#Sonoff Matter module troubleshooting
1 messages · Page 1 of 1 (latest)
thx
@twin pivot what type of IP are you pinging (first 4 digits)?
what's the best way to get debug log for the next time the problem occurs ?
You can also get the logs from last night, e.g. this command run from the web terminal gets the last 10000 lines of log:
ha host logs --identifier addon_core_matter_server -n 10000 > /config/matter.log
i can only ping it on ipv4 and public V6, not link local
Yeah hardware wise it looks really nice!
But we know from other brands that this does not say anything about software quality 🙈
I have no idea what brand you are talking about 😂
Interrupteur lumière chambre parents became unavailable
1:15:17 AM - 10 hours ago
nothing in this timeframe in the log
here is the matter.log file
Interrupteur lumière chambre parents became unavailable
8:14:20 AM - 4 hours ago
Interrupteur lumière chambre parents became unavailable
9:15:18 AM - 3 hours ago
3 times today
The first timestamp is UTC.
I think Sonoff’s dev is registering discord😆
Looking at your log its indeed doing resubs
There are Timeout waiting for mDNS resolution..
Haha thx, Sonoff and Shelly have one of the smallest offerings for relay
In the meanwhile I have this device added to my development instance, running in the debugger so I will keep an eye on it
sonoff dev
I reproduced the issue here already as it seems.
The device is indeed denying the case session now
Now I'll find out if we're using the correct ip
I think this says it all
feel free to check historical chat on this case, thx for your help in advance
and this happens only moments later
so when the subscription is failing, the link local addresses also are not pingable
ah, yeah that could be true, I think its by default disabled and I have some elevated rights or whatever
to ping link local ipv6 you need to do that with the interface included. our matter server ping does that
but it should be able to ping ipv4
well, matter uses the (link local) ipv6 to communicate
OK, power cycling the sonoff device fixed it so we're probably looking at a device issue here.
@daner could it be that the device is at times denying case sessions ? Or is communicating the wrong IP's on mdns ?
Or perhaps even the entire ipv6 stack not responding ?
It’s 7:25pm here, she might have clocked off 🫣
Yeah, that is imaginable
Well, at least we did find something. I'll poke around a bit more if I can narrow it down
great
i'am available for testing
why i can ping device from my computer
and not the addon :?
I grabbed a unit too, can give it a try once I’m home
ho now it's green (excepted for fe80) 😄
local link w/ matter over wifi is on the same network as you
so icmp pings work
Sorry for the late reply.
Can you reproduce this offline problem stably here? Can you tell us your reproduction path and method?
We need to verify and research this issue. Thank you very much for the information and help you provide. We will synchronize any news.
Well, I'm not 100% sure but it looks like the device is either;
- denying ipv6 communications at the Ip its communicating with mdns
- the device communicates the wrong (link local) ipv6 address
The IPv4 stack of the device sees to be functioning all times, at least a regular ping command.
@elfin leaf @viscid plank here's more log from my side https://filetransfer.io/data-package/DJMV2tW0#link
HAOS on VM (latest OS and Matter server) It is also attached to Google and Samsung fabric
hi, no disconnection since I rebooted the module 4 days ago
That is good to know. The module only sometimes gets into this state as concluded in my limited testing.
Can you ping and paste your device's IPs here? Also your device ID in the eWeLink app. Thx
But now its back available right ?
yep
en fully fonctionnal
and when it's unavailable
if i use the device
it's OK
it trigger my automation
the real problem for me is that when it came back alone, it trigger also my automation
this morning the mini4rm became unavailable on HA AND eweelink 🤔 nothing weird on my (w)lan and no problem with all my others iot
woo
device seems in trouble 😄
3 times unavailable since 00:00
and i can't ping it from my computer
looks not associated to my 2.4ghz wifi
had to reboot it
now it ping from HA, ipv4 and ipv6 (excepted link local)
So it essentially dropped off from the WiFi network then without recovery on it's own? 🤔 Probably worth reporting to Sonoff.
this is the first time i receive the eweelink offline notification
but anyway, every time it starts to fail in ha, rebooting the device help
i'll probably put an old zigbee relay in place 😦
@twin pivot any behavior of the indicator? When it went offline, what’s the blinking pattern?
i'll try to check but it's in my wall :x
I have a unit in debug mode for testing for a few days but not sure if I have the same issue cuz my network and matter setup is messy. It might go offline from matter fabrics but controllable from eWeLink app. A couple of toggle would get it back online to matter controllers, rebooting also helps
We are still trying to capture the log that could help to navigate the cause, sorr for the inconvenience