Linux - ServerThis forum is for the discussion of Linux Software used in a server related context.
Notices
Welcome to LinuxQuestions.org, a friendly and active Linux Community.
You are currently viewing LQ as a guest. By joining our community you will have the ability to post topics, receive our newsletter, use the advanced search, subscribe to threads and access many other special features. Registration is quick, simple and absolutely free. Join our community today!
Note that registered members see fewer ads, and ContentLink is completely disabled once you log in.
If you have any problems with the registration process or your account login, please contact us. If you need to reset your password, click here.
Having a problem logging in? Please visit this page to clear all LQ-related cookies.
Get a virtual cloud desktop with the Linux distro that you want in less than five minutes with Shells! With over 10 pre-installed distros to choose from, the worry-free installation life is here! Whether you are a digital nomad or just looking for flexibility, Shells can put your Linux machine on the device that you want to use.
Exclusive for LQ members, get up to 45% off per month. Click here for more info.
The metadata is not going to be compatible unless the hardware RAID card was actually running the Linux md code. (Who knows?) What you would need to do is convince md to assemble the RAID without having the metadata. You would need to know the drive order, the chunk size, the offset of the first chunk on each disk, and which way the parity goes on consecutive stripes. Don't do create, which would wipe out your data. Don't do any operation which writes to the disk, which would wipe out your data. Assemble is fairly safe because it preserves what is there. Not sure if this is possible. Good luck.
If it's a hardware raid then assuming you have the same hardware controller card it should present the disks to the OS as a single device abstracting the underlying RAID, so it would simply appear as a normal device. Personally I wouldn't even think about trying to jam them in a different box and throw any form of software RAID at them, but that's just me.
The metadata is not going to be compatible unless the hardware RAID card was actually running the Linux md code. (Who knows?)
I will look into that.
Quote:
Originally Posted by smallpond
What you would need to do is convince md to assemble the RAID without having the metadata. You would need to know the drive order, the chunk size, the offset of the first chunk on each disk, and which way the parity goes on consecutive stripes. Don't do create, which would wipe out your data. Don't do any operation which writes to the disk, which would wipe out your data. Assemble is fairly safe because it preserves what is there. Not sure if this is possible. Good luck.
Thanks a lot!
I have lost the disk ordering info. May I guess it out?
Last edited by vrrivaro; 11-07-2013 at 01:44 PM.
Reason: Added one comment (I will look...)
There's 6 cases for order x 2 for parity direction (left or right) x N common chunk sizes (4K, 16K, 64K, etc.).
Guessing is painful. As TenTenths suggests, finding a replacement RAID card is going to be much simpler.
It may be possible to figure out by dumping data from the disks using dd. On a 3-disk RAID 5 you should see 2 chunks of data, then one of parity. If you can find a section of a text file, it will be easy to figure out the chunk size and parity order. The parity chunks will look like noise, the text can be reassembled like a jigsaw puzzle.
Each column is consecutive chunks on one disk. Parity either shifts to the left or the right on each row.
Code:
Left Right
D D P D D P
D P D P D D
P D D D P D
D D P D D P
etc. etc.
In fact, the heck with getting it to work with software RAID. If you can figure this out, it would be simple to write a script that skips all the parity and just reads the data in the right order and writes it out to another volume.
There's 6 cases for order x 2 for parity direction (left or right) x N common chunk sizes (4K, 16K, 64K, etc.).
Guessing is painful. As TenTenths suggests, finding a replacement RAID card is going to be much simpler.
Sure it would. That would be so much simpler... unfortunatly, the data is needed for yesterday and we where not given all the info. We have more now, they used the BIOS based RAID 5 feature od an ASUS P8H77-M LE Mothercboard, which means that it perobably is Linux or BSD based.
Quote:
Originally Posted by smallpond
It may be possible to figure out by dumping data from the disks using dd. On a 3-disk RAID 5 you should see 2 chunks of data, then one of parity. If you can find a section of a text file, it will be easy to figure out the chunk size and parity order. The parity chunks will look like noise, the text can be reassembled like a jigsaw puzzle.
I asked for 3 spare disks so I can dd out the info and parity for safekeeping while doing the recovery. I was told to go ahead with the recovery anyway. Given those constraints, here is what I have decided to ask for a last more question before I proceed to risk it all out:
Can Linux really mount a RAID 5 disk array that has been made under a Asus BIOS and formatted as NTFS under Windows 2008 server? What are my chances?
md is amazingly versatile but any software has its limits.
mdadm has "build" mode with "--assume-clean" which will safely assemble
an existing RAID array without metadata. However it looks like
it is limited to RAID 0, 1 or 10. So I give the chance of
getting an unknown RAID 5 to assemble in md as small without some hacking.
I did a S.A.R.R.T. hard disk test of each of the drives. I should have started there, shouldn't I?
It turns out, 2 of of the 3 units are full of errors. The extended check is aborted after 30s or so even when it gives an estimated ETA of about 2 hours. The 3rd disk is OK, but I have a RAID5 array where more than one disk failed--so my array is gone.
LinuxQuestions.org is looking for people interested in writing
Editorials, Articles, Reviews, and more. If you'd like to contribute
content, let us know.