ARECA ARC-1222 data recovery *$50 bounty*

Raris

n00b
Joined
Nov 19, 2012
Messages
5
As ARECA has yet to get back to me regarding data recovery, i am posting this here in the hopes that someone can help me recover my data from my 8-disk RAID array.

I am offering $50 via paypal to whomever can guide me through getting my data back.

My machine had suffered a power problem with the MOLEX connectors to the hard drives; one disk saw a failure, the RAID controller began to rebuild, and a second hard drive suddenly dropped out of the array (RAID 5 i believe), 'breaking' the array during a rebuild. i tried level one and level 2 recoveries to no avail. I have all the technical information relating to the array statistics; drive order, sector/cluster size, RAID type and configuration, etc. I tried creating a new array (NO_INIT) uninitialized to try to 'fake' a new array and access my data, to no avail. the hard drives have not been used, formatted, or altered in any way since the failure, so the data is still intact, just not accessible. I would prefer a way to copy the unorganized data and rehash it later instead of risking it with a format etc; i know there were a few people on here that had successfully recovered data in its raw format. English is not my first language, and i am not intimately familiar with ARECA hardware, RAID array failure recovery, etc so please forgive me for any errors in language, technical details, or syntax.

The affected array was initally in a different drive order (something like 5,2,1,6,7,3,4,8) and is now 1,2,3,4,5,6,7,8 if this helps, and i have no way to 'reorder' the drives via ARECA's HTTP GUI interface or the RAID boot configuration.

I have since soldered all the power connectors on the machine, so power problems will no longer be an issue. I should have bought and installed the battery backup module, but i did not have it then. The hardware is an ARECA ARC-1222 8x SATA II card, connected to 8 Samsung HM500JJ 500GB hard drives. They were in a RAID 5 configuration at 2.74TB total size. I do not have a 3+ TB hard drive to copy the data to; if needed, i can purchase one to dump the raw data on if it can be correctly reassembled later. I would prefer this in fact, to copy and rehash the data, rather than risk losing it. It is footage and photos spanning 4 continents and 20 years, so it is very important to me.

I can provide all the specs to anyone interested; i NEED the data recovered, but i do not need it ASAP; i have at least a month until i need the data recovered, so i can wait for a bit, but i do NOT want to risk losing it. ARECA has not gotten back to me in a week now, which worries me. I do not know enough to attempt data recovery otherwise, and would prefer to be guided through this by someone with experience; and i am willing to pay $50 USD as previously stated.



Anyone interested i can either post or email the system and configuration specs; my email address is [email protected]; please email me if you believe you can help.

Thank you,

Panagiotis Raris
 
Last edited:
And so goes another chapter in the book of "RAID IS NOT A BACKUP". Can you please post your entire logfile. What firmware version is on the card? What OS and FS did you use? I can't believe you have 20 years of data that you have never backed up. You could get an external 3TB drive for backup for about $150 which would have saved you a lot of worry.
Why did you reorder your drives after the failure, that is NOT the thing to do. Can you be more descriptive of the power problem you experienced? Was it a faulty PS? What did you solder?
Next, you said you have 8x 500GB drives in a RAID5 yielding ~2.75TB. 8x 500GB should have yielded you ~3.5TB formatted (or ~3.25 after GiB-=>GB conversion)
Before you try screwing with the data, my first suggestion is going to be to make mirror images of each of the drives (either with dd to 8x 500GB additional drives) or at least making bit perfect images of each drive to a disk file and copying them somewhere safe in case you need to go back to current baseline in case something goes wrong and makes things even worse.
We'll all try and help you here, if we can help and you feel like paying $50 please contact Kyle (The Big Guy) and see where he would like you to donate the cash to charity)
 
Last edited:
As ARECA has yet to get back to me regarding data recovery, i am posting this here in the hopes that someone can help me recover my data from my 8-disk RAID array.

I am offering $50 via paypal to whomever can guide me through getting my data back.

