Do you know if this issue Is present in beta fw?
Yesterday i was locked out of my house
I have a finger print Reader that call home assistant script to open the Door… Luckily my wife was in the area and she had the Key in her bag.
It was my fault for forgetting my cell phone at home otherways i was able to open with Nuki app
No, I don’t see malformed packages. It just disconnects.
I‘m having the same issue with the latest firmware using Mosquitto MQTT Broker. Unfortunately this breaks all my automations with my KNX home. So any hint on the timeline of a fix would be highly appreciated.
Hi, this morning another disconnection.
The log of mosquitto mqtt broker reports:
1709881823: Client Nuki_XXXX closed its connection.
Remove and replug the battery pack make mqtt working for about 48hours
Hi, I have the same problem!
How to activate mosquitto log? I don’t have any logs under /var/log
I use HA 2024.3.0 and 6.4.0 broker
thanks
Luca
We’ve been able to observe the issue with one of our systems too and have a potential fix for it, which is included in todays 4.2.1 beta.
Please update your Smart Lock and test it and report back if the problem still occurs for you.
Installed beta fw. I Will report you in next days
Installed 4.2.1 Beta today and after the Update it first didnt connected to my Broker until I have removed the Battery. Since then no new malformed package errors in the Mosquitto Log:
2024-03-08T16:06:58: New client connected from 192.168.1.103:51951 as Nuki_39C748EC (p2, c1, k300, u'loxberry').
2024-03-08T16:06:59: Client Nuki_39C748EC disconnected due to malformed packet.
2024-03-08T16:07:51: New client connected from 192.168.1.103:49867 as Nuki_39C748EC (p2, c1, k300, u'loxberry').
2024-03-08T16:08:26: Client Nuki_39C748EC already connected, closing old connection.
2024-03-08T16:08:26: New client connected from 192.168.1.103:51601 as Nuki_39C748EC (p2, c1, k300, u'loxberry').
root@loxberry:/opt/loxberry
I don’t have new 4.2.1 firmware! The check report the 4.2.0 beta Is the last firmware. Any suggestioni?
Thanks
I forst had to update to 4.2.0 and then the 4.2.1 showed up.
I already have 4.2.0
Luca metti la 4.20 e dopo ti trova la 4.2.1, da me ha funzionato…
Paolo ho già la 4.2.0 ma non mi propone la 4.2.1.
Se vedi lo screen sopra ho la 4.2.0 e mi dice che è l’ultima disponibile.
We discovered another issue which can lead to the malformed packages problem. It only occurs if “Home Assistant auto discovery” is turned off. We therefore advise you to turn it on if you are using 4.2.1 beta. Issue will be fixed in the next beta.
Today my SL4P - running with Beta Firmware 4.2.1 - doesnt responded again to MQTT commands, although according to the Nuki App it was connected to my Broker:
I reconnected it manually (this time with “Home Assistant auto discovery” is turned on, even when I dont need this) and this time this immediately worked and MQTT commands worked again.
Update:
1 of the 2 locks with the new firmware (4.2.1) was unavailable starting 11-03-2024 01:35:59.
I needed to disconnect the power around 07:18:05 to get it working again.
Timestamps are from Home Assistant.
Best regards,
Jacob
In my case the reason for this where the malformed packages, which was fixed/workarounded with Firmware 4.2.1.
Now the Connection is way more stable and when the problem of not accepting MQTT commands is occuring again, an manuell reconnect or battery pull/push will resut in an immediately successfull reconnect.
Long Story short, for me (as an Thread opener) the core / “nightmare” problem is fixed, thanks!
@Juergen Hi, good morning. Bad news: after about 60 hours i got again disconnection.
Mosquitto logs:
1710148006: Client Nuki_xxxxx closed its connection
So in my case 4.2.1 does not solve the issue
Before to buy nuki 4 pro i had nuki 3 pro for 12 months and i never never never got any disconnection
@Juergen, MQTT Commands are not handled again from the Nuki even with 4.2.1 beta:
Again Confirmed: MQTT Reconnecting via App triggered (although Nuki App and my Tracking/Monitoring said MQTT is still connected [1] ) and the SL4P is immediately accepting MQTT commands again:
[1]
According to our logs, the message arrived at the Smart Lock and was denied because of
lockAction = 30
was sent. Your screenshot also says 3.000. Maybe there is a problem with the “.” and how the numbers are sent in general and/or interpreted by the Smart Lock. Can you please DM me the pcap from your screen shot?