| DPRIVE | T. Reddy |
| Internet-Draft | D. Wing |
| Intended status: Experimental | P. Patil |
| Expires: June 3, 2017 | Cisco |
| November 30, 2016 |
Specification for DNS over Datagram Transport Layer Security (DTLS)
draft-ietf-dprive-dnsodtls-13
DNS queries and responses are visible to network elements on the path between the DNS client and its server. These queries and responses can contain privacy-sensitive information which is valuable to protect.
This document proposes the use of Datagram Transport Layer Security (DTLS) for DNS, to protect against passive listeners and certain active attacks. As latency is critical for DNS, this proposal also discusses mechanisms to reduce DTLS round trips and reduce DTLS handshake size. The proposed mechanism runs over port 853.
This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.
Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at http://datatracker.ietf.org/drafts/current/.
Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress."
This Internet-Draft will expire on June 3, 2017.
Copyright (c) 2016 IETF Trust and the persons identified as the document authors. All rights reserved.
This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (http://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License.
The Domain Name System is specified in [RFC1034] and [RFC1035]. DNS queries and responses are normally exchanged unencrypted and are thus vulnerable to eavesdropping. Such eavesdropping can result in an undesired entity learning domains that a host wishes to access, thus resulting in privacy leakage. The DNS privacy problem is further discussed in [RFC7626].
This document defines DNS over DTLS (DNS-over-DTLS) which provides confidential DNS communication between stub resolvers and recursive resolvers, stub resolvers and forwarders, forwarders and recursive resolvers. DNS-over-DTLS puts an additional computational load on servers. The largest gain for privacy is to protect the communication between the DNS client (the end user's machine) and its caching resolver. DNS-over-DTLS might work equally between recursive clients and authoritative servers, but this application of the protocol is out of scope for the DNS PRIVate Exchange (DPRIVE) Working Group per its current charter. This document does not change the format of DNS messages.
The motivations for proposing DNS-over-DTLS are that
Note: DNS-over-DTLS is an experimental update to DNS, and the experiment will be concluded when the specification is evaluated through implementations and interoperability testing.
DNS queries can be sent over UDP or TCP. The scope of this document, however, is only UDP. DNS over TCP can be protected with TLS, as described in [RFC7858]. DNS-over-DTLS alone cannot provide privacy for DNS messages in all circumstances, specifically when the DTLS record size is larger than the path MTU. In such situations the DNS server will respond with a truncated response (see Section 5). Therefore DNS clients and servers that implement DNS-over-DTLS MUST also implement DNS-over-TLS in order to provide privacy for clients that desire Strict Privacy as described in [I-D.ietf-dprive-dtls-and-tls-profiles].
DNS Security Extensions (DNSSEC [RFC4033]) provides object integrity of DNS resource records, allowing end-users (or their resolver) to verify legitimacy of responses. However, DNSSEC does not provide privacy for DNS requests or responses. DNS-over-DTLS works in conjunction with DNSSEC, but DNS-over-DTLS does not replace the need or value of DNSSEC.
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in [RFC2119] .
By default, DNS-over-DTLS MUST run over standard UDP port 853 as defined in Section 8, unless the DNS server has mutual agreement with its clients to use a port other than 853 for DNS-over-DTLS. In order to use a port other than 853, both clients and servers would need a configuration option in their software.
The DNS client should determine if the DNS server supports DNS-over-DTLS by sending a DTLS ClientHello message to port 853 on the server, unless it has mutual agreement with its server to use a port other than port 853 for DNS-over-DTLS. Such another port MUST NOT be port 53 but MAY be from the "first-come, first-served" port range (User Ports [RFC6335], range 1024- 49151) . This recommendation against use of port 53 for DNS-over-DTLS is to avoid complication in selecting use or non-use of DTLS and to reduce risk of downgrade attacks.
A DNS server that does not support DNS-over-DTLS will not respond to ClientHello messages sent by the client. If no response is received from that server, and the client has no better round-trip estimate, the client SHOULD retransmit the DTLS ClientHello according to Section 4.2.4.1 of [RFC6347]. After 15 seconds, it SHOULD cease attempts to re-transmit its ClientHello. The client MAY repeat that procedure to discover if DNS-over-DTLS service becomes available from the DNS server, but such probing SHOULD NOT be done more frequently than every 24 hours and MUST NOT be done more frequently than every 15 minutes. This mechanism requires no additional signaling between the client and server.
DNS clients and servers MUST NOT use port 853 to transport cleartext DNS messages. DNS clients MUST NOT send and DNS servers MUST NOT respond to cleartext DNS messages on any port used for DNS-over-DTLS (including, for example, after a failed DTLS handshake). There are significant security issues in mixing protected and unprotected data, therefore UDP connections on a port designated by a given server for DNS-over-DTLS are reserved purely for encrypted communications.
DNS client initiates DTLS handshake as described in [RFC6347], following the best practices specified in [RFC7525]. After DTLS negotiation completes, if the DTLS handshake succeeds according to [RFC6347] the connection will be encrypted and is now protected from eavesdropping.
