Hi all,

 

I'm about to update from 6.10.3 to 6.12.6 and was told to change the "Docker custom network type:" to ipvlan before updating. I tried it and i kind of broke a couple of things.

 

I have a setup of a few containers and swag that are set to br01 and have the extra parameter "--sysctl net.ipv6.conf.all.disable_ipv6=0 --sysctl net.ipv6.conf.eth0.use_tempaddr=2" as I had issues with accessing the container via ipv6 in my version and was the workaround. So every container has a local ip. All containers not related to swag are set to Bridge (i guess thats the standard?)

 

So after switching to ipvlan i had issues with my containers in br0:

- Nextcloud would not load

- photoview somehow loaded but not fully

- Shinobi was not working

 

That was the case when accessing them directly (via local ip) or externally via the swag reverse proxy.

So my questions: How would you change the setup so I can upgrade? I Don't need the br01 workaround if ipv6 access to bridged containers works just fine in the currrent version.

 

Thanks a lot

Sorry, but does anyone have an idea how to solve this the best way?

You can try updating first and keep using macvlan, you need to disable bridging for eth0, see the release notes, also make a backup of your flash drive first in case you need to go back.

ok thanks. But more generally. Is the setzup with the br0 and ipv6 workaround still necessary or could i just switch them all to Bridge (Thats the standard, right?). Only thing i would have to do is change the swag settings - correct?

If you want to keep using macvlan you should only need to disable bridging after booting with v6.12, nothing else, all another settings should keep.

@JorgeBThanks. And with turnin off you mean in each individual container settings, correct?

The effekt would be that the container afterwards does not have it's own ip adress anymore but can be reached with the unraids ip + specific port

 

proxy congis in swag look like this:

server {
    listen 443 ssl;
    listen [::]:443 ssl;

    server_name nextcloud.*;

    include /config/nginx/ssl.conf;

    client_max_body_size 0;

    location / {
        include /config/nginx/proxy.conf;
        resolver 127.0.0.11 valid=30s;
        set $upstream_app nextcloud;
        set $upstream_port 443;
        set $upstream_proto https;
        proxy_pass $upstream_proto://$upstream_app:$upstream_port;

        proxy_max_temp_file_size 2048m;
    }
}

so i assume they point to the container and not the ip anyway!? Only issue I see is the port which is 443 for swag and nextcloud, which doesn't matter now as they have different ips

Edited January 25, 20242 yr by dehein2

15 minutes ago, dehein2 said:

And with turnin off you mean in each individual container settings, correct?

No, for the NIC, see the v6.12.4 release notes for more info

11 hours ago, JorgeB said:

No, for the NIC, see the v6.12.4 release notes for more info

ah ok, so the vontainers are still bridged afterwards (have their own ip?)

They should.

Thanks.

But is the workaround still required in the current version or could i (after updating) switch all containers back to Bridge and and still access them via ipv6?

If I'm understanding correctly what you are asking it should not be.

ok, thanks again