Hey guys, I'm sorry.. Im desperate ..
I'm using 2 Piholes (v6 beta, up2date) + unbound for local DNS resolution + split horizon.
Locally, I have true DualStack available, so I'd like to use both IPv4 and IPv6 to get a better feel for it. Therefore, every device here currently has both an internal IPv4 and a v6 address... This will be important in a moment...
- Devices freshly installed, Pihole installed, unbound installed. Did the same on the other device (same unbound config on both) - also installed keepalived for HA and that's about it.
Configured Pi number 1: set up local DNS names, groups, etc., exported settings, and IMPORTED them onto Pi number 2. Tested everything, and it works as usual... But when Pi number 1 briefly goes offline and number 2 takes over, it suddenly responds with IPv6 addresses - I've only done the name resolution locally in IPv4, and some services still depend just on v4 here..
I did not change any hosts file. Everything is configured via Pi GUI / unbound config.
The settings are the same on both Piholes, including the OS' of the hosts. The only difference is that one is connected via LAN and the other via WLAN... Oh, and number 2 also has Docker running. Could it be because of that?! I've been looking for ages and just can't see what the issue is.
Maybe someone here has an idea of what I could look into. I'm really at a loss...
Or how I can set it so that in such cases it responds with IPv4. Because in general, I don't want to completely disable IPv6 in the config...
Thanks!
The .99 is master and respnds with ipv4, the .11 slave obviously and returns an ipv6. The .254 is keepalived's vIP for both of em.
Edit: While reading this text again I wonder... Could it be possible this happens because Pi number 2 is answering for a c name of its own dns name?
Only services which have a c-name of "madPi" are the ones becoming ""unresolvable"" when Pi 1 goes down.
Log Nr. 1: https://tricorder.pi-hole.net/ahGpc3aq/
Log Nr. 2: https://tricorder.pi-hole.net/codl7BUv/