Lacie 4Big Quadra Raid5 parameters 4 x 1.5tb

[Gordon Bulger]

I am not having luck with this HPS+array. Although the user says the 4 drives were arranged 0,1,2,3, winhex shows 2,3,0,1. I have tried both.

These are the parameters I am using in UFS Explorer:
Left Symmetric
custom stripe size (sectors) 512
parity delay 1
asynchronous I/o - no

Any suggestions?

Thx
Gordon

[Dmitri]

Should you find yourself out of ideas, we'll be happy to look into this RAID 5 remotely.

[lcoughey]

Did you try all 24 permutations of the drive order?

[data-medics]

I am not having luck with this HPS+array. Although the user says the 4 drives were arranged 0,1,2,3, winhex shows 2,3,0,1. I have tried both.

These are the parameters I am using in UFS Explorer:
Left Symmetric
custom stripe size (sectors) 512
parity delay 1
asynchronous I/o - no

Any suggestions?

Thx
Gordon

Try using 512 byte block size, not 512 sector. Lacie often uses single sector block size.

[lcoughey]

Try using 512 byte block size, not 512 sector. Lacie often uses single sector block size.

Nice catch...I read it that he had it set for 1 sector block size.

[Gordon Bulger]

Thanks Luke (and others). How about this Ontario weather??

My bad. I meant 512 bytes and I have been using this. What aggravates me is that one of my data recovery associates has recovered long movies from this setup. I will try different drive arrangements. Anything else you can suggest?

Gordon

[DR-Kiev]

Try to miss drives one by one to find "old" culprit. Define HFS+ partition from 409640 offset what color it gaves you?

[S.Haran]

From my notes on a LaCie 4Big Quadra recovery...

The stripe size was 512 bytes but in UFSexplorer you need to enter a custom stripe with the value of "1" not 512.

[Gordon Bulger]

Thanks for your suggestions DR Kiev and S. Haran. HFS+ partition seen as yellow only so far. Yes I have been using 512 bytes which equals 1 sector. When I arrange the 4 drives backwards (3,2,1,0) I get the HFS+ partition starting at 409640. other arrangements either start at 409639 or 409641. Do these Lacie 4Big Quadras have asynchronous I/O and is parity always left symmetric? Are there any other stripe sizes to be tried?

[S.Haran]

Here's a UFSexplorer screenshot from my notes. Hope it helps.... http://picpaste.com/CHgood-stripe-is-one-LYDlXJkO.PNG

[ROb]

have you tried madam in linux ?

[mr_spokk]

Deleted

[lcoughey]

...are you sure it is RAID 5? Give me a call if you'd like me to look at it remotely and give you a second opinion.

[DataPros]

S.Haran, Hi, is it possible you could re-post that picture?
I have a Lacie 4Big Quadra Raid5 with 4 x 3TB and one has failed.
Not sure what client did afterwards, but they say the RAID was
not seen on the network after the failure.

I've imaged the 3 remaining drives (which one of them had lots of
bad sectors) and am trying to recover the RAID on UFS Explorer -
(there's a new RAID-specific edition) with no definitive results yet.

At this stage, facing 28 hours is what UFSx says for every parameter
change, I'll take a shortcut to "discovery".

Thanks!

[S.Haran]

Here's the screenshot... ftp://ftp.freedatarecovery.us/images/CH ... is-one.PNG

[Martin]

have you tried madam in linux ?

Better to try monsieur ... It's a joke, sorry ROb.
You've meant mdadm, I think.

As it was mentioned, better to try to miss drives one by one using config - Raid 5, Left Asynchronous, Block Size = 512 Byte (1 sector).

[ROb]

Typing error

[deffie83]

Hello everyone, this is my first post on HDD guru and I'm resuming this old post as it has been useful to me and I want to add some infos that may help others.

This is about a 4big quadra RAID5 with 3 of 4 disks available and an XFS filesystem (no partition table) on it.

The missing disk on the quadra was the fourth looking from the back and starting from left.

After many tries I've found the right disk order by looking at the XFS internals:

https://cdn.app.compendium.com/uploads/ ... _final.pdf

Page 8, primary AGs.

SB AGF AGI AGFL were on the first blocks of every disk so it was easy to find the right order.
I spent many time trying combinations with the missing disk as the last one but it was not.

Check the attached screenshot for details.

As you already said and it was very useful to me the stripe size is set to one sector.
Parity distribution as Left-Asymmetric.

Once configured this way UFSExplorer showed me the green (healthy) XFS partition and it was immediately able to let me browse the directory structure.

As I'm with the trial version i'm dumping an image of this to a file and I'm going to mount it on a linux machine. Four days left. In many other combinations with was able to see the XFS partition but without access to directory structure and compromised files data trying to recover with ufsexplorer.

A note about mdadm which looked almost perfect for this and I previously tried (working on copied versions of the disks i have to avoid damage to the original ones): unfortunately mdadm doesn't allow for 512bytes stripe size, so it's a no-no.



Thats all folks,
thank you for your help.

Attachments

[S.Haran]

Good job recovering your data. And thank you for the XFS link.

Regarding mdadm not working with a 512 byte chunk size, I've looked into this problem at length as I use a Linux platform to perform my remote RAID recovery work. Anyway it seems the minimum chunk size mdadm can use is equal to the Linux kernel page size. And Linux uses a 4096 byte page size. I've been looking for a Linux version that uses a 512 byte page size but so far I have not been able to find one that meets my needs.