G-RAID 4TB Gen 4 Recovery

[managerharry]

Working on a G-RAID 4TB Gen 4 raid box. Drive refused to mount. SMART reporting on both drives shows one OK, the other with some bad sectors. Failing drive was imaged with minimal corruption. Now trying to reconstruct RAID in UFS Explorer 5.6 but cant seem to find the correct parameters. Have identified first disk by EFI part in hexedit but running rebuild as RAID 0 with stripe size set to 64KB, 512KB, or 256KB (mentioned elsewhere as a possible for G-RAID models) does not result in a valid rebuild. Thinking perhaps it's a problem with the start sector or some other parameter?

Thank you!

[DR-Kiev]

NOP . You just don't know how to cook them. Solution is very nearest. Keep patience and continue trying another stripe sizes. The correct one is very unusual, and you will find it in the very end

[mr_spokk]

NOP . You just don't know how to cook them. Solution is very nearest.

[pcimage]

NOP . You just don't know how to cook them. Solution is very nearest. Keep patience and continue trying another stripe sizes. The correct one is very unusual, and you will find it in the very end

Yep, it's very unusual on these!

[lcoughey]

Working on a G-RAID 4TB Gen 4 raid box. Drive refused to mount. SMART reporting on both drives shows one OK, the other with some bad sectors. Failing drive was imaged with minimal corruption. Now trying to reconstruct RAID in UFS Explorer 5.6 but cant seem to find the correct parameters. Have identified first disk by EFI part in hexedit but running rebuild as RAID 0 with stripe size set to 64KB, 512KB, or 256KB (mentioned elsewhere as a possible for G-RAID models) does not result in a valid rebuild. Thinking perhaps it's a problem with the start sector or some other parameter?

Thank you!

Just to confirm, you only mirrored the one drive that was reported bad? If so, be aware that your assumption that there are no issues with the other drives may be incorrect and your attempts to recover using original drives may result in destroying any chances of recovery.

[managerharry]

Hmmm... my cooking skills are still not sufficient for this recipe, perhaps you could help with some more specific details? HFS+ volume label is at sector 131137 on first drive. Tried running the rebuild with all the preset stripe sizes from 1KB to 1024KB with no success. Also I'm not sure about start sector settings for each drive when rebuilding, should they be the same?

@lcoughey -- Both drive have been imaged. Working off the disk images, not original drives.

[dmarques]

Try to check for footer as well as header.
Also, are you sure it's a RAID0, not a RAID5 ?

[hddguy]

rebuild with all the preset stripe sizes from 1KB to 1024KB with no success.

Yep, it's very unusual on these!

[DR-Kiev]

Tried running the rebuild with all the preset stripe sizes from 1KB to 1024KB with no success. Also I'm not sure about start sector settings for each drive when rebuilding, should they be the same?

Hmm.... was very close , 1 step beside.

[hddguy]

Alternatively, if you cant figure it out, just put the cloned images to the unit in the relevant order and work through the enclosure. Assuming it mounts, read off your data, if not recover it with software as the unit will build it for you.

Also, is much faster to recover this RAID via the enclosure...

[managerharry]

mmm, stripe 512b aka 1 sector, correct still this soufflé does not rise. maybe I need to manually define partition start?

I will try putting the imaged drives back in the enclosure and see if it rebuilds but I'd still like to figure out the correct parameters.

[DR-Kiev]

mmm, stripe 512b aka 1 sector, correct

Bingo!!
Bonus game:

still this soufflé does not rise. maybe I need to manually define partition start?

Answer here:

HFS+ volume label is at sector 131137 on first drive.

[managerharry]

Even with the correct stripe and partition start UFS is unable to autodetect the partition as it usually does. Error size on the bad drive was only 1197kb but perhaps that was enough to damage the partition table... Maybe UFS's scan feature would do the trick. I think I will try & put the imaged drives in the enclosure and see if it rebuilds. Thanks for all the pointers!

[DR-Kiev]

Even with the correct stripe and partition start UFS is unable to autodetect the partition as it usually does. Error size on the bad drive was only 1197kb but perhaps that was enough to damage the partition table... Maybe UFS's scan feature would do the trick. I think I will try & put the imaged drives in the enclosure and see if it rebuilds. Thanks for all the pointers!

You gave up too early. UFS doesn't know well oxRI raid serviceMeta standart. It keeps S/N of all drives which give you troubles when you connect non original drives(clones) to the G-raid box .

May be you did mistake with order as addiotion?

[managerharry]

Indeed DR-Kiev you are correct -- G-RAID enclosure does not appear to like the cloned drives. Pretty sure I've got the drive order correct; first drive is the one with the GPT signature at start and HFS+ further on. I'm missing some ingredient...

[hddguy]

Maybe partition is bad. Create region in r-studio starting at start of HFS+ partition for direct access to data

[DR-Kiev]

Indeed DR-Kiev you are correct -- G-RAID enclosure does not appear to like the cloned drives. Pretty sure I've got the drive order correct; first drive is the one with the GPT signature at start and HFS+ further on. I'm missing some ingredient...

Yes you are. You forgot , stripe size is the 512bytes. And where do you think sector with EFI header have to start ?

[lcoughey]

Next step will be to understand how HFS+ works, then you will be able to figure out the drive order...or just try all possible options and compare the results.