About a year and a half ago, I made a HTPC/NAS for my apartment that we hooked to our TV and use to download and watch movies, TV shows, listen to music, etc.
In making it, I used 3 1TB hard drives I had laying around, so they have all seen use in the part. I somehow corralled them into a single 2.7TB volume, taking advantage of Ubuntu's ability to have volumes constituting multiple physical volumes (logical volumes? I forget the technical term).
Bottom line is, all of the media is stored on this drive, and of course one of the hard drives seems to be beginning to fail (the volume will not mount, if it does mount it is unusably slow, or content is "missing). Obviously this is an expected downside of the way I chose to structure the storage.
Does anyone have a good suggestion on a way to restructure the storage so as to avoid one drive failing jeopardizing all the files? Obviously one way would be to have three separate volumes and partition the content across them. Does anyone have any more elegant solutions?
Also, any advice on preserving the individual drives for as long as possible? I know that when the drives are failing on the hardware side, not much can be done on the software side. But are there any utilities built into Ubuntu or that I could download? I personally have had trouble finding any.
Thank you!
You are looking for raid. A raid5 array can suffer the failure of a single drive and keep on chugging. You only get (n-1)/n as much storage though, or 2/3 the space for a 3 disk array.
As for how to recover now, the first thing you want to do is get a fresh drive and clone the contents of the failing drive onto the fresh one using dd or dd_rescue.
On the other hand, raid is no substitute for proper backups, since there are plenty of other things that can eat your data besides a failing drive, so whether you use raid or not, you still need to keep proper backups if you would mind terribly much if you lost the data.
No comments:
Post a Comment