draft-ietf-dnsop-5966bis-01.txt   draft-ietf-dnsop-5966bis-02.txt 
dnsop J. Dickinson dnsop J. Dickinson
Internet-Draft Sinodun Internet Technologies Internet-Draft S. Dickinson
Obsoletes: 5966 (if approved) R. Bellis Obsoletes: 5966 (if approved) Sinodun
Intended status: Standards Track Nominet Intended status: Standards Track R. Bellis
Expires: September 10, 2015 A. Mankin Expires: January 7, 2016 ISC
A. Mankin
D. Wessels D. Wessels
Verisign Labs Verisign Labs
March 9, 2015 July 6, 2015
DNS Transport over TCP - Implementation Requirements DNS Transport over TCP - Implementation Requirements
draft-ietf-dnsop-5966bis-01 draft-ietf-dnsop-5966bis-02
Abstract Abstract
This document specifies the requirement for support of TCP as a This document specifies the requirement for support of TCP as a
transport protocol for DNS implementations and provides guidelines transport protocol for DNS implementations and provides guidelines
towards DNS-over-TCP performance on par with that of DNS-over-UDP. towards DNS-over-TCP performance on par with that of DNS-over-UDP.
Status of This Memo Status of This Memo
This Internet-Draft is submitted in full conformance with the This Internet-Draft is submitted in full conformance with the
skipping to change at page 1, line 36 skipping to change at page 1, line 37
Internet-Drafts are working documents of the Internet Engineering Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF). Note that other groups may also distribute Task Force (IETF). Note that other groups may also distribute
working documents as Internet-Drafts. The list of current Internet- working documents as Internet-Drafts. The list of current Internet-
Drafts is at http://datatracker.ietf.org/drafts/current/. Drafts is at http://datatracker.ietf.org/drafts/current/.
Internet-Drafts are draft documents valid for a maximum of six months Internet-Drafts are draft documents valid for a maximum of six months
and may be updated, replaced, or obsoleted by other documents at any and may be updated, replaced, or obsoleted by other documents at any
time. It is inappropriate to use Internet-Drafts as reference time. It is inappropriate to use Internet-Drafts as reference
material or to cite them other than as "work in progress." material or to cite them other than as "work in progress."
This Internet-Draft will expire on September 10, 2015. This Internet-Draft will expire on January 7, 2016.
Copyright Notice Copyright Notice
Copyright (c) 2015 IETF Trust and the persons identified as the Copyright (c) 2015 IETF Trust and the persons identified as the
document authors. All rights reserved. document authors. All rights reserved.
This document is subject to BCP 78 and the IETF Trust's Legal This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents Provisions Relating to IETF Documents
(http://trustee.ietf.org/license-info) in effect on the date of (http://trustee.ietf.org/license-info) in effect on the date of
publication of this document. Please review these documents publication of this document. Please review these documents
skipping to change at page 2, line 12 skipping to change at page 2, line 13
to this document. Code Components extracted from this document must to this document. Code Components extracted from this document must
include Simplified BSD License text as described in Section 4.e of include Simplified BSD License text as described in Section 4.e of
the Trust Legal Provisions and are provided without warranty as the Trust Legal Provisions and are provided without warranty as
described in the Simplified BSD License. described in the Simplified BSD License.
Table of Contents Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2
2. Requirements Terminology . . . . . . . . . . . . . . . . . . 3 2. Requirements Terminology . . . . . . . . . . . . . . . . . . 3
3. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 3 3. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 3
4. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . 3 4. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . 4
5. Transport Protocol Selection . . . . . . . . . . . . . . . . 4 5. Transport Protocol Selection . . . . . . . . . . . . . . . . 5
6. Connection Handling . . . . . . . . . . . . . . . . . . . . . 5 6. Connection Handling . . . . . . . . . . . . . . . . . . . . . 6
7. Query Pipelining . . . . . . . . . . . . . . . . . . . . . . 6 6.1. Current practices . . . . . . . . . . . . . . . . . . . . 6
8. Response Reordering . . . . . . . . . . . . . . . . . . . . . 7 6.1.1. Clients . . . . . . . . . . . . . . . . . . . . . . . 6
9. TCP Fast Open . . . . . . . . . . . . . . . . . . . . . . . . 8 6.1.2. Servers . . . . . . . . . . . . . . . . . . . . . . . 7
10. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 8 6.2. Recommendations . . . . . . . . . . . . . . . . . . . . . 7
11. Security Considerations . . . . . . . . . . . . . . . . . . . 8 6.2.1. Connection Re-use . . . . . . . . . . . . . . . . . . 7
12. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 9 6.2.1.1. Query Pipelining . . . . . . . . . . . . . . . . 8
13. References . . . . . . . . . . . . . . . . . . . . . . . . . 9 6.2.2. Concurrent connections . . . . . . . . . . . . . . . 8
13.1. Normative References . . . . . . . . . . . . . . . . . . 9 6.2.3. Idle Timeouts . . . . . . . . . . . . . . . . . . . . 8
