Network Working Group P. Srisuresh
Request for Comments: 2694 Consultant
Category: Informational G. Tsirtsis
BT Laboratories
P. Akkiraju
Cisco Systems
A. Heffernan
Juniper Networks
September 1999
DNS extensions to Network Address Translators (DNS_ALG)
Status of this Memo
This memo provides information for the Internet community. It does
not specify an Internet standard of any kind. Distribution of this
memo is unlimited.
Copyright Notice
Copyright (C) The Internet Society (1999). All Rights Reserved.
Abstract
Domain Name Service (DNS) provides name to address mapping within a
routing class (ex: IP). Network Address Translators (NATs) attempt to
provide transparent routing between hosts in disparate address realms
of the same routing class. Typically, NATs exist at the border of a
stub domain, hiding private addresses from external addresses. This
document identifies the need for DNS extensions to NATs and outlines
how a DNS Application Level Gateway (DNS_ALG) can meet the need.
DNS_ALG modifies payload transparently to alter address mapping of
hosts as DNS packets cross one address realm into another. The
document also illustrates the operation of DNS_ALG with specific
examples.
1. IntrodUCtion
Network Address Translators (NATs) are often used when network's
internal IP addresses cannot be used outside the network either for
privacy reasons or because they are invalid for use outside the
network.
Ideally speaking, a host name uniquely identifies a host and its
address is used to locate routes to the host. However, host name and
address are often not distinguished and used interchangeably by
applications. Applications embed IP address instead of host name in
payload. Examples would be e-mails that specify their MX server
address (ex: user@666.42.7.11) instead of server name (ex:
user@private.com) as sender ID; Html files that include IP address
instead of names in URLs, etc. Use of IP address in place of host
name in payload represents a problem as the packet traverses a NAT
device because NATs alter network and transport headers to suit an
address realm, but not payload.
DNS provides Name to address mapping. Whereas, NAT performs address
translation (in network and transport headers) in datagrams
traversing between private and external address realms. DNS
Application Level Gateway (DNS_ALG) outlined in this document helps
translate Name-to-Private-Address mapping in DNS payloads into Name-
to-external-address mapping and vice versa using state information
available on NAT.
A Network Address Port Translator (NAPT) performs address and
Transport level port translations (i.e, TCP, UDP ports and ICMP query
IDs). DNS name mapping granularity, however, is limited to IP
addresses and does not extend to transport level identifiers. As a
result, the DNS_ALG processing for an NAPT configuration is
simplified in that all host addresses in private network are bound to
a single external address. The DNS name lookup for private hosts
(from external hosts) do not mandate fresh private-external address
binding, as all private hosts are bound to a single pre-defined
external address. However, reverse name lookups for the NAPT external
address will not map to any of the private hosts and will simply map
to the NAPT router. Suffices to say, the processing requirements for
a DNS_ALG supporting NAPT configuration are a mere subset of Basic
NAT. Hence, the discussion in the remainder of the document will
focus mainly on Basic NAT, Bi-directional NAT and Twice NAT
configurations, with no specific reference to NAPT setup.
Definitions for DNS and related terms may be found in [Ref 3] and
[Ref 4]. Definitions for NAT related terms may be found in [Ref 1].
2. Requirement for DNS extensions
There are many ways to ensure that a host name is mapped to an
address relevant within an address realm. In the following sections,
we will identify where DNS extensions would be needed.
Typically, organizations have two types of authoritative name
servers. Internal authoritative name servers identify all (or
majority of) corporate resources within the organization. Only a
portion of these hosts are allowed to be Accessed by the external
world. The remaining hosts and their names are unique to the private
network. Hosts visible to the external world and the authoritative
name server that maps their names to network addresses are often
configured within a DMZ (De-Militarized Zone) in front of a firewall.
We will refer the hosts and name servers within DMZ as DMZ hosts and
DMZ name servers respectively. DMZ host names are end-to-end unique
in that their FQDNs do not overlap with any end node that
communicates with it.
\ /
+-----------------------+
Service Provider Router
+-----------------------+
WAN
Stub A .........\....
+-----------------+
Stub Router w/NAT
+-----------------+
DMZ - Network
------------------------------------------------------------
+--+ +--+ +--+ +--+ +----------+
__ __ __ __ Firewall
/____\ /____\ /____\ /____\ +----------+
DMZ-Host1 DMZ-Host2 ... DMZ-Name DMZ-Web
Server Server etc.
Internal hosts (Private IP network)
------------------------------------------------------------
+--+ +--+ +--+ +--+
__ __ __ __
/____\ /____\ /____\ /____ Int-Host1 Int-Host2 ..... Int-Hostn Int-Name Server
Figure 1: DMZ network configuration of a private Network.
Figure 1 above illustrates configuration of a private network which
includes a DMZ. Actual configurations may vary. Internal name servers
are accessed by users within the private network only. Internal DNS
queries and responses do not cross the private network boundary. DMZ
name servers and DMZ hosts on the other hand are end-to-end unique
and could be accessed by external as well as internal hosts.
Throughout this document, our focus will be limited to DMZ hosts and
DMZ name servers and will not include internal hosts and internal
name servers, unless they happen to be same.
2.1. DMZ hosts assigned static external addresses on NAT
Take the case where DMZ hosts are assigned static external addresses
on the NAT device. Note, all hosts within private domain, including
the DMZ hosts are identified by their private addresses. Static
mapping on the NAT device allows the DMZ hosts to be identified by
their public addresses in the external domain.
2.1.1. Private networks with no DMZ name servers
Take the case where a private network has no DMZ name server for
itself. If the private network is connected to a single service
provider for external connectivity, the DMZ hosts may be listed by
their external addresses in the authoritative name servers of the
service provider within their forward and in-add.arpa reverse zones.
