One specific domain (and all its subdomains) return no result

[nanos]

The issue I am facing:

I am unable to resolve one domain, and any of its sub domains (this is my personal domain, so I've removed it. It's a .xyz domain, but other .xyz domains resolve fine)

$ dig example.xyz

; <<>> DiG 9.10.3-P4-Raspbian <<>> example.xyz
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 4188
;; flags: qr rd ra; QUERY: 1, ANSWER: 0, AUTHORITY: 0, ADDITIONAL: 1

;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags:; udp: 4096
;; QUESTION SECTION:
;example.xyz. IN A

;; Query time: 1 msec
;; SERVER: 127.0.0.1#53(127.0.0.1)
;; WHEN: Mon Mar 08 08:50:48 GMT 2021
;; MSG SIZE rcvd: 41

I'm running cloudflared as my upstream name server, and if I dig the domain directly @ that name server, it responds fine:

$ dig example.xyz @127.0.0.1 -p 5053

; <<>> DiG 9.10.3-P4-Raspbian <<>> example.xyz @127.0.0.1 -p 5053
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 2456
;; flags: qr rd ra ad; QUERY: 1, ANSWER: 2, AUTHORITY: 0, ADDITIONAL: 1

;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags:; udp: 1232
;; QUESTION SECTION:
;example.xyz. IN A

;; ANSWER SECTION:
example.xyz. 300 IN A 104.21.83.80
example.xyz. 300 IN A 172.67.217.156

;; Query time: 40 msec
;; SERVER: 127.0.0.1#5053(127.0.0.1)
;; WHEN: Mon Mar 08 08:53:08 GMT 2021
;; MSG SIZE rcvd: 97

Anyone got any suggestion how to debug this?

I have tried restarting the dns server with

pihole restartdns

as well as restarting the whole pihole-FTL service:

systemctl restart pihole-FTL

to see if flushing the cache would help, but neither seems to help.

Here is the entry from the pihole interface for the query:

Details about my system:

It's a fairly standard system, except I'm running Apache as http server, and cloudflared as a DoH upstream.

What I have changed since installing Pi-hole:

I'm not aware of having changed anything for months.

[Coro]

So what is special about .xyz domains for your? Do you use Pi-hoe as DHCP server? Is xy maybe your local domain?

What happens the first time after a restart of pihole-FTL ? It cannot be cached when the cache is still empty. Also please check what /var/log/pihole.log says for the first (and then also for seconds etc.) queries so the developers can see what is going on.

[nanos]

@Coro

Do you use Pi-hoe as DHCP server? Is xy maybe your local domain?

Aha! That's indeed the issue. I have example.xyz as my "Pi-hole domain name" in the Advanced DHCP settings. If I change that it works.

Interestingly this used to work fine until quite recently. But, I guess I'll just change that name. I always found it quite nice that I could use rpi.example.xyz and have it resolve to my local IP while at home, and to my external IP while out and about. But not mission critical (and can be resolved with an entry in the hosts file anyway).

Thanks for the pointer!

[Coro]

This is a security improvement added in some recent version of Pi-hole (beginning 2021). The idea is to never forward queries to local domains upstream. Imagine you have

paul-smiths-iphone.example.xyz

When this domain isn't available locally, you probably don't want to get your full name sent to the upstream server.

[nanos]

Yeah, I certainly get why it's not actually a good idea to set it up the way I had .