Hi Mike,
"What the heck happens to legacy ATM-SONET MAN/WAN networks if a carrier starts to rollout GE in the WAN/MAN?"
IMO, in most instances nothing measurable will happen immediately in the negative to the legacy nets. The legacy nets might grow less rapidly, or at some point in time flatten out entirely and then be replaced over the next half-decade or so, but the new GbE nets (as well as those which are founded on fibre channel and other non-ITU sanctioned WAN prtocols) aren't supplanting older forms of networking just yet, as much as they are revolutionizing how work flows take place a la ecommerce, and through techniques that take advantage of distance neutralization.
Thus, these new networks support new applications thanks to the enabling characteristics of optical. And many firms will begin their exploits in this direction by installing a "pilot" network for experimental use and proving in, first, which could take anywhere between a six months and a year or more to fully test and evaluate.
In other words, while I see some cannibalization in the works here, these newer networks are largely incremental, rather than being immediately supplantive, for now. For example, optical will allow SANs to operate over vast distances now, whereas they were once relegated to in-building, or in-campus distances in the past.
Likewise, servers no longer need to be local to the same building in which they are accessed, thus allowing centralized server locations to service mulitple enterprise sites within a given MAN, and beyond (WAN) through the use of IEEE and ANSI LAN / FC protocols, instead of those which are strictly tied down to SONET/ATM/ISDN/etc.
To be sure, howewver, over time they will overtake many of the legacy forms of SONET/ATM networking that you referenced.
"I recently read some PR about both of them [gbe and atm] being compatible on the latest incarnation of metro-DWDM/ metro-switching platforms."
I don't know what the p-r might have meant by that. I could only guess that they meant that both GbE and ATM are able to be encapsulated in SONET, which is true, and that DWDM units use SONET framing on their i/o's. Syllogistically, then, the statement about compatibility would be accurate if all of the premises I've presented here are true, and you are so inclined to come to that conclusion. No?
Maybe you can find that p-r, just to be sure...?
FAC |
