SIGTERM when using unbound forward-zone

[yacobucci]

Versions

• Pi-hole: v6.3
• Web: v6.4
• FTL: v6.4.1

Platform

• OS and version: Debian 13 trixie

• Platform: Raspberry Pi 4B

Expected behavior

Pihole uses unbound with local forward zone configuration to resolve LAN DHCP names.
DNS: 127.0.0.1#5335

forward unbound snippet:

private-domain: "lan"
domain-insecure: "lan"
forward-zone:
    name: "lan"
    forward-addr: 192.168.0.2

When resolving name.lan request is forwarded to upstream unbound and dhcp server and an answer is received and delivered to client.

Actual behavior / bug

Intermittent pihole crash with SIGTERM.

Steps to reproduce

Steps to reproduce the behavior:

• Run two pihole instances.
• Pihole A is DHCP server running unbound: 127.0.0.1#5335
• Pihole A configuration:
  • dns.domain.name: lan
  • dhcp.active: true
• Pihole B is running unbound, NO DHCP: 127.0.0.1#5335
  • dns.domain.name: ''
  • dhcp.active: false
  • unbound forward zone:

    • domain-insecure: "lan"
      forward-zone:
      name: "lan"
      forward-addr: 192.168.0.2
      

• Query Pihole B for internal DHCP identities.

Pihole periodically stops running:

  [✗] DNS service is NOT running

sudo tail -f /var/log/pihole/pihole.log
Dec 12 10:10:04 dnsmasq[649397]: query[A] beta.lan from 192.168.0.2
Dec 12 10:10:04 dnsmasq[649397]: cached-stale beta.lan is 192.168.0.200
Dec 12 10:10:04 dnsmasq[649397]: forwarded beta.lan to 127.0.0.1#5335
Dec 12 10:10:04 dnsmasq[649397]: reply beta.lan is 192.168.0.200
Dec 12 10:10:05 dnsmasq[649397]: exiting on receipt of SIGTERM

Debug Token

• URL:

Screenshots

If applicable, add screenshots to help explain your problem.

** Edits for formatting

Additional context

Add any other context about the problem here.

[DL6ER]

Sorry for the delay in replying.

Dec 12 10:10:05 dnsmasq[649397]: exiting on receipt of SIGTERM

is highly unlikely to be happening because of a crash.

I tried to use the information you provided to reproduce this in a container but am not seeing the dysfunctional state. Could you please provide also the output fro /var/log/pihole/FTL.log at the time this is happening? It should contain somewhat more details who sent such a signal or - if indeed triggered internally - why.

[yacobucci]

I'll need to reproduce myself, the FTL logs don't go back that far.

I was running nebula-sync when the errors occurred, I had it syncing everything but a handful of excludes at a 5m interval. However, since I turned nebula-sync off I haven't seen the reported behavior.