My machine had suffered a power problem with the MOLEX connectors to the hard drives; one disk saw a failure, the RAID controller began to rebuild, and a second hard drive suddenly dropped out of the array (RAID 5 i believe), 'breaking' the array during a rebuild. i tried level one and level 2 recoveries to no avail. I have all the technical information relating to the array statistics; drive order, sector/cluster size, RAID type and configuration, etc. I tried creating a new array (NO_INIT) uninitialized to try to 'fake' a new array and access my data, to no avail. the hard drives have not been used, formatted, or altered in any way since the failure, so the data is still intact, just not accessible. I would prefer a way to copy the unorganized data and rehash it later instead of risking it with a format etc; i know there were a few people on here that had successfully recovered data in its raw format. English is not my first language, and i am not intimately familiar with ARECA hardware, RAID array failure recovery, etc so please forgive me for any errors in language, technical details, or syntax.

The affected array was initally in a different drive order (something like 5,2,1,6,7,3,4,8) and is now 1,2,3,4,5,6,7,8 if this helps, and i have no way to 'reorder' the drives via ARECA's HTTP GUI interface or the RAID boot configuration.

I have since soldered all the power connectors on the machine, so power problems will no longer be an issue. I should have bought and installed the battery backup module, but i did not have it then. The hardware is an ARECA ARC-1222 8x SATA II card, connected to 8 Samsung HM500JJ 500GB hard drives. They were in a RAID 5 configuration at 2.74TB total size. I do not have a 3+ TB hard drive to copy the data to; if needed, i can purchase one to dump the raw data on if it can be correctly reassembled later. I would prefer this in fact, to copy and rehash the data, rather than risk losing it. It is footage and photos spanning 4 continents and 20 years, so it is very important to me.

I can provide all the specs to anyone interested; i NEED the data recovered, but i do not need it ASAP; i have at least a month until i need the data recovered, so i can wait for a bit, but i do NOT want to risk losing it. ARECA has not gotten back to me in a week now, which worries me. I do not know enough to attempt data recovery otherwise, and would prefer to be guided through this by someone with experience; and i am willing to pay $50 USD as previously stated.



Anyone interested i can either post or email the system and configuration specs; my email address is [email protected]; please email me if you believe you can help.

Thank you,

Panagiotis Raris


I have extensive experience in recovering arrays on Areca, 3ware, and LSI. That being said having the full event history (since the array was in a normal state) and the original disk order is really a must in order to recover.

The only way to re-create the array (rescue/no init) when the disk order has changed is to *physically* re-order the disks then delete the array/re-create with rescue/no init. You have to make sure raid level, stripe size, volume set size and everything is exact though. You can' not create them on a specific order on the Areca web/bios utility.

It does not bode well to me when it seems like you don't even know if it was in raid5 or not. From the size you listed and if truly raid5 I would guess it was raid5 with a hot spare. And that would make sense why the disk order changed. This is exactly the reason I recommend against hot-spares as it makes recovery more difficult.
 
Step 1 is creating a non-compressed image of every disk with R-Studio. I usually tell people to do that first and if they don't bother or don't respond or want to argue it, I'm out because its a lost cause which this one already smells like.

That said, if he's done no more and gone no further than what he described then the data is totally recoverable. The biggest challenge is always how committed a person is to recovering their data, their ability to have patience and follow instructions exactly without deviating and experimenting.
 
Last edited:
just out of interest, how does the drive-order change when the hot-spare kicks in?
I'd assume that the hot-spare takes the place of the failed drive, which does not change the order?

good luck!
 
R-Studio has done very well for me in the past. It will let you pick the disk order but you will probably have to experiment with that which will be a pain with that many disks. It'll give you a color-coded representation of what it finds on the disks though, so I usually find that easy enough to use as an indicator of whether I have the right order or not. Just make sure nothing tries to write to the disks before you get to R-Studio with them. Assuming you don't just send them out somewhere anyway...
 
Last edited:
You never operate on the original disks unless you don't care about the data. If you care about the data from day 1 then you begin day 1 with it on redundant disks. But if you didn't know that then, now is the time to start!
 
R-Studio has done very well for me in the past. It will let you pick the disk order but you will probably have to experiment with that which will be a pain with that many disks. It'll give you a color-coded representation of what it finds on the disks though, so I usually find that easy enough to use as an indicator of whether I have the right order or not. Just make sure nothing tries to write to the disks before you get to R-Studio with them. Assuming you don't just send them out somewhere anyway...

