NETGEAR is aware of a growing number of phone and online scams. To learn how to stay safe click here.
Forum Discussion
can data be recovered from a single failed drive?
The question is almost, but not quite, as dumb as it sounds. I have an Ultra 6+ with a single x-raid-2 volume comprised of four 3TB drives. One drive failed, and is no longer recognized by the R...
Not exactly - and it depends on what you mean by "recoverable". Mdgm's equations should not be taken literally :D
xeltros wrote: ...As far as I know raid 5 is totally unrecoverable if you have more than one drive failing. If you have four drives they'll have about 1/3 of you data (the fourth drive being a backup).
So what they'll have is that equation : x+y+z=s...
RAID-5 with 4 drives uses block striping. Blocks are grouped into 3 data blocks and one parity block (one placed on each drive). The data blocks are not mathematically coded. The parity block is the XOR of the other 3.
On SPARC v1 units (which use hardware raid), all the parity blocks are organized onto drive 4 - on the so-called "parity disk" (sometimes this arrangement is called RAID-4 btw).
The other ReadyNAS (which use software RAID) distribute the parity blocks evenly across the 4 disks. You can see a picture here: http://en.wikipedia.org/wiki/RAID if you scroll down.
In this case we are talking about an ultra, so the parity is distributed. Each disk in the volume set holds 1/4 of the data blocks (perfectly recoverable) and 1/4 of the parity blocks (not recoverable). The recoverable portion is 2.25TB of the total 9 TB volume, the parity is the other 0.75 TB. Sorting out what these datablocks contain would be difficult, but one could search them for data patterns, etc. The attacker wouldn't know the parity organization, but can simply search those blocks along with all the rest.
I think the risk is fairly low, particularly with an RMA to the manufacturer. Generally I attempt to zero the drive with the manufacturer's diags before returning (actually seagate's write zeros test works on WDC drives, and vice versa). Though if the drive failure prevents that, I don't worry about it. One reason for the zeroing is that it often does fail, and gives me an error code for the RMA.
The ATA secure erase command is available on all drives manufactured after 2001, so it is another option. If the drive is hardware-encrypted, then the enhanced secure erase command is extremely fast (it just changes the encryption key). Otherwise, it commands the drive to overwrite every sector. BTW, third party programs to overwrite the drive are not better than this (and are generally less effective) - there is a tutorial here if you are interested: http://cmrr.ucsd.edu/people/Hughes/docu ... torial.pdf
NETGEAR Academy
Boost your skills with the Netgear Academy - Get trained, certified and stay ahead with the latest Netgear technology!
Join Us!
