RFC 1918

RFC 1918 (rfc1918)

Address AlLocation for Private Internets

 

 

 

 

Network Working Group                                         Y. Rekhter
Request for Comments: 1918                                  Cisco Systems
Obsoletes: 1627, 1597                                            B. Moskowitz
BCP: 5                                                                Chrysler Corp.
Category: Best Current Practice                             D. Karrenberg
                                                                         RIPE NCC
                                                                         G. J. de Groot
                                                                         RIPE NCC
                                                                         E. Lear
                                                                         Silicon Graphics, Inc.
                                                                         February 1996

                Address AlLocation for Private Internets

Status of this Memo

   This document specifies an Internet Best Current Practices for the
   Internet Community, and requests discussion and suggestions for
   improvements.  Distribution of this memo is unlimited.

1. Introduction

   For the purposes of this document, an enterprise is an entity
   autonomously operating a network using TCP/IP and in particular
   determining the addressing plan and address assignments within that
   network.

   This document describes address alLocation for private internets. The
   alLocation permits full network layer connectivity among all hosts
   inside an enterprise as well as among all public hosts of different
   enterprises. The cost of using private internet address space is the
   potentially costly effort to renumber hosts and networks between
   public and private.

2. Motivation

   With the proliferation of TCP/IP technology worldwide, including
   outside the Internet itself, an increasing number of non-connected
   enterprises use this technology and its addressing capabilities for
   sole intra-enterprise communications, without any intention to ever
   directly connect to other enterprises or the Internet itself.

   The Internet has grown beyond anyone's expectations. Sustained
   exponential growth continues to introduce new challenges.  One
   challenge is a concern within the community that globally unique
   address space will be exhausted. A separate and far more pressing
   concern is that the amount of routing overhead will grow beyond the

   capabilities of Internet Service Providers. Efforts are in progress
   within the community to find long term solutions to both of these
   problems. Meanwhile it is necessary to revisit address alLocation
   procedures, and their impact on the Internet routing system.

   To contain growth of routing overhead, an Internet Provider obtains a
   block of address space from an address registry, and then assigns to
   its customers addresses from within that block based on each customer
   requirement. The result of this process is that routes to many
   customers will be aggregated together, and will appear to other
   providers as a single route [RFC1518], [RFC1519].  In order for route
   aggregation to be effective, Internet providers encourage customers
   joining their network to use the provider's block, and thus renumber
   their computers. Such encouragement may become a requirement in the
   future.

   With the current size of the Internet and its growth rate it is no
   longer realistic to assume that by virtue of acquiring globally
   unique IP addresses out of an Internet registry an organization that
   acquires such addresses would have Internet-wide IP connectivity once
   the organization gets connected to the Internet. To the contrary, it
   is quite likely that when the organization would connect to the
   Internet to achieve Internet-wide IP connectivity the organization
   would need to change IP addresses (renumber) all of its public hosts
   (hosts that require Internet-wide IP connectivity), regardless of
   whether the addresses used by the organization initially were
   globally unique or not.

   It has been typical to assign globally unique addresses to all hosts
   that use TCP/IP. In order to extend the life of the IPv4 address
   space, address registries are requiring more justification than ever
   before, making it harder for organizations to acquire additional
   address space [RFC1466].

   Hosts within enterprises that use IP can be partitioned into three
   categories:

      Category 1: hosts that do not require access to hosts in other
                  enterprises or the Internet at large; hosts within
                  this category may use IP addresses that are
                  unambiguous within an enterprise, but may be
                  ambiguous between enterprises.

      Category 2: hosts that need access to a limited set of outside
                  services (e.g., E-mail, FTP, netnews, remote login)
                  which can be handled by mediating gateways (e.g.,
                  application layer gateways). For many hosts in this
                  category an unrestricted external access (provided

                  via IP connectivity) may be unnecessary and even
                  undesirable for privacy/security reasons. Just like
                  hosts within the first category, such hosts may use
                  IP addresses that are unambiguous within an
                  enterprise, but may be ambiguous between
                  enterprises.

      Category 3: hosts that need network layer access outside the
                  enterprise (provided via IP connectivity); hosts in
                  the last category require IP addresses that are
                  globally unambiguous.

