I'm currently an AdGuard Home user and am considering switching to pi-hole, IF there is some performance advantage in doing so. I understand functionally both have ~80% overlap, so performance would be a major differentiator for me
Does anyone have any metrics that compare both?
Pi-hole says it has optimized several things to make the DNS resolution "way faster", so wondering if anyone actually had the chance to test it somehow...
I use DNS-over-HTTPS on AdGuardHome, so I'd be looking at a setup where pi-hole is used with cloudflared to achieve the same DNS resolution privacy.
I doubt you will see any significant speed differences between the two, but the only way to know is to install Pi-hole and see how the speed compares.
Pi-hole is quite fast - adding Pi-hole into the DNS resolution path adds a few msecs delay perhaps, but it also caches so cached queries are served essentially instantly.
Actual DNS usage may vary wildly among different users.
Still, for typical home usage, differences in DNS performance would hardly be noticeable, so I wonder what kind of DNS traffic you are facing that would make performance a major differentiator.
Also, to be of practical use to you, you'd have to devise a set of DNS benchmarks that would cover your personal typical usage and at the same time, you'd have to be careful that you would test Pi-hole's/Adguard's performance, and not the performance of the whole DNS resolution chain up to your public resolvers.
Yeah, I'm not worried about pi-hole adding any substantial overhead, just as AGH's is quite negligible... The question is more "could pi-hole's optimized DNS forwarder be even faster by some meaningful percentage compared to AGH's intrinsic forwarder", everything else remaining equal... it doesnt have to be a precise scientific study, just anecdotal based on some approx numbers
my usage scenario is not that heavy. Mostly home use, accessing social media, some news, YT, gmail, streaming, etc... normally I perform <200k DNS requests / week give or take... also, as said in the other reply, it doesnt have to be a scientific study, but anything that shows any apples-to-apples comparison, perhaps even in a different usage scenario
In my specific case, Id be looking at pihole and AGH each running on a dedicated physical rpi 4GB RAM, raspiOS or dietpi, DoH on 1.1.1.1 for upstream resolver, same block lists. My usage scenario is such that the 80% of overlapping functionality between pi-hole and AGH fully cover my needs, so the only potential major difference would be if performance would be measurably better with pihole...
In my case I have 2 forwarders in my network. All devices use their default gateway as DNS, and the gateway uses AGH as its upstream. Overhead is not an observable issue