13.2. Informative References . . . . . . . . . . . . . . . . . 10 6.2.4. Tear Down . . . . . . . . . . . . . . . . . . . . . . 9
7. Response Reordering . . . . . . . . . . . . . . . . . . . . . 9
8. TCP Message Length Field . . . . . . . . . . . . . . . . . . 10
9. TCP Fast Open . . . . . . . . . . . . . . . . . . . . . . . . 10
10. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 11
11. Security Considerations . . . . . . . . . . . . . . . . . . . 11
12. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 12
13. References . . . . . . . . . . . . . . . . . . . . . . . . . 12
13.1. Normative References . . . . . . . . . . . . . . . . . . 12
13.2. Informative References . . . . . . . . . . . . . . . . . 13
Appendix A. Summary of Advantages and Disadvantages to using TCP Appendix A. Summary of Advantages and Disadvantages to using TCP
for DNS . . . . . . . . . . . . . . . . . . . . . . 11 for DNS . . . . . . . . . . . . . . . . . . . . . . 13
Appendix B. Changes -00 to -01 . . . . . . . . . . . . . . . . . 11 Appendix B. Changes -01 to -02 . . . . . . . . . . . . . . . . . 14
Appendix C. Changes to RFC 5966 . . . . . . . . . . . . . . . . 12 Appendix C. Changes -00 to -01 . . . . . . . . . . . . . . . . . 15
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 12 Appendix D. Changes to RFC 5966 . . . . . . . . . . . . . . . . 15
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 16
1. Introduction 1. Introduction
Most DNS [RFC1034] transactions take place over UDP [RFC0768]. TCP Most DNS [RFC1034] transactions take place over UDP [RFC0768]. TCP
[RFC0793] is always used for full zone transfers (AXFR) and is often [RFC0793] is always used for full zone transfers (AXFR) and is often
used for messages whose sizes exceed the DNS protocol's original used for messages whose sizes exceed the DNS protocol's original
512-byte limit. 512-byte limit. The growing deployment of DNSSEC and IPv6 has
increased response sizes and therefore the use of TCP. The need for
increased TCP use has also been driven by the protection it provides
against address spoofing and therefore exploitation of DNS in
reflection/amplification attacks. It is now widely used in Response
Rate Limiting [RRL] Response Rate Limiting [RRL].
Section 6.1.3.2 of [RFC1123] states: Section 6.1.3.2 of [RFC1123] states:
DNS resolvers and recursive servers MUST support UDP, and SHOULD DNS resolvers and recursive servers MUST support UDP, and SHOULD
support TCP, for sending (non-zone-transfer) queries. support TCP, for sending (non-zone-transfer) queries.
However, some implementors have taken the text quoted above to mean However, some implementors have taken the text quoted above to mean
that TCP support is an optional feature of the DNS protocol. that TCP support is an optional feature of the DNS protocol.
The majority of DNS server operators already support TCP and the The majority of DNS server operators already support TCP and the
default configuration for most software implementations is to support default configuration for most software implementations is to support
TCP. The primary audience for this document is those implementors TCP. The primary audience for this document is those implementors
whose failure to support TCP restricts interoperability and limits whose limited support for TCP restricts interoperability and hinders
deployment of new DNS features. deployment of new DNS features.
This document therefore updates the core DNS protocol specifications This document therefore updates the core DNS protocol specifications
such that support for TCP is henceforth a REQUIRED part of a full DNS such that support for TCP is henceforth a REQUIRED part of a full DNS
protocol implementation. protocol implementation.
There are several advantages and disadvantages to the increased use There are several advantages and disadvantages to the increased use
of TCP as well as implementation details that need to be considered. of TCP as well as implementation details that need to be considered.
This document addresses these issues and therefore extends the This document addresses these issues and therefore extends the
content of [RFC5966], with additional considerations and lessons content of [RFC5966], with additional considerations and lessons
learned from new research and implementations learned from research, developments and implementation in DNS and in
[Connection-Oriented-DNS]. other internet protocols.
Whilst this document makes no specific requirements for operators of Whilst this document makes no specific requirements for operators of
DNS servers to meet, it does offer some suggestions to operators to DNS servers to meet, it does offer some suggestions to operators to
help ensure that support for TCP on their servers and network is help ensure that support for TCP on their servers and network is
optimal. It should be noted that failure to support TCP (or the optimal. It should be noted that failure to support TCP (or the
blocking of DNS over TCP at the network layer) may result in blocking of DNS over TCP at the network layer) may result in
resolution failure and/or application-level timeouts. resolution failure and/or application-level timeouts.
2. Requirements Terminology 2. Requirements Terminology
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
document are to be interpreted as described in [RFC2119]. document are to be interpreted as described in [RFC2119].
