I need a complete storage solution for a medium sized corporation. We have one main facility and two offsite locations all connected together through an internal ds1 WAN.
I am wondering what would be the most cost effective way to provide a complete storage solution solution for all facilities, and provides a fast, reliable, and scalable means for users to store data.
Should there be one centralized data server at the main facility that is accessible by all locations? Or should each location have their own data servers? What are the advantages or disadvantages of doing something like this? I would like to start out with close to a half of a terabyte of storage available for all facilities, and be able to scale beyond that if neccesary in the future. This could be easily and cost effectively achieved using large IDE drives, but how greatly is that going to effect access speeds, especially across the WAN? For the most part the data is only going to be written to the server, and not really read or executed so do I need the extra speedy access times of a SCSI solution?
What about the server that houses these drives? I would like to run a large raid with each drive having its own bus, so I would assume that I would need a large double wide type server to house all these drives. From what I know most of those servers are more oriented around scsi solutions. So how would I go about getting the extra ide buses if I went the ide route? Just add extra ide controllers? Is that the best way to approach it?
I am basically looking for any advice on how to approach this. Perhaps you can list solutions you have implemented in similar situations or refer me to some links so I can do some research on my own. Any advice wouldbe appreciated. Thanks for your time.
Jim