Virtual JBOD (Span) Recovery
A disk on a virtual JBOD (Span) (managed by windows using dynamic disks) has blown away.
I guess if it's possible to recover some data from the other disks.
Thanks.
To actually recover from a RAID0 is possible using software like File Scavenger from QueTek (it's a microsoft Partner, anyways). But you must know the correct disk order and block size.
January 31st, 2011 10:25pm
Highly unlikely, if not impossible.
A RAID 0 (also known as a stripe set or
striped volume) splits data evenly across two or more disks (striped) with no
parity information for redundancy. RAID 0 was not one of the original RAID levels and provides no
data redundancy. RAID 0 is normally used to increase performance, although it can also be used as a way to create a small number of large virtual disks out of a large number of small physical ones.
See illustration @ https://secure.wikimedia.org/wikipedia/en/wiki/Standard_RAID_levelsPlease remember to click Mark as Answer on the post that helps you, and to click Unmark as Answer if a marked post does not actually answer your question. This can be beneficial to other community members reading the thread.
Free Windows Admin Tool Kit Click here and download it now
January 31st, 2011 10:39pm
On Mon, 31 Jan 2011 19:39:43 +0000, Rick Dee wrote:
RAID 0 is normally used to increase performance,
In theory. But in my experience, it hardly ever does, and if it does,
the improvement is so small as to be almost unnoticeable.
Ken Blake
February 1st, 2011 2:13am
Ok, let me explain better:
I have 4 hdd:
- 2x 320 GB - Raid0
- 1x 500 GB - Raid0
- 1x 200 GB (blown) which was split with 1 partition of 100 GB for Windows 7 and the other 100 GB on the Raid0
The raid had about 600 GB occupied. That means that if all files were split between the hard drives up to 400 GB would be lost. If each file was on a separate drive, up to 100 GB was lost.
What i ask if it's possible to recover that 200-500 GB of data, which is what mainly interests me because this raid has some years on it's back and so important data *should* be on the end of it. I know that windows does not do things that way, and so it
might be that the data i want is unrecovarable. But besides that, there should be something that i could get back.
How do i do that?
Thanks.
PD: It seems the controller it's burned, i'll try to replace it. If i get a working drive i'll do an image of it and throw it away - because i do not know how much will it last -. On this last, there's any way to tell windows that the 100 GB of data that's
looking is on that image (provided that i restore that image on a new, larger, hdd).
It doesn't make any difference how much you explain your setup. RAID 0 spreads the writing across the disks in the RAID, so recovering data is extremely difficult or highly unlikely because you litterally have pieces of everythoing written spread across
each of the disks in the RAID. It is litteraly a puzzle and with one disk out of the picture, you are missing pieces of the puzzle. The probability of recovering data depends on exactly which pieces of the puzzle are missing.
If the data is extremely important, then you need to search the internet for a company that specializes in recovering data from fialed hard drives. The service these companies provide is not cheap.
There is absolutely no capability in Windows to recover your data.Please remember to click Mark as Answer on the post that helps you, and to click Unmark as Answer if a marked post does not actually answer your question. This can be beneficial to other community members reading the thread.
Free Windows Admin Tool Kit Click here and download it now
February 2nd, 2011 4:21pm
On Wed, 2 Feb 2011 13:21:37 +0000, Rick Dee wrote:
It doesn't make any difference how much you explain your setup. RAID 0 spreads the writing across the disks in the RAID, so recovering data is extremely difficult or highly unlikely because you litterally have pieces of everythoing written spread across
each of the disks in the RAID. It is litteraly a puzzle and with one disk out of the picture, you are missing pieces of the puzzle. The probability of recovering data depends on exactly which pieces of the puzzle are missing.
Exactly. And that's why I think using RAID 0 is a poor idea. It
substantially increases the risk of data loss. And despite its
sounding like it should improve performance, in my experience with it
any performance improvement is so slight as to be unnoticeable.
Ken Blake, Microsoft MVP
February 2nd, 2011 5:55pm
And, to support your view, RAID 0 is not really RAID. RAID stands for
Redundant Array of Independent Disks and Redundant means Able to be omitted without loss of meaning or function, which is defiitely not the case with RAID 0!
I think of RAID 0 as an oxymoron like Military Intelligence!Please remember to click Mark as Answer on the post that helps you, and to click Unmark as Answer if a marked post does not actually answer your question. This can be beneficial to other community members reading the thread.
Free Windows Admin Tool Kit Click here and download it now
February 2nd, 2011 7:15pm
I'm sorry to have posted this thread. What i actually have is spanning, as RAID can't have disks of diferent sizes.
Is there any way to recover a spanning? I've used several tools, but the first disk was actually the one that was damaged, and so they can't find the folder structure.
