I run two instances of Pi-Hole + unbound recursive DNS via Proxmox and everything works fine with auto boot enabled, even when there is a power outage, everything gets back online on its own.
All I do is wait 3 minutes or so, this happened last week in the middle of the night.
Keep in mind that his has been the case for years until today.
Today however, there was a planned power outage so I did shutdown everything, 3h later, turn everything back online, 3 minutes top and everything should be golden.
Well, Pi-holes were beyond broken like 100%, Pi-hole service ( systemctl status pihole-FTL ) was up but with errors:
I restarted those VMs like 5 times or so each and then it finally went back online at some point.
I never ever have to touch them when power cycling everything, I just noticed that there was 2 updates once everything was finally online, Core 6.2.2 and FTL version 6.2.3.
Was there just coincidence or what???
Restarting the service was killing the entire Pihole, no web server access doing that at all so why the reboots.
I cannot find one single root cause of why I had to restart those VMs 5 times to get Pi-hole service to behave.
Details:
Core v6.2.2 and FTL v6.3.2 have only a few minor adjustments. FTL has actually a small preventive memory handling fix (nothing was actually broken), but the issue could only affect a few devices (I only saw the issue in armv7 devices).
Very bad coincidence then coz that status=1/FAILURE just doesn’t happen.
I lost the count of how many times I have installed Pi-Hole + Unbound either manually or automated with Ansible, it is the most straightforward process ever.
Just a heads up I have had random startup failures in once and a great while with pi-hole and Proxmox. It seems related when you apply updates to the hosting Proxmox node. Don’t know which in particular but the last time I saw this was about 6 months ago.
The strange thing it I run two copies of Pi-hole one on each node the the Proxmox cluster. Pi-hole startup failed on one node but not the the other.
I see, I do not update Proxmox, I am running VE9.0.3 which I did a clean install, I am not a big fan of “upgrading” version coz sh1t always go sideways, that is why I use Ansible to automate everything and perform a clean install of my setup as needed.
Overall I have no issues at all, I have been running Pi-Hole + Unbound as recursive DNS via Proxmox for a few years now and it just works.
I know my drama was 100% Pi-Hole coz OPNSense was connected to the outside just fine but my network could not resolve any DNS with errors, Pi-Hole service was completely dead.
That is where I noticed the service pihole-FTL even tho it was up, it had status=1/FAILURE when running systemctl status pihole-FTL
To add to the drama, If I manually restarted the service to fix whatever was broken, Pi-Hole as a whole stopped working altogether and no more browser access even tho the service was up, so why I ended up restarting the VMs like 5 times and status=1/FAILURE eventually went away.
Each instances runs on a different Proxmox server, both Pi-Holes had the very same problem and were down.
I am 100% sure it was not Proxmox.
The only major difference is that I was running the updates PRIOR to 6.2.2/6.3.2 which I updated a few days back. That is the only change and yes, the unplanned power outage happened before that and everything was back online just fine.
Anyway, it is back now and I will try to improve my troubleshooting process for the next time a few years from now haha everything just works and I love it.