Unless you're talking about recovering from motherboard RAID with 2-3 disks, or from hardware raid where drive order hasn't been changed yet (which the OP apparently did), then this guy's recovery goes beyond what any recovery software can simply scan or auto-detect on its own as far as drive order/raid level/stripe size/partition offset, but can be done manually with R-Studio though requires expertise, compounded by proprietary block patterns depending on manufacturer. Such is the case with Areca - proprietary block pattern.

In any case nobody's going to do this for $50 - it will cost him a couple hundred to have someone remote-in to his system and manually analyze each disk image to determine original drive order, stripe size and raid level (since he's apparently confused on what he had). It'll likely take a few hours of analysis in WinHex to determine the manual parameters to lay into R-Studio in order to reconstruct the array virtually. So he'll need 2 x 3TB drives - one to hold the disk images and the other to recover the data to. Place the Areca into JBOD mode to create UNCOMPRESSED disk images from each raidset member disk.

An 8 disk recovery isn't bad, two years ago I managed a near-100% recovery from a 24-disk Areca RAID6, despite a NO-INIT having been erroneously performed on disks that were out of their original order, *and* the controller decided to begin a new rebuild operation on the new (scrambled) drive order, essentially beginning to fill the disks with garbage data (aye caramba). A near worst case scenario, but did require 8-9 manhours of analysis in Winhex to untangle.

Feel free to PM me if you want a referral to someone that can assist with this.
 
Last edited:
Thank you all for your input. The array was RAID5 with a hot spare, all specs below. I had copied all the specs, minus drive serial numbers, on a piece of paper and left it taped inside the PC case, just in case i needed it as i do now. I did find out via phone that my ex gf's brother DID in fact reconnect the HDD's most likely in a different order; i did have the initial order and have put them all back in order as they were. i have also posted the log from the ARECA controller below.

Normally i do have backup external drives, but broke up with my girlfriend and moved out, and put all her stuff on the externals and let her keep them, so all my stuff was on the RAID5 array. Ive been abroad last few days and my PC just arrived, so i didnt have access to it or ability to post here. Anyways here are all the specs i have.

I assume it would be easiest to physically reorder the drives and try recreating exact RAID set with same specs and correct order.

I have an Areca ARC-1222 RAID controller running v1.49 Firmware and Boot ROM with SAS Firmware 4.7.3.0, as well as 8 Samsung HM500JJ 2.5" 500GB hard drives installed in a RAID 5 configuration. All the hardware has been thoroughly checked and reports hard drives functioning normally (S.M.A.R.T)
Here are the specifics regarding the array.
ARC-1222 (Serial Number Y103CABRAR200079) RAID SET # 000
(8) Samsung HM500JJ Hard Disk Drives
RAID5
Stripe Size 64kb
Block Size 4096
RAID Set Hierarchy
E1Slot#8
E1Slot#2
E1Slot#7
E1Slot#6 (listed as FAILED due to faulty power connector)
E1Slot#5
E1Slot#4
E1Slot#3
E1Slot#1


The array suffered a power failure due to a faulty power connector, and reported as FAILED upon reboot. After trying several recovery and Rescue RAID Set options, I am lost as to how to proceed to recover my data. E1Slot#6 I had listed as a Hot Spare.

I had tried 'RESCUE RAID SET' option, only to get the 'One or more volumes or controllers have failed' upon reboot multiple times, same when I attempted a Level 2 Rescue command.
As of now, I have deleted both the RAID Set and Volume Set, and attempted to create both a new RAID Set and Volume Set with identical specifications, however I am unable to change the RAID Set Hierarchy, which has changed to
E1Slot#1
E1Slot#2
E1Slot#3
E1Slot#4
E1Slot#5
E1Slot#6
E1Slot#7
E1Slot#8
I made sure the 'NO-INIT' option was selected so that the volume was not initialized. To my understanding the data is still on the disks, but the Array order appears to need to be changed, which I cannot seem to modify via either ARCHTTP GUI or via BIOS, the drive now appears correctly as 'R' drive in Windows 7 x64 Professional, but is not accessible and requires formatting, which leads me to believe again the hierarchy is incorrect.

