Another network-device goes offline as soon openHAB is booting up

Setup, Configuration and Use Runtime
1
  • Platform information:
    • Synology Docker/Container
    • openHAB version: 4.2.0
  • Issue of the topic:

I have a very strange problem:
as soon as I start openHAB, another device goes offline (no ping anymore). It’s a Weblog monitoring device for a photovoltaic system.
I have already changed the IP of the PV-Weblog-device several times: I can rule out a duplicate IP in the network.
The subnet mask and gateway are also correct.

The weblog-device has been running for years, only when openHAB starts on Synology in Docker and I reboot OH the weblog device go offline. You can watch the ping stop replying during OH-reboot.
In night time (~2:00h am) the WebLog comes online again by itself.

For testing I installed a new docker-container with an totally empty openHAB - but the same thing happens.

What could be the reason?
This does not necessarily have to be realated to openHAB - but I am confused why this happens. Any idea is appreciated.

2

All I can suggest is that there is something going on on the network. When OH starts up it does send out some UPnP packets to try and discover other services/devices which it can integrate with. It uses the responses to populate the “suggested add-ons” list on the add-on store. Maybe the Weblog device is receiving that packet and treating it as some sort of command to go offline?

Beyond that all I can think of is there some issue in the router/gateway on your LAN. You;'ve already confirmed the obvious problem of multiple devices with the same IP/MAC isn’t happening. Maybe there will be something useful in the logs on the router?

3

Thanks Rich for your feedack.
The logs of my Router (Ubiquiti Unifi Dream Machine) only shows that the Weblog-device got disconnected. Nothing more.

It is 100% reproducable: I am rebooting (an empty) openHAB and WebLog stops answering to pings - and goes offline.

I am faced with a mystery - I have no more ideas what to check or how to solve this.

4

Since you are running with Docker, what if you don’t use host networking and only expose port 8080? That will prevent any network discovery packets from reaching your LAN as they will all be trapped in the Docker network.

That could at least rule out that as the source of the problem.

5

That’s a great idea! :ok_hand:
I made a copy of my openHAB-settings and installed another openHAB-container with bridged networking (instead of host networking).
I exposed all 4 ports: 5007, 8080, 8101, 8443 and changed the persistence from 127.0.0.1:8086 to the IP of the NAS.

openHAB with bridged network started without problems - and the Weblog-device still answers to ping!
Isn’t that interesting? :thinking:

So the root-cause seems to be related to host- versus bridged-networking of containers running on Synology NAS?
Ideas how to find out what’s going on, when running with host-network?

6

All I can guess is it has to do with the multicast packets that OH sends out when it starts out to discover devices on the network. When using bridged mode those packets never leave the docker network so are never seen by the Weblog.

There is nothing that OH sends out that should impact the Weblog so that device must be doing something weird.

7

What about if you check network traffic with wireshark

8

good idea Wolf, but…
Although I am an IT-guy, I must confess that I do not know how to use or understand Wireshark. I tried some time ago - and it was not helpful for me :roll_eyes: :innocent: