North American Network Operators Group

Date Prev | Date Next | Date Index | Thread Index | Author Index | Historical

Re: "Bad bgp identifier"

  • From: Owen DeLong
  • Date: Fri Mar 31 06:17:51 2006

Unicast currently ends at is multicast and I believe that are listed as reserved for experimental purposes.


--On March 31, 2006 5:06:54 AM -0500 Joe Maimon <[email protected]> wrote:

4271 specifies that bgp identifier must be a valid unicast ip address

So what is the larget 32 bit value expressed as a dotted quad that meets
this requirement?

Is it the last address in class c? class e? Can 255.x.x.x be used?

Do all vendors implement this?

I understand that  draft-ietf-idr-bgp-identifier-06.txt does away with
the above requirements. Is this something I should ask vendors if they
will support it?



If this message was not signed with gpg key 0FE2AA3D, it's probably
a forgery.

Attachment: pgp00033.pgp
Description: PGP signature