Review: IP Next Generation Overview

 

Reviewer: Kevin Hofstra

 

The current internet protocol (IPv4) is quickly running out of address space for all the exponentially expanding number of hosts, lacks authentication, and does not allow for creation of efficient addressing hierarchies. 

 

The article enumerates the requirements and the steps towards the efficient deployment of IPng (Ipv6).  The shortage of hosts problem is new, and was somewhat unseen due to the tremendous exponential growth of the internet.  Even today the problem is still around.  Although IPv6 was proposed as a solution many years ago and was seen as a viable solution, many companies have been reluctant to switch over because the incentives are not enough to outweigh the initial rollout and learning curve however small it may be.  This problem will continue to grow in importance as the world uses up more and more host addresses and temporary solutions such as CIDR, internal addressing, proxys, and address translation begin to take more effort to support than the transition to IPv6. 

 

When IPv6 was first introduced it was almost uniformly seen as necessary for the continuation of growth, but I think this article is a significant contribution (4) because it recognizes the temporary solutions that companies are enacting and has a much more realistic outlook on the incentives of adapting the new standard.  Hinden points out that there are ways of remaining on the old version and that most companies are reluctant to change something that is working.  The only way that there will be significant support for the transition is to make it more maintenance and cost effective to use the new version or to have new added capabilities that entice new users.  The only thing that I believe that the article is lacks is more emphasis on the performance enhancements that a new global addressing scheme and routing hierarchy will achieve.  I believe this to be one of the greatest improvements in the new version, and one thing that the temporary solutions in use today will be unable to solve.

 

Hinden’s requirements for the transition are very convincing and backed by both scientific reason and a realistic viewpoint.  Flexibility of deployment and integration with legacy systems are the most important requirements of the successor.  The new version must be compatible with the older version to ease the transition between the two.  Many companies will be unwilling to switch to the new version until it is widespread, but the only way to begin the migration is to allow the new version to be introduced into symbiotic relationship with IPv4 and slowly work its way into widespread acceptance as the dominant version.  The flexibility of deployment is just as crucial because companies are not willing to bear the transition costs of a difficult migration if the old standard is working fine.  The transition must be seen as easy enough that it outweighs one of the temporary solutions.

 

Researchers have much to learn from this article, but even more so from the problem.  The addressing problem has been around for years and the solution of IPv6 has also, but corporations are going to be unwilling to make a major transition if a temporary solution will be just as effective and more cost efficient.  The only way to change a standard when the transition is voluntary is to make the move as easy as possible and to make the enhancements of the new version an incentive to change.