If the network is connected to multiple service providers, the DMZ
host names may be listed by their external address(es) within the
authoritative name servers of each of the service providers. This is
particularly significant in the case of in-addr.arpa reverse zones,
as the private network may be assigned different address prefixes by
the service providers.
In both cases, externally generated DNS lookups will not reach the
private network. A large number of NAT based private domains pursue
this option to have their DMZ hosts listed by their external
addresses on service provider's name servers.
2.1.2. Private networks with DMZ name servers
Take the case where a private network opts to keep an authoritative
DMZ name server for the zone within the network itself. If the
network is connected to a single service provider, the DMZ name
server may be configured to obviate DNS payload interceptions as
follows. The hosts in DMZ name server must be mapped to their
statically assigned external addresses and the internal name server
must be configured to bypass the DMZ name server for queries
concerning external hosts. This scheme ensures that DMZ name servers
are set for exclusive access to external hosts alone (not even to the
DMZ hosts) and hence can be configured with external addresses only.
The above scheme requires careful administrative planning to ensure
that DMZ name servers are not contacted by the private hosts directly
or indirectly (through the internal name servers). Using DNS-ALG
would obviate the administrative ordeals with this approach.
2.2. DMZ hosts assigned external addresses dynamically on NAT
Take the case where DMZ hosts in a private network are assigned
external addresses dynamically by NAT. While the addresses issued to
these hosts are fixed within the private network, their externally
known addresses are ephemeral, as determined by NAT. In such a
scenario, it is mandatory for the private organization to have a DMZ
name server in order to allow access to DMZ hosts by their name.
The DMZ name server would be configured with private addresses for
DMZ hosts. DNS Application Level Gateway (DNS_ALG) residing on NAT
device will intercept the DNS packets directed to or from the DMZ
name server(s) and perform transparent payload translations so that a
DMZ host name has the right address mapping within each address realm
(i.e., private or external).
3. Interactions between NAT and DNS_ALG
This document operates on the paradigm that interconnecting address
realms may have overlapping address space. But, names of hosts within
interconnected realms must be end-to-end unique in order for them to
be accessed by all hosts. In other Words, there cannot be an overlap
of FQDNs between end nodes communicating with each other. The
following diagram illustrates how a DNS packet traversing a NAT
device (with DNS_ALG) is subject to header and payload translations.
A DNS packet can be a TCP or UDP packet with the source or
destination port set to 53. NAT would translate the IP and TCP/UDP
headers of the DNS packet and notify DNS-ALG to perform DNS payload
changes. DNS-ALG would interact with NAT and use NAT state
information to modify payload, as necessary.
Original-IP
packet
vv
+---------------------------------+ +-----------------------+
DNS Appl. Level Gateway
Network Address Translation (NAT)---> (DNS_ALG)
*IP & Transport header mods <--- *DNS payload mods
+---------------------------------+ +-----------------------+
vv
Translated-IP
packet
Figure 2: NAT & DNS-ALG in the translation path of DNS packets
3.1. Address Binding considerations
We will make a distinction between "Temporary Address Binding" and
"Committed Address Binding" in NATs. This distinction becomes
necessary because the DNS_ALG will allow external users to create
state on NAT, and thus the potential for denial-of-service attacks.
Temporary address binding is the phase in which an address binding is
reserved without any NAT sessions using the binding. Committed
address binding is the phase in which there exists at least one NAT
session using the binding between the external and private addresses.
Both types of bindings are used by DNS_ALG to modify DNS payloads.
NAT uses only the committed address bindings to modify the IP and
Transport headers of datagrams pertaining to NAT sessions.
For statically mapped addresses, the above distinction is not
relevant. For dynamically mapped addresses, temporary address binding
often precedes committed binding. Temporary binding occurs when DMZ
name server is queried for a name lookup. Name query is likely a
pre-cursor to a real session between query originator and the queried
host. The temporary binding becomes committed only when NAT sees the
first packet of a session between query initiator and queried host.
A configurable parameter, "Bind-holdout time" may be defined for
dynamic address assignments as the maximum period of time for which a
temporary address binding is held active without transitioning into a
committed binding. With each use of temporary binding by DNS_ALG (to
modify DNS payload), this Bind-holdout period is renewed. A default
Bind-holdout time of a couple of minutes might suffice for most DNS-
ALG implementations. Note, it is possible for a committed address
binding to occur without ever having to be preceded by a temporary
binding. Lastly, when NAT is ready to unbind a committed address
binding, the binding is transitioned into a temporary binding and
kept in that phase for an additional Bind-holdout period. The binding
is freed only upon eXPiry of Bind-holdout time. The Bind-holdout time
preceding the committed-address-binding and the address-unbinding are
required to ensure that end hosts have sufficient time in which to
initiate a data session subsequent to a name lookup.
For example, say a private network with address prefix 10/8 is mapped
to 198.76.29/24. When an external hosts makes a DNS query to host7,
bearing address 10.0.0.7, the DMZ name server within private network
responds with an A type RR for host7 as:
host7 A 10.0.0.7
DNS_ALG would intercept the response packet and if 10.0.0.7 is not
assigned an external address already, it would request NAT to create
a temporary address binding with an external address and start Bind-
holdout timer to age the binding. Say, the assigned external address
is 198.76.29.1. DNS-ALG would use this temporary binding to modify
the RR in DNS response, replacing 10.0.0.7 with its external address
and reply with:
host7 A 198.76.29.1
When query initiator receives DNS response, only the assigned
external address is seen. Within a short period (presumably before
the bind-holdout time expires), the query initiator would initiate a
session with host7. When NAT notices the start of new session
directed to 198.76.29.1, NAT would terminate Bind-holdout timer and
transition the temporary binding between 198.76.29.1 and 10.0.0.7
into a committed binding.
