All things related to multicast

[Bill Owens <>]

On Wed, Feb 12, 2014 at 11:18:00AM -0500, Bill Owens wrote:
> Is from MIT! Aren't you glad they're using so much multicast? Let's see 
> what it is:
> 
> 224.16.19.195   18.7.32.13 (w92-gen-arch-3.mit.edu) RP: 18.255.255.48
> 224.16.22.7     18.14.84.12 (m48-exterior-cam-2.mit.edu) RP: 18.255.255.48
> 224.16.22.13    18.14.84.11 (m48-exterior-cam-1.mit.edu) RP: 18.255.255.48
> 224.16.17.172   18.14.89.13 (m9-058-cam-2.mit.edu) RP: 18.255.255.48
> 224.16.27.229   18.14.245.38 (w7-100lb-cam-2.mit.edu) RP: 18.255.255.48
> 224.16.27.217   18.14.245.49 (w7-ext-cam-7.mit.edu) RP: 18.255.255.48
> 
> Oh, dear. Those look like security cameras. Perhaps someone didn't 
> understand the concept of the organization-local scope, and decided they'd 
> just use IANA reserved group addresses instead. 447 of them, by my count, 
> out of a total of 2025 SAs. The good news is that they don't appear to be 
> leaking traffic, that would be really embarassing. As opposed to only being 
> embarassing in front of network geeks.
> 
> BTW, I noticed this because I'm finally fed up with all the crap SAs (and 
> traffic) floating around, and am starting to tighten down our MSDP filters 
> to exclude the reserved ranges. I wasn't going to do 
> 224.5.0.0-224.251.255.255 because it's a pain to write it out in Cisco ACL 
> format, but maybe I will make an exception thanks to MIT's example in this 
> area.

It's been mentioned to me that the attempt at sarcasm and humor (albeit dark 
humor) didn't come through in this note, and it could easily be interpreted 
as a slap at MIT. I should point out, especially for anyone who hasn't been 
on this group for very long, that I've been finding and pointing fingers at 
misconfigurations of multicast like this for years now, including a number 
that were inside NYSERNet. And the sarcasm - that MIT was singlehandedly 
responsible for 22% of global multicast traffic - is perhaps more apparent 
for those who have read my notes questioning the value proposition for 
interdomain multicast. 

Regardless, however, it was not my intention to cause any hurt, and I'm sorry 
if I did. Rest assured that if I trip across another example of multicast 
problems, I'll do my best to convince the responsible parties to fix it at 
the origin rather than holding them up in front of their colleagues.

Bill.