)
→ what is your network configuration ? switches used ?
Vincèn
→ is it your only esphome device ? is it only one going offline ?
I have two, both go offline at the same time.
→ how are you running ha ? hardware platform too ?
HA is in Docker on a powerful NAS - no issues there
→ your ESP and HA have both fixed IP (real fixed ip, not fixed ip in dhcp leases !)
Yes. Both are on their own network (with other devices that are fine)
→ what is your network configuration ? switches used ?
Config is fine. Nothing at all has changed here since the devices have been installed.
Only change was me turning on the “Motion Detection” in Fully Kiosk Browser
The only thing I can see that is totally different now, is the Temp in the room…
But I have no idea why this would take out both my ESPs
which confirms quite sur a network/wifi issue !
In Docker with its own LAN IP for HA ? no nat in docker ? HA doesn’t log any disconnection of network randomly ?
I’m sorry but Config is fine is not an answer ! It is surely not fine considering you have some problems so if you don’t detail your network configuration it will make it impossible to debug…
Side question: have you some integrations in HA that use network ? and do they show some disconnection issues ?
I’ll do some further debugging on my network as your right, something is a miss there…
Although, I do like to know the ‘trigger’ - Why has it been stable for months? Why now?
Over the past couple of months I’ve overloaded my network and never had this type of issue.
Might be a Red Herring but the temps spike dramatically, I wonder if that has some affect on the network if it goes offline?
Done further investigating around the network and can’t spot anything other than my ISP is a little flaky - but not taking anything offline and the WiFi signal stays strong on both frequencies (although esp only runs 2.4)
I have a feeling that coincidently the dht22’s might be having an impact.
The ESP is still unstable, but recently the temp/humid has stopped populating when the other sensors remain online.
Mine “fixed” themselves when I deleted the browser mod add on.
I’ve been reading through this topic WARNING Disconnected from API: Timeout while waiting for message response! · Issue #1194 · esphome/issues · GitHub along with many others and it almost feels very ‘random’.
Loads are blaming network, but loads have also ruled that out.
I thought it was the DHT so I commented that out and nothing changed - still dropped out.
I removed Fully Kiosk (as that was the most recent install) … still nothing.
Changed my WiFi & modem … still nothing.
Unplugged several times … still nothing.
Then randomly the device connectivity become a little more stable.
The final opinion is that the ESP boards along with some DHT sensors just turn out to be bad hardware!
ESP been strongly connected almost 24hrs
However temp/humid have occasional drop outs