DNS privacy requires encrypting the query (and response) from passive attacks. Such encryption typically provides integrity protection as a side-effect, which means on-path attackers cannot simply inject bogus DNS responses. However, to provide stronger protection from active attackers pretending to be the server, the server itself needs to be authenticated. To authenticate the server providing DNS privacy, DNS client MUST use the authenication mechanisms discussed in [I-D.ietf-dprive-dtls-and-tls-profiles]. This document does not propose new ideas for authentication.
In DTLS, all data is protected using the same record encoding and mechanisms. When the mechanism described in this document is in effect, DNS messages are encrypted using the standard DTLS record encoding. When a user of DTLS wishes to send a DNS message, the data is delivered to the DTLS implementation as an ordinary application data write (e.g., SSL_write()). A single DTLS session can be used to send multiple DNS requests and receive multiple DNS responses.
To mitigate the risk of unintentional server overload, DNS-over-DTLS clients MUST take care to minimize the number of concurrent DTLS sessions made to any individual server. It is RECOMMENDED that for any given client/server interaction there SHOULD be no more than one DTLS session. Similarly, servers MAY impose limits on the number of concurrent DTLS sessions being handled for any particular client IP address or subnet. These limits SHOULD be much looser than the client guidelines above, because the server does not know, for example, if a client IP address belongs to a single client, is multiple resolvers on a single machine, or is multiple clients behind a device performing Network Address Translation (NAT).
In between normal DNS traffic while the communication to the DNS server is quiescent, the DNS client MAY want to probe the server using DTLS heartbeat [RFC6520] to ensure it has maintained cryptographic state. Such probes can also keep alive firewall or NAT bindings. This probing reduces the frequency of needing a new handshake when a DNS query needs to be resolved, improving the user experience at the cost of bandwidth and processing time.
A DTLS session is terminated by the receipt of an authenticated message that closes the connection (e.g., a DTLS fatal alert). If the server has lost state, a DTLS handshake needs to be initiated with the server. For the client, state should be destroyed when disconnecting from the network (e.g., associated IP interface is brought down). For the server, to mitigate the risk of unintentional server overload, it is RECOMMENDED that the default DNS-over-DTLS server application-level idle time out be on the order of several seconds, but no particular value is specified. When no DNS queries have been received from the client after idle time out, the server MUST send a DTLS fatal alert and then destroy its DTLS state. The DTLS fatal alert packet indicates the server has destroyed its state, signaling to the client if it wants to send a new DTLS message it will need to re-establish cryptographic context with the server (via full DTLS handshake or DTLS session resumption). In practice, the idle period can vary dynamically, and servers MAY allow idle connections to remain open for longer periods as resources permit.
Figure 1 shows DTLS handshake and issuing new session ticket for session resumption.
Client Server
------ ------
ClientHello -------->
(empty SessionTicket extension)
ServerHello
(empty SessionTicket extension)
Certificate*
ServerKeyExchange*
CertificateRequest*
<-------- ServerHelloDone
Certificate*
ClientKeyExchange
CertificateVerify*
(ChangeCipherSpec)
Finished -------->
NewSessionTicket
(ChangeCipherSpec)
<-------- Finished
DNS Request --------->
<--------- DNS Response
Figure 1: Message Flow for Full Handshake Issuing New Session Ticket
DTLS protocol profile for DNS-over-DTLS is discussed in Section 11 of [I-D.ietf-dprive-dtls-and-tls-profiles]. To reduce the number of octets of the DTLS handshake, especially the size of the certificate in the ServerHello (which can be several kilobytes), DNS clients and servers can use raw public keys [RFC7250] or Cached Information Extension [RFC7924]. Cached Information Extension avoids transmitting the server's certificate and certificate chain if the client has cached that information from a previous TLS handshake. TLS False Start [RFC7918] can reduce round-trips by allowing the TLS second flight of messages (ChangeCipherSpec) to also contain the (encrypted) DNS query.
It is highly advantageous to avoid server-side DTLS state and reduce the number of new DTLS sessions on the server which can be done with TLS Session Resumption without server state [RFC5077]. This also eliminates a round-trip for subsequent DNS-over-DTLS queries, because with [RFC5077] the DTLS session does not need to be re-established.
Since responses within a DTLS session can arrive out of order, clients MUST match responses to outstanding queries on the same DTLS connection using the DNS Message ID. If the response contains a question section, the client MUST match the QNAME, QCLASS, and QTYPE fields. Failure by clients to properly match responses to outstanding queries can have serious consequences for interoperability ( [RFC7766], Section 7).
