"I've never used multicasting in a real-world application but will be keeping my eyes open for an opportunity". Don't. Seriously, just don't do it. I personally

"I've never used multicasting in a real-world application but will be keeping my eyes open for an opportunity". Don't. Seriously, just don't do it. I personally know of outages directly attributed to multicast that have cost more than your $widget will ever make. But, you say, I've got the textbook case of media broadcast! Yeah, still probably not worth it. Check out UDP broadcast, or tiered distribution, or service and peer discovery. There's another way to do what you want. No way! I know all about multicast forwarding, and every node really does need all of these packets Really? You're going to build reliable delivery at the application layer? And deliver yet more mcast to every node. And you won't have contention problems, across your entire domain, for the limited byte & namespace resources? And it won't be a problem when every packet must be routed to every TOR? Because you will have heterogenous hosts in each rack subscribed to every available group. So, really, don't do it. It looks cool. It looks efficient. It looks simple. It's not. In grown up land it's just not worth it.