However, now, a new class of applications is on the horizon that will prompt a reevaluation of our backbone networks.

Multicast Maintenance If the last time you considered multicast was during your initial IP 101 training course, then the jabbering in the press about Internet/intranet-based push applications should awaken you from your slumber.

Why? Because your network is out of date. Yes, it's out of date. Chances are you started putting in bridges and routers six or seven years ago. Early versions of many LAN switches, for instance, broadcast all multicast packets. How much of that equipment is still out there plugging away, with just the occasional tweak for maintenance? If your network is like most others, the answer is probably a fair amount.

The bottom line is that much of that gear won't support multicast in its current state, and you'll have a fairly large problem on your hands upgrading firmware, memory capacity and software throughout most of the network.

Don't Get Pushed Around Push technologies have gotten quite a lot of media attention lately, but isn't all this push nonsens e just a way to jam down user's throats the low-quality content that nobody would take the trouble to browse on their own? Sometimes, I liken PointCast to getting hit over the head repeatedly with a rolled-up copy of USA Today. At least with TV, the netw orks don't ask me to pay for it directly. Would-be Internet broadcasters are asking us to pay to upgrade our networks to support their broadcasts. Multicast to desktops on your network depends on you to create the underlying multicast delivery infrastructure. With all of this effort, the resulting content had better contribute to the bottom line, not the intranet equivalent of broadcast TV.

But then again, no one ever got rich being a curmudgeon. You can challenge the value of applications like PointCast all you like, but the fact is that more and more of your users are running it everyday. Video-on-demand is another application that users think means multicast. But video-on-demand means you want some video on your desktop now--a unicast, in other words. Classroom learning could well benefit from multicast. Sometimes even real work gets done. For example, Starburst (it's a candy for those of you who've forgotten) has a software distribution product based on multicast.

Multicast is like nuclear energy: a very useful technology with lots of benefit for all. However, when things go wrong, everyone gets fried. So, paying attention to multicast is well worth the effort. It's also worth spending some time on planning when you begin contemplating multicast implementation on your network. Many multicast software applications' vendors would have you believe that multicast is something you simply switch on in the network. Lesson one: Guess again. Lesson two: Tell the vendors if the content doesn't contribute to the bottom line, they'll need to try again--broadcast of live TV or radio news doesn't really sound like a compelling business imperative to me.