3. Terminology 3. Terminology
o Persistent connection: a TCP connection that is not closed either
by the server after sending the first response nor by the client
after receiving the first response.
o Connection Reuse: the sending of multiple queries and responses o Connection Reuse: the sending of multiple queries and responses
over a single TCP connection. over a single TCP connection.
o Idle DNS-over-TCP session: Clients and servers view application
level idleness differently. A DNS client considers a DNS-over-TCP
session to be idle when it has no pending queries to send and
there are no outstanding responses. A DNS server considers a DNS-
over-TCP session to be idle when it has sent responses to all the
queries it has received on that connection.
o Pipelining: the sending of multiple queries and responses over a o Pipelining: the sending of multiple queries and responses over a
single TCP connection but not waiting for any outstanding replies single TCP connection but not waiting for any outstanding replies
before sending another query. before sending another query.
o Out-Of-Order Processing: The processing of queries in parallel and o Out-Of-Order Processing: The processing of queries concurrently
the returning of individual responses as soon as they are and the returning of individual responses as soon as they are
available, possibly out-of-order. This will most likely occur in available, possibly out-of-order. This will most likely occur in
recursive servers, however it is possible in authoritative servers recursive servers, however it is possible in authoritative servers
that, for example, have different backend data stores. that, for example, have different backend data stores.
4. Discussion 4. Discussion
In the absence of EDNS0 (Extension Mechanisms for DNS 0) (see below), In the absence of EDNS0 (Extension Mechanisms for DNS 0) (see below),
the normal behaviour of any DNS server needing to send a UDP response the normal behaviour of any DNS server needing to send a UDP response
that would exceed the 512-byte limit is for the server to truncate that would exceed the 512-byte limit is for the server to truncate
the response so that it fits within that limit and then set the TC the response so that it fits within that limit and then set the TC
skipping to change at page 4, line 30 skipping to change at page 5, line 9
Since the original core specifications for DNS were written, the Since the original core specifications for DNS were written, the
Extension Mechanisms for DNS (EDNS0 [RFC6891]) have been introduced. Extension Mechanisms for DNS (EDNS0 [RFC6891]) have been introduced.
These extensions can be used to indicate that the client is prepared These extensions can be used to indicate that the client is prepared
to receive UDP responses larger than 512 bytes. An EDNS0-compatible to receive UDP responses larger than 512 bytes. An EDNS0-compatible
server receiving a request from an EDNS0-compatible client may send server receiving a request from an EDNS0-compatible client may send
UDP packets up to that client's announced buffer size without UDP packets up to that client's announced buffer size without
truncation. truncation.
However, transport of UDP packets that exceed the size of the path However, transport of UDP packets that exceed the size of the path
MTU causes IP packet fragmentation, which has been found to be MTU causes IP packet fragmentation, which has been found to be
unreliable in some circumstances. Many firewalls routinely block unreliable in many circumstances. Many firewalls routinely block
fragmented IP packets, and some do not implement the algorithms fragmented IP packets, and some do not implement the algorithms
necessary to reassemble fragmented packets. Worse still, some necessary to reassemble fragmented packets. Worse still, some
network devices deliberately refuse to handle DNS packets containing network devices deliberately refuse to handle DNS packets containing
EDNS0 options. Other issues relating to UDP transport and packet EDNS0 options. Other issues relating to UDP transport and packet
size are discussed in [RFC5625]. size are discussed in [RFC5625].
The MTU most commonly found in the core of the Internet is around The MTU most commonly found in the core of the Internet is around
1500 bytes, and even that limit is routinely exceeded by DNSSEC- 1500 bytes, and even that limit is routinely exceeded by DNSSEC-
signed responses. signed responses.
skipping to change at page 5, line 28 skipping to change at page 6, line 5
Regarding the choice of when to use UDP or TCP, Section 6.1.3.2 of Regarding the choice of when to use UDP or TCP, Section 6.1.3.2 of
RFC 1123 also says: RFC 1123 also says:
... a DNS resolver or server that is sending a non-zone-transfer ... a DNS resolver or server that is sending a non-zone-transfer
query MUST send a UDP query first. query MUST send a UDP query first.
This requirement is hereby relaxed. A resolver MAY elect to send This requirement is hereby relaxed. A resolver MAY elect to send
either TCP or UDP queries depending on local operational reasons. either TCP or UDP queries depending on local operational reasons.
TCP MAY be used before sending any UDP queries. If it already has an TCP MAY be used before sending any UDP queries. If it already has an
open TCP connection to the server it SHOULD reuse this connection. open TCP connection to the server it SHOULD reuse this connection.
In essence, TCP SHOULD be considered as valid a transport as UDP. In essence, TCP should be considered a valid alternative transport to
UDP, not purely a fallback option.
In addition it is noted that all Recursive and Authoritative servers In addition it is noted that all Recursive and Authoritative servers
MUST send responses using the same transport as the query arrived on. MUST send responses using the same transport as the query arrived on.
In the case of TCP this MUST also be the same connection. In the case of TCP this MUST also be the same connection.
6. Connection Handling 6. Connection Handling
One perceived disadvantage to DNS over TCP is the added connection 6.1. Current practices
setup latency, generally equal to one RTT. To amortize connection
setup costs, both clients and servers SHOULD support connection reuse
by sending multiple queries and responses over a single TCP
connection.
DNS currently has no connection signaling mechanism. Clients and
servers may close a connection at any time. Clients MUST be prepared
to retry failed queries on broken connections.
