On 1/7/2020 12:47 PM, Eric Rescorla
wrote:
To address the more general problem I suggest:
“Should the trend away from using ISP managed
resolvers to using a small set of large public
resolvers continue, then an increased proportion
of the global DNS resolution traffic will to be
served by only a few entities. Some potential
impacts of centralisation within the Internet
Infrastructure are outlined in
[I-D.draft-arkko-arch-infrastructure-centralisation]
and include some privacy related considerations..
"
Yeah, my point is that I don't agree with this. Right now
there is a lot of ISP centralization and the move of some of
that traffic to public resolvers potentially decreases
centralization at the margin. I certainly don't agree with
citing draft-arkko, which is not at all a neutral or factual
source, so this is just a way of incorporating by reference
material which doesn't have consensus.
Centralization manifests itself in many ways. EKR is correct that
big ISP do get a huge part of the traffic -- last time I checked,
there was at least one ISP in China and another in India that
served pretty much as many customers as Google DNS.
I’m sure there are examples of certain regions where large ISPs control a large percentage of the traffic. However that does not change the argument that globally there are tens of thousands of ISPs, which users are free to choose between (I accept that in a few places, notably the US, there is sometimes no choice of which cable company you can use). Here in the UK there are over 100 ISPs I can choose from, almost all of whom run their own resolver. That is a vast difference from a small number of global centralised DNS providers.
The text proposed is pretty specific, i.e. "*if* the trend away from using ISP managed resolvers to using a small set of large public resolvers continue”. So if it turns out that there ends up being more public resolvers for end-users to choose from than there are ISPs running their own resolver, then great. However is anyone here really suggesting that there will be? The economics just don’t add up. There is also
centralization at work due to outsourcing of the DNS service by
ISP. This is a classic concentration path: an outsourcer that
serves many ISP will achieve economies of scale and may be able to
monetize the data flow, making outsourcing a viable option for the
ISP.
Outsourcing of the DNS service by an ISP is surely part of the “trend away from using ISP managed resolvers to a small set of large public resolvers”, and thus proves the point? Experience predicts that competition between these
outsourcers will exhibit "winners take all" dynamics leading to
concentration. As EKR says, the move to third party resolvers may
well counter concentration in the back end of the network. It
could also achieve the opposite, but there are risks on both sides
of this issue. I don't see how we can achieve consensus that one
side of the risk is more dangerous than the other.
I see both of these being the same. I don’t see how they counter each other. Both are examples of centralisation.
_______________________________________________ dns-privacy mailing list dns-privacy@xxxxxxxxhttps://www.ietf.org/mailman/listinfo/dns-privacy
|
--
last-call mailing list
last-call@xxxxxxxx
https://www.ietf.org/mailman/listinfo/last-call
