Thanks, Zorb.
I asked several times if anyone knew what a HW RAID controller does to the 2 drives when asked to create an RAID0 array, then to delete same. Only got a couple of answers along the lines of "don't know, don't care".
I have reason to believe (until proven otherwise) that the RAID controller will actually do the exact same things to the drives every time you ask for those operations.
I want to set up an experiment - get 2 expendable drives, write 0's to every writable location on both, then let the RAID controller do its thing and look for 1's on each of the drives using DiskExplorer or similar tool.
To make the experiment complete, I'd have to then write F's to every writable location on both, do the same thing and see how many bits turned to 0's, but even without that I should get a decent idea of what happens and where. I am going to hypothesize that unlike SW, the RAID controller doesn't bother with the READ-MODIFY-WRITE type stuff that you're describing, but rather overwrites entire sectors that it wants to write to with no regard for existing data (which should make it easier to look for changes), and that the sectors most likely to be affected would be in the beginning of the drive, possibly at the stripe boundaries and possibly at the end of the drive.
This may or may not ultimately lead me to getting my bits back, but it will at least answer the question I posed.
I can post the results if anyone cares.

A case of obscurity trying to pass for profundity . . . . ?

I can simplify matters greatly for you at this point, vasechek. Send it to a pro in your region. I don't want you to take this personally, but there are problems with your experimental approach and technique thus far. Why continue to risk your data playing around-- if indeed your data is important?

That's all I'll say on this topic. Good luck.

Jono

_________________
http://www.datasaversllc.com

I really don't see what you are getting at. You can't write units less than a sector through software or hardware. The hard drive doesn't understand what a byte is, it understands what a sector is. It can't work with less than that. It's because of this that if a sector is damaged at all, you get nothing from it.

The end result of read, modify, write is the same as simple writing. The drive writes at the sector level. For example, if you have a sector with the first 256 bytes of 00 and the last 256 bytes of FF, and you write one sentence of ASCII text less than 256 bytes onto that sector, the last 256 bytes would still be changed to zeroes. The 1s would all be gone, because no matter how you write to a drive, the smallest you can write is a sector. Now, if you were to read the existing content from that sector, insert your changes into it, replacing material from the beginning of the sector through the end of the sentence, then write the modified sector to the drive, the high bits at the end would still be there.

I see what you are thinking, but it's not possible.

I have to agree though with Jono. It's time to seek expert help if you want your information. You're going to zork it all if you keep this up, that is, if you haven't already.

Zorb,
The assumption is that the drives are ok and all sectors can still be read and written successfully. If my little experiment does yield some concrete findings, i.e. tells me which sectors got overwritten by the RAID controller at the very least that would narrow it down from 2 corrupted drives to several known corrupted sectors. If I am lucky and the RAID controller only overwrote those sectors that are not user-writable during normal use of the LaCie system (the MBR, the O/S - the things that stay the same no matter how much data you put on the user shares) then I should be able to copy those sectors from the drives I fished out of the working EBD and that just might be it. I never let Windows or any SW touch anything on these drives, so I am fairly certain the only sectors that are corrupted are those stomped on by the RAID controller unless the failed EBD HW did some damage. This Disk Explorer tool supports manual overwriting of data in any sector (yes, via the read-modify-write if the change is less than 512 bytes), so with a bit of patience I should be able to manually copy the sectors that have been butchered by the RAID (at 1k characters per sector I hope there aren't too many sectors affected) from the good drives. It goes without saying that both of the damaged drives will first get backed up sector-by-sector prior to this so no more irreversible damage will be possible. There is a Seagate util that does that for any drive, not just Seagate, so I will try it out.
I know it's a long shot, but I think there is a chance and this time I will take precautions not to make things worse. Worst case I will figure out the RAID controller behavior
Thanks for the helpful info.

Update: Got most of my data back quickly and easily. Missing about 512 bytes out of the 700GB. Is there a way to figure out which file is damaged if I know where on the 2 drives that make up the soft JBOD array the lost sector is located? Would that be in the partition table? Any clues on what to look for?
Thx

I don't quite follow you... How do you know you are missing 512 bytes? It's odd to get apparently all your data back but be missing a single sector. It may just be the sector at the end of the volume or something and probably nothing to worry about.

If it's a JBOD and not a RAID, and the disks (or sections of disks) are simply joined end to end, then it would be easy to tell where your missing sector is missing from. It would be in the file system data, depending on what file system your drive is formatted with. I assume you know what the missing sector is by number, in this case, just use a program that understands your file system. There are lots of them.

