Pi-hole admin page not loading, DNS service not running

[jkaplon]

Expected Behaviour:

• Pi-hole admin page responding at http://192.168.0.118/admin/index.php?login
• pihole status returns happy messages

Actual Behaviour:

• http://192.168.0.118/admin/index.php?login loading with spinners and "Lost connection to API" messages
• pihole status returns DNS service is NOT running
• pihole restartdns succeeds, but subsequent pihole status sill returns DNS error

Debug Token:

b8v07pvvwa

Additional specs and info:

Raspberry Pi 2, running Ubuntu Mate 16.04.5.
I'm not quite sure when pi-hole started having problems, it's run so well for so long, I stopped checking on it regularly.
Last pi-hole upgrade may have started the problem, but I'm not sure.

[Mcat12]

What's the output of this command? sudo service pihole-FTL status -a

[jkaplon]

output of sudo service pihole-FTL status -a:

pihole-FTL.service - LSB: pihole-FTL daemon
Loaded: loaded (/etc/init.d/pihole-FTL; bad; vendor preset: enabled)
Active: active (exited) since Sat 2019-02-02 15:43:14 EST; 1h 20min ago                                                Docs: man:systemd-sysv-generator(8)
Process: 509 ExecStart=/etc/init.d/pihole-FTL start (code=exited, status=0/SUCCESS)

Feb 02 15:43:04 rpi2 systemd[1]: Starting LSB: pihole-FTL daemon...
Feb 02 15:43:04 rpi2 pihole-FTL[509]: Not running
Feb 02 15:43:13 rpi2 su[989]: Successful su for pihole by root
Feb 02 15:43:13 rpi2 su[989]: + ??? root:pihole
Feb 02 15:43:13 rpi2 su[989]: pam_unix(su:session): session opened for user pihole by (uid=0)
Feb 02 15:43:14 rpi2 pihole-FTL[509]: dnsmasq: failed to create listening socket for port 53: Address already in use Feb 02 15:43:14 rpi2 systemd[1]: Started LSB: pihole-FTL daemon.

Just in case it helps, it looks to me like dnsmasq stole port 53 from itself (but I'm nearing the limits of my hobbyist-linux-admin skills...and pasting from powershell-ssh is no fun ;).
sudo lsof -i -P -n | grep LISTEN returns:

sshd       776     root    3u  IPv4  14829      0t0  TCP *:22 (LISTEN)
sshd       776     root    4u  IPv6  14831      0t0  TCP *:22 (LISTEN)
lighttpd   810 www-data    5u  IPv4  13378      0t0  TCP *:80 (LISTEN)
lighttpd   810 www-data    6u  IPv6  13379      0t0  TCP *:80 (LISTEN)
dnsmasq    840   nobody    5u  IPv4  11825      0t0  TCP 127.0.1.1:53 (LISTEN)                                       cupsd     3242     root   10u  IPv6  17053      0t0  TCP [::1]:631 (LISTEN)
cupsd     3242     root   11u  IPv4  17054      0t0  TCP 127.0.0.1:631 (LISTEN)

[Mcat12]

You'll have to stop dnsmasq or whatever is using port 53 before you can use FTL. FTL embeds dnsmasq, so you don't need the standalone dnsmasq running.

[jkaplon]

@Mcat12, thanks for the help.

I'm not sure how I missed it previously, but this appeared in tonight's frantic searching:

Algo/Ubuntu starting dnsmasq

You can, but I think it's a dependency of network manager and it will be installed again on a system (auto) update. I posted this info here: Due to the nature of your environment and the presence of Network Manager, after FTLDNS is installed, you need to make sure that Network Manager does NOT enable dnsmasq (at reboot) before FTLDNS. In order to achieve that, you need to edit /etc/NetworkManager/NetworkManager.conf And comment out or delete the line that has dns=dnsmasq. Then you need …

After trying a bunch of stuff, it was that edit to /etc/NetworkManager/NetworkManager.conf that got everything running again.

On my initial install of pi-hole, I had to remove /etc/dnsmasq.d/network-manager...maybe some recent Ubuntu Mate package updates decided to change up the Network Manager settings again.
I'll certainly think twice before my next apt-get upgrade...