Pihole stopped working suddenly

Expected Behaviour:

Pihole is running in a docker container on a RPi4. Worked fine for weeks, no issues.

Actual Behaviour:

Suddenly stopped working. Getting a 502 Bad Gateway when I try to connect to it. Debug shows FTL and lighttpd daemon inactive.

Debug Token:

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

I wish I had more to go on. I pulled the latest image, and checked all the configurations, but nothing changed that I know of to trigger the break. It was working, then it stopped. I can't connect to the web interface over LAN or WAN.

Apologies for picking this up late.

A 502 HTTP status code would hint at an issue with an HTTP server, not necessarily with Pi-hole's, and specifically not with Pi-hole's DNS server.

Did you verify whether Pi-hole would still respond to DNS requests, e.g. by executing nslookup pi.hole from a client?

Does your issue persist?

If so, could you please provdie a fresh debug token?

The server has been working fine, to include pihole until everything suddenly stopped working. My other sites are still being hosted just fine.

My docker-compose block, which has not changed

 pihole:
    container_name: pihole
    image: pihole/pihole:latest
    ports:
      - "53:53/tcp"
      - "53:53/udp"
#      - "67:67/udp"
      - "8089:80/tcp"
    environment:
      TZ: 'America/Chicago'
#     ServerIP: 192.168.86.10
      WEBPASSWORD: 'xxxxxxx'
      #Volumes store your data between container upgrades
    volumes:
      - '/opt/appdata/pihole:/etc/pihole/'
      - '/opt/appdata/pihole/dnsmasq.d/:/etc/dnsmasq.d/'
    # Recommended but not required (DHCP needs NET_ADMIN)
    #   https://github.com/pi-hole/docker-pi-hole#note-on-capabilities
    cap_add:
      - NET_ADMIN
    restart: unless-stopped

And here is the new token: https://tricorder.pi-hole.net/VGFNO89n/

Your debug log looks normal.

I notice that your Pi-hole is receiving DNS requests from 192.168.86.1.
If that is your router, then you may have configured your router to use Pi-hole as ist upstream. While a valid configuration, you may have to lift Pi-hole's DNS query rate limiting, as your router would aggregate your entire network's DNS traffic and send it to Pi-hole..

Ok, so its normal for lighttdp and pi-holeFTL daemons to be inactive?

Yes, that's my router. It's a Google Wifi, which I haven't been thrilled with. I'm not sure there's any other way to configure a custom DNS.

And strangely enough, it's working again, even though I've changed nothing. Thanks for the help!

Our debug log is a bit dodgy in the Docker environment.

The rate limit may only be triggered occassionally, during high levels of DNS activity. Also, a rate limit expire after a while, once the level drops to lower levels again.

(In virtualised environments like Docker, the information from the debug log is less conclusive than for bare metal. We can still infer that DNS is working by assessing other relevant parts of the log.)

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.