pihole-FTL suddenly started to fail (on docker)

a perfectly fine pihole started barfing all of a sudden.

The FTL log has these lines

[2024-05-01 11:28:23.063 246M] delete_shm(): shm_unlink(FTL-settings) failed: No such file or directory
[2024-05-01 11:28:23.063 246M] delete_shm(): shm_unlink(FTL-dns-cache) failed: No such file or directory
[2024-05-01 11:28:23.063 246M] delete_shm(): shm_unlink(FTL-per-client-regex) failed: No such file or directory
[2024-05-01 11:28:23.063 246M] ########## FTL terminated after 17m 23s (code 1)! ##########
[2024-05-01 11:28:23.063 246M] FATAL ERROR in dnsmasq core: duplicate dhcp-host IP address 192.168.2.215 at line 68 of /etc/dnsmasq.d/20-bitwarp.conf

pihole-FTL

root@gbrix01:/etc/dnsmasq.d# pihole-FTL dnsmasq-test

dnsmasq: duplicate dhcp-host IP address 192.168.2.215 at line 68 of /etc/dnsmasq.d/20-bitwarp.conf.broke
Segmentation fault (core dumped)
root@gbrix01:/etc/dnsmasq.d#

All I did was unplug one machine from the lan being managed. Removing the offending dnsmasq,conf file makes pihole-FTL pass the dnsmasq-test, but then it stil complains about the shm stuff.

/dev/shm directory is EMPTY (on docker)

Here is the pihole -d https://tricorder.pi-hole.net/8aS28N18/

(I just rebuilt the docker image too)
I am stumped. Please help

Your debug log shows 2 different files with the exact same content, in /etc/dnsmasq.d/:

-rw-r--r-- 1 pihole pihole 3.0K May  1 10:58 /etc/dnsmasq.d/20-bitwarp.conf.broke

and

-rw-r--r-- 1 pihole pihole 3.0K Dec 13 17:55 /etc/dnsmasq.d/20-bitwarp.conf~

Try to move those files to a different path (or delete them) and restart the container.

Tried that. didn't work. pihole-FTL still crashes with shm errors

no. it did work. WTF?

Okay, is it that there was a editor backup file? Did that make pihole crash?

I am still very confused.

So it looks like I had duplicate entries in 04-pihole-static-dhcp.conf
that I made through the pihole webui which conflicted with some older settings I had when I imported the outer dnsmasq config to pihole (THAT dnsmasq service is disabled)

This was fine for months and months. Why did it start failing now?

So is it that now it's a fatal error to have duplicate dhcp-host

dhcp-host=MACDDR,HOSTNAME,IPADDR

config lines?

It has always been a fatal error. The crash comes from

and is obviously unintended

O I C.
thank you for the quick response.

1 Like