Section 4.2.2 of [RFC1035] says: Section 4.2.2 of [RFC1035] says:
If the server needs to close a dormant connection to reclaim o The server should assume that the client will initiate connection
closing, and should delay closing its end of the connection until
all outstanding client requests have been satisfied.
o If the server needs to close a dormant connection to reclaim
resources, it should wait until the connection has been idle for a resources, it should wait until the connection has been idle for a
period on the order of two minutes. In particular, the server period on the order of two minutes. In particular, the server
should allow the SOA and AXFR request sequence (which begins a should allow the SOA and AXFR request sequence (which begins a
refresh operation) to be made on a single connection. Since the refresh operation) to be made on a single connection. Since the
server would be unable to answer queries anyway, a unilateral server would be unable to answer queries anyway, a unilateral
close or reset may be used instead of a graceful close. close or reset may be used instead of graceful close.
Other more modern protocols (e.g., HTTP/1.1 [RFC7230]) have support Other more modern protocols (e.g., HTTP/1.1 [RFC7230]) have support
for persistent TCP connections and operational experience has shown by default for persistent TCP connections for all requests.
that long timeouts can easily cause resource exhaustion and poor Connections are then normally closed via a 'connection close' signal
response under heavy load. Intentionally opening many connections from one party.
and leaving them dormant can trivially create a "denial-of-service"
attack.
It is therefore RECOMMENDED that the default application-level idle The description in [RFC1035] is clear that servers should view
period should be of the order of seconds, but no particular value is connections as persistent (particularly after receiving an SOA), but
specified. In practice, the idle period may vary dynamically, and unfortunately does not provide enough detail for an unambiguous
servers MAY allow dormant connections to remain open for longer interpretation of client behaviour for queries other than a SOA.
periods as resources permit. Additionally, DNS does not yet have a signalling mechanism for
connection timeout or close, although some have been proposed.
To mitigate the risk of unintentional server overload, DNS clients 6.1.1. Clients
MUST take care to minimize the number of concurrent TCP connections
made to any individual server. Similarly, servers MAY impose limits
on the number of concurrent TCP connections being handled for any
particular client. It is RECOMMENDED that for any given client -
server interaction there SHOULD be no more than one connection for
regular queries, one for zone transfers and one for each protocol
that is being used on top of TCP, for example, if the resolver was
using TLS. The server MUST NOT enforce these rules for a particular
client because it does not know if the client IP address belongs to a
single client or is, for example, multiple clients behind NAT.
For reasons of efficiency, implementations SHOULD wherever possible There is no clear guidance today in any RFC as to when a DNS client
attempt to coalesce the two byte length field and subsequent DNS should close a TCP connection, and there are no specific
payload data into a single packet. recommendations with regard to DNS client idle timeouts. However it
is common practice for clients to close the TCP connection after
sending a single request (apart from the SOA/AXFR case).
If a server finds that a client has closed a TCP session, or if the 6.1.2. Servers
session has been otherwise interrupted, before all pending responses
have been sent then the server MUST NOT attempt to send those
responses. Of course the server MAY cache those responses.
7. Query Pipelining Many DNS server implementations use a long fixed idle timeout and
default to a small number of TCP connections. They also offer little
by the way of TCP connection management options. The disadvantages
of this include:
Due to the use of TCP primarily for zone transfer and truncated o Operational experience has shown that long server timeouts can
responses, no existing RFC discusses the idea of pipelining DNS easily cause resource exhaustion and poor response under heavy
queries over a TCP connection. load.
o Intentionally opening many connections and leaving them idle can
trivially create a TCP "denial-of-service" attack as many DNS
servers are poorly equipped to defend against this by modifying
their idle timeouts or other connection management policies.
o A modest number of clients that all concurrently attempt to use
persistent connections with non-zero idle timeouts to such a
server could unintentionally cause the same "denial-of-service"
problem.
Note that this denial-of-service is only on the TCP service.
However, in these cases it affects not only clients wishing to use
TCP for their queries for operational reasons, but all clients who
must fall back to TCP from UDP after receiving a TC=1 flag.
6.2. Recommendations
The following sections include recommendations that are intended to
result in more consistent and scalable implementations of DNS-over-
TCP.
6.2.1. Connection Re-use
One perceived disadvantage to DNS over TCP is the added connection
setup latency, generally equal to one RTT. To amortize connection
setup costs, both clients and servers SHOULD support connection reuse
by sending multiple queries and responses over a single persistent
TCP connection.
When sending multiple queries over a TCP connection clients MUST take
care to avoid Message ID collisions. In other words, they MUST not
re-use the DNS Message ID of an in-flight query. This is especially
important if the server could be performing out-of-order processing
(see Section 7).
6.2.1.1. Query Pipelining
Due to the historical use of TCP primarily for zone transfer and
truncated responses, no existing RFC discusses the idea of pipelining
DNS queries over a TCP connection.
