An unfortunate drive failure has cost us our test instances and all archive history for Fedora, CentOS, and Ubuntu. I am partially to blame in this (bad luck is to blame for the other part). When ordering and setting up the drives, I ordered only three drives, and put them in JBOD. I should have scrounged around for a fourth drive to make it a RAID system (then we'd at least still have our data).
In March, I purchased several 1TB hard drives for use in the CHASM project (using my research stipend from the Rensselaer Center for Open Source). Since then, my backup workstation (currently a headless server) has been syncing Fedora and Ubuntu every six hours and making a snapshot of the results. Our goal is to have an accurate profile of the nature of the changes in each mirror archive so that we may decrease latency better.
I'm looking into RMA'ing the drive as quickly as possible. In all my years of purchasing hard drives, this is only the second Western Digital drive I've had to RMA (the first was damaged in shipping). I'm hoping the folks at either Newegg or Western Digital can help with a fast turn around so that we can be back in business.
I'd like to just add that Western Digital makes damn fine products, and it's only because of their products that this work has been possible so far. I personally own seven WD1001FALS (three are used for CHASM), two WD20EADS, two WD6400AAKS, and two WD800 drives. It would not be a stretch to say that I'm a WD fan.
-Robert