Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[mdnosliw]

I was behind my computer table cleaning up some wiring. When I was getting out I tripped over a cable and my dnd 323 fell on the floor(not hard 2' onto a rug). The front plate came off and both drives fell out. I put it back together and hooked it up and all seemed fine. Then I noticed that I am missing files.

I had the drive hooked up to just span across 2 1.5 tb drives. I now seem to only be able to see the one drive.  When I go into the web interface all I can access is the config format page.

I have tried to reset the drive and the disks are in the correct bays.

Any Ideas appreciated.
Mark

[fordem]

I had the drive hooked up to just span across 2 1.5 tb drives. I now seem to only be able to see the one drive

Do you mean a JBOD configuration?  And where are you "seeing the one drive".

With JBOD there are two physical drives, both of which should be shown on the status page, but these appear as a single logical drive - so it would be likely that you never seen more then one drive.  The fact that you can only access the config format page (isn't there a cancel button somewhere) suggests that the unit has somehow lost track of a drive and thinks a new one has been installed.

First - I would avoid the use of JBOD - as failure of either physical disk will result in the logic disk becoming unavailable, and possibly the loss of all your data - so - start by backing up what you can access and then I'd say reformat the disks preferably as standard volumes and move on.

[mdnosliw]

Yeah well that is what I am in the process of doing. Was hoping there may have been a way to rebuild the config or something. After I back up all the data off drive 1 I will put drive 2 into a computer and see if I can recover anything off of it. I have never been a fan of a raid, but maybe time to give it a try.

Thanks
Mark

[fordem]

If you're not a fan of RAID, that would suggest that you don't need it - in which case, give it a miss altogether.  As my signature says, RAID is for disk redundancy, not backup - it's about data continuing to be available even though a disk has failed, not about recovering data after a disaster.

[mdnosliw]

No sure what you mean, a disk failing is sort of a disaster. So in a raid environment what do you do send the disks out for recovery. I just know I have had Dell servers running raid and when the hd crashed I still had to start from scratch. Maybe it was just one of those bad experiences that I had when I was young that you never forget.

Mark

 

[fordem]

What I mean is this - RAID does not protect you from data loss, that's what a back up is for - but, what a backup does not do, is prevent the down time that a disk failure causes, that's what RAID is for.

So if you're not a fan of RAID, it suggests that you probably are not worried about the down time (if you were, you'd be as enthusiastic about RAID as I am), so there's no reason to accept the loss of available storage space that it causes.

The important thing to recognize is that even with RAID you still need to backup to reduce the risk of data loss.

A couple of points I want to emphasize - a disk failure is not a disaster, not if you have a backup.  As long as you have a backup, it's an inconvenience, and without RAID a potentially costly inconvience.  Also your experience with that Dell server is atypical - I've been doing warranty support for Dell for a number of years, and lost track of the number of failed disks that I've replaced - it's not impossible to have a disk fail and have to rebuild from scratch, but it's also quite rare.

[mdnosliw]

Well back to the original problem. If I take the drives out of the DNS 323 and hook them up to an Xp machine
there are two partitions on each drive. A 500 meg and a 698 gb. The partitions are reported healthy, but come up as unknown partition type. I will have to see what utils I may have that will allow me to access the partition.

Yes I do have backups of most of my data, but there is a lot of time required to rebuild the archive . Overall I do not have all that much faith in a 750gb drive for a $100.

I have been using WD Velociraptor  drives for the os and they have not given me any problems. You get what you pay for. I have some old IBM SCCI 10gb drives that I paid $500 never had any problem with them.

[fordem]

Sometimes I wonder why it always comes back to this ...

This is not about the hardware - it's about bad or perhaps unwise decisions - your choice to use a JBOD implementation put your data at unneccessary risk, I don't use D-Link's JBOD implementation because I don't feel that the potential gain (the ability to store a single file or data-set larger than the capacity of a single disk) is worth the risk (losing access to all of the data in the event of a single disk failure) - and no, this is not documented anywhere, I went out and tested it personally, failure of a single drive can cause a loss of access to the data even if it resides entirely on the remaining functional drive, I am not implying here that the implementation is flawed, but simply that D-Link chose to implement JBOD as a linear concatenated volume, which suffers from this problem - there are other ways to implement JBOD, and in fact, what D-Link terms standard volumes is called JBOD by other NAS manufacturers.

You also chose to go under the desk to tidy up the cables and that triggered the mini disaster that you're dealing with - a two foot fall will create significant shock to the HDA, more so if unit is powered up and the heads are "loaded".  The DNS-323 enclosure provides no "cushioning" and what happened there is pretty much the equivalent of dropping a bare drive onto the floor.

Perhaps you should take a look at the drive manufacturer's handling instructions and see what they say on dropping them - that two foot drop exceeds the permissible distance of at least one manufacturer by a factor of four.

For the record - IBM sold their disk drive division many years ago, and the drives you get from IBM have probably been manufactured by Seagate or Maxtor.  Personally I think their pricing is ridiculous (earlier this year I paid $746 for an 18GB drive 10k rpm, U160 drive) and in my experience (my relationship with IBM dates back some 18 years) the drives are no more, or less reliable than any other brand.

Improvements in drive technology is what's responsible for the rapid decline in the cost of storage - and in my experience, the newer, higher capacity drives are no less reliable than the ones of yesterday, in fact.

[gunrunnerjohn]

The only comment on newer hard drive reliability I have is that due to the cheaper prices and volume production, you'll see more DOA and infant mortality issues with newer drives than in the old days.  The testing and burnin that used to be common in the industry is not done nowadays for consumer level drives and it's hard to expect it for the prices they're selling at.  As far as long term reliability, I tend to agree that once a drive has lasted for a few months, it's as likely to last for years as the old iron.