In order to achieve performance on par with UDP DNS clients SHOULD In order to achieve performance on par with UDP DNS clients SHOULD
pipeline their queries. When a DNS client sends multiple queries to pipeline their queries. When a DNS client sends multiple queries to
a server, it should not wait for an outstanding reply before sending a server, it should not wait for an outstanding reply before sending
the next query. Clients should treat TCP and UDP equivalently when the next query. Clients should treat TCP and UDP equivalently when
considering the time at which to send a particular query. considering the time at which to send a particular query.
DNS clients should note that DNS servers that do not both process
pipelined queries concurrently and send out-of-order responses will
likely not provide performance on a par with UDP. IF TCP performance
is of importance, clients may find it useful to use server processing
times as input to server and transport selection algorithms.
DNS servers (especially recursive) SHOULD expect to receive pipelined DNS servers (especially recursive) SHOULD expect to receive pipelined
queries. The server should process TCP queries in parallel, just as queries. The server should process TCP queries concurrently, just as
it would for UDP. The handling of responses to pipelined queries is it would for UDP. The server SHOULD answer all pipelined queries,
covered in the following section. even if they are sent in quick succession. The handling of responses
to pipelined queries is covered in Section 7.
When pipelining queries over TCP it is very easy to send more DNS 6.2.2. Concurrent connections
queries than there are DNS Message ID's. Implementations MUST take
care to check their list of outstanding DNS Message ID's before
sending a new query over an existing TCP connection. This is
especially important if the server could be performing out-of-order
processing. In addition, when sending multiple queries over TCP it
is very easy for a name server to overwhelm its own network
interface. Implementations MUST take care to manage buffer sizes or
to throttle writes to the network interface.
8. Response Reordering To mitigate the risk of unintentional server overload, DNS clients
MUST take care to minimize the number of concurrent TCP connections
made to any individual server. It is RECOMMENDED that for any given
client/server interaction there SHOULD be no more than one connection
for regular queries, one for zone transfers and one for each protocol
that is being used on top of TCP, for example, if the resolver was
using TLS.
Similarly, servers MAY impose limits on the number of concurrent TCP
connections being handled for any particular client. These limits
SHOULD be much looser than the client guidelines above, because the
server does not know if the client IP address belongs to a single
client or is, for example, multiple resolvers on a single machine, or
multiple clients behind NAT.
6.2.3. Idle Timeouts
To mitigate the risk of unintentional server overload, DNS clients
MUST take care to minimize the idle time of DNS-over-TCP sessions
made to any individual server. DNS clients SHOULD close the TCP
connection of an idle session, unless an idle timeout has been
established using some other signalling mechanism.
To mitigate the risk of unintentional server overload it is
RECOMMENDED that the default server application-level idle period be
of the order of seconds, but no particular value is specified. In
practice, the idle period may vary dynamically, and servers MAY allow
idle connections to remain open for longer periods as resources
permit. A timeout of at least a few seconds is advisable for normal
operations to support those clients that expect the SOA and AXFR
request sequence to be made on a single connection as originally
specified in [RFC1035]. Servers MAY use zero timeouts when
experiencing heavy load or are under attack.
6.2.4. Tear Down
Under normal operation clients should initiate connection closing on
idle connections however servers may close the connection if their
local idle timeout policy is exceeded. Connections may be also
closed by either end under unusual conditions such as defending
against an attack or system failure/reboot.
Clients SHOULD retry unanswered queries if the connection closes
before receiving all outstanding responses. No specific retry
algorithm is specified in this document.
If a server finds that a client has closed a TCP session, or if the
session has been otherwise interrupted, before all pending responses
have been sent then the server MUST NOT attempt to send those
responses. Of course the server MAY cache those responses.
7. Response Reordering
RFC 1035 is ambiguous on the question of whether TCP responses may be RFC 1035 is ambiguous on the question of whether TCP responses may be
reordered -- the only relevant text is in Section 4.2.1, which reordered -- the only relevant text is in Section 4.2.1, which
relates to UDP: relates to UDP:
Queries or their responses may be reordered by the network, or by Queries or their responses may be reordered by the network, or by
processing in name servers, so resolvers should not depend on them processing in name servers, so resolvers should not depend on them
being returned in order. being returned in order.
For the avoidance of future doubt, this requirement is clarified. For the avoidance of future doubt, this requirement is clarified.
skipping to change at page 7, line 44 skipping to change at page 10, line 9
support the sending of responses in parallel and/or out-of-order, support the sending of responses in parallel and/or out-of-order,
regardless of the transport protocol in use. Stub and recursive regardless of the transport protocol in use. Stub and recursive
resolvers MUST be able to process responses that arrive in a resolvers MUST be able to process responses that arrive in a
different order to that in which the requests were sent, regardless different order to that in which the requests were sent, regardless
of the transport protocol in use. of the transport protocol in use.
In order to achieve performance on par with UDP, recursive resolvers In order to achieve performance on par with UDP, recursive resolvers
SHOULD process TCP queries in parallel and return individual SHOULD process TCP queries in parallel and return individual
responses as soon as they are available, possibly out-of-order. responses as soon as they are available, possibly out-of-order.
