>My co-author added this statement to the book I'm working on: > >".... an ATM virtual circuit may begin on an OC-3 fiber link, cross over to >a T3 line, pass across a Gigabit Ethernet fiber backbone, and end up going >out through an OC-12 fiber link. This may be an implementation of a single >ATM circuit, however."
> >Could an ATM virtual circuit really span an Ethernet backbone?? This is one of those protocol questions that at first makes me think of Bill Clinton expounding on the definition of "is," but more appropriately, Butler Lampson's observation that "no problem in computer science is insoluble with a sufficient level of indirection." Zeroth, it's not clear what the speed of the ATM VC is, which would influence the technologies that could be used. First, if the statement had referred to 10 Gigabit Ethernet fiber, I could argue that that is simply the same physical medium dependent layer that underlies OC-192 SONET as well as 10 GE. Second, is GMPLS involved? I'm almost certain there is an encapsulation for cells. Third, if there were segmentation, reassembly, and resegmentation, would the customer ever know? Fourth, there may be conversions anyway. By the statement "begin on an OC-3 fiber link," I wouldn't interpret that as raw OC-3 over fiber, but, since it's described as subsequently traversing a T3 line, then is there SONET encapsulating one DS-3 or slower signal within the OC-3? > >Thanks! I don't want to be one of those authors that propagates >misinformation. ;-) Seriously, some mistakes are unavoidable, but this one >seems avoidable (if it is a mistake) > >Priscilla > >________________________ > >Priscilla Oppenheimer >http://www.priscilla.com Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=28799&t=28774 -------------------------------------------------- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]