Open wosiu opened 1 month ago
Hey there @cmroche, mind taking a look at this issue as it has been labeled with an integration (gree
) you are listed as a code owner for? Thanks!
(message by CodeOwnersMention)
gree documentation gree source (message by IssueLinks)
I have the exact same issue. Gree AC devices were working fine for months. Yesterday, they all simultaneously dropped from the network, with behavior described exactly as the original post. I will run the discovery outside of HA with different firewall settings to confirm as well. In the meantime, I am using https://github.com/rapi3/HA-OS-Gree, which allows manual IP definitions, but even then, 2 of my 5 units partially work (switches available, but climate unavailable). Does anyone know a way to fix that issue?
So even though I was able to workaround adding them as described in my main post, all the devices become eventually unavailable. So I can only guess that the discovery process is run from time to time? And if devices do not respond during the discovery, they are dropped (?)
The problem
I'm running latest stable HA. When I'm trying to add my Gree AC via bultin Gree integration I'm getting:
FWIW The host running HA container is in the same subnet as Gree devices. Ping from the host to devices IP works fine.
This is all I get from
tail -f home-assistant.log | grep -i gree
:So I started to run manually the discovery task gree.py script as described at the end here from different hosts in my network. Depending on the setup I get in a console:
case 1: some firewall rules which seems to break the auto-discovery broadcast thing (basically blocking incoming traffic it seems):
BUT when I send the packet manually to my Gree AC device (its IP is
192.168.1.129
) using netcat, it works fine:case 2: Added entry to my firewall to whitelist all incoming traffic from one of the AC devices to whole machine (which I don't like):
That said, even with this hole in my firewall and the script returning the above entry, my home-assistant did not discover the device after the first run of adding the Gree integration in the UI. I had to run it for the second time for the device to be discovered.
And just to make it clear - I prefer I do not make this hole in my firewall. I've got plenty other wifi devices which works just fine without any extra holes in the firewall. I also tried to narrow the hole to dst port of my home assistant UI, but then the discovery breaks, because it seems different ports are used for autodiscovery - this is tcpdump on a host running HA during the autodiscovery:
Once the autodiscovery is completed, I was able to remove a hole from my firewall, and controling the AC from HA
seems to work fine. Not sure how it will behave in a long term[EDIT] was working for a bit and then devices become unavailable and HA.So I guess we really need this: https://github.com/home-assistant/core/issues/83048 to be added to the integration as an option. I think many people use static IPs via MAC binding for the IoT devices anyway. And I prefer static IP in the HA config too and avoid any extra broadcasting in my network if not neccessary.
What version of Home Assistant Core has the issue?
core-2024.10.1
What was the last working version of Home Assistant Core?
No response
What type of installation are you running?
Home Assistant Container
Integration causing the issue
Gree Climate
Link to integration documentation on our website
https://www.home-assistant.io/integrations/gree/
Diagnostics information
No response
Example YAML snippet
No response
Anything in the logs that might be useful for us?
No response
Additional information
No response