DSM-G600, DNS-3xx and NSA-220 Hack Forum

Yes, i know, so i wanted a fresh start whit the stability focus, does and don´t . I am using my unit for backup, in my world thats the whole idea whit RAID. I know opinions divers in this, but what is the perfekt backup, really?

Well, i aint expecting inahilation any time soon... but i gueass your right. Still, i have a reasonable level of paranoia, so i dubbel back-ups, as it is now on elements external usb hard drives. But if the DNS-323 is as reliable as it looks, it might just as well be the final solution.

blbrown wrote:

For another opinion...  I decided the DNS-323 is reliable if I don't use it in RAID mode--I really want to be able to pull a drive out of the DNS-323 and 100% know that I can stick it in any Linux computer to retrieve the data.  With RAID I just can't be sure.  In addition to that I found and embellished a method of backing up between 2 disks on the DNS-323 that is much nicer.  Check out my work at http://backupnetclone.sourceforge.net/

You can do that with RAID1

blbrown wrote:

fordem wrote:

blbrown wrote:

For another opinion...  I decided the DNS-323 is reliable if I don't use it in RAID mode--I really want to be able to pull a drive out of the DNS-323 and 100% know that I can stick it in any Linux computer to retrieve the data.  With RAID I just can't be sure.  In addition to that I found and embellished a method of backing up between 2 disks on the DNS-323 that is much nicer.  Check out my work at http://backupnetclone.sourceforge.net/

You can do that with RAID1

Good.  I wasn't sure about that when I started, so I didn't go that route.  In the end I'm still happy I chose rsync instead of RAID 1 since I get backup snapshots instead of just one exact duplicate.  Also RAID 1 tends to keep both disks spinning the same amount, which might contribute to both drives failing around the same time since they're identical.

if both drives are identical that is no indication that they will or even could fail around the same time

even if the drives were only 1 SN off from eachother like they came off the assembly line one after the other the chances of both dying at the same time or even relativity close to each other    unless there was something wrong with the drives, but if both are in normal working order then its not going to happen     there are so many factors at play here that just because they are identical means nothing about reliability

it is a common way of thinking since cars and other things that need repair sorta function that way, but with HD's all bets are off     they can die for no reason or can continue to work for decades without problems while the ones around it all die earlier      boy do I wish that was the case, if all you had to do to determine if a HD was about to fail was to have another identical drive fail first as a warning (sorta is to most people, if a drive in my RAID dies I replace all of them eventually) but alas they fail with no rhyme or reason to them really

just look at the google report on hard drive reliability     just do a google search for it  ha ha ha

mig wrote:

This seems to support that identical drives WILL have similar failure characteristics

I agree with mig and the report.  I used to work for a disk drive manufacturer in development and testing.  I feel pretty qualified in saying the report is accurate.  Leftkidney, you're right in saying there isn't necessarily a direct correlation, but the chances of any 2 drives failing around the same time are definitely higher when they are from the same mfg, lot, s/n.

So rather than increase (even if only slightly) my chances of having both drives fail around the same time, I decided to run them at different load levels--the second disk (B) only runs at night during the A-to-B backup, while the first disk (A) runs throughout the day any time data is changed.