nextdns stops giving results for various minutes

[tomFlowee]

Context

• CLI Version: nextdns version 1.39.4
• Platform: ArchLinux, running the released binary from your github pages

<Describe the bug>

As a paying customer I have this client running on all machines in the house, sometimes at seemingly random times, dns resolves just stop working. A ping or a webrequest simply don't return for a long time.

Running 'nextdns log' gives me loads of lines like;

Query 127.0.0.1 UDP AAAA www.googleapis.com. (qry=36/res=12) cache fallback HTTP/2.0: doh resolve: context deadline exceeded

today this started directly after a line like this: (I'm in CET, btw)
dec 04 21:13:29 mylinux nextdns[69001]: Connected 93.189.61.195:443 (con=27ms tls=36ms, TCP, TLS13)

It happens more often and not always directly after such a connection. Let me know if you want more logs.

I'm surprised that the client doesn't detect this issue but does spew loads of those 'query' warnings. I typically need to do a nextdns restart to make things work again. Though waiting a couple of minutes reportedly works too.

Please consider making the nextdns client less prone to such stalls.

[lavendercolor]

this is very annoying. sometimes restart is not enough:

$ sudo nextdns restart
Error: systemctl restart nextdns.service: signal: killed:
$ sudo nextdns stop
$ sudo nextdns start

especially when changing networks

[lavendercolor]

If others are experiencing the same issue, I've discovered a workaround that significantly reduces its occurrence. I've noticed that when nextdns-cli stops working, a simple restart command instantly fixes it. I'm planning to write a shell script to automate this restart whenever DNS requests fail. I'll continue testing it for a bit longer and will report the results.

If you have two networks at home (for example, 2.4 GHz and 5 GHz), you can choose one for NextDNS. First, uninstall NextDNS, reinstall it, but answer no to the prompt that says:

Changes DNS settings of the host automatically when NextDNS is started.
If you say no here, you will have to manually configure DNS to 127.0.0.1.
Automatically setup local host DNS? [Y|n]: n

Then, in the settings of the chosen network, set the DNS as follows:

DNS ipv4 127.0.0.1
DNS ipv6 ::1

This will point to the nextdns-cli proxy.
For the other network, manually set the IPv4 and IPv6 DNS to those provided on https://my.nextdns.io/YOUR USER ID/setup (or any other backup resolver of your preference). Keep in mind that this backup connection will not be encrypted DNS.

Everything in /etc/systemd/resolved.conf should be commented out.