All things related to multicast

["Joe St Sauver" <>]

Gary commented on Steve Wallace's note:

#> I had no idea multicast was being used for usenet. That's cool. Still, I
#> suspect the bit rate for usenet is way south of 3Gb/s. 

As of January 2011, Usenet was running just under 9TB/day (see for example
http://www.ghacks.net/2011/01/26/usenet-traffic-growth-to-almost-9tb-per-day/)

The 9-10TB/day number appears to be surprisingly consistent over time 
(although I'd be interested in hearing about any public statistics that 
show something substantially above that value on a consistent basis now).

Depending on your assumptions (e.g., is the load perfectly buffered and 
uniform, or does it see peaks and troughs in throughput volume?), and the
extent to which a site peers with other news servers, you can estimate the 
volume of traffic that you'd see on the wire from Usenet in in Gp/s.

#Sorry Steve, I didn't make myself clear. Usenet doesn't, afaik, use 
multicast.

There was one NNTP over multicast implementation, Newscaster, that I'm aware
of, but to the best of my knowledge that's not a currently active effort.

#What I was getting at is that, end-to-end, nntp doesn't guarantee delivery of
#data from the poster to the reader, so the poster can employ FEC in the form 
of
#PAR2 chunks to ens ure that the reader can recreate any missing blocks of 
data
#that didn't get through to their local usenet node. I see the same method 
being
#potentially useful when you can't guarantee delivery of data via UDP for 
large
#datasets over multicast. 

Usenet's flooding model, when used with multple redundant peers in a well-
connected mesh, can be surprisingly efficient at efficiently and completely
distributing large volumes of content.

Regards,

Joe