Since responses may arrive out-of-order, clients must take care to Since pipelined responses may arrive out-of-order, clients must take
match responses to outstanding queries, using the ID field, port care to match responses to outstanding queries, using the ID field,
number, query name/type/class, and any other relevant protocol port number, query name/type/class, and any other relevant protocol
features. features. Failure by clients to properly match responses to
outstanding queries can have serious consequences for inter-
operability.
8. TCP Message Length Field
For reasons of efficiency, DNS clients and servers SHOULD transmit
the two-octet length field, and the message described by that length
field, in a single TCP segment. This additionally avoids problems
due to some DNS servers being very sensitive to timeout conditions on
receiving messages (they may abort a TCP session if the first TCP
segment does not contain both the length field and the entire
message).
9. TCP Fast Open 9. TCP Fast Open
This section is non-normative. This section is non-normative.
TCP fastopen [RFC7413] (TFO) allows data to be carried in the SYN TCP fastopen [RFC7413] (TFO) allows data to be carried in the SYN
packet. It also saves up to one RTT compared to standard TCP. packet. It also saves up to one RTT compared to standard TCP.
TFO mitigates the security vulnerabilities inherent in sending data TFO mitigates the security vulnerabilities inherent in sending data
in the SYN, especially on a system like DNS where amplification in the SYN, especially on a system like DNS where amplification
attacks are possible, by use of a server-supplied cookie. TFO attacks are possible, by use of a server-supplied cookie. TFO
clients request a server cookie in the initial SYN packet at the clients request a server cookie in the initial SYN packet at the
start of a new connection. The server returns a cookie in its SYN- start of a new connection. The server returns a cookie in its SYN-
ACK. The client caches the cookie and reuses it when opening ACK. The client caches the cookie and reuses it when opening
subsequent connections to the same server. subsequent connections to the same server.
The cookie is stored by the client's TCP stack (kernel) and persists The cookie is stored by the client's TCP stack (kernel) and persists
if either the client or server processes are restarted. TFO also if either the client or server processes are restarted. TFO also
falls back to a regular TCP handshake gracefully. falls back to a regular TCP handshake gracefully.
Adding support for this to existing name server implementations is
relatively easy, but does require source code modifications. On the
client, the call to connect() is replaced with a TFO aware version of
sendmsg() or sendto(). On the server, TFO must be switched into
server mode by changing the kernel parameter (net.ipv4.tcp_fastopen
on Linux) to enable the server bit (Set the integer value to 2
(server only) or 3 (client and server)) and setting a socket option
between the bind() and listen() calls.
DNS services taking advantage of IP anycast [RFC4786] may need to DNS services taking advantage of IP anycast [RFC4786] may need to
take additional steps when enabling TFO. From [RFC7413]: take additional steps when enabling TFO.From [RFC7413]:
Servers that accept connection requests to the same server IP Servers that accept connection requests to the same server IP
address should use the same key such that they generate identical address should use the same key such that they generate identical
Fast Open Cookies for a particular client IP address. Otherwise a Fast Open Cookies for a particular client IP address. Otherwise a
client may get different cookies across connections; its Fast Open client may get different cookies across connections; its Fast Open
attempts would fall back to regular 3WHS. attempts would fall back to regular 3WHS.
10. IANA Considerations 10. IANA Considerations
This memo includes no request to IANA. This memo includes no request to IANA.
skipping to change at page 9, line 11 skipping to change at page 11, line 23
Some DNS server operators have expressed concern that wider use of Some DNS server operators have expressed concern that wider use of
DNS over TCP will expose them to a higher risk of denial-of-service DNS over TCP will expose them to a higher risk of denial-of-service
(DoS) attacks. (DoS) attacks.
Although there is a higher risk of such attacks against TCP-enabled Although there is a higher risk of such attacks against TCP-enabled
servers, techniques for the mitigation of DoS attacks at the network servers, techniques for the mitigation of DoS attacks at the network
level have improved substantially since DNS was first designed. level have improved substantially since DNS was first designed.
Readers are advised to familiarise themselves with [CPNI-TCP]. Readers are advised to familiarise themselves with [CPNI-TCP].
To mitigate the risk of DoS attacks, DNS servers should engage in TCP
connection management. This may include maintaining state on
existing connections, re-using existing connections and controlling
request queues to enable fair use. It is likely to be advantageous
to provide configurable connection management options, for example:
o total number of TCP connections
o maximum TCP connections per source IP address
o TCP connection idle timeout
o maximum DNS transactions per TCP connection
o maximum TCP connection duration
No specific values are recommended for these parameters.
Operators are advised to familiarise themselves with the
configuration and tuning parameters available in the operating system
TCP stack. However detailed advice on this is outside the scope of
this document.
Operators of recursive servers should ensure that they only accept Operators of recursive servers should ensure that they only accept
connections from expected clients, and do not accept them from connections from expected clients, and do not accept them from
unknown sources. In the case of UDP traffic, this will help protect unknown sources. In the case of UDP traffic, this will help protect
against reflector attacks [RFC5358] and in the case of TCP traffic it against reflector attacks [RFC5358] and in the case of TCP traffic it
will prevent an unknown client from exhausting the server's limits on will prevent an unknown client from exhausting the server's limits on
the number of concurrent connections. the number of concurrent connections.