To minimize denial of service attacks, where a malicious user keeps
attempting name resolutions, without ever initiating a connection,
NAT would have to monitor temporary address bindings that have not
transitioned into committed bindings. There could be a limit on the
number of temporary bindings and attempts to generate additional
temporary bindings could be simply rejected. There may be other
heuristic solutions to counter this type of malicious attacks.
We will consider bi-directional NAT to illustrate the use of
temporary binding by DNS_ALG in the following sub-sections, even
though the concept is applicable to other flavors of NATs as well.
3.2. Incoming queries
In order to initiate incoming sessions, an external host obtains the
V4 address of the DMZ-host it is trying to connect to by making a DNS
request. This request constitutes prelude to the start of a
potential new session.
The external host resolver makes a name lookup for the DMZ host
through its DNS server. When the DNS server does not have a record
of IPv4 address attached to this name, the lookup query is redirected
at some point to the Primary/Backup DNS server (i.e., in DMZ) of the
private stub domain.
Enroute to DMZ name server, DNS_ALG would intercept the datagram and
modify the query as follows.
a) For Host name to Host address query requests:
Make no change to the DNS payload.
b) For Host address to Host name queries: Replace the external V4
address octets (in reverse order) preceding the string "IN-
ADDR.ARPA" with the corresponding private V4 address, if such
an address binding exists already. However, if a binding does
not exist, the DNS_ALG would simply respond (as a name server
would) with a response code (RCODE) of 5 (REFUSED to respond
due to policy reasons) and set ANCOUNT, NSCOUNT and ARCOUT to 0
in the header section of the response.
In the opposite direction, as DNS response traverses from the DNS
server in private network, DNS_ALG would once again intercept the
packet and modify as follows.
a) For a host name to host address query requests, replace the
private address sent by DMZ name server with a public address
internally assigned by the NAT router. If a public address is
not previously assigned to the host's private address, NAT
would assign one at this time.
b) For host address to host name queries, replace the private
address octets preceding the string "IN-ADDR.ARPA" in response
RRs with their external address assignments. There is a chance
here that by the time the DMZ name server replies, the bind-
holdout timer in NAT for the address in question has expired.
In such a case, DNS_ALG would simply drop the reply. The sender
will have to resend the query (as would happen when a router
enroute drops the response).
For static address assignments, the TTL value supplied in the
original RR will be left unchanged. For dynamic address assignments,
DNS_ALG would modify the TTL value on DNS resource records (RRs) to
be 0, implying that the RRs should only be used for transaction in
progress, and not be cached. For compatibility with broken
implementations, TTL of 1 might in practice work better.
Clearly, setting TTL to be 0 will create more traffic than if the
addresses were static, because name-to-address mapping is not cached.
Specifically, network based applications will be required to use
names rather than addresses for identifying peer nodes and must use
DNS for every name resolution, as name-to-address mapping cannot be
shared from the previously run applications.
In addition, NAT would be requested to initiate a bind-holdout timer
following the assignment. If no session is initiated to the private
host within the Bind-holdout time period, NAT would terminate the
temporary binding.
3.3. Outgoing Queries
For Basic and bi-directional NATs, there is no need to distinguish
between temporary and committed bindings for outgoing queries. This
is because, DNS_ALG does not modify the DNS packets directed to or
from external name servers (used during outbound sessions), unlike
the inbound DNS sessions.
Say, a private host needs to communicate with an external host. The
DNS query goes to the internal name server (if there exists one)
and from there to the appropriate authoritative/cache name server
outside the private domain. The reply follows the same route but
neither the query nor the response are subject to DNS_ALG
translations.
This however will not be the case with address isolated twice NAT
private and external domains. In such a case, NAT would intercept all
DNS packets and make address modifications to payload as discussed in
the previous section. Temporary Private to external address bindings
are created when responses are sent by private DNS servers and
temporary external to private address bindings are created when
responses are sent by external DNS servers.
4. DNS payload modifications by DNS-ALG
Typically, UDP is employed as the transport mechanism for DNS queries
and responses and TCP for Zone refresh activities. In either case,
name servers are accessed using a well-known DNS server port 53
(decimal) and all DNS payloads have the following format of data [Ref
4]. While NAT is responsible for the translation of IP and TCP/UDP
headers of a DNS packet, DNS-ALG is responsible for updating the DNS
payload.
The header section within the DNS payload is always present and
includes fields specifying which of the remaining sections are
present. The header identifies if the message is a query or a
response. No changes are required to be made by DNS-ALG to the Header
section. DNS_ALG would parse only the DNS payloads whose QCLASS is
set to IN (IP class).
+---------------------+
Header
+---------------------+
Question the question for the name server
+---------------------+
Answer RRs answering the question
+---------------------+
Authority RRs pointing toward an authority
+---------------------+
Additional RRs holding additional information
+---------------------+
4.1. Question section
The question section contains QDCOUNT (usually 1) entries, as
specified in Header section, with each of the entries in the
following format:
1 1 1 1 1 1
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5
+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+
/ QNAME /
/ /
+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+
QTYPE
+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+
QCLASS
+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+
4.1.1. PTR type Queries
DNS_ALG must identify all names, whose FQDNs (i.e., Fully Qualified
Domain Names) fall within IN-ADDR.ARPA domain and replace the address
octets (in reverse order) preceding the string "IN-ADDR.ARPA" with
the corresponding assigned address octets in reverse order, only if
the address binding is active on the NAT router. If the address
preceding the string "IN-ADDR.ARPA" falls within the NAT address map,
but does not have at least a temporary address binding, DNS_ALG would
simply simply respond back (as a DNS name server would) with a
response code (RCODE) of 5 (REFUSED to respond due to policy reasons)
and set ANCOUNT, NSCOUNT and ARCOUT to 0 in the header section of the
response.
Note that the above form of host address to host name type queries
will likely yield different results at different times, depending
upon address bind status in NAT at a given time.
