I was wondering if anyone can comment on the scalability and efficiency of MLPPP. I would like to direct this question to those who work at telcos, as well as Mr. Berkowitz, could you please please provide me with some information? For example, I have noticed Cisco documentation about the 10000 ESR that states that you can link 10 DS-1's. Why just 10, what happens if you try to bundle more? And exactly how efficient is it when 10 DS-1's are linked together like that - as I have to believe there must be significant overhead that reduces the throughput below the theoretical limit of 10*DS1= 15.44 Mbps. But exactly how much bandwidth could you expect - 10 M? 12M? More? http://www.cisco.com/univercd/cc/td/doc/product/aggr/10000/techedge/ch1ovw.h tm Also, in a discussion with one of the head guys at UUNet, he scoffed at the notion of being able to use multilink to deal with very high-capacity future networks. For example, he gave the example of having 100 cross-country optical links, each using DWDM with 100 wavelengths, for a grand total of 10,000 links, and he completely dismissed the idea of using multilink to bond 10,000 links together. Now 10,000 is indeed ridiculous, but then that begs the question - what's reasonable? 10 (like the Cisco ESR)? 100? 1000? Unfortunately, I don't have the link of that guy's interview on me, I'll find it soon. Finally, I've noticed that Cisco recommends for async links, that PPP not be used for more than 3 links. Why? Exactly what happens when you bundle more than 3 async links? Thanks in advance. Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=7685&t=7685 -------------------------------------------------- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
