Dead Drive in RAID 0 Array

[barmstrong]

I have a gigabyte 8iexp and one of the hard drives in my RAID 0 array died. is there any way to recover the data or do i just unplug the dead one and reinstall everything on the good hard drive alone? im not sure as to the extent of the hard drive's death, is it possible to image the hard drives if they are still functional at all? can you image a RAID array? does the OS control where the data is in the array on each drive or is it the controller that does it? like if you boot into DOS, will the array still be functional? i was hoping that if one of the drives are functional, they could be put in another machine with an 8IEXP and imaged then extract what data i can from them. please reply, thanks.

 

[JC]

Sorry, but I think your data's hosed 🙁 I'm pretty sure only mirroring or parity would let you rebuild the array. Booting into DOS won't help you.

Bummer, RAID 0 is like flying without a net, though 🙁

 

[stevewm]

When you have a RAID array the data is divided up between the drives in the array.

Say you save a 64KB file and your stripe size is set to 32KB. the first 32KB of that file will be written to drive 1 while the last 32KB will be written to drive 2.

When one drive goes half of your data is gone, therefore no chance in recovering it.

Remember, backups backups backups!

 

[barmstrong]

thanks guys, turns out the hard drive isnt dead dead, the system boots but then blue screens at a certain point, so the drive is still readable in parts. if i throw a drive in the machine on the regular ide controller, would it be possible to try to image the array to the new drive? or could i put another drive in there, install an OS on it and get some of the files that way? the MOST important stuff was backed up so its not a total loss 😀 thanks.

 

[JC]

Originally posted by: barmstrong
thanks guys, turns out the hard drive isnt dead dead, the system boots but then blue screens at a certain point, so the drive is still readable in parts. if i throw a drive in the machine on the regular ide controller, would it be possible to try to image the array to the new drive? or could i put another drive in there, install an OS on it and get some of the files that way? the MOST important stuff was backed up so its not a total loss 😀 thanks.

Sure, I'd give it a try....couldn't hurt 😉

Good luck recovering the data 🙂

 

[Zepper]

Boilerplate-02:
Data Recovery-
. If your drive isn't physically damaged perhaps one of the following will help:
Lost & Found from PowerQuest (the Partition Magic folks). They don't make it any more, but I have seen it listed as "In Stock" here: MSBCD . I've used this to save the bacon for myself and others several times. Be sure to download and install the updates from the PQ web site before using. They would have sold a lot more if they hadn't tried to limit its use to one computer and their protection scheme so easy to circumvent .
. This next is new and even claims to recover date from fragged RAID 0 drives...
. R-Studio You can DL a demo which will tell you what the full pkg will be able to recover--so you can see what you would get before you plunk down your $$$s (I just DLd the demo myself for insurance <g> ). Can't beat that!
.bh.
😎

 

[barmstrong]

k, ran drive fitness test and the drives both came back with bad sectors. i looked up the error code and it said to run erase disk and if it comes back with no errors then the drive is fine. ran erase disk, now they have no errors. does this mean its safe to use them or just IBM's way of not having to take a return?

 

[Zepper]

All the drives I have zeroed out with the mfr's software have been fine afterwards, but I always check it again with other surface testing software. I suggest that when you are running RAID-0 you should be using ECC memory and on a quality UPS. ECC because the IDE RAID (Promise, AMI, HighPoint) on most mobos is software based - quite memory dependent. UPS for obvious reasons. It's a PITA to rebuild a RAID setup so you should take every precaution. Belt and suspenders...
.bh.
😉