   We will refer to the hosts in the first and second categories as
   "private".  We will refer to the hosts in the third category as
   "public".

   Many applications require connectivity only within one enterprise and
   do not need external (outside the enterprise) connectivity for the
   majority of internal hosts. In larger enterprises it is often easy to
   identify a substantial number of hosts using TCP/IP that do not need
   network layer connectivity outside the enterprise.

   Some examples, where external connectivity might not be required,
   are:

         - A large airport which has its arrival/departure displays
           individually addressable via TCP/IP. It is very unlikely
           that these displays need to be directly accessible from
           other networks.

         - Large organizations like banks and retail chains are
           switching to TCP/IP for their internal communication. Large
           numbers of local workstations like cash registers, money
           machines, and equipment at clerical positions rarely need
           to have such connectivity.

         - For security reasons, many enterprises use application
           layer gateways to connect their internal network to the
           Internet.  The internal network usually does not have
           direct access to the Internet, thus only one or more
           gateways are visible from the Internet. In this case, the
           internal network can use non-unique IP network numbers.

         - Interfaces of routers on an internal network usually do not
           need to be directly accessible from outside the enterprise.

3. Private Address Space

   The Internet Assigned Numbers Authority (IANA) has reserved the
   following three blocks of the IP address space for private internets:

     10.0.0.0        -   10.255.255.255  (10/8 prefix)
     172.16.0.0      -   172.31.255.255  (172.16/12 prefix)
     192.168.0.0     -   192.168.255.255 (192.168/16 prefix)

   We will refer to the first block as "24-bit block", the second as
   "20-bit block", and to the third as "16-bit" block. Note that (in
   pre-CIDR notation) the first block is nothing but a single class A
   network number, while the second block is a set of 16 contiguous
   class B network numbers, and third block is a set of 256 contiguous
   class C network numbers.

   An enterprise that decides to use IP addresses out of the address
   space defined in this document can do so without any coordination
   with IANA or an Internet registry. The address space can thus be used
   by many enterprises. Addresses within this private address space will
   only be unique within the enterprise, or the set of enterprises which
   choose to cooperate over this space so they may communicate with each
   other in their own private internet.

   As before, any enterprise that needs globally unique address space is
   required to obtain such addresses from an Internet registry. An
   enterprise that requests IP addresses for its external connectivity
   will never be assigned addresses from the blocks defined above.

   In order to use private address space, an enterprise needs to
   determine which hosts do not need to have network layer connectivity
   outside the enterprise in the foreseeable future and thus could be
   classified as private. Such hosts will use the private address space
   defined above.  Private hosts can communicate with all other hosts
   inside the enterprise, both public and private. However, they cannot
   have IP connectivity to any host outside of the enterprise. While not
   having external (outside of the enterprise) IP connectivity private
   hosts can still have access to external services via mediating
   gateways (e.g., application layer gateways).

   All other hosts will be public and will use globally unique address
   space assigned by an Internet Registry. Public hosts can communicate
   with other hosts inside the enterprise both public and private and
   can have IP connectivity to public hosts outside the enterprise.
   Public hosts do not have connectivity to private hosts of other
   enterprises.

   Moving a host from private to public or vice versa involves a change
   of IP address, changes to the appropriate DNS entries, and changes to
   configuration files on other hosts that reference the host by IP
   address.

   Because private addresses have no global meaning, routing information
   about private networks shall not be propagated on inter-enterprise
   links, and packets with private source or destination addresses
   should not be forwarded across such links. Routers in networks not
   using private address space, especially those of Internet service
   providers, are expected to be configured to reject (filter out)
   routing information about private networks. If such a router receives
   such information the rejection shall not be treated as a routing
   protocol error.

   Indirect references to such addresses should be contained within the
   enterprise. Prominent examples of such references are DNS Resource
   Records and other information referring to internal private
   addresses. In particular, Internet service providers should take
   measures to prevent such leakage.

4. Advantages and Disadvantages of Using Private Address Space

   The obvious advantage of using private address space for the Internet
   at large is to conserve the globally unique address space by not
   using it where global uniqueness is not required.

