Here a few more items to add to Chris' very sensible comments... All our materials are backed up on CDs. We established redundancy for the"mission critical servers" (all beginning and intermediate materials) - so PC and Mac users can go to either one of two locations on the network. We do keep track of the various drives on our servers and replace drives about three months before the warranty period is over. The one thing which helped us up here in the backwoods of New Hampshire are UPS's - uninterruptible power supplies - that take care of power failures that happen rather often (for about 20 minutes!). We do keep spare power supplies and spare drives on hand for emergencies. We also supply faculty and students with a small set of instructions for log-in procedures and redundant services as well as emergency numbers to call. Our most challenging problem iin this context has been to make our clients aware of the two different locations on the network and keeping these locations, their names, and their IP's over the years. I do check the servers via remote log-in from at home on weekends - not a recommended procedure! But I live only about 10 minutes from my office and can walk over easily in case of trouble. Fortunately, this was not necessary over the last two years. Our streaming services are slightly easier to manage for redundancy purposes, since we keep our home page on the main Dartmouth www machine which should be up 365/7/24..... One things that can be very helpful for the purpose of getting out of a major catastrophe is a backed up system installation or at least a complete record of the system and server software installation procedure. Also - mission critical servers should have an absolute minimum of accounts to make rebuilding easier.... especially when a catastrophe occurs. Feel free to give me a call, if you have any questions..... Otmar Foelsche (llti-editor)