Also, I thought you said it wasn't jbod?

How did you piece it back together?

It turns out that in the LaCie EBD the 2nd drive has data right at the start in sector 0 (I did claim otherwise before, but I was "speaking from my memory" and it let me down that time). I also dug up a "success story" on another forum about the Big Disk coughing up the user's data after the last partition on the first drive and the (only) partition on the second drive were simply concatenated. That's JBOD. I noticed that the RAID controller likes to wipe out sector 0 on the drives it touches. Sure enough my corrupted drive 2 had 512 bytes of 0's in sector 0 and random-looking data starting in sector 1 and going on forever. The "good" drive 2 had random-looking data starting from sector 0 and going forever. The data appears to be a part of some file that just happened to be stored at the stitch between the 2 drives. The data in these same few sectors at the start of drive 2 doesn't match at all between the bad and the good drives and the partitions are set up the same way so if it was directory-structure-related I would have found some correlation at least. Everything else the RAID controller touches seemed to be in the sections of the drives that are not readily user-writable and the sectors could be copied from the good drives with high probability of success since the surrounding sectors matched exactly on both drives. Sure enough, when I threw the fixed 2 drives into the enclosure they worked as if nothing had happened. Sector 0 on drive 2 is lost, however, because it apparently is a part of my file, so I couldn't just copy it from the good drive's. To verify that I want to find which file it is and whether or not it's visibly affected. It's possible that I have a backup of that particular file somewhere, too.
In any case, despite my modest abilities I am much better off than last week and it didn't take that much effort or resources, plus I now know what the RAID controller messes with and got exposure to some of the tools that help examine, tinker with, backup or fix drives.
Thanks for your help

2vasechek
Your issue is very simple.
Your both HDD-s work just fine. For LaCie box most typical failure - break Power Supply ( if it work NON-STOP), when it happens, drives just clicking because 5V voltage is too small. You may just CHANGE (buy new) power supply , or repair it, its very easy.
Get back drives in to the LaCie BOX.

If I helped, send me a beer to Kiev.


P.s. Sorry for my English, I only teach it

_________________
Angel Data Recovery

Hi, I have a Lacie firewire 500GB Big Disk that has just failed on me. Blue light still on. Disks sound like they spin up fine, but nobody home. Running OS X 10.5.8 on imac G5. No longer mounting or showing up in disk utility. I should make it quite clear that I am pretty out of my depth here.

What I have done using what I have available to me at home:

cracked open the lacie box.
placed the 2x 250GB Maxtor drives inside my old G4 quicksilver 933mhz.
(which uses both IDE cables so no boot drive) connected two machine with firewire cable
So Started up the imac G5 in target disk mode. Then started G4 holding the option key in order to boot from the imac G5.

Now disk utility does see the drives. and a volume called disk 1s6 under one of them
233.8 GB Maxtor 6L250R
233.8 GB Maxtor 6L250R
disk 1s6

they are not mountable or readable of course, oh and on startup there was a system message to say i had inserted unreadable disks do i want to Ignore/Reject/Initialize to which I did not respond.

I have not and will not be writing anything what so ever to these disks or attempting to reformat them or reconstruct the array, as I have no idea what I am doing. Which is why I am here.

At this point all I want to do is get the 400 odd GB of data off of these drives with folder and file names in tact. Then that Lacie enclosure can go to the trash for all I care, as I am suspecting it is this at fault, now the drives do not appear dead.. By the way no clicking sounds have been heard either.

Currently I am trying to evaluate the only two software applications I can find for OSX that are capable of scanning or attempting to virtually reconstruct the Raid 0 array now the disks are no longer in their original enclosure.

Currently scanning (at 90% after 8hrs) is Virtual Lab from binarybiz which has a fully automatic Raid Array sensing module so I only had to choose the disks and it seems to instantly recognize a volume called LaCie Disk (HFS) which it is now scanning. Although this is not the volume name I know it to be as it should be called Mumbomike.

Still i'm using this more as a test to see iof the files are visible really, as its still gonna be hundreds of dollars to recover 400GB with them.

The other software I found is R-Studio. But it needs me to enter sector sizes etc which I dont know. How can I know this?

The answers to any of these questions and any suggestions would be very much appreciated.
anyone?

To update... The scan completed with Virtual Lab, and looks to have been successful, in displaying the exact folder hierarchy even with the coloured labels still intact. I used the free 1mb quota to get a PDF document back. And having used them before I believe this would work.

