i believe that the ripe ncc tried this allocation idea initially, but discovered that the ncc ended up with bits of address space here and there which was just not the right size, so the ncc would end up allocating more non-contiguous space anyway. with that experience the ncc has abandoned that practice. perhaps, daniel, geert-jan or someone else from the ripe ncc could speak about their experience. --elise
Dorian Rysling Kim writes:
On Fri, 18 Aug 1995, Hank Nussbacher wrote:
Ah. here is the rub. When you ISP buddies come back, you should ask them to return the origianal /22 for a /20. That way, the total size of the routing system stays the same!
Great idea. Know ANYONE who does that? The best I can do is give them a /20 (in addition to the original /22) if their growth warrants it.
Sure. We are starting to do that. We are trying to put into practice following allocation strategy:
1) when allocating, leave enough holes so the address space can be grown within that /16. For example, leaving next three spaces unallocated when assigning /24. We try to guess at who'll be growing and how much. It's not perfect, but it's better than nothing.
2) if the site returns /24, or /23, or /22, and we can't grow that address block, we try to see if we can recycle the old block and assign a new larger block.
This way, if our internal routing changes, and our old aggregation scheme breaks, we keep renumbering to the minimum.
-dorian ______________________________________________________________________________ Dorian Kim Email: dorian@cic.net 2901 Hubbard Drive Network Engineer Phone: (313)998-6976 Ann Arbor MI 48105 CICNet Network Systems Fax: (313)998-6105 http://www.cic.net/~dorian