does adaptive QOS or cake QOS work when using a DoT DNS service

[jata]

I have just switched DNS provider to Control D and setup on my router using the DoT configuration.

Everything seems to be working well but my internet seems a bit slower/sluggish since making the change. Also my buffer bloat scores/latency are not great (from A+ to B with latency +20-30ms

I don't really understand much about this stuff but I think DoT encrypts DNS traffic for privacy.

So does DoT interfere with Asus Adaptive QOS and/or cake QOS?

Is DoT really a good thing and I should use it or privacy overkill?

 

[Unisoft]

DoT is good and I don't believe it affects Adaptive QOS or CAKE, aprt from classification of traffic type possibly.

Buffer bloat can be made A++ by being more conservative under the values in Adaptive QOS for download speed maximum and upload maximum on a stable line. I don't use AUTOMATIC as never found it to work properly.

 

[bbunge]

I have just switched DNS provider to Control D and setup on my router using the DoT configuration.

Everything seems to be working well but my internet seems a bit slower/sluggish since making the change. Also my buffer bloat scores/latency are not great (from A+ to B with latency +20-30ms

I don't really understand much about this stuff but I think DoT encrypts DNS traffic for privacy.

So does DoT interfere with Asus Adaptive QOS and/or cake QOS?

Is DoT really a good thing and I should use it or privacy overkill?

Depending upon where in the world you are located the Control D upstream resolvers, also know as DNS servers, may not be close and thus will take a bit more time to get a response. Control D, like other DNS providers, uses the anycast system which uses a single IP address which is routed to one of several servers. Your ISP usually has this coded into its system and can send the request to other than the closest server. For example, the closest Quad9 servers to me ar 90 miles away but my ISP sends my request to servers 1100 miles away.

 

[Crimliar]

I have CAKE QOS set up and ready to work when needed (most of the time it's not needed and so not enabled). I also use DOT to connect to Clourflare servers 1.0.0.2 & 1.1.1.2
Those servers 135kM away sit in the same data centre as my VDSL gateway, and they're fractionally (less than 1ms difference) faster than those provided by my ISP

 

[Morris]

Unrelated technologies. No effect

 

[jata]

Big thanks for all the info/input on this topic.

So I think the summary is that buffer bloat and DNS resolution are totally unrelated and therefore me thinking it's DNS provider is noise. Can check by switching to google DNS and re-testing.

Not happy with my current buffer bloat scores but there are a number of other factors that might be impacting them so I will check them all one by one.

1. Merlin to stock firmware
2. Cake QOS (merlin) to Adaptive QOS (stock)
3. NextDNS CLI install to Control D using DoT

Note that my internet connection is 100/40 (NBN fibre to the node) and is very stable. Bandwidth set to 94/34.

For context, I have gone from Buffer bloat A+ with around +2ms latency on DL and UL to B with around +20ms latency on DL and UL.

I have a feeling this is related to Cake vs Adaptive but testing this will require me going back to Merlin (I prefer merlin over stock by a long way) but I have issues with merlin and my Aimesh setup. Stuck between a rock and a hard place