All things related to multicast

[James Harr <>]

On the topic of traffic levels across internet2, I’ve been periodically pulling data about traffic volume and active streams.

 

https://docs.google.com/document/d/1R0mfeUc4MWZd-9uC_nYbYssY0rIOgTbhdvCxhT63W8Y/edit

 

The details of the reports are technically public (pulled from a public looking glass), however please don’t share this outside of this group for now.

 

If someone notices I forgot to update the sheet, bug me (email/Slack) and I’ll do another run and include it. If you want to share it outside this group, message me and I’ll make sure it’s OK and keep tabs on who it’s shared with.

 

--

James Harr

 

 

On 3/23/21, 14:54, "" <> wrote:

Greetings, multicast-wg friends:

 

Thank for a good conversation this afternoon.  The key points I'd like

to summarize at this point are:

 

The NGI project is effectively clean-slating network configurations, and

all configurations and protocols need to be re-justified;

 

Internet2 has, as part of its mission, a commitment to develop

technologies that advance the state of networking technology;

 

Production multicast traffic (to put another way, traffic that, if

disrupted, would result in a call to the NOC) on the I2 network is light.

 

The request I made at the end of the call: let's document the use cases.

  The 'hard' uses cases of current, or imminent (if known) science uses

workflows that use multicast on the backbone; and

the 'softer' use case of we, the I2 community, has a responsibility to

maintain a testbed for protocol/application development and testing.

 

The second point I think specifically asked the interested industry

partners and Internet2 peers to think about and communicate.

 

Thank you.

 

 

--

________________________________

Andrew Gallo

The George Washington University