[Valeri.Markushin@psi.ch: Data Analysis on Merlin Cluster]

[Frederick Gray <fegray@berkeley.edu>]

For your information...

----- Forwarded message from Valeri Markushin <Valeri.Markushin@psi.ch> -----

From: Valeri Markushin <Valeri.Markushin@psi.ch>
Organization: Paul Scherrer Institute
To: David Webber <dwebber@npl.uiuc.edu>,
	Frederick Gray <fegray@berkeley.edu>
Subject: Data Analysis on Merlin Cluster

Hi David and Fred

I've talked to our network people about the potential bottleneck with the 
data
transfer to Merlin.  They suggest to open TroubleTicket(s) to deal with it
on per-experiment basis.
Assuming that each group can use about 10 CPUs continuously and at least 2
groups can run simultaneously, one estimate the needed data transfer rate.
Having your tape stations in the computer room is not an option, but
other locations, including the experimental hall, may be discussed.
When you submit a TroubleTicket, please, specify the expected data transfer
rate (average and peak), the amount of data to be transfered (per 
week/month/...),
and the networks, on which you have your data.

Best regards,
Valeri

----- End forwarded message -----