Network
Address Translation (NAT) Behavioral Requirements UpdatesCisco Systems, Inc.170 West Tasman DriveSan JoseCalifornia95134USArepenno@cisco.comJive CommunicationsCanadasperreault@jive.comOrangeRennes35000Francemohamed.boucadair@orange.comCisco Systems, Inc.United Statesssenthil@cisco.comNTTTokyoJapank.naito@nttv6.jp
Transport Area
TSVWGaddress sharingIPv4 service continuityCarrier Grade NATCGNLSNNAT traversalRFC4787RFC5382RFC5508DS-LiteNAT64Address depletionThis document clarifies and updates several requirements of RFC4787,
RFC5382, and RFC5508 based on operational and development experience.
The focus of this document is NAT44.This document updates RFCs 4787, 5382, and 5508., , and
contributed to enhance Network Address
Translation (NAT) interoperability and conformance. Operational
experience gained through widespread deployment and evolution of NAT
indicates that some areas of the original documents need further
clarification or updates. This document provides such clarifications and
updates.The goal of this document is to clarify and update the set of
requirements listed in , , and . The
document focuses exclusively on NAT44.The scope of this document has been set so that it does not create
new requirements beyond those specified in the documents cited
above.Carrier-Grade NAT (CGN) related requirements are defined in .The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
document are to be interpreted as described in .The reader is assumed to be familiar with the terminology defined
in: ,,, and .In this document, the term "NAT" refers to both "Basic NAT" and
"Network Address/Port Translator (NAPT)" (see Section 3 of ). As a reminder, Basic NAT and NAPT are two
variations of traditional NAT, in that translation in Basic NAT is
limited to IP addresses alone, whereas translation in NAPT is extended
to include IP address and Transport identifier (such as TCP/UDP port
or ICMP query ID) (refer to Section 2 of ). specifies TCP timers associated with
various connection states but does not specify the TCP state machine a
NAT44 should follow as a basis to apply such timers. The TCP state machine depicted in , adapted from ,
SHOULD be implemented by a NAT for TCP session tracking
purposes.The transitory connection idle-timeout is defined as the minimum
time a TCP connection in the partially open or closing phases must
remain idle before the NAT considers the associated session a
candidate for removal (REQ-5 of ). But
does not clearly state whether these
can be configured separately.This document clarifies that a NAT
SHOULD provide different configurable parameters for configuring
the open and closing idle timeouts.To
accommodate deployments that consider a partially open timeout of
4 minutes as being excessive from a security standpoint, a NAT MAY
allow the configured timeout to be less than 4 minutes. However, a
minimum default transitory connection idle-timeout of 4 minutes is
RECOMMENDED. leaves the handling of TCP RST
packets unspecified.This document adopts a similar default
behavior as in . Concretely, when
the NAT receives a TCP RST matching an existing mapping, it MUST
translate the packet according to the NAT mapping entry. Moreover,
the NAT SHOULD wait for 4 minutes before deleting the session and
removing any state associated with it if no packets are received
during that 4 minutes timeout. Notes:Admittedly, the NAT has to verify whether received TCP RST
packets belong to a connection. This verification check is
required to avoid off-path attacks.If the NAT removes immediately the NAT mapping upon receipt
of a TCP RST message, stale connections may be maintained by
endpoints if the first RST message is lost between the NAT and
the recipient.REQ-1 from and REQ-1 from specify a specific port overlapping behavior;
that is the external IP address and port can be reused for connections
originating from the same internal source IP address and port
irrespective of the destination. This is known as endpoint-independent
mapping (EIM). This document clarifies that this port
overlapping behavior may be extended to connections originating from
different internal source IP addresses and ports as long as their
destinations are different. The following
mechanism MAY be implemented by a NAT:If destination addresses and ports are different for outgoing
connections started by local clients, a NAT MAY assign the same
external port as the source ports for the connections. The port
overlapping mechanism manages mappings between external packets
and internal packets by looking at and storing their 5-tuple
(protocol, source address, source port, destination address,
destination port).This enables concurrent use of a
single NAT external port for multiple transport sessions, which
allows a NAT to successfully process packets in an IP address
resource limited network (e.g., deployment with high address space
multiplicative factor (refer to Appendix B.
