NForce 4 Raid 0 Stability (General)

[Dowfen]

I've got a raid 0 setup with two WD 120 GB SATA 7200 RMP drives on a NForce 4 Asus A8N-SLI motherboard. It runs great, but what kind of stability should be expected of it? In other words, is it less stable than just running a single SATA/PATA drive?

Thanks in advance for any help,

Eric

 

[MDE]

If one drive goes, it all goes. That said you shouldn't expect either drive to die. Just make frequent backups (should do this regardless of RAID) and you'll be fine in the end. I'm running two 36GB Raptors in a RAID0 on my Epox nForce3 Ultra board (and formerly on an Intel 865 board and nForce2 board) and haven't had any problems related to the RAID array once everything was up and running.

 

[ribbon13]

Nothing wrong with RAID0 if you keep regular back ups.

A nice Seagate or WD IDE drive in one of these is worth thier weight in gold.
AMS DS-2316CBK
AMS DS-2316B2BK

 

[Dowfen]

I'm basically just asking whether or not they are inheritantly unstable. Obviously, if I had a single drive installed and it died, the data would be gone too. So, is the chance of the RAID failing , not the physical disk (you could still use the disk after reformatting/rebuilding array), very large?

 

[ribbon13]

the only way a RAID0 would 'fail' is if the of the member disks fail. A RAID0 can become corrupted (logical, soft error because of bad data) but that is very rare with good controllers.

 

[elecrzy]

well, i guess you can say with RAID 0, you're twice as likely to lose data. This means you just cut your system's or data storage reliability in half.

Example: http://forums.anandtech.com/messageview...atid=27&threadid=1542729&enterthread=y

 

[boomerang]

Originally posted by: elecrzy
well, i guess you can say with RAID 0, you're twice as likely to lose data. This means you just cut your system's or data storage reliability in half.

Example: http://forums.anandtech.com/messageview...atid=27&threadid=1542729&enterthread=y

Got any statistics to back that statement up?

 

[JBT]

Statistics? Its just plain simple math... if a drive has a 5 % chance its going to die with in a year then you add another one that also has a 5% chance... its almost twice as likely one of them will die IE a little less than 10%. I forget the exact formula but it IS true.

Anyhow I'm not saying that it is a 5% chance one will go down I was just using that as an example. It is likely much less than that.

 

[ribbon13]

Statistical analysis of the empirical data... duh! Where n == # of member disks -> (MTBF / n) == MTBF

 

[Dowfen]

Thanks guys!