Re: [rrg] Renumbering needs work
Olivier Bonaventure <[email protected]>
| Newsgroups | gmane.ietf.ops |
|---|---|
| Message-ID | <[email protected]> |
Brian, > A draft on this topic has been updated: > http://www.ietf.org/internet-drafts/draft-carpenter-renum-needs-work-01.txt > > Comments and discussion are invited on the OPS Area list, > [email protected] I completely agree that renumbering needs more work. For IPv4, I guess that this will always remain difficult. For IPv6, as most networks have still not deployed IPv6 and there are more addresses, there is still some hope. I think that a key point for renumbering is that it should be planned from day one when the first lines of the configuration of a network or network service are written. Every IPv6 operator should expect that he will have to renumber his network one day and should plan for this event. In most cases, there numbering will be due to economic factors such as merging or acquisition. An important recommendation in the draft (section 5.3.4) is that the network configuration should be based on databases and scripts/automated tools. The database stores all the information about the network and the tools derive the configuration from the database. As you note, this way of configuring networks is not the most widespread, but is being used by more network operators. It is also possible to prepare network configurations so that renumbering will be much easier. We have recently proposed a tool based on macros that allows to ease the reconfiguration of a network after a renumbering event. We applied the tool on real DNS, firewall and DHCP configuration as case studies, but it could also be applied to the configurations of routers or other services. More detailed informations about this work are available in http://inl.info.ucl.ac.be/publications/preparing-network-configurations-ipv6-renumbering Best regards, Olivier Bonaventure -- http://inl.info.ucl.ac.be , UCLouvain, Belgium