For example, a resolver that wanted to find out the hostname
corresponding to address 198.76.29.1 (externally) would pursue a
query of the form:
QTYPE = PTR, QCLASS = IN, QNAME = 1.29.76.198.IN-ADDR.ARPA.
DNS_ALG would intervene and if the address 198.76.29.1 is internally
mapped to a private address of 10.0.0.1, modify the query as below
and forward to DMZ name server within private network.
QTYPE = PTR, QCLASS = IN, QNAME = 1.0.0.10.IN-ADDR.ARPA
Presumably, the DMZ name server is the authoritative name server for
10.IN-ADDR.ARPA zone and will respond with an RR of the following
form in answer section. DNS_ALG translations of the response RRs will
be considered in a following section.
1.0.0.10.IN-ADDR.ARPA PTR host1.fooboo_org.provider_domain
An example of Inverse translation is e-mail programs using inverse
translation to trace e-mail originating hosts for spam prevention.
Verify if the address from which the e-mail was sent does indeed
belong to the same domain name the sender claims in sender ID.
Query modifications of this nature will likely change the length of
DNS payload. As a result, the corresponding IP and TCP/UDP header
checksums must be updated. In case of TCP based queries, the sequence
number deltas must be tracked by NAT so that the delta can be applied
to subsequent sequence numbers in datagrams in the same direction and
acknowledgement numbers in datagrams in the opposite direction. In
case of UDP based queries, message sizes are restricted to 512 bytes
(not counting the IP or UDP headers). Longer messages must be
truncated and the TC bit should be set in the header.
Lastly, any compressed domain names using pointers to represent
common domain denominations must be updated to reflect new pointers
with the right offset, if the original domain name had to be
translated by NAT.
4.1.2. A, MX, NS and SOA type Queries
For these queries, DNS_ALG would not modify any of the fields in the
query section, not even the name field.
4.1.3. AXFR type Queries
AXFR is a special zone transfer type query. Zone transfers from
private address realm must be avoided for address assignments that
are not static. Typically, TCP is used for AXFR requests.
When changes are made to a zone, they must be distributed to all name
servers. The general model of automatic zone transfer or refreshing
is that one of the name servers is the master or primary for the
zone. Changes are coordinated at the primary, typically by editing a
master file for the zone. After editing, the administrator signals
the master server to load the new zone. The other non-master or
secondary servers for the zone periodically check the SERIAL field of
the SOA for the zone for changes (at some polling intervals) and
obtain new zone copies when changes have been made.
Zone transfer is usually from primary to backup name servers. In the
case of NAT supported private networks, primary and backup servers
are advised to be located in the same private domain (say,
private.zone) so zone transfer is not across the domain and DNS_ALG
support for zone transfer is not an issue. In the case a secondary
name server is located outside the private domain, zone transfers
must not be permitted for non-static address assignments. Primary and
secondary servers are required to be within the same private domain
because all references to data in the zone had to be captured. With
dynamic address assignments and bindings, it is impossible to
translate the axfr data to be up-to-date. Hence, if a secondary
server for private.zone were to be located external to the domain, it
would contain bad data. Note, however, the requirement outlined here
is not in confirmence with RFC2182, which recommends primary and
secondary servers to be placed at topologically and geographically
dispersed locations on the Internet.
During zone transfers, DNS_ALG must examine all A type records and
replace the original address octets with their statically assigned
address octets. DNS_ALG could also examine if there is an attempt to
transfer records for hosts that are not assigned static addresses and
drop those records alone or drop the whole transfer. This would
minimize misconfiguration and human errors.
4.1.4. Dynamic Updates to the DNS.
An authoritative name server can have dynamic updates from the nodes
within the zone without intervention from NAT and DNS-ALG, so long as
one avoids spreading a DNS zone across address realms. We recommend
keeping a DNS zone within the same realm it is responsible for. By
doing this, DNS update traffic will not cross address realms and
hence will not be subject to consideration by DNS-ALG.
Further, if dynamic updates do cross address realms, and the updates
must always be secured via DNSSEC, then such updates are clearly out
of scope for DNS-ALG (as described in section 7).
4.2. Resource records in all other sections
The answer, authority, and additional sections all share the same
format, with a variable number of resource records. The number of RRs
specific to each of the sections may be found in the corresponding
count fields in DNS header. Each resource record has the following
format:
The TTL value supplied in the original RRs will be left unchanged for
static address assignments. For dynamic address assignments, DNS_ALG
will modify the TTL to be 0, so the RRs are used just for the
transaction in progress, and not cached. RFC2181 requires all RRs
in an RRset (RRs with the same name, class and type, but with
different RDATA) to have the same TTL. So if the TTL of an RR is set
to 0, all other RRs within the same RRset will also be adjusted by
the DNS-ALG to be 0.
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5
+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+
/ /
/ NAME /
+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+
TYPE
+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+
CLASS
+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+
TTL
+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+
RDLENGTH
+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--
/ RDATA /
/ /
+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+
4.2.1. PTR type RRs
The considerations specified in the Question section is equally valid
with names for PTR type RRs. Private address preceding the string
"IN-ADDR.ARPA" (in reverse order of octets) must be replaced by its
external address assignment (in reverse order), if a binding exists.
The remaining fields for this RR remain unchanged.
4.2.2. A type RRs
The RDATA for A records is a 4-byte IP address. DNS_ALG would simply
replace the original address in RDATA with its externally assigned IP
address, if it succeeded in finding an address binding. Successful
address translation should cause no changes to payload length. Only
the transport header checksum would need updating. In case of failure
to find an address binding, DNS_ALG would have to drop the record and
decrement the corresponding COUNT field in the header section.
4.2.3. CNAME, MX, NS and SOA type RRs
No changes required to be made by DNS_ALG for these RRs, as the RDATA
does not contain any IP addresses. The host names within the RDATA
remain unchanged between realms.