2012-11-27 19:15:31 ARC-1222-VOL#000 Create Volume
2012-11-27 19:14:03 RAID5 Delete Volume
2012-11-27 19:12:49 RAID5 Create Volume
2012-11-27 19:11:42 Raid Set # 000 Create RaidSet
2012-11-27 19:11:14 Raid Set # 000 Delete RaidSet
2012-11-27 19:11:02 Proxy Or Inband HTTP Log In
2012-11-27 18:57:53 Proxy Or Inband HTTP Log In
2012-11-27 18:56:56 H/W Monitor Raid Powered On
2012-11-27 18:37:08 ARC-1222-VOL#000 Create Volume
2012-11-27 18:36:47 ARC-1222-VOL#000 Delete Volume
2012-11-27 17:41:09 Proxy Or Inband HTTP Log In
2012-11-27 17:37:17 H/W Monitor Raid Powered On
2012-11-27 17:36:28 ARC-1222-VOL#000 Create Volume
2012-11-27 17:35:57 Raid Set # 000 Create RaidSet
2012-11-27 17:35:21 H/W Monitor Raid Powered On
2012-11-27 17:34:13 Raid Set # 000 Delete RaidSet
2012-11-27 17:33:57 Raid Set # 000 Create RaidSet
2012-11-27 17:33:38 Raid Set # 000 Delete RaidSet
2012-11-27 17:06:45 Proxy Or Inband HTTP Log In
2012-11-27 17:05:26 H/W Monitor Raid Powered On
2012-11-27 17:04:17 ARC-1222-VOL#000 Create Volume
2012-11-27 17:03:24 Raid Set # 000 Create RaidSet
2012-11-27 17:02:45 Raid Set # 000 Delete RaidSet
2012-11-27 17:01:47 Raid Set # 000 Create RaidSet
2012-11-27 17:01:09 Raid Set # 000 Delete RaidSet
2012-11-27 16:59:51 H/W Monitor Raid Powered On
2012-11-27 16:33:19 H/W Monitor Raid Powered On
2012-11-27 16:31:53 H/W Monitor Raid Powered On
2012-11-27 16:30:43 H/W Monitor Raid Powered On
2012-11-27 15:26:53 Proxy Or Inband HTTP Log In
2012-11-27 14:00:37 H/W Monitor Raid Powered On
2012-11-27 13:50:39 H/W Monitor Raid Powered On
2012-11-14 15:10:16 H/W Monitor Raid Powered On
2012-11-14 14:27:43 H/W Monitor Raid Powered On
2012-11-14 14:45:44 H/W Monitor Raid Powered On
2012-09-23 16:20:37 H/W Monitor Raid Powered On
2012-09-21 14:25:11 H/W Monitor Raid Powered On
2012-09-13 18:34:02 H/W Monitor Raid Powered On
2012-09-10 20:17:21 H/W Monitor Raid Powered On
2012-09-09 04:48:48 H/W Monitor Raid Powered On
2012-09-03 11:14:22 H/W Monitor Raid Powered On
2012-09-03 04:08:51 H/W Monitor Raid Powered On
2012-09-02 03:28:33 H/W Monitor Raid Powered On
2012-08-30 20:06:41 H/W Monitor Raid Powered On
2012-08-29 20:51:42 H/W Monitor Raid Powered On
2012-08-25 01:23:05 H/W Monitor Raid Powered On
2012-08-24 18:07:53 H/W Monitor Raid Powered On
2012-08-24 17:34:48 ARC-1222-VOL#000 Create Volume
2012-08-24 17:34:24 ARC-1222-VOL#000 Delete Volume
2012-08-24 17:34:05 Proxy Or Inband HTTP Log In
2012-08-24 17:32:15 ARC-1222-VOL#000 Create Volume
2012-08-24 17:31:27 ARC-1222-VOL#000 Delete Volume
2012-08-24 17:08:20 ARC-1222-VOL#000 Create Volume
2012-08-24 17:07:42 ARC-1222-VOL#000 Delete Volume
2012-08-24 17:05:05 Proxy Or Inband HTTP Log In
2012-08-24 17:03:44 H/W Monitor Raid Powered On