   Enterprises themselves also enjoy a number of benefits from their
   usage of private address space: They gain a lot of flexibility in
   network design by having more address space at their disposal than
   they could obtain from the globally unique pool. This enables
   operationally and administratively convenient addressing schemes as
   well as easier growth paths.

   For a variety of reasons the Internet has already encountered
   situations where an enterprise that has not been connected to the
   Internet had used IP address space for its hosts without getting this
   space assigned from the IANA. In some cases this address space had
   been already assigned to other enterprises. If such an enterprise
   would later connects to the Internet, this could potentially create
   very serious problems, as IP routing cannot provide correct
   operations in presence of ambiguous addressing. Although in principle
   Internet Service Providers should guard against such mistakes through
   the use of route filters, this does not always happen in practice.
   Using private address space provides a safe choice for such
   enterprises, avoiding clashes once outside connectivity is needed.

   A major drawback to the use of private address space is that it may
   actually reduce an enterprise's flexibility to access the Internet.
   Once one commits to using a private address, one is committing to
   renumber part or all of an enterprise, should one decide to provide
   IP connectivity between that part (or all of the enterprise) and the
   Internet.  Usually the cost of renumbering can be measured by
   counting the number of hosts that have to transition from private to
   public. As was discussed earlier, however, even if a network uses
   globally unique addresses, it may still have to renumber in order to
   acquire Internet-wide IP connectivity.

   Another drawback to the use of private address space is that it may
   require renumbering when merging several private internets into a
   single private internet. If we review the examples we list in Section
   2, we note that companies tend to merge. If such companies prior to
   the merge maintained their uncoordinated internets using private
   address space, then if after the merge these private internets would
   be combined into a single private internet, some addresses within the
   combined private internet may not be unique. As a result, hosts with
   these addresses would need to be renumbered.

   The cost of renumbering may well be mitigated by development and
   deployment of tools that facilitate renumbering (e.g.  Dynamic Host
   Configuration Protocol (DHCP)). When deciding whether to use private
   addresses, we recommend to inquire computer and software vendors
   about availability of such tools.  A separate IETF effort (PIER
   Working Group) is pursuing full documentation of the requirements and
   procedures for renumbering.

5. Operational Considerations

   One possible strategy is to design the private part of the network
   first and use private address space for all internal links. Then plan
   public subnets at the Locations needed and design the external
   connectivity.

   This design does not need to be fixed permanently. If a group of one
   or more hosts requires to change their status (from private to public
   or vice versa) later, this can be accomplished by renumbering only
   the hosts involved, and changing physical connectivity, if needed. In
   Locations where such changes can be foreseen (machine rooms, etc.),
   it is advisable to configure separate physical media for public and
   private subnets to facilitate such changes.  In order to avoid major
   network disruptions, it is advisable to group hosts with similar
   connectivity needs on their own subnets.

   If a suitable subnetting scheme can be designed and is supported by
   the equipment concerned, it is advisable to use the 24-bit block
   (class A network) of private address space and make an addressing
   plan with a good growth path. If subnetting is a problem, the 16-bit
   block (class C networks), or the 20-bit block (class B networks) of
   private address space can be used.

   One might be tempted to have both public and private addresses on the
   same physical medium. While this is possible, there are pitfalls to
   such a design (note that the pitfalls have nothing to do with the use
   of private addresses, but are due to the presence of multiple IP
   subnets on a common Data Link subnetwork).  We advise caution when
   proceeding in this area.

   It is strongly recommended that routers which connect enterprises to
   external networks are set up with appropriate packet and routing
   filters at both ends of the link in order to prevent packet and
   routing information leakage. An enterprise should also filter any
   private networks from inbound routing information in order to protect
   itself from ambiguous routing situations which can occur if routes to
   the private address space point outside the enterprise.

   It is possible for two sites, who both coordinate their private
   address space, to communicate with each other over a public network.
   To do so they must use some method of encapsulation at their borders
   to a public network, thus keeping their private addresses private.

   If two (or more) organizations follow the address alLocation
   specified in this document and then later wish to establish IP
   connectivity with each other, then there is a risk that address
   uniqueness would be violated.  To minimize the risk it is strongly
   recommended that an organization using private IP addresses choose
   randomly from the reserved pool of private addresses, when allocating
   sub-blocks for its internal alLocation.

