RE: Why does Sprint have address filters again?

The last CIDR report I have in my in-box states there are 3606 AS numbers in
the routing table. So, my first question is at what point does ARIN reclaim
AS numbers?

(comparison: 2680 last September and 3075 at the beginning of the year)

That's a lot of money Arin rakes in on un-used AS numbers.

I understand that ARIN probably doesn't want to get in the middle of routing
policies, but by limiting when you can receive an AS, I don't understand why
they couldn't limit when you lose it.

Second Q: How many AS numbers are available in total?

Best regards,

Jamie Scheinblum - FASTNET(tm) / You Tools Corporation (888)321-FAST(3278)
FASTNET - Business and Personal Internet Solutions

The views stated above are mine and do not reflect those
of my employer.

Currently an ASN is a 16-bit number.

Second Q: How many AS numbers are available in total?

Currently an ASN is a 16-bit number.

And a whole lot (~1/2) are reserved to IANA. Specifically:

  32768-64511 IANA-RSVD
  64512-65535 IANA-RSVD2

You can find this on (even if it hasn't
been February ;).

The second block is the one you have to worry about since those numbers are
used for things like BGP confederations. I believe the first block could
be allocated to the general public, but you'd have to check with someone
more cluefull to be sure.

Karl Denninger (karl@MCS.Net)| MCSNet - Serving Chicagoland and Wisconsin


Well, other than the definition of an ASN as a "short" in router software
and BGP4, there's no *reason* an ASN has to be a short integer.

That is, it wouldn't be difficult *at all* to define BGP4.1 in which an ASN
was either defined as a "long" or as a "numeric string of arbitrary length".

Its not like an ASN is in the header of an IP packet (where field lengths
are limited) you know.

I suspect the first "reserved" block is due to suspected buggy
implementations that defined an ASN as a *signed* short. Obviously that's
not an issue any longer, or the internal "reserved" numbers wouldn't work.