12. Acknowledgements 12. Acknowledgements
The authors would like to thank Francis Dupont for his detailed The authors would like to thank Francis Dupont and Paul Vixie for
review, Liang Zhu, Zi Hu, and John Heidemann for extensive DNS-over- detailed review, Andrew Sullivan, Tony Finch, Stephane Bortzmeyer and
TCP discussions and code and Lucie Guiraud and Danny McPherson for the many others who contributed to the mailing list discussion. Also
Liang Zhu, Zi Hu, and John Heidemann for extensive DNS-over-TCP
discussions and code. Lucie Guiraud and Danny McPherson for
reviewing early versions of this document. We would also like to reviewing early versions of this document. We would also like to
thank all those who contributed to RFC 5966. thank all those who contributed to RFC 5966.
13. References 13. References
13.1. Normative References 13.1. Normative References
[RFC0768] Postel, J., "User Datagram Protocol", STD 6, RFC 768, [RFC0768] Postel, J., "User Datagram Protocol", STD 6, RFC 768,
August 1980. August 1980.
skipping to change at page 10, line 33 skipping to change at page 13, line 22
for DNS (EDNS(0))", STD 75, RFC 6891, April 2013. for DNS (EDNS(0))", STD 75, RFC 6891, April 2013.
[RFC7230] Fielding, R. and J. Reschke, "Hypertext Transfer Protocol [RFC7230] Fielding, R. and J. Reschke, "Hypertext Transfer Protocol
(HTTP/1.1): Message Syntax and Routing", RFC 7230, June (HTTP/1.1): Message Syntax and Routing", RFC 7230, June
2014. 2014.
13.2. Informative References 13.2. Informative References
[CPNI-TCP] [CPNI-TCP]
CPNI, "Security Assessment of the Transmission Control CPNI, "Security Assessment of the Transmission Control
Protocol (TCP)", 2009, <http://www.cpni.gov.uk/Docs/ Protocol (TCP)", 2009, <http://www.gont.com.ar/papers/
tn-03-09-security-assessment-TCP.pdf>. tn-03-09-security-assessment-TCP.pdf>.
[Connection-Oriented-DNS] [Connection-Oriented-DNS]
Zhu, L., Hu, Z., Heidemann, J., Wessels, D., Mankin, A., Zhu, L., Hu, Z., Heidemann, J., Wessels, D., Mankin, A.,
and N. Somaiya, "T-DNS: Connection-Oriented DNS to Improve and N. Somaiya, "Connection-Oriented DNS to Improve
Privacy and Security (extended)", Privacy and Security",
<http://www.isi.edu/publications/trpublic/files/ <http://www.isi.edu/~johnh/PAPERS/Zhu15b.pdf>.
tr-693.pdf>.
[RFC6824] Ford, A., Raiciu, C., Handley, M., and O. Bonaventure, [RFC6824] Ford, A., Raiciu, C., Handley, M., and O. Bonaventure,
"TCP Extensions for Multipath Operation with Multiple "TCP Extensions for Multipath Operation with Multiple
Addresses", RFC 6824, January 2013. Addresses", RFC 6824, January 2013.
[RFC7413] Cheng, Y., Chu, J., Radhakrishnan, S., and A. Jain, "TCP [RFC7413] Cheng, Y., Chu, J., Radhakrishnan, S., and A. Jain, "TCP
Fast Open", RFC 7413, December 2014. Fast Open", RFC 7413, December 2014.
[RRL] Vixie, P. and V. Schryver, "DNS Response Rate Limiting
(DNS RRL)", ISC-TN 2012-1-Draft1, April 2012.
[fragmentation-considered-poisonous] [fragmentation-considered-poisonous]
Herzberg, A. and H. Shulman, "Fragmentation Considered Herzberg, A. and H. Shulman, "Fragmentation Considered
Poisonous", May 2012, <http://arxiv.org/abs/1205.4011>. Poisonous", May 2012, <http://arxiv.org/abs/1205.4011>.
Appendix A. Summary of Advantages and Disadvantages to using TCP for Appendix A. Summary of Advantages and Disadvantages to using TCP for
DNS DNS
The TCP handshake generally prevents address spoofing and, therefore, The TCP handshake generally prevents address spoofing and, therefore,
the reflection/amplification attacks which plague UDP. the reflection/amplification attacks which plague UDP.
skipping to change at page 11, line 37 skipping to change at page 14, line 28
prepared to re-establish connections or otherwise retry outstanding prepared to re-establish connections or otherwise retry outstanding
queries. It may also possible for TCP Multipath [RFC6824] to allow a queries. It may also possible for TCP Multipath [RFC6824] to allow a
server to hand a connection over from the anycast address to a server to hand a connection over from the anycast address to a
unicast address. unicast address.
