smaller packets than they otherwise could. Some hosts get noticeably higher performance when they are able to use FDDI- sized packets compared to Ethernet-sized packets, and restricting the packet size to 1500 bytes will put a limit on the maximum
Some hard figures on this would be interesting. Ie, % of packets with > 1500 MTU, % performance degradation if fragmented, etc. I suspect that other backbone design issues (like congestion) dominate any fragmentation issue.
I'm not in a position to give you that figure, sorry (I couldn't even if I would). Do however note that this is not so much an issue about (IP) fragmentation happening in routers as it is about efficiency of host interfaces, since most of these guys use (or should use, anyway) path MTU discovery. (It is conceivable that this argument also has weakened over time.)
I'm not sure a few people trying to get a little extra throughput should dictate the design of a NAP (unless they want to pay for it).
In some instances that would be a fair trade-off. But do note that the "packet size" issue is a side issue -- the main argument for using FDDI-style switches with "backpressure" via token stealing is appropriate amounts and handling of buffering. - Havard