Discovery of
Encrypted DNS Resolvers: Deployment ConsiderationsOrangeRennes35000Francemohamed.boucadair@orange.comNokiaIndiakondtir@gmail.comCitrix Systems, Inc.USAdwing-ietf@fuggles.comOpen-XchangeUKneil.cook@noware.co.ukMicrosoftUSAtojens@microsoft.comThe document discusses some deployment considerations of the various
options to discover encrypted DNS resolvers (e.g., DNS-over-HTTPS,
DNS-over-TLS, or DNS-over-QUIC). In particular, the document describes
how Discovery of Network-designated Resolvers (DNR) and Discovery of
Designated Resolvers (DDR) can be used in typical deployment
contexts.This document does not intend to provide deployment recommendations,
but is meant to exemplify how operators can enable the encrypted DNS
discovery mechanisms. In addition, the document illustrates the
feasibility of hosting encrypted DNS forwarders in Customer Premises
Equipment (CPEs).Discovery of Network-designated Resolvers (DNR) specifies how a local encrypted DNS
resolver can be discovered by connected hosts by means of DHCP , DHCPv6 , and
IPv6 Router Advertisement (RA) options.
These options are designed to convey the following information: the DNS
Authentication Domain Name (ADN), a list of IP addresses, and a set of
service parameters. The ADN is used as a reference identifier for
authentication purposes, while the list of IP addresses designate where
to locate the resolver without relying upon an external resolver. The
service parameters provide additional information to characterize a DNS
resolver (e.g., supported encrypted DNS, customized DNS port number, or
URI Template for DNS-over-HTTPS (DoH)). Such an information is used by a
DNS client for DNS resolver selection and session establishment.This document discusses some considerations to make use of the
discovery of encrypted DNS resolvers such as DoH , DNS-over-TLS (DoT) , or DNS-over-QUIC (DoQ) in local networks.Sample target deployment scenarios are discussed in ; both managed and unmanaged Customer Premises
Equipment (CPEs) are covered. It is out of the scope of this document to
provide an exhaustive inventory of deployments where Encrypted DNS
options can be used.Considerations related to hosting a DNS forwarder in a local network
are described in . In contexts where
CPEs can't be upgraded to support DNR, Discovery of Designated Resolvers
(DDR) can be used. See Sections
and for more details.Techniques, such as the one defined in , can be enabled
together with to feed the
Encrypted DNS options. However, the document does not make any
assumption about the internal behavior at the network side to feed the
Encrypted DNS options that are supplied to requesting hosts; only the
external observed behavior is detailed in the following sections.Policies to guide the activation and selection of encrypted DNS can
be configured by users using implementation specific means (e.g., CPE
management interface).This document is not setting deployment recommendations or claiming
to share best current practices. It is purposely scoped to exemplify how
encrypted DNS discovery mechanisms can be enabled in typical networks. A
set of considerations are specifically drawn to assist Internet Service
Providers (ISPs), CPE vendors, and home network security service
providers.Concretely, generalizing the use of encrypted DNS while preserving
services that are offered to users (especially, those services that
require a local DNS forwarder) depend on many actors:ISPs
who need to investigate and elaborate plans about how their managed
CPEs will be upgraded to support encrypted DNS forwarders and
whether home network security mechanisms will still be required to
enforce per-device policies. Some ISPs may
also need to investigate plans to offer encrypted DNS services even
for CPE models whose firmware cannot be updated. For example, ISPs
may consider updating the CPE configuration to point to the ISP's
Do53 resolver for DDR to work.ISPs will
also need to assess the impacts of bypassing local DNS forwarders on
their DNS infrastructure and the services they are offering to their
subscribers.to help them assess the feasibly of CPEs
to host an encrypted DNS forwarder. To that aim, the document
sketches some realization approaches. For example, CPE vendors may
