RFC 5771 - Global Multicast Addresses

Sat Aug 3 03:06:35 UTC 2019

Good Evening,

I'm looking for some insight into the usage of a few of the blocks defined in RFC 5771 (and IPv6 Multicast Addressing as described in RFC 4291 and 7346) , specifically regarding their use on the public internet. I know multicast isn't routed  on the public internet. However, it appears that this may be due to operational issues and concerns or that it was simply never implemented. The specific blocks I'm looking at are (as defined in RFC 5771 Section 3-10:

Internetwork Control Block - 224.0.1.0/24
> Addresses in the Internetwork Control Block are used for protocol control traffic that MAY be forwarded through the Internet

AD-HOC Blocks (I, II, and III) - 224.0.2.0 - 224.0.255.255, 224.3.0.0 - 224.4.255.255, and 233.252.0.0 - 233.255.255.255
> These addresses MAY be globally routed

GLOP Block (233/8) - Global routing is never mentioned in RFC 5771, but given the context and the use of ASNs, I'm not sure if the intention was for these to be publicly routable or to simply to guarantee that the address would be unique within your AS (are large telcos and webscale companies exhausting 239.0.0.0/8?)

IPv6 Multicast Addresses with scope 0xE - The RFC doesn't really go into detail on how these would be used.

As a young network engineer (no historic perspective) and only SMB and enterprise experience. It seems like the intention was to allow these to be publicly routed, but it would be a nightmare to implement so it never was.

You'd probably require PIM Sparse Mode (we can't flood traffic to the entire internet), RPs would need to be advertised somehow (maybe BSR could be implemented with RP advertisements coming from the providers edge?). RPF would be a constant process and shortest path trees would change constantly.

That's all without mentioning 224.0.1.0/24 is tiny and the AD-HOC and GLOP blocks aren't exactly huge given the size of the internet.

I'd love to hear what others have to say about this, maybe get some historic perspective and thoughts on whether or not any of this will change as IPv6 adoption increases. I'd also love to see any guidance on actually implementing multicast on the internet from IANA or the IETF (or guidance that says that it should not or can not be done) as I wasn't able to find any.

Regards,

Brandon
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nanog.org/pipermail/nanog/attachments/20190803/091f6a1d/attachment.html>