Please follow the below template, it will help us to help you!
Expected Behaviour:
I should be able to connect through the web client and SSH. Devices should be able to access the internet normally while having PiHole set as their DNS. The RPi/PiHole should be responding to all of these interactions.
Actual Behaviour:
The RPi cannot be accessed through the web client and SSH. The SSH times out, and Chrome simply shows that the site cannot be reached. During this time, any clients with the PiHole set as their
DNS address are unable to access the internet at all.
Manually unplugging and replugging in the RPi fixes the issue temporarily. After this reboot, looking at the graph of the queries during downtime show no queries, aside from one or two instances. This unresponsiveness has happened twice over a 24h period, once between 19:00 and 21:00, and from ??(early in the morning) to 8:00 this morning.
The RPi is connected to a surge protector of good quality, and the power supply does not use a separately bought USB cable. The PSU is rated for an output of 5.1V, 2.4A.
Are you running this on a Zero with a third party USB-ethernet cable? This has the appearance of a hardware problem.
One probably unrelated note from your debug log. You are forwarding PTR requests for local domains to your upstream DNS server, and they won't be able to resolve these. On the web admin GUI > settings > DNS, select "Never forward reverse lookups for private IP ranges"
-rw-r--r-- 1 pihole pihole 2885358 Sep 21 13:06 /var/log/pihole.log
-----head of pihole.log------
Sep 21 00:00:05 dnsmasq[533]: query[PTR] 107.0.168.192.in-addr.arpa from 127.0.0.1
Sep 21 00:00:06 dnsmasq[533]: forwarded 107.0.168.192.in-addr.arpa to 8.8.8.8
While looking at that description of the dongle, it notes that for compatibility with RPi's, the hub must be set to USB 1.0 speed (and it includes instructions). I have not done that yet, so it is possible that this may be the cause of the issue. I'll update in a couple hours if it crashes again.
As for the PTR request forwarding, that setting was already enabled, as seen in the image below:
11 hours later, my PiHole has become unresponsive again.
Pinging it results in timed out requests, while the web admin GUI tries to load for ~10 seconds and then says that is has timed out as well.
I swapped out the ethernet cable when I made the changes noted in my previous reply.
The PiHole did not respond after my first attempt to disconnect and reconnect the power.
I then tried that again, except I disconnected/reconnected the USB Ethernet dongle, both ends of the ethernet cable, and blew into the RPi ports (you never know).
Following this, the RPi responded and PiHole was reachable through both interfaces.