learn from the effort that was conducted by some DNS providers to
optimize the encrypted DNS forwarder to run in a container in home
routers and how this may be integrated with home network security
service agents.may want to avoid depending on the capabilities
of their ISP-supplied CPE. They may consider deploying an unmanaged
CPE that uses DNR to advertise the local encrypted DNS information
to connected devices. discusses
how DNR can be used in such contexts.which need to support the
Discovery of Designated Resolvers (DDR) or the Discovery of
Network-designated Resolvers (DNR) procedures.This document is meant to assist future deployments and (hopefully)
accelerate the network deployment of encrypted DNS servers.This document makes use of the terms defined in .The following additional terms are used: refers to both DHCPv4 and DHCPv6.refers to unencrypted DNS.refers to the Discovery of Network-designated
Resolvers procedure defined in .refers to the Discovery of Designated Resolvers
procedure defined in .refers to a scheme where DNS exchanges
are transported over an encrypted channel. Examples of encrypted DNS
are DoT, DoH, or DoQ.refers to the options defined
in .refers to a CPE that is managed by an
ISP.refers to a CPE that is not managed by
an ISP.ISPs usually provide DNS resolvers to their customers. To that aim,
ISPs deploy the following mechanisms to advertise a list of DNS
Recursive DNS server(s) to their customers:Protocol Configuration Options in cellular networks .DHCPv4 (Domain Name Server Option)
or DHCPv6 (OPTION_DNS_SERVERS).IPv6 Router Advertisement (Type 25 (Recursive DNS Server
Option)).The communication between a customer's device (possibly via a CPE)
and an ISP-supplied DNS resolver takes place by using cleartext DNS
messages (Do53). Some examples are depicted in cases (a) and (c) of
. In the case of cellular networks, the
cellular network will provide connectivity directly to a host (e.g.,
smartphone, tablet) or via a CPE. Do53 mechanisms used within the Local
Area Network (LAN) are similar in both fixed and cellular CPE-based
broadband service offerings.Some ISPs rely upon external resolvers (e.g., outsourced service or
public resolvers); these ISPs provide their customers with the IP
addresses of these external DNS resolvers. An example is depicted in
cases (b) and (d) of .The IP addresses of the DNS resolver can also be configured on CPEs
using dedicated management tools. As such, users can modify the default
DNS configuration of their CPEs (e.g., supplied by their ISP) to
configure their favorite DNS servers. This document permits such
deployments.This section focuses on CPEs that are managed by ISPs.ISPs have developed an expertise in managing service-specific
configuration information (e.g., CPE WAN Management Protocol ). For example, these tools may be used to
provision the DNS server's ADN and additional service parameters to
managed CPEs if an encrypted DNS is supported by a network similar
to what is depicted in .For example, DoH-capable DNS clients establish the DoH session
with the discovered DoH server.When the CPE supports DNR, the DNS client discovers whether the
network-designated DNS resolver supports a given encrypted DNS
scheme (e.g., DoT or DoH) by using the "alpn" service parameter
().
Otherwise, the DNS client uses DDR with the Do53 resolver advertised
by the CPE and upgrades to encrypted DNS if that succeeds.
Otherwise, the DNS client may fall back to using unencrypted DNS to
the IP address advertised by the CPE or use some other configuration
it has.DNR is attempted first because it requires fewer round trips to
any network peer because all of the necessary information to use
encrypted DNS is presented directly by the CPE. DDR requires the DNS
client to receive Do53 resolver configuration from the CPE and then
further query for encrypted DNS support from the DNS resolver before
any encrypted DNS can be attempted. shows the scenario where the CPE
relays the list of encrypted DNS resolvers that it learns from the
network by using, e.g., DNR. Direct encrypted DNS sessions will be
established between a host serviced by a CPE and an ISP-supplied
encrypted DNS resolver. shows the
example of exchanges that occur for an encrypted DNS capable host.
