Re: SWSL cost model (draft 0.01) from Bruce Barkstrom on 2003-10-14 (www-ws@w3.org from October 2003)

From: Bruce Barkstrom <Bruce.R.Barkstrom@nasa.gov>
Date: Tue, 14 Oct 2003 18:04:21 -0400
To: www-ws@w3.org
Message-Id: <5.1.0.14.2.20031014180038.0206c008@express.larc.nasa.gov>

There are some more serious issues that come out of the grid scheduling
community - particularly those that relate to the cost of reserving some
computational or storage capacity at a future date, possibly with the
presence of competition and the possibility of backing out of or cancelling
the reservation at the last moment.

A serious cost model for pricing may need to include a model for amortizing
the investment in hardware/software/management, as well as a method for
dealing with the cost of operations.  It also has to include some method of
forecasting future sales (or projected future sales), as well as deciding 
on what
kind of feedback is needed.

Bruce R. Barkstrom

Received on Tuesday, 14 October 2003 18:13:28 UTC