of )).The "IP address pooling" behavior of "Paired" (APP) was recommended
in REQ-2 from , but the behavior when an
external IPv4 runs out of ports was left undefined.This document clarifies that if APP is
enabled, new sessions from a host that already has a mapping
associated with an external IP that ran out of ports SHOULD be
dropped. A configuration parameter MAY be provided to allow a NAT to
starting using ports from another external IP address when the one
that anchored the APP mapping ran out of ports. Tweaking this
configuration parameter is a trade-off between service continuity
and APP strict enforcement. Note, this behavior is sometimes
referred as 'soft-APP'. As a reminder, the
recommendation for the particular case of a CGN is that an
implementation must use the same external IP address mapping for all
sessions associated with the same internal IP address, be they TCP,
UDP, ICMP, something else, or a mix of different protocols .This behavior SHOULD apply also for TCP.REQ-1 from and REQ-1 from do not specify whether EIM are
protocol-dependent or protocol-independent. For example, if an outbound
TCP SYN creates a mapping, it is left undefined whether outbound UDP
packets can reuse such mapping.EIM mappings SHOULD be protocol-dependent. A
configuration parameter MAY be provided to allow protocols that
multiplex TCP and UDP over the same source IP address and port
number to use a single mapping. The default value of this
configuration parameter MUST be protocol-dependent EIM.This update is consistent with the stateful NAT64
that clearly specifies three binding
information bases (TCP, UDP, ICMP).REQ-8 from and REQ-3 from do not specify whether mappings with
endpoint-independent filtering (EIF) are protocol-independent or
protocol-dependent. For example, if an outbound TCP SYN creates a
mapping, it is left undefined whether inbound UDP packets matching that
mapping should be accepted or rejected.EIF filtering SHOULD be protocol-dependent. A
configuration parameter MAY be provided to make it
protocol-independent. The default value of this configuration
parameter MUST be protocol-dependent EIF.This behavior is aligned with the update in . Applications that can
be transported over a variety of transport protocols and/or support
transport fall back schemes won't experience connectivity failures
if the NAT is configured with protocol-independent EIM and
protocol-independent EIF.The NAT mapping Refresh direction may have a "NAT Inbound refresh
behavior" of "True" according to REQ-6 from , but does not
clarify how this behavior applies to EIF mappings. The issue in question
is whether inbound packets that match an EIF mapping but do not create a
new session due to a security policy should refresh the mapping
timer.This document clarifies that even when
a NAT has an inbound refresh behavior set to 'TRUE', such packets
SHOULD NOT refresh the mapping. Otherwise a simple attack of a
packet every 2 minutes can keep the mapping indefinitely.This behavior SHOULD apply also for TCP.In the case of NAT outbound refresh
behavior, ICMP Errors or TCP RST outbound packets, sent as
response to inbound packets, SHOULD NOT refresh the mapping. Other
packets which indicate the host is not interested in receiving
packets MAY be configurable to also not refresh state, such as
STUN error response or IKE
INVALID_SYNTAX .A NAT MAY disable port parity preservation for
all dynamic mappings. Nevertheless, A NAT SHOULD support means to
explicitly request to preserve port parity (e.g., ).Note: According to , dynamic
mappings are said to be dynamic in the sense that they are created
on demand, either implicitly or explicitly:Implicit dynamic mappings refer to mappings that are created
as a side effect of traffic such as an outgoing TCP SYN or
outgoing UDP packet. Implicit dynamic mappings usually have a
finite lifetime, though this lifetime is generally not known to
the client using them.Explicit dynamic mappings refer to mappings that are created
as a result, for example, of explicit Port Control Protocol
(PCP) MAP and PEER requests. Explicit dynamic mappings have a
finite lifetime, and this lifetime is communicated to the
client.A NAT SHOULD follow the recommendations
specified in Section 4 of ,
especially: "A NAPT that does not implement port preservation [RFC4787]
[RFC5382] SHOULD obfuscate selection of the ephemeral port of a
packet when it is changed during translation of that packet. A
NAPT that does implement port preservation SHOULD obfuscate the
ephemeral port of a packet only if the port must be changed as a
result of the port being already in use for some other session.
A NAPT that performs parity preservation and that must change
the ephemeral port during translation of a packet SHOULD
obfuscate the ephemeral ports. The algorithms described in this
document could be easily adapted such that the parity is
preserved (i.e., force the lowest order bit of the resulting
port number to 0 or 1 according to whether even or odd parity is
desired)."A NAT SHOULD handle the Identification field
of translated IPv4 packets as specified in Section 5.3.1 of .Section 3.1 of specifies that ICMP
Query Mappings are to be maintained by a NAT. However, the specification
doesn't discuss Query Mapping timeout values. Section 3.2 of only discusses ICMP Query Session Timeouts.ICMP Query Mappings MAY be deleted once the
last session using the mapping is deleted.REQ-7 from specifies that a NAT
enforcing 'Basic NAT' must support traversal of hairpinned ICMP Query
sessions.This implicitly means that address
mappings from external address to internal address (similar to
Endpoint Independent Filters) must be maintained to allow inbound
ICMP Query sessions. If an ICMP Query is received on an external
address, a NAT can then translate to an internal IP.REQ-7 from specifies that all
NATs must support the traversal of hairpinned ICMP Error messages.This behavior requires a NAT to
maintain address mappings from external IP address to internal IP
address in addition to the ICMP Query Mappings described in Section
3.1 of .This document does not require any IANA action.NAT behavioral considerations are discussed in , , and .Because some of the clarifications and updates (e.g., ) are inspired from NAT64, the security
considerations discussed in Section 5 of
apply also for this specification.The update in allows for an
optimized NAT resource usage. In order to avoid service disruption, the
NAT must not invoke this functionality unless the packets are to be sent
to distinct destination addresses.Some of the updates (e.g., , , and ) allow
for an increased security compared to ,
, and .
Particularly:The updates in and prevent an illegitimate node to maintain
mappings activated in the NAT while these mappings should be
cleared.Port randomization ()
complicates tracking hosts located behind a NAT. and
propose updates that increase the serviceability of a host located
behind a NAT. These updates do not introduce any additional security
concerns to , , and .The updates in and allow for a better NAT transparency from an
application standpoint. Hosts that require a restricted filtering
behavior should enable specific policies (e.g., access control list
(ACL)) either locally or by soliciting a dedicated security device
(e.g., firewall). How a host updates its filtering policies is out of
scope of this document.The update in induces security concerns
that are specific to the protocol used to interact with the NAT. For
example, if PCP is used to explicitly request parity preservation for a
given mapping, the security considerations discussed in should be taken into account.The update in may have undesired effects on
the performance of the NAT in environments in which fragmentation is
massively experienced. Such issue may be used as an attack vector
against NATs.Thanks to Dan Wing, Suresh Kumar, Mayuresh Bakshi, Rajesh Mohan, Lars
Eggert, Gorry Fairhurst, Brandon Williams, and David Black for their
review and discussion. Many thanks to Ben
Laurie for the secdir review, and Dan Romascanu for the Gen-ART review.
Dan Wing proposed some text for the
configurable errors in .The following individual contributed text to the document: Sarat Kamiset, Insieme Networks, United States