The DNR exchanges that occur at the CPE WAN may be terminated by a
centralized DHCP server or a router that is located at the edge of
the ISP's network. shows various network setups where
the CPE embeds a caching DNS forwarder. Cases (b) and (d) involves a
host (called legacy host) that does not support DNR. discusses the applicability of DDR as a
function of the address used by the CPE for the verification of
ownership.For all the cases shown in , the
CPE advertises itself as the default DNS server to the hosts it
serves in the LAN. The CPE relies upon DHCP or RA to advertise
itself to internal hosts as the default encrypted DNS (cases (a) and
(c)) or Do53 resolver (cases (b) and (d)). When receiving a DNS
request it cannot handle locally, the CPE forwards the request to an
upstream encrypted DNS. The upstream encrypted DNS can be hosted by
the ISP (cases (a) and (b)) or provided by a third party (cases (c)
and (d)).Such a forwarder presence is required for IPv4 service continuity
purposes (e.g., Section 3.1 of ) or
for supporting advanced services within a local network (e.g.,
malware filtering, parental control, Manufacturer Usage Description
(MUD) to only allow intended
communications to and from an IoT device). When the CPE behaves as a
DNS forwarder, DNS communications can be decomposed into two
legs:The leg between an internal host and the CPE.The leg between the CPE and an upstream DNS resolver.An ISP that offers encrypted DNS to its customers may enable
encrypted DNS in one or both legs as shown in . Additional considerations related to this
setup are discussed in .Customers may decide to deploy unmanaged CPEs (assuming the CPE
is compliant with the network access technical specification that is
usually published by ISPs). Upon attachment to the network, an
unmanaged CPE receives from the network its service configuration
(including the network-designated DNS information) by means of,
e.g., DHCP. That DNS information is shared within the LAN following
the same mechanisms as those discussed in . A host can then establish encrypted DNS
sessions with encrypted DNS resolvers similar to what is depicted in
or .Customers may also decide to deploy internal routers (called
hereafter, Internal CPEs) for a variety of reasons that are not
detailed here.Absent any explicit configuration on the internal CPE to override
the DNS configuration it receives from the ISP-supplied CPE, an
Internal CPE relays the DNS information it receives via DHCP/RA from
the ISP-supplied CPE to connected hosts. Encrypted DNS sessions can
be established by a host with the DNS resolvers that are supplied by
the ISP (see ).Similar to managed CPEs, a user may modify the default DNS
configuration of an unmanaged CPE to use his/her favorite encrypted
DNS resolvers instead. Encrypted DNS sessions can be established
directly between a host and a 3rd Party DNS resolver (see ). discusses considerations
related to hosting a forwarder in the Internal CPE.This section discusses some deployment considerations to host an
encrypted DNS forwarder within a local networkDDR requires proving possession of an IP address, as the DDR
certificate contains the server's IPv4 and IPv6 addresses and is
signed by a certificate authority. DDR is constrained to public IP
addresses because WebPKI certificate authorities will not sign
special-purpose IP addresses , most
notably IPv4 private-use , IPv4 shared
address , or IPv6 Unique-Local address space. A tempting
solution is to use the CPE's WAN IP address for DDR and prove
possession of that IP address. However, the CPE's WAN IPv4 address
will not be a public IPv4 address if the CPE is behind another layer
of NAT (either Carrier Grade NAT (CGN) or another on-premise NAT),
reducing the success of this mechanism to CPE's WAN IPv6 address. If
the ISP renumbers the subscriber's network suddenly (rather than slow
IPv6 renumbering described in )
encrypted DNS service will be delayed until that new certificate is
acquired.DNR requires proving possession of an FQDN as the encrypted
resolver's certificate contains the FQDN. The entity (e.g., ISP,
network administrator) managing the CPE would assign a unique FQDN to
the CPE. There are two mechanisms for the CPE to obtain the
certificate for the FQDN: using one of its WAN IP addresses or
requesting its signed certificate from an Internet-facing server used
for remote CPE management (e.g., the Auto Configuration Server (ACS)
in the CPE WAN Management Protocol ). If
using a CPE's WAN IP address, the CPE needs a public IPv4 or a global
unicast IPv6 address together with DNS A or AAAA records pointing to
that CPE's WAN address to prove possession of the DNS name to obtain a
WebPKI CA-signed certificate (that is, the CPE fulfills the DNS or
HTTP challenge discussed in ACME ).
