We commonly use /32 subnets on routers when we have multiple (redundant) paths to the equipment and therefore need to know the difference between an interface down event and a device down event.
While PAT and NAT addressing uses a single IP address, we commonly bind it to the router's Ethernet interface with a /30. RFC 3021 is not currently a 'Standard,' but rather a 'Proposed Standard' and as I understand it, the change to a Standard will occur when and if it is widely adopted. Right now it's a toss up. On the other hand, use of /32 is a Standard. Having read RFC 3021, I can see why some might want to use it. Though some of use might feel it's the camel's nose under the tent, it seems to have some traction. We use /30's on our point to point connections. To be honest we have more useable IP in reserve now than we did five years ago as company after company as returned large IP assignments to use as they have installed firewall systems. In theory ubiquitous use of firewall NAT/PAT translations should release sufficient IP space for a good long time. It's really a matter of economics for those who really do need more address space and who's RIR annual fee will increase by thousands of dollars a year if they can't maximize their existing IP space and must seek a larger assignment. We use subnet-zero /30 assignments on routers and have done so for many years. IM has no problem with this. Mike Lieberman -----Original Message----- From: [email protected] [mailto:[EMAIL PROTECTED] On Behalf Of Richard E. Brown Sent: Friday, March 02, 2007 11:14 AM To: [email protected] Subject: RE: [IM-Talk] /31 network masks Folks: Thanks for the spirited discussion of how you use various kinds of subnets. It's helpful to hear all this so that InterMapper can "do the right thing" for our customers. In pursuit of this goal, I'd like to expand the discussion so that we can understand all the possibilities. 1) /31 subnets. It seems as if there are proponents of both sides: people who never intend to deploy a /31 subnet, (and want a facility to warn them if they've done so); and people who regularly use them on point-to-point links as described in RFC 3021. 2) What about /32 subnets? Is anyone using them? (see the links below that discuss how they're used...) 3) Are there other unusual subnets/related network configurations that we should take into account? My question to the group: What should/could *InterMapper* do when presented with equipment that uses any of these configurations? I've googled around and come up with a few citations on the subject. See the links at the bottom of this message. Many thanks! Rich Brown [EMAIL PROTECTED] Dartware, LLC http://www.dartware.com 10 Buck Road, PO Box 130 Telephone: 603-643-9600 Hanover, NH 03755-0130 USA Fax: 603-643-2289 --------- Sources of info ----------- /31 subnets: RFC 3021 (31-bit prefixes): http://www.faqs.org/rfcs/rfc3021.html Cisco's "Using 31-bit prefixes..." note: http://www.cisco.com/univercd/cc/td/doc/product/software/ios122/122newft/122 t/12 2t2/ft31addr.htm /32 subnets: Subnet Zero and All-ones Subnet: http://www.cisco.com/warp/public/105/40.html Stanford's Experience: http://rescomp.stanford.edu/staff/manual/studevelop/projects/iprequest/netwo rk_c onfiguration.html ____________________________________________________________________ List archives: http://www.mail-archive.com/intermapper-talk%40list.dartware.com/ To unsubscribe: send email to: [EMAIL PROTECTED] ____________________________________________________________________ List archives: http://www.mail-archive.com/intermapper-talk%40list.dartware.com/ To unsubscribe: send email to: [EMAIL PROTECTED]
