(UPDATE) Can a Customer take their IP's with them? (Court says yes!)

Michel Py michel at arneill-py.sacramento.ca.us
Thu Jul 1 00:46:09 UTC 2004


> Matthew Crocker wrote:
> From my understanding the customer has their own IP space
> allocated by ARIN and has had that space for over a year.
> They have already had adequate time to transition to their
> own space.

Agree: they could have their ARIN space routed to their equipment and
begun the renumbering process a year ago. In my experience, ARIN does
allocate enough space to get the job done, and they also do allow for
enough time using both the current PA block and the newly allocated
space to complete a migration that has been reasonably planned.

It is the customer's responsibility to plan for renumbering, not the
provider's. The fact that non-portable addresses can't be kept is not
new nor secret.

> The Internet routing table should not suffer due to the
> laziness of one customer.

My point also. People tend to think that one more route in the routing
table is no big deal. This is absolutely wrong, because these onesies
and twotsies are what makes the routing table too big already.

> Black holing is a drastic step but I think decisive action needs
> to be taken the Internet at large to protect the routing table.

Agree.

> I know I would *love* to gain ownership of some of my space I
> have from Sprint. I'm too lazy to move out of that space but I
> do continue to by bandwidth from Sprint (have been doing so for
> 10 years now). If this holds up,  maybe I'll try and sue
> Sprint ;) *this is a joke.... I'm not that irresponsible to
> the 'net*

Problem is that not everybody thinks like you, which is why we must act
before this gets out of hand.


> Sabri Berisha wrote:
> This is more than the court prohibiting NAC from
> terminating this customer (on their request!)..

Indeed.

> The TRO says:
 >> (f) NAC shall permit UCI to continue utilization through any
 >> carrier or carriers of UCI's choice of any IP addresses that
 >> were utilized by, through or on behalf of UCI under the April
 >> 2003 Agreement during the therm thereof (the "Prior UCI
 >> Addresses) and shall not interfere in any way with the use of
 >> the Prior UCI Addresses.
> I don't know about you, but I read this as a court converting
> non-portable address space to portable address space,

This is clear to me as well.

> without looking at the consequenses for ARIN, NAC and network
> operators around the world who need to utilize more RAM in
> their routers to store additional routes simply because they
> are to ignorant to follow the technical guidelines and policies
> of ARIN.

Or too lazy. The issue is not about moving, the issue is about UCI not
willing to renumber because they were stupid enough to a) build a
business on PA address that can't be renumbered b) sign a contract that
allowed NAC to quick them out in 45 days. If NAC tariffs and services
were as alleged pricey and bad, UCI should have found another provider
and moved a long time ago. Instead, they want a free ride on renumbering
and are trying to have a court that does not even have jurisdiction over
the matter to legislate from the bench in order to transform PA space
into PI. This is doubly unacceptable, first because the court's role is
to interpret the law, not to make them and second because it would
trigger large number of copycat lawsuits from people that want portable
space, but don't qualify for it or don't want to pay for it.


> Doug White wrote:
> Hopefully this whole affair will be a wake-up call to providers
> who put revenue ahead of sound policy enforcement, assuming they
> have enforceable policies in effect.

Indeed. This is valid for providers who do have sound policy enforcement
as well: detecting an undesirable customer before signing a contract
might be difficult, as the sales force will typically not perform sanity
checks before signing new customers. And then, as we can see, getting
rid of the customer might prove more challenging than originally
thought.


> Patrick W Gilmore wrote:
> The TRO of and by itself is ... troublesome, but not a
> threat to the Internet.  If this becomes common practice,
> we are all in trouble.

That's why it _is_ a threat, Patrick.

Michel.




More information about the NANOG mailing list