Besides that, i can't find the correct disk order, it's taking me a long time to search over the possible combinations of disks. Thanks.
PD: It's possible to do JBOD + Raid 1 using windows lvm? I'm going to add a mirror to the current jbod (once i format it anew) so that it will be most difficult to loose data (again). PD2: It's possible for windows to just store single files on each disk?
I've read that some OSes have the possibility to do that, just as JBOD true single disk (if you loose a disk you'll loose each file that's in there).
February 3rd, 2011 12:20pm
It doesn't make any difference how much you explain your setup.
The probability of recovering data depends on exactly which pieces of the puzzle are missing.
If the data is extremely important, then you need to search the internet for a company that specializes in recovering data from failed
hard drives. The service these companies provide is not cheap.
There is absolutely no capability in Windows to recover your data.
Please remember to click “Mark as Answer” on the post that helps you, and to click “Unmark as Answer” if a marked post does not actually answer your question. This can be beneficial to other community members reading the thread. ”
Free Windows Admin Tool Kit Click here and download it now
February 3rd, 2011 3:31pm
Thanks for your answers.
I found a software that does the job for you called "DiskInternals RAID Recovery". So far it has detected most folders and files that were on the jbod. Of course that some has gone lost, but it's better to pay 30 $ than 2000$ for a professional recovery.
Anyways i've also been suggested to just use a bulk disk to replace the other one, with the same partitions so that windows should recognize the volume as full.
Thanks again!
February 3rd, 2011 3:49pm
I strongly suggest you do some research about RAID configurations on the internet. I did a quick search and I don't think your JBOD + RAID 1 is feasible because most RAID controllers will not permit you to mix configurations as this.
This forum is not about RAID configurations. I don't think there are any forums for Windows 7 to discuss RAID configurations. You might find something in a Server forum where RAID is used much more than in Windows 7.Please remember to click Mark as Answer on the post that helps you, and to click Unmark as Answer if a marked post does not actually answer your question. This can be beneficial to other community members reading the thread.
Free Windows Admin Tool Kit Click here and download it now
February 3rd, 2011 6:32pm
I believe the explanation of the setup was very useful and extremely relevant. As yushir has indicated since the first post, he (or she) is trying to recover data from a disk in a JBOD spanned configuration (which is NOT the same as RAID0) that were members
of the JBOD array but were not the failed disk in question.
For anyone referencing this forum for assistance on a similar problem, it should noted that JBOD and RAID0 are significantly different in the way that data is written and when trying to solve a problem with a JBOD array, do not approach it as if
it is a RAID0 - this is incorrect.
JBOD array is a configuration in which multiple (dynamic) disks which may or may not be the same size are smashed or sewn together to act as one drive and data is subsequently written sequentially (in theory) to each disk. For instance, suppose you have
three disks (100GB; 50GB; 300GB) and you set them up as a JBOD array - the array will be a size of 450GB. When data is written to this new JBOD array, it writes the first 100GB to the first disk and then the next 50GB to the second disk and the last 300GB
to the last disk. Data is not written in stripes across disk as in a RAID setup. That is to say, if you write a file of 10MB to the 450GB array, it will not separate the file into chunks of data (e.g. 512K blocks) and write it to disk 1, disk 2, and disk 3
in an alternating fashion. With a JBOD array, it will write the whole 10MB file to the first available space in a contiguous fashion (in theory).
I wanted to clear up the difference between a RAID0 array and JBOD array so that any user referencing this forum does not get incorrect information.
As for a solution, there are a variety of software products out there that can help to recover data from the failed disks. However, at the time of writing, I know of no utility or configuration alteration within Windows that will allow you to access data
from member disks of a JBOD array if the array itself is not in tact.
However, from some primarily proding I have done using virtual machines and virtual disks in a JBOD array, the solution seems to lie in the partition tables at the beginning of each separate disk as well as some tag information
near the end of each independent disk. Unfortunately, I have not spent enough time trying to decipher these tables and correct them to act once again as normal independent disks.
June 26th, 2011 7:34am
The fact is that there are some OS'es that thread JBOD disks (even if it's the own OS that handles the virtual array) as logical disks. That means that data won't be distributed the way it should be, because data is fragmented and so it's spared between
the disks (the OS may even write contiguous data to each disk in order to speed up r&w). As i discovered by myself, windows seems to do this (write data on each disk so taht seek times are lower), and thus if you loose a disk you loose almost all data.
So for those looking in jbod under windows, i'd recommend a fake jbod instead of using windows built-in. It's true that with windows you'll get a speed boost, but that implementation doesn't keep contiguous data in one disk and so you may be able to recover
some data (or do a raid1/raid5 with windows and you'll be fine)
Free Windows Admin Tool Kit Click here and download it now
June 26th, 2011 11:54am