Still I dont feel that the data is damaged or particularly difficult to get at, its just this Lacie raid controller/enclosure that is failing, so I guess if I had another identical model I could swap out the drives. But i dont want to invest in another one obviously, and just need a cheap way to bypass it with software.
Still I want it back as virtual lab would give it to me, not just a useless list of grouped and anonymously numbered files. Any ideas?

Thanks
Mike

contact a local dta recovery company. Maybe a lot cheaper than buying software off the net which isnt guaranteed to work.

mikerider,

You say you "pretty much out of your depth here" then you proceed to do something

Like you go to a dentist and he starts to smack your head with a hammer, or gives you a haircut.

Hope you get my meaning - yes, you completely out of your depth. You need help.

There's someone near you on this forum - pcimage - I don't know him/her but I believe he/she does not work with a hammer and a pair of scissors.

contact him

hi derp

thanks for response, I knew someone on here would give me verbal for my actions.
I'm pretty sure I've not done anything destructive to hardware or data though.
I will give pcimage a go, but I'm trying to do this on the cheap, and am happy to try this myself with advice.

From what I've said can anyone see what is the issue here? Is it the failed lacie controller? surely?
i have the demo of R-studio and with some basic advice from someone with experience of this application would happily follow instructions to correctly create virtual images of the raid array and scan it etc etc.
i'm half way there?

of course you are probably right derp... and maybe I'll pay for being a cheap skate, but if anyone wants to help me further in this direction before I pay more money towards this damn drive, I'm all ears.

Ta
Mike

http://www.insanelymac.com/forum/lofive ... 84073.html

I was acting on the discussion and tip offered by jackanova in the forum link above. but I dont have a PC.

mikerider,

You don't have a PC? Windows? an extra 500GB hard drive?
Now the dentist don't even have a chair or a drill,
or a sexy assistant (oral hygienist)

derp.
i have a 2 macs. OSX. R-studio and plenty of hard drives

what I need is someone who knows something about these components and their ability to rescue from a f@ckin lacie raid controller thats spazzed out.

same thing has happened at my workplace with another lacie 1000gb by the way. probably purchased around the same time in 2005.... looks like there's a time limit on these pieces o crapola

enough of the dentist metaphor dude, its not helpful. I did try PC Image, he kindly got back with an offer. But still more money than I was willing to spend...

mikerider,

Anyone who knows how to recover these lacie's must examine each drive and establish
how the two are configured(this does vary) and if the configuration is not EXACTLY
and CORRECTLY established then you get no data.

So. why not hook up one of your drives(destination) and Run this Studio program you have
on the two drives, and try and try and guess and guess and guess until maybe something pops up,
who knows maybe you get lucky and you get your data.
I am not sure how many permutations there could be, but it could be quite a few, so good luck to you.

I don't think anyone can help you more than this. Why? Cause, as I said, they need to see your two drives.

Yep, Derp is correct.
Lacie uses many different strategies to build their raid's, so without having seen and studied these disks it is impossible to help.

I've recovered allready quite a few of these lacie's and they can be quite painfull to recover.


Best regards,

Dobre

_________________
Murphy was an optimist

Datarecovery in Belgium, Holland, France and Germany
Datarecoverytools http://www.drtools.eu

Do you really need 2 MAC's? So you own 2 MAC's but don't have the funds to recover your data or you don't want to spend money on a data recovery service which takes years of training and research? Seems a little disrepectful to us DR Gurus.

Why not sell one of your MAC's to get the data recovered?

If you want your data, you gonna have to pay.

Or give me your MAC and I'll do it for free.

Well I've had this old G4 for years and years, and I just acquired the G5 imac from my kindly boss.
I was in the process of transferring things. and of course this drive goes and fails

Far from disrespecting Gurus, there a lots of things that require years of training in this world, and with the internet, we have an opportunity to share a wealth of knowledge with each other. I have used the internet many times in this way, and have managed to do a hell of a lot of stuff myself, that would have otherwise been absolutely no chance. I've taken complex computers to pieces and fixed/replaced stuff. There is so much very helpful advice and demonstrations out there for so many things, it is not always necessary to pay through the nose, or go to college to study the whole subject in full.
The attitude of some people on here is pretty disgusting, jumping down peoples throats who dare mention they are here for the reasons mentioned above. Its some weird power thing, that I hope they get over some day, its a pretty revealing display into the inadequacies they obviously suffer in most areas of their sad lives.

with respect to helpful Gurus
Mike