Poor Steve (Mr. Furrygoat). Like many geeks out there he bought a RAID5 system thinking it was a silver bullet providing massive reliable storage for all his precious data (precious to him, due to the time it took to back up all his DVDs, not necessarily because the content iteself is precious).
A drive failed in the array. The rebuild took over a day for only 300GB of stuff; that’s in addition to the time it took to get the replacement time. Steve was smart enough not to risk total loss by not using the array while it was in its compromised state.
We are facing a brave new world with the combination of massive amounts of storage in the home and horrible technologies for managing and dealing with it all. A RAID 5 rebuild of 300GB takes over a day. Copying 1TB over a 1.5Mbps broadband connection will take 2 months. Over a 100Mbps LAN it will take a day. Over Gigabit Ethernet still takes 2.2hours. Think about it.
It’s true; RAID5 is far from the perfect solution. You still will have a heck of a time, as Steve found out, if a drive fails.
I find it interesting that he points out in his post on the topic that he couldn’t imagine an average user going through the rebuild process, or even using RAID5 for that matter. Both are very valid points. The average home user (I know a LOT of people doing this) will go out and buy a big drive, maybe 300 or 400 GB, and fill it full of all their "precious" data; digital pics, ripped DVDs/CDs, and so on. While RAID5 is not necessarily convenient to work with if a drive fails, you will be in much better shape with it than with the setup of the average home user. It is very true though that it wouldn’t be realistic to expect the average user to put together a RAID set… so what is the answer?
Hey Josh, here’s some fairy dust for you to sprinkle in the magical utopia you live in.