2012-08-24 17:02:54 ARC-1222-VOL#000 Create Volume
2012-08-24 17:01:08 Raid Set # 000 Create RaidSet
2012-08-24 16:59:50 Raid Set # 000 Delete RaidSet
2012-08-24 16:57:55 ARC-1222 Delete Volume
2012-08-24 16:54:52 H/W Monitor Raid Powered On
2012-08-24 16:38:22 Proxy Or Inband HTTP Log In
2012-08-24 16:37:21 H/W Monitor Raid Powered On
2012-08-24 16:36:08 H/W Monitor Raid Powered On
2012-08-24 16:34:37 Proxy Or Inband HTTP Log In
2012-08-24 16:33:30 H/W Monitor Raid Powered On
2012-08-24 16:21:01 Proxy Or Inband HTTP Log In
2012-08-24 16:18:32 Proxy Or Inband HTTP Log In
2012-08-24 16:18:32 Proxy Or Inband HTTP Log In
2012-08-24 16:17:24 H/W Monitor Raid Powered On
2012-08-24 16:16:10 H/W Monitor Raid Powered On
2012-08-24 16:11:15 H/W Monitor Raid Powered On
2012-08-24 16:08:52 H/W Monitor Raid Powered On
2012-08-24 16:07:18 Proxy Or Inband HTTP Log In
2012-08-24 16:06:14 H/W Monitor Raid Powered On
2012-08-24 16:01:00 H/W Monitor Raid Powered On
2012-08-24 15:59:15 Proxy Or Inband HTTP Log In
2012-08-24 15:57:55 H/W Monitor Raid Powered On
2012-08-24 15:56:21 H/W Monitor Raid Powered On
2012-08-24 15:55:13 Raid Set # 000 RaidSet Degraded
2012-08-24 15:55:11 ARC-1222 Volume Failed
2012-08-24 15:55:11 H/W Monitor Raid Powered On
2012-08-24 15:53:33 H/W Monitor Raid Powered On
2012-08-24 15:53:24 Enc#1 Slot#8 Device Removed
2012-08-24 15:53:24 Enc#1 Slot#7 Device Removed
2012-08-24 15:52:02 Raid Set # 000 Rebuild RaidSet
2012-08-24 15:52:02 Enc#1 Slot#5 Device Inserted
2012-08-24 15:52:00 Enc#1 Slot#5 Device Removed
2012-08-24 15:52:00 Raid Set # 000 RaidSet Degraded
2012-08-24 15:51:58 ARC-1222 Volume Failed
2012-08-24 15:50:22 Raid Set # 000 Rebuild RaidSet
2012-08-24 15:50:22 Enc#1 Slot#4 Device Inserted
2012-08-24 15:49:59 Enc#1 Slot#4 Device Removed
2012-08-24 15:49:58 Raid Set # 000 RaidSet Degraded
2012-08-24 15:49:56 Raid Set # 000 Rebuild RaidSet
2012-08-24 15:49:56 ARC-1222 Volume Failed
2012-08-24 15:49:56 Enc#1 Slot#4 Device Inserted
2012-08-24 15:49:53 Raid Set # 000 Rebuild RaidSet
2012-08-24 15:49:53 ARC-1222 Failed Volume Revived
2012-08-24 15:49:52 Enc#1 Slot#4 Device Inserted
2012-08-24 15:49:34 Enc#1 Slot#4 Device Removed
2012-08-24 15:49:30 Enc#1 Slot#4 Device Removed
2012-08-24 15:49:29 Raid Set # 000 RaidSet Degraded
2012-08-24 15:49:27 ARC-1222 Volume Failed
2012-08-24 03:06:27 H/W Monitor Raid Powered On
2012-08-24 03:02:44 000:000030CFBD00 Lost Rebuilding/Migration LBA
2012-08-24 03:02:44 000:000030CFB600 Lost Rebuilding/Migration LBA
2012-08-24 03:02:43 Raid Set # 000 RaidSet Degraded
2012-08-24 03:02:41 ARC-1222 Volume Degraded
2012-08-24 03:02:41 H/W Monitor Raid Powered On
2012-08-24 02:29:19 ARC-1222 Start Rebuilding
