Intermittent DNS Timeouts

Help
1

Expected Behaviour:

Uninterrupted DNS lookups with no network interruption

Actual Behaviour:

I am seeing consistent, intermittent DNS timeouts - when browsing on multiple devices on the network, wired and wireless, there is a 5-10 second timeout every now and then, across all devices at the same time. This is especially noticeable on google meet, where it drops me temporarily from the meeting 2-3 times in the span of 45 minutes, every single meeting.

I replaced my ASUS router with a RT-AX86U Pro running Merlin-WRT, replaced my older Raspberry Pi with a Pi 5, fresh install of everything - Pi-Hole running DHCP with DNS set to Cloudflare, OpenDNS, Google, I have tried them all. The issue persists. When I circumvent the Pi-Hole it does not seem to occur. I've tried whitelisting some google domains to alleviate the meet issues, but no luck.

Debug Token:

https://tricorder.pi-hole.net/sz4GHzDy/

Google Meet Error
Google Meet Error1110Ă—381 8.07 KB

11-28-24-ASUS_Wireless_Router_RT-AX86U_Pro_-DHCP_Server—_-kdnMI
11-28-24-ASUS_Wireless_Router_RT-AX86U_Pro_-DHCP_Server—_-kdnMI768×737 23.4 KB

11-28-24-ASUS_Wireless_Router_RT-AX86U_ProDNS_Director_—_Mo-22UIT
11-28-24-ASUS_Wireless_Router_RT-AX86U_ProDNS_Director_—_Mo-22UIT762×680 32.4 KB

11-28-24-Pi-hole_-raspberrypi—_Mozilla_Firefox-0jBDC
11-28-24-Pi-hole_-raspberrypi—_Mozilla_Firefox-0jBDC995×848 41.2 KB

11-28-24-Pi-hole_-raspberrypi—_Mozilla_Firefox-qxD3G
11-28-24-Pi-hole_-raspberrypi—_Mozilla_Firefox-qxD3G1009×405 16.9 KB

2

Your debug looks normal, apart from suggesting that you'd run a tracing tool (LTTng).

Why do you suspect DNS time outs?
Would your Pi-hole's Query Log suggest that?

Please share some log excerpts from /var/log/pihole/pihole.log* covering just the time frames where you did observe network connection losses.

3

A thought as I have a similar router and setup. Have you tried turning DNS Director off or global redirecting everything to user defined DNS 1 (the Pi) in the dropdown? It's currently set to redirect everything to the router and that may be keeping clients from getting to the Rasperry Pi.

1 Like
4

I disabled DNS Director as per bluzfanmr1

Here is an excerpt of the log when the issue occurred again just now, around 1:06-1:08

log.txt (56.8 KB)

5

The log excerpt doesn't indicate any occurrences of time-outs, but there is a ~3 second period where upstream 1.1.1.1 and 1.0.0.1 are unresponsive for some of your queries, e.g.:

13:09:06: query[A] notifications.bitwarden.com from 192.168.1.134
13:09:06: forwarded notifications.bitwarden.com to 1.0.0.1
13:09:06: query[A] notifications.bitwarden.com from 192.168.1.134
13:09:06: forwarded notifications.bitwarden.com to 1.1.1.1
13:09:06: forwarded notifications.bitwarden.com to 1.0.0.1
(…)
13:09:07: query[A] notifications.bitwarden.com from 192.168.1.134
13:09:07: forwarded notifications.bitwarden.com to 1.1.1.1
13:09:07: forwarded notifications.bitwarden.com to 1.0.0.1
(…)
13:09:09: query[A] notifications.bitwarden.com from 192.168.1.134
13:09:09: forwarded notifications.bitwarden.com to 1.1.1.1
13:09:09: forwarded notifications.bitwarden.com to 1.0.0.1
13:09:09: reply notifications.bitwarden.com is <CNAME>

This would suggest an issue with upstream DNS servers providing replies not as fast as usually.
Pi-hole wouldn't be involved, as it just waits for the replies to arrive.

You could try to switch to a different set of upstreams and see if the issue persists.