when & how could policy be changed

[Jeff Williams]

Jeremy,

Jeremy Porter wrote:
> 
> In message <v0310281aafdeee0588e2@[10.11.12.33]>, Michael Dillon writes:
> >At 5:55 PM -0400 6/30/97, Gordon Cook wrote:
> >
> >> is there a view from the arin board that with stringent
> >>dampening it could agree to give everyone a 19/?   or are you saying that
> >>it would not have do this because the 'big boys' would all agree to route
> >>20/s?
> ...
> >I personally would like to see some PI space opened up with longer prefixes
> >than /19. This could be a new /8 like 210/8 that we all agree to allocate
> >in /20 blocks. Or we could use reclaimed space from the swamp and allocate
> >it in /20 and/or /21 sizes. In the case of 210/8 we need providers to agree
> >to adjust their filters. But before we can decide just how this should be
> >done we need some hard numbers, especially on how many additional routes
> >the new PI space would add. And we also need some more thorough analysis of
> >the prefixes that appear to be eligible for aggregation in the weekly CIDR
> >reports.
> 
> The only problem is that going from /19 to /20 doubles the total
> table size, assuming lack of greater aggregation, worst case.

  Exactly what I was intimating earlier on another thread.  And, hence,
the reason that if a Medium size ISP startup in being considered,
which I am sure many "Recional" ones will, this argument supports
what I had indicated earlier in some alloance for allocations
of /19's right off.

> There are issues with flap, and dampening helps but does not solve
> the problem.  The problem seens to growth with order N^M,
> where N is the number of prefixes and M is the number of peering sessions
> or views.  Also for fun the cost of upgrading networks grows at N^R, in this
> case R is the number of routers. We have some data to suggest
> that todays hardware could handle the load generated by /19 aggregation,
> but also seems to indicate that we cannot as a general rule freely
> allocate /20s, without severly imparing network perfomance in the near
> term.

  Good point here.  ANd again supports my original argument and
question.
> 
> Now if we had some real data on the per flap costs, pathology of route
> flap, effectiveness of flap dampening, etc.  Right now we are seriously
> lacking data on flap. We need to ask where does flap orginate?
> Can we dampen it at the source?  (Vadim's suggestion of link bounce
> dampening might be useful.)  Also there is some hint of evidence to suggest
> that some part of route flap is caused by policy changes.  Changes
> to allow for soft reconfigurate can help, but there is the router upgrade
> problem again.

  Yep.  But if there has been reasonable hardware overdesign in place
the stressing of routing hardware should not be that big of a problem.
However this is not usually the case, as you indicate here.

> Backbone providers do not have a economic motive to dampen flap that
> is customer originated, compared to dampening flap at the peer level.
> 
> Having renumbered several /20s and a /19, I don't see they need to
> create PI /20 space.  There is this ideal out there that the playing
> field should be completely flat, however, in the real world, this
> isn't the case.  Smaller providers have a number of cost advantages over
> the larger players, and I see this as a way to offset the cost of
> renumbering.  Since new allocations involve customer interaction, and
> the customer interaction is the primary cost of renumbering, and renumbering
> is fairly painless if the network is design properly.

Very well put.  I agree compleately.  Hence my earlier argument or
suggestion
that initialy allocating /20 or /21's is even worse in a initial startup
Medium ISP situation.

>  All modern hardware
> and software can support dynamic assignment for networks.  With a small
> bit of planning and intergration, one change can renumber DNS A and PTR
> records, and change the assigneds when the DHCP leases expire.

  This should not be necessary at all.
> 
> If you have people in PA /20, /21, etc., already you are going to have to
> renumber, if you haven't started yet renumbering can be without any
> distruption in services, and with a small finacial impact.  Thus renumbering
> once does not seem to be a huge issue to me.  So far the only response
> to this I have seen is that well, @Home got a large assignment,
> which is completely different because @Home went to IANA, not to the
> registries.

  Seems that now you are agreeing with me.  Where earlier you were not.
I agree that in most cases an allocation of /20 or 21's for small ISP's
me be ok in the short term, this is not true for Medium or larger ISP's.
IMHO, this is something that should be considered a policy issue for
most of the reasons you state here.
> 
> With this said, I do support the allocation of a routeable /19 to
> providers that are a. Multihomed b. Have a history of efficient utilization
> of addresses c. Are willing to renumber customers into their allocations
> to maintain efficient utilization of routeable prefixes.
> 
> I have yet to see an objection to this proposal.  Other than unfounded
> complaints of "Its not fair."

  We have seen them right here on this list.
> 
> If there is enough intereste I will work up this proposal to put
> before the ARIN membership at the first suitable time.

  I would be happy to calaborate with you on this.
> 
> ---
> Jeremy Porter, Freeside Communications, Inc.      jerry at fc.net
> PO BOX 80315 Austin, Tx 78708  |  1-800-968-8750  |  512-458-9810
> http://www.fc.net

Regards,
-- 
Jeffrey A. Williams
DIR. Internet Network Eng/SR. Java Development Eng.
Information Eng. Group. IEG. INC. 
Phone :913-294-2375 (v-office)
E-Mail jwkckid1 at ix.netcom.com