Hi folks, I'm busily writing up my next lecture for my TCP/IP course at CUNY Hunter (text is Comer vol. 1 4th edition), the topic is the base Internet Protocol datagram format. I come to the TOS field which, as all know, was originally specified in the base IP specification RFC 791 then attempt to extend and reconcile it with later RFCs made in RFC 1349 followed subsequently in RFC 2474 in late 1998 which completely changed it to support DiffServ. In RFC 791 on page 11, Jon Postel writes: Several networks offer service precedence, which somehow treats high precedence traffic as more important than other traffic (generally by accepting only traffic above a certain precedence at time of high load). The major choice is a three way tradeoff between low-delay, high-reliability, and high-throughput. Was this something actually supported in the Internet? Widely? any examples of who? Around when did it stop being supported? Did anyone ever actually support RFC1349 in a host or router? How about DiffServ as specified in 2474 and 2475? I believe there are router implementations but is anyone actually got it turned on ? Possibly only in an IP "intranet"? Dana Hudes CSCI Dept CUNY Hunter College
Dana Hudes wrote:
Was this something actually supported in the Internet? Widely? any examples of who? Around when did it stop being supported? Did anyone ever actually support RFC1349 in a host or router?
Yes, on the half-dozen or so routers that I worked on, the low delay bit was supported. This was especially important for dial-up links. (NetBlazer, Lan'sEnd, etc., none of which are in much use today.) I have also _set_ the low delay bit for telnet traffic on those boxen, but you don't telnet out of routers very often. I'd have to check the source, but I'm pretty sure I put at least some of that stuff in Qualcomm/Sony cell phones and base stations, so it might still be in use today. I have also used the TOS bits in a weighted fair queuing scheme. I never figured out how "high reliability" would be implemented. I just tried to never have low reliability. :-) WSimpson@UMich.edu Key fingerprint = 17 40 5E 67 15 6F 31 26 DD 0D B9 9B 6A 15 2C 32
thanks for the information all, and to Ping Pan for reminding me that we used to support TOS on the Milford router. I vaguely recall now that was a feature added late in the product lifecycle, so may have only been available on the IBM Global Network. It is a trivia problem at this point. I have sufficient material to revise my lecture notes. Although I want to point out that low delay is RFC 791 back in 1981. It had precedence and TOS specified. I know all routers support the precedence field, and its interesting about the use of TOS and low delay to avoid dial-up links where possible. Dana ----- Original Message ----- From: "William Allen Simpson" <wsimpson@greendragon.com> To: "Dana Hudes" <dhudes@panix.com> Cc: <nanog@merit.edu> Sent: Tuesday, February 22, 2000 5:32 AM Subject: Re: TOS history?
Dana Hudes wrote:
Was this something actually supported in the Internet? Widely? any examples of who? Around when did it stop being supported? Did anyone ever actually support RFC1349 in a host or router?
Yes, on the half-dozen or so routers that I worked on, the low delay bit was supported. This was especially important for dial-up links. (NetBlazer, Lan'sEnd, etc., none of which are in much use today.)
I have also _set_ the low delay bit for telnet traffic on those boxen, but you don't telnet out of routers very often.
I'd have to check the source, but I'm pretty sure I put at least some of that stuff in Qualcomm/Sony cell phones and base stations, so it might still be in use today.
I have also used the TOS bits in a weighted fair queuing scheme.
I never figured out how "high reliability" would be implemented. I just tried to never have low reliability. :-)
WSimpson@UMich.edu Key fingerprint = 17 40 5E 67 15 6F 31 26 DD 0D B9 9B 6A 15 2C 32
Dana, Here is perhaps another useful historical reference, at least in the context of IP Precedence: Mitigating the coming Internet crunch: multiple service levels via Precedence R. Bohn, H.-W. Braun, K. Claffy and S. Wolff, submitted for publication, November 1993. http://www.nlanr.net/Papers/mcic.html - paul At 11:59 PM 02/21/2000 -0500, Dana Hudes wrote:
Hi folks, I'm busily writing up my next lecture for my TCP/IP course at CUNY Hunter (text is Comer vol. 1 4th edition), the topic is the base Internet Protocol datagram format. I come to the TOS field which, as all know, was originally specified in the base IP specification RFC 791 then attempt to extend and reconcile it with later RFCs made in RFC 1349 followed subsequently in RFC 2474 in late 1998 which completely changed it to support DiffServ. In RFC 791 on page 11, Jon Postel writes:
participants (3)
-
Dana Hudes -
Paul Ferguson -
William Allen Simpson