Upgrade from 3.2 to 4.1 -> web requests from local network to 8080 are denied

I tried this while sshā€™ing on the synology:

image

and in the container:

openhabian-config is NOT available in a standard docker installation. Would be new to me.

1 Like

so I cannot use openhabian-config right?

that is correct.

yes, sorry 1031.
Any error messages when starting openhab.service from command line?

If you look into the log you added in your first post (openhab.txt), youā€˜ll see that docker did not find user openhab and was trying to add this user what somehow failed but somehow succeeded as there were 2 openhab users. That somehow messed up your container, I think.

Does the instance work without your specific configuration files ?

What does it show in the Docker Logs? I think thats ā€œProtokollā€. :slight_smile:

Hi Hannes,
Iā€™ve got exactly the same problem as you have, also running OH in the official docker container on Synology. Since I upgraded from OH 4.0.4 to 4.1.0 the Webserver vanished.

In my configuration I do not have a specific user Openhab on Syno.
But for the docker image Iā€™ve GROUP_ID=9001 and USER_ID=9001 as recommended in docker registry.
Further I didnā€™t get the warning in the update.log on the user.

Try the following inside the container:
curl -f http://localhost:8080
Does it give a response? I always get connection refused.

Further I could not see any listeners on expected ports using
netstat -nap (use apt update; apt install net-utils to enrich your container temporarily)

So I came to the conclusion that the Webserver does not come up for any reasons.

same here curl -f http://localhost:8080 returns connection refused

all docker logs are empty

regarding the question : does the instance run without m special thank instance?

do you mean to setup an additional openhab container with separate clean directories for openhab?

reagrding ā€œAny error messages when starting openhab.service from command line?ā€

How would start ā€œopenhab.serviceā€ from command line? I would assume to do that within the docker container (as openhab is not visible outside).

I tried this:

But I received an error message

sudo systemctl stop openhab.service
sudo systemctl start openhab.service

unfortunately there is no sudo or systemctl command within the docker container

sorry, I am not familiar too much with docker and tried to apply an approach of a typical installation. Forget what I have posted here. You need to wait for somebody more experienced. Sorry for confusing you

Yes I just managed to run openhab 4.1 in a seperate container ā€¦

Although at the end of of the upgrade you get JSON database updated successfully.
It could be that there are some problems with the jsondb as there are some error messages like

[main] ERROR org.openhab.core.storage.json.internal.JsonStorage - Couldn't deserialize value

Do you have a backup of your data before the update? If so Iā€˜d recommend to restore that data, make sure everything is working fine and do an update to 3.4 first as there have been some major changes coming along with 3.4. There is also a run once script executed because of those changes.
Then do another update to v4.1
I am sure your problem in the other thread will be solved, too, as, like Wolfgang pointed out, even if you manage to get access to MainUI, you will see a lot of things which you most probably need to delete and re-add them.

Yes there is an backup. But I now decided to give up on this one. I setup a new and fresh 4.1 system and spent the last hours to do a manual configuration of all my devices from scratch :frowning: It seems that I am already done 90% so there I donā€™t want to mess this new installation with backupfiles which (as you pointed out) would only work in steps to 3.4. and then up ā€¦ And then still you cannot be sure if you keep everything.
The last days were quite frustrating as I am almost 100% sure that by updting the firmware of my Tapos those devices cannot be handled right now by OH anymore. (I already posted this separate issue: P110 via TapoControl provides unknown API error (-1)

Then I made the mistake to think: Well itā€™s time to update anyways. And this crashed my whole system.

The learning from that is: Always try to keep your systems uptodate. From what I understand, that means, you need to constantly check whethter there are new openhab images releases, then kill your old container and install a new one and mount the old directories (conf/userdata/addons) right?
Is there a chance to get automatically notified whether there is a new image available?
And even more interesting: Is there a way to automate all this? Although I think that this is something which does not any longer belong to this thread (as itā€™ll be handled in general by docker? Right)

Thanks anyway for your help.

Meanwhile I fixed my OH installation by setting up a brand new OH 4.1 container. Further I copied the old directories /openhab/conf and /openhab/userdata over to the new volumes.
Luckily OH did start again and I didnā€™t loose my long history of data.
No single clue nor any traces in logs what went wrong.