There are many "Middleboxes" in use today that interfere with TCP There are many "Middleboxes" in use today that interfere with TCP
over port 53 [RFC5625]. This document does not propose any over port 53 [RFC5625]. This document does not propose any
solutions, other than to make it absolutely clear that TCP is a valid solutions, other than to make it absolutely clear that TCP is a valid
transport for DNS and must be supported by all implementations. transport for DNS and must be supported by all implementations.
Appendix B. Changes -00 to -01 A more in-depth discussion of connection orientated DNS can be found
elsewhere [Connection-Oriented-DNS].
Appendix B. Changes -01 to -02
o Added more text to Introduction as background to TCP use.
o Added definitions of Persistent connection and Idle session to
Terminology section.
o Separated Connection Handling section into Current Practice and
Recommendations. Provide more detail on current practices and
divided Recommendations up into more granular sub-sections.
o Add section on Idle time with new text on recommendations for
client idle behaviour.
o Move TCP message field length discussion to separate section.
o Removed references to system calls in TFO section.
o Added more discussion on DoS mitigation in Security Considerations
section.
o Added statement that servers MAY use 0 idle timeout.
o Re-stated position of TCP as an alternative to UDP in Discussion.
o Updated text on server limits on concurrent connections from a
particular client.
o Added text that client retry logic is outside the scope of this
document.
o Clarified that servers should answer all pipelined queries even if
sent very close together.
Appendix C. Changes -00 to -01
o Changed updates to obsoletes RFC 5966. o Changed updates to obsoletes RFC 5966.
o Improved text in Section 4 Transport Protocol Selection to change o Improved text in Section 4 Transport Protocol Selection to change
"TCP SHOULD NOT be used only for the transfers and as a fallback" "TCP SHOULD NOT be used only for the transfers and as a fallback"
to make the intention clearer and more consistent. to make the intention clearer and more consistent.
o Reference to TCP FASTOPEN updated now that it is an RFC. o Reference to TCP FASTOPEN updated now that it is an RFC.
o Added paragraph to say that implementations MUST NOT send the TCP o Added paragraph to say that implementations MUST NOT send the TCP
skipping to change at page 12, line 13 skipping to change at page 15, line 42
o Added Terminology section. o Added Terminology section.
o Changed should and RECOMMENDED in reference to parallel processing o Changed should and RECOMMENDED in reference to parallel processing
to SHOULD in sections 7 and 8. to SHOULD in sections 7 and 8.
o Added text to address what a server should do when a client closes o Added text to address what a server should do when a client closes
the TCP connection before pending responses are sent. the TCP connection before pending responses are sent.
o Moved the Advantages and Disadvantages section to an appendix. o Moved the Advantages and Disadvantages section to an appendix.
Appendix C. Changes to RFC 5966 Appendix D. Changes to RFC 5966
This document differs from RFC 5966 in four additions: This document differs from RFC 5966 in four additions:
1. DNS implementations are recommended not only to support TCP but 1. DNS implementations are recommended not only to support TCP but
to support it on an equal footing with UDP to support it on an equal footing with UDP
2. DNS implementations are recommended to support reuse of TCP 2. DNS implementations are recommended to support reuse of TCP
connections connections
3. DNS implementations are recommended to support pipelining and out 3. DNS implementations are recommended to support pipelining and out
skipping to change at page 12, line 40 skipping to change at page 16, line 22
John Dickinson John Dickinson
Sinodun Internet Technologies Sinodun Internet Technologies
Magdalen Centre Magdalen Centre
Oxford Science Park Oxford Science Park
Oxford OX4 4GA Oxford OX4 4GA
UK UK
Email: jad@sinodun.com Email: jad@sinodun.com
URI: http://sinodun.com URI: http://sinodun.com
Ray Bellis Sara Dickinson
Nominet Sinodun Internet Technologies
Edmund Halley Road Magdalen Centre
Oxford OX4 4DQ Oxford Science Park
Oxford OX4 4GA
UK UK
Phone: +44 1865 332211 Email: sara@sinodun.com
Email: ray.bellis@nominet.org.uk URI: http://sinodun.com
URI: http://www.nominet.org.uk/
Ray Bellis
Internet Systems Consortium, Inc
950 Charter Street
Redwood City CA 94063
USA
Phone: +1 650 423 1200
Email: ray@isc.org
URI: http://www.isc.org
Allison Mankin Allison Mankin
Verisign Labs Verisign Labs
12061 Bluemont Way 12061 Bluemont Way
Reston, VA 20190 Reston, VA 20190
US US
Phone: +1 703 948-3200 Phone: +1 703 948-3200
Email: amankin@verisign.com Email: amankin@verisign.com
Duane Wessels Duane Wessels
Verisign Labs Verisign Labs
12061 Bluemont Way 12061 Bluemont Way
Reston, VA 20190 Reston, VA 20190
US US
Phone: +1 703 948-3200 Phone: +1 703 948-3200
Email: dwessels@verisign.com Email: dwessels@verisign.com
 End of changes. 41 change blocks. 
119 lines changed or deleted 302 lines changed or added

This html diff was produced by rfcdiff 1.42. The latest version is available from http://tools.ietf.org/tools/rfcdiff/