5. Illustration of DNS_ALG in conjunction with Bi-directional NAT
The following diagram illustrates the operation of DNS_ALG in a a
bi-directional NAT router. We will illustrate by walking through how
name lookup and reverse name lookup queries are processed.
.
________________ . External.com
( ) .
( ) . +-------------+
+--+ ( Internet )-.---Border Router
__------ ( ) . +-------------+
/____\ (________________) .
Root .
DNS Server . ---------------
+---------------+ .
Provider Router . +--+ +--+
+---------------+ . __ __
. /____\ /____ . DNS Server Host X
External domain . 171.68.1.1 171.68.10.1
...........................................................
Private domain
Private.com
+--------------------------------------+
Bi-Directional NAT router with DNS_ALG
Private addresses: 172.19/16
External addresses: 131.108.1/24
+--------------------------------------+
---------- ----------
DNS Server
+--+ +--+ Authoritative
__ __ for private.com
/____\ /____ Host A DNS Server
172.19.1.10 172.19.2.1
(Mapped to 131.108.1.8)
Figure 3: DNS-ALG operation in Bi-Directional NAT setup
The above diagram depicts a scenario where a company private.com
using private address space 172.19/16 connects to the Internet using
bi-directional NAT. DNS_ALG is embedded in the NAT device to make
necessary DNS payload changes. NAT is configured to translate the
private addresses space into an external address block of
131.108.1/24. NAT is also configured with a static translation for
private.com's DNS server, so it can be referred in the external
domain by a valid address.
The company external.com is located in the external domain, using a
registered address block of 171.68/16. Also shown in the topology is
a root DNS server.
Following simplifications are made to the above configuration:
* private.com is not multihomed and all traffic to the external
space transits a single NAT.
* The DNS server for private.com is authoritative for the
private.com domain and points to the root server for all other
DNS resolutions. The same is true for the DNS server in
external.com.
* The internal name servers for private.com and external.com are
same as their DMZ name servers. The DNS servers for these
domains are configured with addresses private to the
organization.
* The name resolvers on host nodes do not have recursion
available on them and desire recursive service from servers.
All name servers are assumed to be able to provide recursive
service.
5.1. Outgoing Name-lookup queries
Say, Host A in private.com needs to perform a name lookup for host X
in external.com to initiate a session with X. This would proceed as
follows.
1. Host A sends a UDP based name lookup query (A record) for
"X.External.Com" to its local DNS server.
2. Local DNS server sends the query to the root server enroute NAT.
NAT would change the IP and UDP headers to reflect DNS server's
statically assigned external address. DNS_ALG will make no
changes to the payload.
3. The root server, in turn, refers the local DNS server to query the
DNS server for External.com. This referal transits the NAT enroute
to the local DNS server. NAT would simply translate the IP and
UDP headers of the incoming packet to reflect DNS server's private
address. No changes to the payload by DNS_ALG.
4. Private.com DNS server will now send the query to the DNS server
for external.com, once again, enroute NAT. Just as with the query
to root, The NAT router would change the IP and UDP headers to
reflect the DNS server's statically assigned external address.
And, DNS_ALG will make no changes to the payload.
5. The DNS server for external.com replies with the IP address
171.68.10.1. This reply also transits the NAT. NAT would
translate the IP and UDP headers of the incoming packet to reflect
DNS server's private address. Once again, no changes to the
payload by DNS_ALG.
6. The DNS server in Private.com replies to host A.
When Host A finds the address of Host X, A initiates a session with
host X, using a destination IP address of 171.68.10.1. This datagram
and any others that follow in this session will be translated as
usual by NAT.
Note, DNS_ALG does not change the payload for DNS packets in either
direction.
5.2. Reverse name lookups originated from private domain
This scenario builds on the previous case by having host A in
Private.com perform a reverse name lookup on 171.68.10.1, which is
host X's global address. Following is a sequence of events.
1. Host A sends a UDP based inverse name lookup query (PTR record)
for "1.10.68.171.IN-ADDR.ARPA." to its local DNS server.
2. Local DNS server sends the query to the root server enroute NAT.
As before, NAT would change the IP and UDP headers to reflect DNS
server's statically assigned external address. DNS_ALG will make
no changes to the payload.
3. The root server, in turn, refers the local DNS server to query the
DNS server for External.com. This referal transits the NAT enroute
to the local DNS server. NAT would simply translate the IP and
UDP headers of the incoming packet to reflect DNS server's private
address. No changes to the payload by DNS_ALG.
4. Private.com DNS server will now send the query to the DNS server
for external.com, once again, enroute NAT. Just as with the query
to root, The NAT router would change the IP and UDP headers to
reflect the DNS server's statically assigned external address.
And, DNS_ALG will make no changes to the payload.
5. The DNS server for external.com replies with the host name of
"X.External.Com.". This reply also transits the NAT. NAT would
translate the IP and UDP headers of the incoming packet to reflect
DNS server's private address. Once again, no changes to the
payload by DNS_ALG.
6. The DNS server in Private.com replies to host A.
Note, DNS_ALG does not change the payload in either direction.
5.3. Incoming Name-lookup queries
This time, host X in external.com wishes to initiate a session with
host A in Private.com. Below are the sequence of events that take
place.
1. Host X sends a UDP based name lookup query (A record) for
"A.Private.com" to its local DNS server.
2. Local DNS server in External.com sends the query to root server.
3. The root server, in turn, refers the DNS server in External.com to
query the DNS server for private.com,
4. External.com DNS server will now send the query to the DNS server
for Private.com. This query traverses the NAT router. NAT would
change the IP and UDP headers of the packet to reflect the DNS
server's private address. DNS_ALG will make no changes to the
payload.
5. The DNS server for Private.com replies with the IP address
172.19.1.10 for host A. This reply also transits the NAT. NAT
would translate the IP and UDP headers of the outgoing packet from
the DNS server.