However, a CPE's WAN address will not be a public IPv4 address if the
CPE is behind another layer of NAT (either a CGN or another on-premise
NAT), reducing the success of this mechanism to a CPE's WAN IPv6
address. If the subscribers IPv4 or IPv6 address is included in the
certificate name (e.g., "dyn-192-0-2-1.example.net") then DNR will
experience IP renumbering complications identical to DDR, described
above. The former mechanism has the following limitations when ACME
protocol is used for certificate issuance:Each CPE would have to create a different account for ordering
a certificate. When a large scale of CPEs request certificate
issuance for a large number of subdomains, it could be treated as
an attacker by the certificate authorities to overwhelm it.The CPE would have to host an Internet-facing HTTP server or a
DNS authoritative server to complete the HTTP or DNS
challenge.The section discusses mechanisms that can be used to host an
encrypted DNS forwarder in a managed CPE ().The managed CPE should support a configuration parameter to
instruct the CPE whether it has to relay the encrypted DNS resolver
received from the ISP's network or has to announce itself as a
forwarder within the local network. The default behavior of the CPE
is to supply the encrypted DNS resolver received from the ISP's
network.The ISP can assign a unique FQDN (e.g., "cpe1.example.com") and a
domain-validated public certificate to the encrypted DNS forwarder
hosted on the CPE.Automatic Certificate Management Environment (ACME) can be used by the ISP to automate
certificate management functions such as domain validation
procedure, certificate issuance, and certificate revocation.The approach specified in does
not apply for hosting a DNS forwarder in an unmanaged CPE.The unmanaged CPE administrator can host an encrypted DNS forwarder
on the unmanaged CPE. This assumes the following:The encrypted DNS resolver certificate is managed by the entity
in-charge of hosting the encrypted DNS forwarder. Alternatively, a security service provider can
assign a unique FQDN to the CPE. The encrypted DNS forwarder will
act like a private encrypted DNS resolver only be accessible from
within the local network.The encrypted DNS forwarder will either be configured to use
the ISP's or a 3rd party encrypted DNS resolver.The unmanaged CPE will advertise the encrypted DNS forwarder
ADN using DHCP/RA to internal hosts as per . illustrates an example of
an unmanaged CPE hosting a forwarder which connects to a 3rd party
encrypted DNS resolver. In this example, the DNS information received
from the managed CPE (and therefore from the ISP) is ignored by the
Internal CPE hosting the forwarder. The internal CPE may support a
mechanism (e.g., ) to resolve
split-horizon domains (e.g., provider's private name discussed in
).An unmanaged CPE can be used to host an encrypted DNS forwarder
even if the managed CPE does not support DNR. In the example depicted
in , the ISP uses DHCP to
provision Do53 resolvers to managed CPEs, while DNR is enabled between
the internal CPE and the hosts it services. The internal CPE ignores
the DNS configuration that it receives from the managed CPE.Hosts serviced by legacy CPEs that can't be upgraded to support the
options defined in Sections 4, 5, and 6 of won't be able to learn the encrypted
DNS resolver hosted by the ISP, in particular. If the ADN is not
discovered using DHCP/RA, such hosts will have to fall back to use
discovery using the resolver IP address as defined in to discover the designated
resolvers.The guidance in Sections 4.1 and 4.2 of related to the designated resolver
verification has to be followed in such a case.DNR-related security considerations are discussed in . Likewise, DDR-related
security considerations are discussed in .This document does not require any IANA action.This text was initially part of .Thanks to Eliot Lear for the ISE review.CPE WAN Management ProtocolMobile radio interface Layer 3 specification; Core network
protocols; Stage 3 (Release 16)3GPP