Compared to normal DNS, DTLS adds at least 13 octets of header, plus cipher and authentication overhead to every query and every response. This reduces the size of the DNS payload that can be carried. DNS client and server MUST support the EDNS0 option defined in [RFC6891] so that the DNS client can indicate to the DNS server the maximum DNS response size it can reassemble and deliver in the DNS client’s network stack. If the DNS client does set the EDNS0 option defined in [RFC6891] then the maximum DNS response size of 512 bytes plus DTLS overhead will be well within the Path MTU. If the Path MTU is not known, an IP MTU of 1280 bytes SHOULD be assumed. The client sets its EDNS0 value as if DTLS is not being used. The DNS server MUST ensure that the DNS response size does not exceed the Path MTU i.e. each DTLS record MUST fit within a single datagram, as required by [RFC6347]. The DNS server must consider the amount of record expansion expected by the DTLS processing when calculating the size of DNS response that fits within the path MTU. Path MTU MUST be greater than or equal to [DNS response size + DTLS overhead of 13 octets + padding size ([RFC7830]) + authentication overhead of the negotiated DTLS cipher suite + block padding (Section 4.1.1.1 of [RFC6347]]. If the DNS server's response were to exceed that calculated value, the server MUST send a response that does fit within that value and sets the TC (truncated) bit. Upon receiving a response with the TC bit set and wanting to receive the entire response, the client behaviour is governed by the current Usage profile [I-D.ietf-dprive-dtls-and-tls-profiles]. For Strict Privacy the client MUST only send a new DNS request for the same resource record over an encrypted transport (e.g. DNS-over-TLS [RFC7858]). Clients using Opportunistic Privacy SHOULD try for the best case (an encrypted and authenticated transport) but MAY fallback to intermediate cases and eventually the worst case scenario (clear text) in order to obtain a response.
DNS servers are often configured with anycast addresses. While the network is stable, packets transmitted from a particular source to an anycast address will reach the same server that has the cryptographic context from the DNS-over-DTLS handshake. But when the network configuration or routing changes, a DNS-over-DTLS packet can be received by a server that does not have the necessary cryptographic context. Clients using DNS-over-DTLS need to always be prepared to re-initiate DTLS handshake and in the worst case this could even happen immediately after re-initiating a new handshake. To encourage the client to initiate a new DTLS handshake, DNS servers SHOULD generate a DTLS fatal alert message in response to receiving a DTLS packet for which the server does not have any cryptographic context. Upon receipt of an un-authenicated DTLS fatal alert, the DTLS client validates the fatal alert is within the replay window (Section 4.1.2.6 of [RFC6347]). It is difficult for the DTLS client to validate that the DTLS fatal alert was generated by the DTLS server in response to a request or was generated by an on- or off-path attacker. Thus, upon receipt of an in-window DTLS fatal alert, the client SHOULD continue re-transmitting the DTLS packet (in the event the fatal alert was spoofed), and at the same time it SHOULD initiate DTLS session resumption. When the DTLS client receives an authenticated DNS response from one of those DTLS sessions, the other DTLS session should be terminated.
Two Usage Profiles, Strict and Opportunistic are explained in [I-D.ietf-dprive-dtls-and-tls-profiles]. Using encrypted DNS messages with an authenticated server is most preferred, encrypted DNS messages with an unauthenticated server is next preferred, and plain text DNS messages is least preferred.
This specification uses port 853 already allocated in the IANA port number registry as defined in Section 6 of [RFC7858].
The interaction between a DNS client and DNS server requires Datagram Transport Layer Security (DTLS) with a ciphersuite offering confidentiality protection. The guidance given in [RFC7525] MUST be followed to avoid attacks on DTLS. DNS clients keeping track of servers known to support DTLS enables clients to detect downgrade attacks. To interfere with DNS-over-DTLS, an on- or off-path attacker might send an ICMP message towards the DTLS client or DTLS server. As these ICMP messages cannot be authenticated, all ICMP errors should be treated as soft errors [RFC1122]. If the DNS query was sent over DTLS then the corresponding DNS response MUST only be accepted if it is received over the same DTLS connection. This behavior mitigates all possible attacks described in Measures for Making DNS More Resilient against Forged Answers [RFC5452]. Security considerations in [RFC6347] and [I-D.ietf-dprive-dtls-and-tls-profiles] are to be taken into account.
A malicious client might attempt to perform a high number of DTLS handshakes with a server. As the clients are not uniquely identified by the protocol and can be obfuscated with IPv4 address sharing and with IPv6 temporary addresses, a server needs to mitigate the impact of such an attack. Such mitigation might involve rate limiting handshakes from a certain subnet or more advanced DoS/DDoS techniques beyond the scope of this paper.
Thanks to Phil Hedrick for his review comments on TCP and to Josh Littlefield for pointing out DNS-over-DTLS load on busy servers (most notably root servers). The authors would like to thank Simon Josefsson, Daniel Kahn Gillmor, Bob Harold, Ilari Liusvaara, Sara Dickinson, Christian Huitema, Stephane Bortzmeyer, Alexander Mayrhofer, Allison Mankin, Jouni Korhonen and Geoff Huston for discussions and comments on the design of DNS-over-DTLS. The authors would like to give special thanks to Sara Dickinson for her help.