DNS_ALG will request NAT to (a) setup a temporary binding for Host
A (172.19.1.10) with an external address and (b) initiate Bind-
holdout timer. When NAT successfully sets up a temporary binding
with an external address (say, 131.108.1.12), DNS_ALG would modify
the payload to replace A's private address with its external
assigned address and set the Cache timeout to 0.
6. The server in External.com replies to host X
When Host X finds the address of Host A, X initiates a session with
A, using a destination IP address of 131.108.1.12. This datagram and
any others that follow in this session will be translated as usual by
the NAT.
Note, DNS_ALG changes only the response packets from the DNS server
for Private domain.
5.4. Reverse name lookups originated from external domain
This scenario builds on the previous case (section 5.3) by having
host X in External.com perform a reverse name lookup on 131.108.1.12,
which is host A's assigned external address. The following sequence
of events take place.
1. Host X sends a UDP based inverse name lookup query (PTR record)
for "12.1.108.131.IN-ADDR.ARPA." to its local DNS server.
2. Local DNS server in External.com sends the query to the root
server.
3. The root server, in turn, refers the local DNS server to query the
DNS server for Private.com.
4. External.com DNS server will now send the query to the DNS server
for Private.com. This query traverses the NAT router. NAT would
change the IP and UDP headers to reflect the DNS server's private
address.
DNS_ALG will enquire NAT for the private address associated with
the external address of 131.108.1.12 and modify the payload,
replacing 131.108.1.12 with the private address of 172.19.1.10.
5. The DNS server for Private.com replies with the host name of
"A.Private.Com.". This reply also transits the NAT. NAT would
translate the IP and UDP headers of the incoming packet to reflect
DNS server's private address.
Once again, DNS_ALG will enquire NAT for the assigned external
address associated with the private address of 172.19.1.10 and
modify the payload, replacing 172.19.1.10 with the assigned
external address of 131.108.1.12.
6. The DNS server in External.com replies to host X.
Note, DNS_ALG changes the query as well as response packets from DNS
server for Private domain.
6. Illustration of DNS_ALG in conjunction with Twice-NAT
The following diagram illustrates the operation of DNS_ALG in a Twice
NAT router. As before, we will illustrate by walking through how name
lookup and reverse name lookup queries are processed.
.
________________ . External.com
( ) .
( ) . +-------------+
+--+ ( Internet )-.---Border Router
__------ ( ) . +-------------+
/____\ (________________) .
Root .
DNS Server . ---------------
+---------------+ .
Provider Router . +--+ +--+
+---------------+ . __ __
. /____\ /____ . DNS Server Host X
External domain . 171.68.1.1 171.68.10.1
...........................................................
Private domain
Private.com
+-------------------------------------------+
Twice-NAT router with DNS_ALG
Private addresses: 171.68/16
Assigned External addresses: 131.108.1/24
External addresses: 171.68/16
Assigned Private addresses: 10/8
+-------------------------------------------+
---------- ----------
DNS Server
+--+ +--+ Authoritative
__ __ for private.com
/____\ /____ Host A DNS Server
171.68.1.10 171.68.2.1
(Mapped to 131.108.1.8)
Figure 4: DNS-ALG operation in Twice-NAT setup
In this scenario, hosts in private.com were not numbered from the RFC
1918 reserved 172.19/16 space, but rather were numbered with the
globally-routable 171.68/16 network, the same as external.com. Not
only does private.com need translation service for its own host
addresses, but it also needs translation service if any of those
hosts are to be able to exchange datagrams with hosts in
external.com. Twice-NAT accommodates the transition by translating
the overlapping address space used in external.com with a unique
address block (10/8) from RFC1918 address space. Routes are set up
within the private domain to direct datagrams destined for the
address block 10/8 through Twice-NAT device to the external global
network space.
Simplifications and assumptions made in section 5.0 will be valid
here as well.
6.1. Outgoing Name-lookup queries
Say, Host A in private.com needs to perform a name lookup for host X
in external.com (host X has a FQDN of X.external.com), to find its
address. This would would proceed as follows.
1. Host A sends a UDP based name lookup query (A record) for
"X.External.Com" to its local DNS server.
2. Local DNS server sends the query to the root server enroute NAT.
NAT would change the IP and UDP headers to reflect DNS server's
statically assigned external address. DNS_ALG will make no
changes to the payload.
3. The root server, in turn, refers the local DNS server to query the
DNS server for External.com. This referal transits the NAT enroute
to the local DNS server. NAT would simply translate the IP and
UDP headers of the incoming packet to reflect DNS server's private
address.
DNS_ALG will request NAT for an assigned private address for the
referral server and replace the external address with its assigned
private address in the payload.
4. Private.com DNS server will now send the query to the DNS server
for external.com, using its assigned private address, via NAT.
This time, NAT would change the IP and UDP headers to reflect the
External addresses of the DNS servers. I.e., Private.com DNS
server's IP address is changed to its assigned external address
and External.Com DNS server's assigned Private address is changed
to its external address.
DNS_ALG will make no changes to the payload.
5. The DNS server for external.com replies with the IP address
171.68.10.1. This reply also transits the NAT. NAT would once
again translate the IP and UDP headers of the incoming to reflect
the private addresses of the DNS servers. I.e., Private.com DNS
server's IP address is changed to its private address and
External.Com DNS server's external address is changed to its
assigned Private address.
DNS_ALG will request NAT to (a) set up a temporary binding for
Host X (171.68.10.1) with a private address and (b) initiate
Bind-holdout timer. When NAT successfully sets up temporary
binding with a private address (say, 10.0.0.254), DNS_ALG would
modify the payload to replace X's external address with its
assigned private address and set the Cache timeout to 0.
6. The DNS server in Private.com replies to host A.
When Host A finds the address of Host X, A initiates a session with
host X, using a destination IP address of 10.0.0.254. This datagram
and any others that follow in this session will be translated as
usual by Twice NAT.