2012-08-24 02:29:18 Enc#1 Slot#5 Device Removed
2012-08-24 02:29:17 Raid Set # 000 Rebuild RaidSet
2012-08-24 02:29:17 Raid Set # 000 Rebuild RaidSet
2012-08-24 02:29:17 Enc#1 Slot#6 Device Removed
2012-08-24 02:29:17 Raid Set # 000 RaidSet Degraded
2012-08-24 02:29:17 Raid Set # 000 RaidSet Degraded
2012-08-24 02:29:08 ARC-1222 Failed Volume Revived
2012-08-24 02:29:08 Enc#1 Slot#6 Device Inserted
2012-08-24 02:29:08 Enc#1 Slot#5 Device Inserted
2012-08-24 02:29:02 ARC-1222 Volume Failed
2012-08-24 02:29:02 ARC-1222 Volume Degraded
2012-08-23 19:13:57 ARC-1222 Complete Check 002:02:23 0
2012-08-23 17:11:46 ARC-1222 Start Checking
2012-08-22 17:11:03 ARC-1222 Complete Check 002:02:28 0
2012-08-22 15:08:46 ARC-1222 Start Checking
2012-08-21 15:07:57 ARC-1222 Complete Check 002:02:22 0
2012-08-21 13:05:46 ARC-1222 Start Checking
2012-08-20 13:04:56 ARC-1222 Complete Check 002:02:21 0
2012-08-20 11:02:46 ARC-1222 Start Checking
2012-08-19 11:02:09 ARC-1222 Complete Check 002:02:34 0
2012-08-19 08:59:46 ARC-1222 Start Checking
2012-08-18 08:57:16 ARC-1222 Complete Check 002:15:41 0
2012-08-18 06:41:46 ARC-1222 Start Checking
2012-08-17 06:38:56 ARC-1222 Complete Check 002:03:21 0
2012-08-17 04:35:46 ARC-1222 Start Checking
2012-08-16 04:35:02 ARC-1222 Complete Check 002:02:28 0
2012-08-16 02:32:46 ARC-1222 Start Checking
2012-08-16 00:15:02 H/W Monitor Raid Powered On
2012-08-14 12:04:09 ARC-1222 Complete Check 002:04:05 0
2012-08-14 10:00:15 ARC-1222 Start Checking
2012-08-13 09:59:42 ARC-1222 Complete Check 002:02:38 0
2012-08-13 07:57:15 ARC-1222 Start Checking
2012-08-13 05:57:30 H/W Monitor Raid Powered On
2012-08-13 05:51:55 H/W Monitor Raid Powered On
2012-08-12 11:59:21 H/W Monitor Raid Powered On
2012-08-12 11:54:45 H/W Monitor Raid Powered On
2012-08-12 11:52:40 H/W Monitor Raid Powered On
2012-08-12 11:51:02 H/W Monitor Raid Powered On
2012-08-12 11:49:54 H/W Monitor Raid Powered On
2012-08-07 20:30:34 ARC-1222 Stop Checking 001:41:58 0
2012-08-07 18:48:48 ARC-1222 Start Checking
2012-08-07 17:43:03 H/W Monitor Raid Powered On
2012-08-07 12:15:29 ARC-1222 Complete Check 002:02:44 0
2012-08-07 10:12:57 ARC-1222 Start Checking
2012-08-07 08:28:12 H/W Monitor Raid Powered On
2012-08-06 06:23:10 ARC-1222 Complete Check 002:02:19 0
2012-08-06 04:21:03 ARC-1222 Start Checking
2012-08-06 00:15:18 H/W Monitor Raid Powered On
2012-08-06 00:08:09 H/W Monitor Raid Powered On
2012-08-06 00:07:01 H/W Monitor Raid Powered On
2012-08-06 00:04:53 Ctrl 3.3V Under Voltage
2012-08-06 00:04:53 H/W Monitor Raid Powered On
2012-08-05 23:53:45 Ctrl 3.3V Under Voltage
2012-08-05 23:53:45 H/W Monitor Raid Powered On
2012-08-05 23:51:52 Ctrl 3.3V Under Voltage
2012-08-05 23:51:52 H/W Monitor Raid Powered On
 
