Reachability (IP vs. DNS) was Re: ARIN Proposal

[Stephen Satchell]

At 4:43 AM 1/23/97, Howard C. Berkowitz wrote:
>>From a user perspective and what IMHO should be a business perspective, the
>>issue is whether a given node is reachable or not.  As long as my browser
>>can find www.foo.com, I have solved my problem.
>
>www.foo.com is not, as people will notice, an IP address.  It is a DNS name.

As I recall the threads (note the plural) there are actually two different
problems that have been addresses:  renumbering when you change providers,
and the problem of reachability to a multi-homed ISP when that ISP is
*supposed* to get its numbers from one of the upstream providers.  Two
different problems.


>[snip]
>Yes, in a provider change in a provider-based addressing scheme, some
>hard-coded IP addresses will need to change.  But good practices can
>minimize that.

Let's say that you are a mid-level provider and have 100 customers buying
address space from you.  Each of those 100 customers are ISPs with name
servers -- by NSI decree, that's two hard-coded IP addresses per customer.
(I don't kno any ISP that delegates all its name-service load to another
provider -- I know I wouldn't if I wanted to be able to control third-level
names myself.)  Most ISPs also have "public" ftp and www servers, and
because they are addressed by name THEY have to have hard-coded IP
addresses that don't change over time.  Then you have the engine that does
dynamic address assignment for dial-in, which is usually a separate box.
That's now a *minimum* of four hard-coded IP addresses.

With four hundred addresses that have to be changed by 100 people outside
of your organization, that's a problem.

The DNS issue is a small one, and I seem to recall that some of the
time-to-live issues were being addressed.  The political issue -- the time
it takes NSI to repoint records -- may well take care of itself as the
process is better automated.  If it ever is.  In any case, the DNS
discussion takes us even more off-topic.

You mention "good practices".  Can you be more explicit, or provide a
pointer to good practices that would minimize the downstream-change
problem?

---
Stephen Satchell, Satchell Evaluations
http://www.accutek.com/~satchell for contact info
Opinions expressed are my own PERSONAL opinions.