   If an enterprise uses the private address space, or a mix of private
   and public address spaces, then DNS clients outside of the enterprise
   should not see addresses in the private address space used by the
   enterprise, since these addresses would be ambiguous.  One way to
   ensure this is to run two authority servers for each DNS zone
   containing both publically and privately addressed hosts.  One server
   would be visible from the public address space and would contain only
   the subset of the enterprise's addresses which were reachable using
   public addresses.  The other server would be reachable only from the
   private network and would contain the full set of data, including the
   private addresses and whatever public addresses are reachable the
   private network.  In order to ensure consistency, both servers should
   be configured from the same data of which the publically visible zone

   only contains a filtered version. There is certain degree of
   additional complexity associated with providing these capabilities.

6. Security Considerations

   Security issues are not addressed in this memo.

7. Conclusion

   With the described scheme many large enterprises will need only a
   relatively small block of addresses from the globally unique IP
   address space. The Internet at large benefits through conservation of
   globally unique address space which will effectively lengthen the
   lifetime of the IP address space. The enterprises benefit from the
   increased flexibility provided by a relatively large private address
   space. However, use of private addressing requires that an
   organization renumber part or all of its enterprise network, as its
   connectivity requirements change over time.

8. Acknowledgments

   We would like to thank Tony Bates (MCI), Jordan Becker (ANS), Hans-
   Werner Braun (SDSC), Ross Callon (BayNetworks), John Curran (BBN
   Planet), Vince Fuller (BBN Planet), Tony Li (cisco Systems), Anne
   Lord (RIPE NCC), Milo Medin (NSI), Marten Terpstra (BayNetworks),
   Geza Turchanyi (RIPE NCC), Christophe Wolfhugel (Pasteur Institute),
   Andy Linton (connect.com.au), Brian Carpenter (CERN), Randy Bush
   (PSG), Erik Fair (Apple Computer), Dave Crocker (Brandenburg
   Consulting), Tom Kessler (SGI), Dave Piscitello (Core Competence),
   Matt Crawford (FNAL), Michael Patton (BBN), and Paul Vixie (Internet
   Software Consortium) for their review and constructive comments.

9. References

   [RFC1466] Gerich, E., "Guidelines for Management of IP Address
       Space", RFC 1466, Merit Network, Inc., May 1993.

   [RFC1518] Rekhter, Y., and T. Li, "An Architecture for IP Address
       AlLocation with CIDR", RFC 1518, September 1993.

   [RFC1519] Fuller, V., Li, T., Yu, J., and K. Varadhan, "Classless
       Inter-Domain Routing (CIDR): an Address Assignment and
       Aggregation Strategy", RFC 1519, September 1993.

10. Authors' Addresses

   Yakov Rekhter
   Cisco systems
   170 West Tasman Drive
   San Jose, CA, USA
   Phone: +1 914 528 0090
   Fax: +1 408 526-4952
   EMail: yakov@cisco.com

   Robert G Moskowitz
   Chrysler Corporation
   CIMS: 424-73-00
   25999 Lawrence Ave
   Center Line, MI 48015
   Phone: +1 810 758 8212
   Fax: +1 810 758 8173
   EMail: rgm3@is.chrysler.com

   Daniel Karrenberg
   RIPE Network Coordination Centre
   Kruislaan 409
   1098 SJ Amsterdam, the Netherlands
   Phone: +31 20 592 5065
   Fax: +31 20 592 5090
   EMail: Daniel.Karrenberg@ripe.net

   Geert Jan de Groot
   RIPE Network Coordination Centre
   Kruislaan 409
   1098 SJ Amsterdam, the Netherlands
   Phone: +31 20 592 5065
   Fax: +31 20 592 5090
   EMail: GeertJan.deGroot@ripe.net

   Eliot Lear
   Mail Stop 15-730
   Silicon Graphics, Inc.
   2011 N. Shoreline Blvd.
   Mountain View, CA 94043-1389
   Phone: +1 415 960 1980
   Fax:   +1 415 961 9584
   EMail: lear@sgi.com