Hello everyone,
since 2 days ago, when I updated my hassos and hassio to 0.104.3 and hassos 3.8, I can’t seem to get any of the nodes to connect. I read in the release notes that there are some breaking changes with the MQTT but I don’t know what should I do now in order for the nodes to connect.
If you had the same problem or still have it please let me know what should I do to fix it.
Thank you in advance.
hi, i have the same. they work over mqtt but are offline in the esphome interface
did you already found something?
try
“status_use_ping”: true
in addon config
in which addon? and where?
I will try that
@SagiRo in your sidebar you go to hass.io, there you find yout addons like esphome.
If you open esphome and scroll down you get configuration options and you can change configuration to:
{
"status_use_ping": true
}
ok tried it just now,
and nothing… still all the nodes appear to be offline… 
I’m pretty sure it’s something to do with the MQTT… but I don’t know what should I change in there.
also I checked right now and the MQTT broker is working like it should, sent a message and saw it on another MQTT software. So the broker is working… maybe it’s something that was changed in the MQTT and wasn’t changed in ESPhome? Do I need to reflash all the nodes? if so then it’s not going to be good or rilable if anytime there is an update I’ll need to reflash my nodes.
Just wondering why you are using MQTT with ESPHome instead of the API? Is there any benefit?
hmmmm I’m not sure how to do it like that. if you could help me that wouls be great.
For me it is the other way around, I was unaware of the MQTT option… details here: https://esphome.io/components/mqtt.html ( so drop MQTT: and add API: )
But yeah seems you be needing to reflash the nodes and reconfigure Home Assistant 
well that’s a bummer… I really hope it won’t happen again… I’m trying with 1 node. if it’ll work I’ll do it with the others.
Note that the API: entry best has a sub PASSWORD: entry, that password is what you need to enter when pairing the node in Home assistant - Intergrations
Currently I’m not on site to check the nodes… so I’ll need to wait for after the weekend. this is why it’s a bummer… we’ll see after the weekend.
by the way I always had the “api:” in there, didn’t removed it… but also had MQTT.
We’ll see in few days I guess.
I agree, just reflash with ESPHome firmware and bypass the need for MQTT completely.
Same issue - i am not using MQTT though.
I am using the ESPHome API. In the ESPHome screen, most of my devices are showing as offline, both D1 mini devices and flashed plugs. The strange thing is that they are still reporting their status and can be controlled by home assistant in the lovelace cards and if i access their web gui they are fine. It just seems that the ESPHome screen in home assistant cant find them so i can’t OTA flash a new config.
Is it possible they were assigned new IP addresses by your router that doesn’t match up to what ESPHome is expecting?
status_use_ping
By default the dashboard uses mDNS to check if nodes are online. This does
not work across subnets unless your router supports mDNS forwarding or avahi.
Setting this to true will make ESPHome use ICMP ping requests to get the node status. Use this if all nodes always have offline status even when they’re connected.
Thanks @SimonPth, I suspect that status_use_ping may cure some problems, it is not the issue here though. Some devices are working great, others disapear. I have some devices with a fixed IP specified, some without - they all have a fixed IP assigned at the router. I suspect it may be an issue with Unifi USG router. They show up on my wireless network and if I select to ‘reconnect’ in the unifi gui then they get booted of the wireless, then reconnect and they show up again in ESPHome. As i mentioned, it isnt an issue as they still function regardless but it does make it difficult to upload a new config.
ok so apperently my router got stuck (it’s an old one) when I resetted it everything started working even tho it showed as offline on ESPhome… I reflashed all my nodes and now they run on API and shows as online, but now it created another entity for each of the nodes… it’s a mess… I really hope it won’t break again in the future.
anyways, thank you to all those that helped.
This fixed my case.