Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watnot!watmath!clyde!rutgers!ames!ucbcad!ucbvax!CS.UTAH.EDU!peterson
From: peterson@CS.UTAH.EDU.UUCP
Newsgroups: comp.sys.workstations
Subject: Distributed computations for computer animation
Message-ID: <8704110124.AA13781@cs.utah.edu>
Date: Fri, 10-Apr-87 20:24:39 EST
Article-I.D.: cs.8704110124.AA13781
Posted: Fri Apr 10 20:24:39 1987
Date-Received: Thu, 16-Apr-87 01:14:46 EST
Sender: daemon@ucbvax.BERKELEY.EDU
Distribution: world
Organization: The ARPA Internet
Lines: 28
Keywords: workstations networks ray tracing animation

As part of a local seminar project, we're working with a system that
uses idle time on a network of workstations for generating computer
animation.  I'm interested in hearing about other systems people may
have developed or worked with to perform similar sorts of tasks.  The
kinds of information I'm interested in:

 - How was the work to be done specified?  What sort of
   logs/statistics were kept?
 - What sort of granularity (CPU time/frames/scanlines/pixels) did
   the system operate with?
 - What sort of fault tolerance was incorporated into the system?
   How did it deal with worker nodes or the main server going away,
   network glitches, etc?
 - How was the input data distributed and updated?
   How was collecting the results of the computation handled?
 - What criteria was used to determine if a machine was available or
   not? (server status, somebody logged in, time of day, etc..)
 - Any interesting statistics?  (CPU utilization, total time,
   reliability, etc)
 - Any other interesting experiences.

If people are interested, I'll be happy to summarize the replies for
the net in a couple weeks.

Thanks,
John Peterson
{seismo,decvax,ihnp4}!utah-cs!peterson
peterson@cs.utah.edu