Re: [Doh] DNS over HTTP/3?

["Ralf Weber" <[e-mail]>]

Moin!

On 19 Nov 2018, at 11:09, bert hubert wrote:
> I've observed that the thousands of users of doh.powerdns.org (I also 
> do not
> know how this happened) take around 22 packets per DNS query/response.
Wow. Good for manufactures of 10GB ethernet cards ;-).

> Larger scale adoption of TLSv1.3 might improve this somewhat, but it 
> is a
> big number.
>
> I've also personally observed that a "slightly suboptimal" network
> absolutely kills browsing performance in Firefox Nightly using DoH.  A 
> naive
> calculation shows that 0.5% packet loss turns into a 5% failure rate 
> per DoH
> query, which then can cause Head of Line blocking for further queries, 
> which
> cascades into "blank pages" getting rendered.
Hmm doh.powerdns.org resolves to one IP address while 
mozilla.cloudflare-dns.com resolves to two IP addresses. Not sure if 
that would make a difference as I don’t know what the Mozilla stub 
does. I also didn’t find anything in the RFC wrt to the 
failure/failover behaviour other than DoH follows standard HTTP 
mechanisms, but I have no idea what that means for unresponsive servers, 
and more important when a server is declared unresponsive.

[…]
> And, perhaps somewhat more provocatively, should we maybe not start 
> pushing
> DoH/2 if it leaves people with a sub-standard experience, causing them 
> to
> disable DoH?  DoH/3 might be somewhat of a wait but it might prevent 
> that
> sour taste from developing.
Maybe there is a need for dohop (DNS over HTTP Operations) working 
group. I think there are still a lot of unknowns in the operations of 
DoH clients and servers, but the same is true for regular DNS and I’ve 
been working on that for decades now. Maybe a draft or some measurements 
would also help to get a discussion started.

So long
-Ralf
—--
Ralf Weber