Note, the DNS_ALG has had to change payload in both directions.
6.2. Reverse name lookups originated from private domain
This scenario builds on the previous case by having host A in
Private.com perform a reverse name lookup on 10.0.0.254, which is
host X's assigned private address. Following is a sequence of events.
1. Host A sends a UDP based inverse name lookup query (PTR record)
for "254.0.0.10.IN-ADDR.ARPA." to its local DNS server.
2. Local DNS server sends the query to the root server enroute NAT.
As before, NAT would change the IP and UDP headers to reflect DNS
server's statically assigned external address.
DNS_ALG will translate the private assigned address 10.0.0.254
with its external address 171.68.10.1.
3. The root server, in turn, refers the local DNS server to query the
DNS server for External.com. This referal transits the NAT enroute
to the local DNS server. NAT would simply translate the IP and
UDP headers of the incoming packet to reflect DNS server's private
address.
As with the original query, DNS_ALG will translate the private
assigned address 10.0.0.254 with its external address 171.68.10.1.
In addition, DNS_ALG will replace the external address of the
referal server (i.e., the DNS server for External.com) with its
assigned private address in the payload.
4. Private.com DNS server will now send the query to the DNS server
for external.com, using its statically assigned private address,
via NAT. This time, NAT would change the IP and UDP headers to
reflect the External addresses of the DNS servers. I.e.,
Private.com DNS server's IP address is changed to its assigned
external address and External.Com DNS server's assigned Private
address is changed to its external address.
As with the original query, DNS_ALG will translate the private
assigned address 10.0.0.254 with its external address 171.68.10.1.
5. The DNS server for external.com replies with the FQDN of
"X.External.Com.". This reply also transits the NAT. NAT would
once again translate the IP and UDP headers of the incoming to
reflect the private addresses of the DNS servers. I.e.,
Private.com DNS server's IP address is changed to its private
address and External.Com DNS server's external address is changed
to its assigned Private address.
Once again, DNS_ALG will translate the query section, replacing
the external address 171.68.10.1 with its assigned private address
of 10.0.0.254
6. The DNS server in Private.com replies to host A.
Note, the DNS_ALG has had to change payload in both directions.
6.3. Incoming Name-lookup queries
This time, host X in external.com wishes to initiate a session with
host A in Private.com. Below are the sequence of events that take
place.
1. Host X sends a UDP based name lookup query (A record) for
"A.Private.com" to its local DNS server.
2. Local DNS server in External.com sends the query to root server.
3. The root server, in turn, refers the DNS server in External.com to
query the DNS server for private.com,
4. External.com DNS server will now send the query to the DNS server
for Private.com. This query traverses the NAT router. NAT would
change the IP and UDP headers to reflect the private addresses of
the DNS servers. I.e., Private.com DNS server's IP address is
changed to its private address and External.Com DNS server's
external address is changed to assigned Private address.
DNS_ALG will make no changes to the payload.
5. The DNS server for Private.com replies with the IP address
171.68.1.10 for host A. This reply also transits the NAT. NAT
would once again translate the IP and UDP headers of the incoming
to reflect the external addresses of the DNS servers. I.e.,
Private.com DNS server's IP address is changed to its assigned
external address and External.Com DNS server's assigned private
address is changed to its external address.
DNS_ALG will request NAT to (a) set up temporary binding for Host
A (171.68.1.10) with an external address and (b) initiate Bind-
holdout timer. When NAT succeeds in finding an external address
(say, 131.108.1.12) to temporarily bind to host A, DNS_ALG would
modify the payload to replace A's private address with its
external assigned address and set the Cache timeout to 0.
6. The server in External.com replies to host X
When Host X finds the address of Host A, X initiates a session with
A, using a destination IP address of 131.108.1.12. This datagram and
any others that follow in this session will be translated as usual by
the NAT.
Note, DNS_ALG changes only the response packets from the DNS server
for Private domain.
6.4. Reverse name lookups originated from external domain
This scenario builds on the previous case (section 6.3) by having
host X in External.com perform a reverse name lookup on 131.108.1.12,
which is host A's assigned external address. The following sequence
of events take place.
1. Host X sends a UDP based inverse name lookup query (PTR record)
for "12.1.108.131.IN-ADDR.ARPA." to its local DNS server.
2. Local DNS server in External.com sends the query to the root
server.
3. The root server, in turn, refers the local DNS server to query the
DNS server for Private.com.
4. External.com DNS server will now send the query to the DNS server
for Private.com. This query traverses the NAT router. NAT would
change the IP and UDP headers to reflect the private addresses of
the DNS servers. I.e., Private.com DNS server's IP address is
changed to its private address and External.Com DNS server's
external address is changed to assigned Private address.
DNS_ALG will enquire NAT for the private address associated with
the external address of 131.108.1.12 and modify the payload,
replacing 131.108.1.12 with the private address of 171.68.1.10.
5. The DNS server for Private.com replies with the host name of
"A.Private.Com.". This reply also transits the NAT. NAT would once
again translate the IP and UDP headers of the incoming to reflect
the external addresses of the DNS servers. I.e., Private.com DNS
server's IP address is changed to its assigned external address
and External.Com DNS server's assigned private address is changed
to its external address.
Once again, DNS_ALG will enquire NAT for the assigned external
address associated with the private address of 172.19.1.10 and
modify the payload, replacing 171.68.1.10 with the assigned
external address of 131.108.1.12.
6. The DNS server in External.com replies to host X.
Note, DNS_ALG changes the query as well as response packets from DNS
server for Private domain.
7. DNS-ALG limitations and Future Work
NAT increases the probability of mis-addressing. For example, same
local address may be bound to different public address at different
times and vice versa. As a result, hosts that cache the name to
address mapping for longer periods than the NAT router is configured
to hold the map are likely to misaddress their sessions. Note, this
is mainly an issue with bad host implementations that hold DNS
records longer than the TTL in them allows and is not directly
attributable to the mechanism described here.
