Actually, they did use all 52 machines for email. The set up looks like a real nightmare of strangely split and mixed functions. Here are the clustered configuration details of one very high TCO set up:
Three overpriced computers "on the front end". Four overpriced computers "on the back end", two for staff and one for students. Two overpriced computers on some other end run SMTP, as if the other seven can not. One overpriced computer is used for backups. One overpriced computer is used for testing. The rest of the forty one overpriced computers must have gone into the the "storage area network". The report mentions a "Dell/EMC CX2000" with three TB of storage, one of which is actually used for email, and home directories, the rest of which must be used for administrative work or one copy of Longhorn. It is interesting to note that Google now offers five times as much storage as these 52 machines delivered, (5000 x 1 GB/account). Of course the big buy saved them money, right? Dell claims the new mail system stopped sobig and so saved the University all the time that would have been required to repair 1,200 university desktops. I'll bet 100% of those desktops have some kind of mal/spyware on them by now. Imagine how much the university would have saved if it would simply use a desktop that needs less protection. River Bend had a few of these overpriced computers and they sucked the life out of the people who had to keep them up. I also remember that their network storage was not a very good place to try to save your data because the servers regularly failed. I don't think there was anything wrong with the hardware. Claims of efficiency are clearly nuts when I read here that email can be served to many more students and staff by a couple of Linux boxes. A few other boxes could surely deal with all the chores Dell crows about: class assignments, billing, registration, "and more". On Thursday 17 February 2005 08:53 pm, Shannon Roddy wrote: >... > Here is the case study (read the pdf): > http://www1.us.dell.com/content/topics/global.aspx/casestudies/en/2004_uscs >?c=us&l=en&s=corp > >
