IETF-85 Proceedings
Introduction | Area, Working Goup & BoF Reports | Plenaries | Training | Internet Research Task Force
Additional information is available at tools.ietf.org/wg/6renum
Chair(s):SecretaryOperations and Management Area Area Director(s):Operations and Management Area Advisor |
As outlined in RFC 5887, renumbering, especially for medium to large
sites and networks, is currently viewed as an expensive, painful, and
error-prone process, avoided by network managers as much as possible.
As that RFC describes, there are triggers that mean some cases of
renumbering are unavoidable, so it should be considered a given that a
site may need partial or complete renumbering at some stage. It is thus
important to implement and deploy techniques that facilitate simpler
IPv6 site renumbering, so that as IPv6 becomes universally deployed,
renumbering can be viewed as a more routine event. This includes
consideration of how the initial assignment and subsequent management of
address information is performed, as this will affect future renumbering
operations.
If IPv6 site renumbering continues to be considered difficult, network
managers will turn to PI addressing for IPv6 to attempt to minimise the
need for future renumbering. However, widespread use of PI may create
very serious BGP4 scaling problems. It is thus desirable to develop
tools and practices that may make renumbering a simpler process to
reduce demand for IPv6 PI space.
Renumbering, or partial renumbering, can be complicated in an enterprise
site where a short prefix is divided into subnets with longer prefixes.
Aggregation, synchronisation, coordination, etc., need to be carefully
managed, and the use of manually inserted address literals minimised.
Other factors such as applications binding long-term to low level IP
addresses may add constraints to what can be realistically achieved, but
identifying and documenting such factors is a useful objective. In some
scenarios, consideration may also need to be made for 'flag day'
renumbering (in contrast to the procedure described in RFC4192).
The task of the 6RENUM working group is to document existing renumbering
practices for enterprise site networks and to identify specific
renumbering problems or 'gaps' in the context of partial or site-wide
renumbering. Completion of these tasks should provide a basis for
future work to identify and develop point solutions or system solutions
to address those problems or to stimulate such development in other
working groups as appropriate.
6RENUM is chartered to perform an analysis of IPv6 site renumbering. If
the analysis leads to conclusions that are also applicable to IPv4 that
will be an advantage, but it is not an objective of the WG to make its
outputs more widely available than IPv6. Similarly the WG is targeting
enterprise networks, but the analysis may also be applicable to SOHO or
other (e.g. ad-hoc) scenarios.
It may be that for site renumbering to become more routine, a systematic
address management approach will be required. By documenting current
practices and undertaking a gap analysis, we should become better
informed of the requirements of such an approach. Post-analysis
rechartering may lead to further work in this area. RFC 4192, RFC 5887,
and draft-jiang-ipv6-site-renum-guideline are starting points for this
work.
Goals/deliverables
------------------
The goals of the 6RENUM working group are:
1. To undertake scenario descriptions, including documentation of
current capability inventories and existing BCPs, for enterprise
networks, including managed and unmanaged elements. These texts should
contribute towards a gap analysis and provide an agreed basis for
subsequent WG rechartering towards development of solutions (which may
be more appropriate for other WGs to undertake) and improved practices.
Operator input will be of high value for this text.
2. To perform a gap analysis for renumbering practices, to identify
generic issues for network design, network management, address
management and renumbering operations. The methodology for the WG will
be to begin building the enterprise scenario description while in
parallel constructing an initial gap analysis drawing on existing work
in (at least) RFC4192 and RFC5887. As the scenario text hardens, its
contributions will be incorporated into the more detailed gap analysis,
which can be published once the scenario text is completed. The
deliverables are thus the scenario and gap analysis texts.
The following topics are out of scope for the working group:
1. Renumbering avoidance; this can perhaps be considered by appropriate
IRTF groups. As documented in RFC5887, renumbering cannot be completely
avoided. The WG is limited to dealing with how to renumber when it is
unavoidable.
2. IPv4 renumbering. While many sites are likely to run dual-stack, IPv6
is the future and, especially given concerns over extensive use of IPv6
PI, the most appropriate place to focus effort.
3. ISP renumbering; this is potentially the most complex renumbering
case. However, more benefit can be achieved by focusing effort on site
renumbering. The enterprise site analysis should include the ISP's role
in the site's renumbering events.
4. Neither SOHO nor manet networks are targeted by the WG. However, if
outputs from the WG are applicable to those scenarios, that would be an
advantage.
A recharter of the WG will be possible once the gap analysis and
scenario description are completed and published. Such rechartering
would identify more specific work items within the 6RENUM WG or
appropriate protocol WGs, and may include a proposal for work on a
systematic address management approach.