Network challenge MAXIMUM AG Event...

[Tony Rimovsky]

This sounds like a great try for an award.

One thing that jumps to mind: If you are planning on participating, please, 
Please, PLEASE warn your networking groups.  If we do this in a single room 
(one multicast address) then every bit that makes it to dallas for the 
contest will also go to every multicast receiver in that room.

I'll go ahead and warn vBNS and Abilene.  :)

Rick -- is it permissible within the rules to do this as multiple 
streams?  ie., can we spread the load across several venues and have the 
aggregate at Dallas count towards the contests?  It will help spread the 
load out at the campus sites.  One site I am worried about, for example, is 
ACCESS-DC.  If they participate, I fully expect this to saturate their DS3 
if they see all of the traffic.  I am even starting to worry about the 
nodes 100bT connections.



At 08:12 PM 11/2/2000 -0600, you wrote:
>Hi AGsters,
>
>I wanted to send a follow up note to Terry's email about the SC2000 
>Network Challenge.
>
>We have signed the Access Grid up as one of the four or five applications 
>at SC that will be trying to MAX OUT the networks.  The SCInet crew and 
>the judges from the Network Challenge will be measuring the bandwidth used 
>for the AG from noon on Tuesday to noon on Weds.  Therefore we would like 
>to try to some very high-bandwidth sessions on the AG network during the 
>monitoring time..
>
>We will have the AG nodes at ANL and in the ANL booth at SC up during this 
>time and are scheduling some high-resolution video experiments and some 
>Voyager playback experiments.. We would like to invite the entire AG 
>community to actively participate in this Network Challenge. It would very 
>cool to try to set a record for simultaneous AG node connections and 
>number of video streams.  More details will be sent to the list as we 
>learn more about how this is going to work..
>
>Cheers,
>
>--Rick
>
>