DNS_ALG cannot support secure DNS name servers in the private domain.
I.e., Signed replies from an authoritative DNS name server in the DMZ
to queries originating from the external world will be broken by the
DNS-ALG. At best, DNS-ALG would be able to transform secure dnssec
data into unprotected data. End-node demanding DNS replies to be
signed may reject replies that have been tampered with by DNS_ALG.
Since, the DNS server does not have a way to find where the queries
come from (i.e., internal or external), it will most likely have to
resort to the common denomination of today's insecure DNS. Both are
serious limitations to DNS_ALG. Zone transfers between DNS-SEC
servers is also impacted the same way, if the transfer crosses
address realms.
The good news, however, is that only end-nodes in DMZ pay the price
for the above limitation in a traditional NAT (or, a bi-directional
NAT), as external end-nodes may not access internal hosts due to DNS
replies not being secure. However, for outgoing sessions (from
private network) in a bi-directional NAT setup, the DNS queries can
be signed and securely accepted by DMZ and other internal hosts since
DNS_ALG does not intercept outgoing DNS queries and incoming replies.
Lastly, zone transfers between DNS-SEC servers within the same
private network are not impacted.
Clearly, with DNS SEC deployment in DNS servers and end-host
resolvers, the scheme suggested in this document will not work.
8. Security Considerations
If DNS packets are encrypted/authenticated per DNSSEC, then DNS_ALG
will fail because it won't be able to perform payload modifications.
Alternately, if packets must be preserved in an address realm,
DNS_ALG will need to hold the secret key to decrypt/verify payload
before forwarding packets to a different realm. For example, if DNS-
ALG, NAT and IPsec gateway (providing secure tunneling service) are
resident on the same device, DNS-ALG will have access to the IPsec
security association keys. The preceding section, "DNS-ALG
limitations and Future Work" has coverage on DNS_ALG security
considerations.
Further, with DNS-ALG, there is a possibility of denial of service
attack from a malicious user, as outlined in section 3.1. Section
3.1 suggests some ways to counter this attack.
REFERENCES
[1] Srisuresh, P. and M. Holdrege, "IP Network Address Translator
(NAT) Terminology and Considerations", RFC2663, August 1999.
[2] Egevang, K. and P. Francis, "The IP Network Address Translator
(NAT)", RFC1631, May 1994.
[3] Rekhter, Y., Moskowitz, B., Karrenberg, D., de Groot, G. and E.
Lear, "Address Allocation for Private Internets", BCP 5, RFC
1918, February 1996.
[4] Mockapetris, P., "Domain Names - Concepts and Facilities", STD
13, RFC1034, November 1987.
[5] Mockapetris, P., "Domain Names - Implementation and
Specification", STD 13, RFC1035, November 1987.
[6] Reynolds J. and J. Postel, "Assigned Numbers", STD 2, RFC1700,
October 1994.
[7] Braden, R., "Requirements for Internet Hosts -- Communication
Layers", STD 3, RFC1122, October 1989.
[8] Braden, R., "Requirements for Internet Hosts -- Application and
Support", STD 3, RFC1123, October 1989.
[9] Baker, F., "Requirements for IP Version 4 Routers", RFC1812,
June 1995.
[10] Carpenter, B., Crowcroft, J. and Y. Rekhter, "IPv4 Address
Behaviour Today", RFC2101, February 1997.
[11] Eastlake, D., "Domain Name System Security Extensions", RFC
2535, March 1999.
[12] Vixie, P., Thompson, S., Rekhter Y. and J. Bound, "Dynamic
Updates in the Domain Name System (DNS UPDATE)", RFC2136, April
1997.
[13] Eastlake, D., "Secure Domain Name System Dynamic Update", RFC
2137, April 1997.
[14] Elz R. and R. Bush, "Clarifications to the DNS specification",
RFC2181, July 1997.
[15] Elz, R., R. Bush, Bradner S. and M. Patton, "Selection and
Operation of Secondary DNS Servers", RFC2182, July 1997.
Authors' Addresses
Pyda Srisuresh
849 Erie Circle
Milpitas, CA 95035
U.S.A.
Phone: +1 (408) 263-7527
EMail: srisuresh@yahoo.com
George Tsirtsis
Internet Transport Group
B29 Room 129
BT Laboratories
Martlesham Heath
IPSWICH
Suffolk IP5 3RE
England
Phone: +44 1473 640756
Fax: +44 1473 640709
EMail: george@gideon.bt.co.uk
Praveen Akkiraju
cisco Systems
170 West Tasman Drive
San Jose, CA 95134 USA
Phone: +1 (408) 526-5066
EMail: spa@cisco.com
Andy Heffernan
Juniper Networks, Inc.
385 Ravensdale Drive.
Mountain View, CA 94043 USA
Phone: +1 (650) 526-8037
Fax: +1 (650) 526-8001
EMail: ahh@juniper.net
Full Copyright Statement
Copyright (C) The Internet Society (1999). All Rights Reserved.
This document and translations of it may be copied and furnished to
others, and derivative works that comment on or otherwise explain it
or assist in its implementation may be prepared, copied, published
and distributed, in whole or in part, without restriction of any
kind, provided that the above copyright notice and this paragraph are
included on all such copies and derivative works. However, this
document itself may not be modified in any way, such as by removing
the copyright notice or references to the Internet Society or other
Internet organizations, except as needed for the purpose of
developing Internet standards in which case the procedures for
copyrights defined in the Internet Standards process must be
followed, or as required to translate it into languages other than
English.
The limited permissions granted above are perpetual and will not be
revoked by the Internet Society or its successors or assigns.
This document and the information contained herein is provided on an
"AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING
TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING
BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION
HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
Acknowledgement
Funding for the RFCEditor function is currently provided by the
Internet Society.
