As everyone has pointed, be very aware of power. Very. One data center I had had enough power to keep the machines running, but not enough to start them all at once. One of the hurricanes knocked power out long enough for the UPS's to drain. When power was restored, everything tried to come up at once. There wasn't enough juice, it tripped the circuit and they all came down again. The circuit reset, everything started to come up, there wasn't enough, reset. You get the idea. It was a very small unmanned datacenter and there was a hurricane.
With the size of the room, wireless would be a big advantage. Less wire clutter and tripping hazards. If specific machines that will be in the room is static, you could go with 802.11a or g and get the speed boost. If outside people will be bringing machines in, go with 802.11b. Someone else mentioned that you may want wired LAN if you're moving large files around. Heed that advice. If you go with 802.11b and users are twiddling their thumbs waiting for file transfers, you'll be pulling UTP soon enough. I worked one place where we had full spectrum fluorescents bulbs. It was so much nicer than the standard fluorescent bulbs. They're a little more expensive than standard bulbs, but still cheaper than incandescent's power bills in the long run. My $.02 Ken On Wed, 2003-03-12 at 16:59, Joe Meador wrote: > Hey guys & gals, > > The group I work for is about to create a new computer lab in a room 24 x > 20 feet. We'll have 6 PC's with some running Windows and some Linux, 3 SGI > workstations, and a Linux cluster in a rack enclosure. We'll probably > have at least 1 networked printer and possibly a scanner, and will need > workspace for people with laptops. The room currently is not networked > but we are going to get 20 Ethernet jacks installed. Are there any > precautions, suggestions, etc for setting a lab up. The room *should* > already have sufficient ventalation. I am unsure on how I should > physically have the room situated, what size/type desks we should get, > etc. Also what kind of lighting we should look at (I know that the people > in it will probably want fairly soft lighting because they will be using > graphics "stereo" for molecule visualization, etc). > > Thanks for any suggestions, > > Joe Meador > [EMAIL PROTECTED] > > _______________________________________________ > TriLUG mailing list > http://www.trilug.org/mailman/listinfo/trilug > TriLUG Organizational FAQ: > http://www.trilug.org/~lovelace/faq/TriLUG-faq.html
signature.asc
Description: This is a digitally signed message part