Last edited:
SET # 000
(8) Samsung HM500JJ Hard Disk Drives
RAID5
Stripe Size 64kb
Block Size 4096
RAID Set Hierarchy
E1Slot#8
E1Slot#2
E1Slot#7
E1Slot#6 (listed as FAILED due to faulty power connector)
E1Slot#5
E1Slot#4
E1Slot#3
E1Slot#1

I assume this was the orignal configuration or after running rescue? If that is the case of seeing that order when trying a rescue then when you delete/recreate with no init (rescue) you must physically change the driver order. it should be:


Move slot 8 -> 1
Leave slot 2 where it is.
Move slot 7 -> 3
Move slot 6 -> 4
Leave slot 5 where it is.
Move slot 4 -> 6
Move slot 3 -> 7
Move slot 1 -> 8

Then delete/re-create array with no init option.

Has this array been dead since 8-24? If so once this happened and assuming the current order of the disks was the original order when it failed on 8-24 I would then the first two disks the initially failed (slot 5/6 now 4/5) I would kick/remove slot 4 or 5 out of the array (degrading it) and seeing what I could get off of it. They failed at the same time so its hard to say which one it revived the array with but I would think it revived with slot 6 (now 4) due to the fact its listed later in the event history.

You said the array has a hot spare but in that case I wouldnt think you would have 8 drives part of the array .. just 7.


just out of interest, how does the drive-order change when the hot-spare kicks in?
I'd assume that the hot-spare takes the place of the failed drive, which does not change the order?

good luck!

Think about it. There is a 'physical' disk order and then a 'logical' order of the data on the disks.

In a 8 disk array during its initial creation you would when always listing the physical disks in order (1-8) the logical order would be:
1 2 3 4 5 6 7 8

Now lets say you had a hot spare on slot 9 and slot 3 fails. Well when it rebuilds now the logical order (compared to physical) would be:

1 2 x 4 5 6 7 8 3

And replacing slot 3 slot 3 would now be your hot spare. Now say slot 7 fails then it would be:

1 2 7 4 5 6 x 8 3

And x would be your new hot spare.. and then slot 1 fails and rebuilds:

x 2 7 4 5 6 1 8 3

And x would be your new hot spare after replacing it. Now look what happened to your logical data order...
 
Unless you're talking about recovering from motherboard RAID with 2-3 disks, or from hardware raid where drive order hasn't been changed yet (which the OP apparently did), then this guy's recovery goes beyond what any recovery software can simply scan or auto-detect on its own as far as drive order/raid level/stripe size/partition offset, but can be done manually with R-Studio though requires expertise, compounded by proprietary block patterns depending on manufacturer. Such is the case with Areca - proprietary block pattern.

In any case nobody's going to do this for $50 - it will cost him a couple hundred to have someone remote-in to his system and manually analyze each disk image to determine original drive order, stripe size and raid level (since he's apparently confused on what he had). It'll likely take a few hours of analysis in WinHex to determine the manual parameters to lay into R-Studio in order to reconstruct the array virtually. So he'll need 2 x 3TB drives - one to hold the disk images and the other to recover the data to. Place the Areca into JBOD mode to create UNCOMPRESSED disk images from each raidset member disk.

An 8 disk recovery isn't bad, two years ago I managed a near-100% recovery from a 24-disk Areca RAID6, despite a NO-INIT having been erroneously performed on disks that were out of their original order, *and* the controller decided to begin a new rebuild operation on the new (scrambled) drive order, essentially beginning to fill the disks with garbage data (aye caramba). A near worst case scenario, but did require 8-9 manhours of analysis in Winhex to untangle.

Feel free to PM me if you want a referral to someone that can assist with this.

PM'd you. I have one 3TB on hand, ARECA in JBOD and making uncompressed copies of each drive at the moment. I have another 3TB drive on the way.
 
Back
Top