pihole doesn't auto-start on lxc reboot (proxmox debian ct)

The issue I am facing:

A freshly installed pihole on a debian proxmox container doesn't auto-start on container boot, and takes several minutes to start via systemctl start pihole-FTL.

There is nothing obvious in the logs like errors, and CPU/RAM usage appears low.

Unfortunately I have power cuts that outlast my UPS, so I need a setup that auto-starts on reboot.

Can anyone help debug this? Thanks!

Details about my system:

https://tricorder.pi-hole.net/t2F8xvxf/

container-configuration:

arch: amd64
cores: 1
features: nesting=1
hostname: nuc-pihole
memory: 512
net0: name=eth0,bridge=vmbr0,firewall=1,hwaddr=BC:<redacted>:39,ip=dhcp,ip6=dhcp,type=veth
onboot: 1
ostype: debian
rootfs: local-lvm:vm-110-disk-0,size=4G
swap: 512
unprivileged: 1
Linux nuc-pihole 6.8.12-4-pve #1 SMP PREEMPT_DYNAMIC PMX 6.8.12-4 (2024-11-06T15:04Z) x86_64 GNU/Linux
PRETTY_NAME="Debian GNU/Linux 12 (bookworm)"

What I have changed since installing Pi-hole:

expandHosts, some custom dns entries.

Might be a silly question but have you checked the startup on boot option in the ct itself.

And check if pihole services are enabled
systemctl is-enabled pihole-FTL

If not
systemctl enable pihole-FTL

1 Like

Thanks for checking - they are both enabled.

There is a warning about NTP - I checked the pre-requisites and installation documentation and didn't see anything about installing an NTP server, when I tried to load the host localtime on LLM-advice I had various complaints so gave up on that.

2025-04-08 09:43:36.596 UTC [1041/T1042] ERROR: Error NTP client: Failed to adjust time during NTP sync: Insufficient permissions
2025-04-08 09:43:36.596 UTC [1041/T1042] INFO: Local time is too inaccurate, retrying in 3596 seconds before launching NTP server

I can't see an NTP server running?

Pi-hole v6 embeds both a simple NTP server as well as a simple NTP client.

You may disable the NTP server by running:

sudo pihole-FTL --config ntp.ipv4.active false
sudo pihole-FTL --config ntp.ipv6.active false

And to disable the NTP client:

sudo pihole-FTL --config ntp.sync.active false

Ok so after another outtage, some frustrating chat with an LLM, old-fashioned Google yielded the issue and answer.

Basically, I noticed that the networking service was hanging, and it turns out it was dhclient -6 which didn't complete. This then prevented the pihole-FTL service from loading. I was fixing it accidentally by killing various processes.

In proxmox, there is a setting to use SLAAC instead of DHCP, as per this post:

This seems to fix the issue (although it persists in the linked post above).

If this doesn't work, there are two more options:

  1. Disable ipv6 on the pihole, which makes sense as I'm only specifying the ipv4 address as the DNS server anyway
  2. Use a static IP address, which again would make sense, although I'm happy with the DHCP reservation on the router.

Hope this helps someone else.