The IP over ATM Mailing List Archive by date[Date Prev][Date Next][Thread Prev][Thread Next] [Date Index][Thread Index][Author Index][Subject Index] IP flatnet <> ATM flatnet
It seems to me that there are some very good reasons why IP (or IPv6) will be needed for a very long time to come (ie, why ATM can't be the only end to end "network" layer protocol): - Incompleteness of the ATM standards. A lot of the parts aren't done yet. (eg, What is the global addressing plan, equivalent to CIDR)? I think that ATM is a long way from being able to scale to 10 times the size of the current Internet. - Set up time and limits on the number of VCs. It doesn't make much sense to set up an SVC for very short term transactions (such as using the www stock server to inquire "how is my companies' stock doing today"). There needs to be some "datagram mode" for short term transactions, and some datagram protocol over ATM (such as IP) is what we have currently. - Virtual networks: If you want to be able to move the physical connection of a host, and leave its IP address the same, then the IP address will need to be mapped to something which states where the host actually is. One way to do this is to use virtual / emulated LANs over ATM (this might not be the best way in theory, but it seems to be the best that the world has to deploy right now). IP over ATM as a solution which is desirable to allow compatibility with IP over other stuff (ethernet, appletalk, etc..). However, IP over ATM is also desireable even if we did have a huge "flatnet" in which all hosts were attached to ATM (not that this would happen, but even if it did we would still need IP over ATM). To me this means that the "IP over ATM" scenario is very realistic. The "IP without ATM" scenario is also realistic. However, the "IP goes away and IPv6 never shows up and ATM works as the only network layer protocol" scenario seems majorly unlikely to me. Ross |
|