[cross cc'd to v6ops as this sounds more like a ops thing] Paul Vixie wrote: >>>>> A site is a network of computers with a single administration, ... > >>>> Where has the IETF redefined the meaning of the word "site"? ... > >>> This has been a longstanding problem in the IETF; in fact, the inability >>> to agree on what "site" means was one of the reasons SLAs were ... > >> why does it mather if it is physical site, administrative domain ... > > indeed, from the rfc that was quoted at me when i asked about this a month > ago, it appears that "site" is meant as "usually the same thing as an > autonomous system but can mean something else if you want it to." From what I understand at least RIPE NCC is using the meaning along the lines of "A network with the same administrative domain confined to a single or a group of buildings". As such a site could be (amongst others): - a host in a datacenter which provides VPN's to other sites over its own link. - a home network, eg where a DSL or cable line or a PPP link terminates - an office in a home network where a VPN/PPP/DSL terminates - a big office complex, where the network is run by an admin group who have an incoming 1Gbit/s link - multiple buildings around a city/country/world all linked together with VPN using one single /48, but eg a /56 per building. And various other cases. Note that I add the way the network gets connected to the rest of the world. Of course there are cases where one does not link up to the Internet or even to other sites. I am pretty sure that the various RIR's are well known enough with a definition that they handle consistently in their region, which in one way or the other most likely matches the above. Maybe somebody should write a mini-RFC labeled "What is a site" and include some examples to show exactly how/what? Guess it would fit in perfectly for the v6ops WG (cross cc'd for that purpose). Greets, Jeroen
Attachment:
signature.asc
Description: OpenPGP digital signature