Nextcloud AIO npm local IP changes during install

a remark about installing AiO behind Nginx Proxy Manager.
To link to the AiO apache port:11000
I use the local_IP adress of the Apache container (the public_IP won’t work in NPM)

At the first domain check, it works fine and AiO starts, and gives me the choice of the containers I want to install, then launches the installation.
the aio-apache container seems to be restarted, cause the local_IP changes, so I need to modify it again in NPM for to point to the apache container correctly.

Am I doing this wrong? Am I supposed to use the public IP in NPM?
Why isn’t it working? (My settings are those recommanded on github, nothing special)


From all-in-one/ at main · nextcloud/all-in-one · GitHub

1 Like

thanks I had read that. You see here in this post I’m mentioning the fact that I have to set the local_IP twice, because the apache container seems to be restarted after domain check.

Attaching containers to the host wouldn’t lower security?

ip a | grep “scope global” | head -1 | awk ‘{print $2}’ | sed ‘s|/.*||’

shows my public_IP
as I was mentioning, using my public_IP in NPM just never worked …

In that case you could either use network_mode: host for the npm container and then point at or join the npm container to the nextcloud-aio network and then point at nextcloud-aio-apache.

I’m will try later tonight to add network_mode: host
Could you tell me where do you put it in the yml?

What about security for this?

It is basically an entry in your docker-compose.yml. Like this:

Security-wise it is fine, just a different way of doing things.

1 Like

linking npm container to nextcloud-aio network is already what I’m doing.
I’ve found some solutions, cause AiO starts, as I’ve mentioned - but I’m just trying and testing to understand why it doesn’t work for me the same as it does for others …

as i’m still trying to have Collabora to identidy the public_IP automatically during install for WOPI requests, and trying to have Talk working… which doesn’t yet …

It is basically an entry in your docker-compose.yml. Like this:

oh ok, this is going in the npm container, not in aio … THANKS!

Well, I didn’t have to change the yml file … after backing up with AiO, restarted the container everything works fine … I don’t no how nor why.
I have a very long list of logs, that i’ll try to fix one by one … (about redis in the first place)