how to recover jumbled(tm) ARC-1210 Raid5 array?

Areca 1210 (4-port), originally 4-disk raid5

Moved from one machine to another, now it says the following during boot:
No BIOS disk found. Raid controller BIOS not installed!

The only explanation I have is that the order of disks got somehow mixed up and what used to be the disk connected to port 0 is not connected to a different port, and so on.

The disks are all recognized when looking at them using the controllers menu:
Main Menu -> Physical Drives -> View Drive Information

Now I could go through and try all possible combinations, but I'd rather avoid such frequent reboots if possible. Is there some way for me to figure out what the proper order of the drives is supposed to be?

Edit:
The Areca FAQ has the following to say about this issue:

Areca FAQ said:

Two conditions will let the system displayed after boot-up:

>>> No BIOS disk Found. RAID Controller BIOS not installed!

(1). Controller does not create any volume set. The message will be far away after you create the volume set and completed initialization.

(2). If your volume is created greater than 2TB, the BIOS will show ‘NO BIOS DISK FOUND’ and ‘RAID CONTROLLER BIOS NOT INSTALLED”. This is caused by the currently M/B BIOS only support MBR table (32-bit).

Click to expand...

I wasn't using an array that's larger than 2TB, those are four 400GB disks. Somehow the volume set info was lost, which leads me to believe (hope!) the order the drives are connected to the controller is wrong.

Edit2:
Unplugging all drives from the controller, and then plugging each drive into the same channel, one at a time (obviously), results in the controller recognizing that each drive is part of a raid set. However, as soon as a second drive is added, both drives' status changes to "Free" meaning the controller doesn't see them as part of the same raid set, or any raid set really.

I am just going to start trying the different combos, see what I can come up with.

Here's a little update, if I plug one drive at a time into the controller, in the order I think they were in originally, they individually all show up as part of Raid Set # 00.

Then going to Raid Set Function -> Raid Set Information -> Raid Set # 0
the following shows up for all drives when one is plugged in at a time:


Then the string for "Member Disk Channels" is different for each drive inserted, unfortunately I can't seem to figure out what the "-" and "x" actually mean.

CH1 connected: 1--x
CH2 connected: -2x-
CH3 connected: x-3x
CH4 connected: ---4

If someone could shed light on the meaning of the "-" and "x", perhaps I'd get a little closer to figuring this out. Also emailed tech support, see what they have to say about it.

So I received an email from Areca support saying that the order in which the drives are connected to the controller ports doesn't matter as long as they are all drives of the same set. The controller will figure out what's what.

That's good news in a sense that I am not disappointed anymore in the manufacturer who should have just created a system that works regardless of the order the drives are inserted.

The bad news is that I took a fully functional array out of a working powered-down, discharged, and electrically disconnected box (no BBU on the controller), put it in a new box, and lost the array in the process.

I replied to the tech support email with as much detail as I have on all of this and just hope they can fix the issue. I do have some more or less recent backups, so I am not totally doomed, but it would sure be nice to get the previous state restored.

Go to 2cpu.com's storage forum, and look for direct help there. There are many persons who have gone through the exact situation you are having, and fully recovered their array, as long as you have not played with it too much. If you can't find enough helpful information there by searching, contact Jus, who gives support for Areca on that board. It's good that you have some backups in case all else fails. Good luck.

Yeah, i can tell you how to fix this. Its a notorious problem with arecas, you move one drive, and it breaks the entire array. Heres how to fix

In the repair section of the web interface do the following sequence of commands, to the T and your array will show back up

#1) RESCUE

reboot like it tells you

#2) SIGNAT
#3) LeVeL2ReScUe (caps is important)

reboot like it tells you

#4) Array should be back now, do a SIGNAT again and itll stick

That should fix it, its a pain that none of this is documented, but as i said, known problem, i ve had it a ahalf dozen times.

Thanks for the suggestions folks. I did have a look at 2cpu, and used jb's suggestion as well. This is where I am at now:

During the first reboot after SIGNAT + LeVeL2ReScUe the raid set # 00 shows up, the beeper is quiet since a raid set has been found. The two volumes that I had also show up.


The info for the second volume looks identical other than a 1 as set name and LUN number.

This controller is now in an ASUS P5W64 WS Professional. Once the raid set was recognized during boot, the mobo tries to boot from the ARC-1210, and fails (it never was my system drive).

If I restart the system after the mobo BIOS tells me there is no system disk, the raid set is again lost on subsequent boot.

If I enter the mobo BIOS setup while the raid set is recognized, it shows up in the mobo BIOS as well, I then set my system drive to be the boot drive, save the changes to the mobo BIOS and restart the system. The raid set is again lost on the next boot.

So it appears what I need to do now is to somehow manage to run SIGNAT + LeVeL2ReScUe and be able to complete the booting into the OS without that the mobo tries to boot from the Areca. A quick look at through the mobo BIOS didn't show any option for excluding add-on cards from being considered as boot drives/devices. Will keep at it.

Alright, I am back in business!

I was able to select the boot device (not in the BIOS, but using the F8 key during boot) to prevent the machine from trying to boot from the Areca.

madcoder, thanks for suggesting 2cpu, jbrukardt, thanks to you as well man, you are a scholar and a gentleman!

The array is back, both volumes are there, no data was lost, and now a backup of all the data is running.

The truly crazy part about this story is that one of the things I do at work is administer and maintain the backups for everyone else. So if this would have happened to any of my users, it would have been no problem at all to recover their nightly backup in much less time than it took me to figure out how to fix my problem.

Just as a reference point should anyone come across this thread and be unable to recover their array using the SIGNAT + LeVeL2ReScUe procedure, here is what Areca tech support suggested to fix the problem I described above:

Areca Tech Support said:

it looks like you had reorder these drives now.
all these member drives been placed in original order.

due to the history missing and configuration in drives corrupted, so we have
to recovered the array by recreate the raidset configuration table. the
procedure will be :
1. power on system with drive 1, 2, 3 connected only.
2. check the raidset information, make sure all three drives belongs to same
raidset.
3. put the drive 4 in. after it recognized, check the raidset information
again. the drive 4 should be added into the array.
4. create a volume with exactly same settings as original volume.
the most IMPORTANT thing is the initialization mode must be "no init for
rescue".
please double check the initialization before confirm the process, if the
volume start initialzation, the data you stored will be cleared.

after that, you will got a raidset and volume with normal status.

Click to expand...

I have no idea whether that works since I didn't recover that way, but considering the source it is reasonable to assume that the above quoted method will work as well.

I never got that method to work personally, but it never hurts to try. If you really really break your raidset, theres one more method too. It involves remaking the raid with all drives with the same parameters, but not initializing like he said. Had to do that once too

Good news that you got the raid back. The approach seems to be trying the steps outlined by jbrukardt, and which I was referring to first. If that doesn't work then the next recommendation seems to be this rebuilding the array configuration using the same parameters and drive ordering as originally existed, but without allowing the initialization. I myself use an Areca 1280ml, and also moved the whole raid from one machine to another at one point. I was somehow lucky and didn't get the issue.

I made the nasty mistake of pulling the wrong drive (while the system was running) when it was in a degraded state in my 12x1TB raid5 array on my areca-1231ML. That one was fun.... I had to shut down the system put the good drive i pulled back in the array add the new drive (in place of the failed one) completely erase the raid and volume sets and then re-create them using the exact same settings as before except chosing no init (rescue) reboot and and don't let the OS try to load but let the cards firmware startup then disconnect the new good drive (blank) and so it was back in its original degraded state, and hook the drive back up which showed a ghost volume, delete that, and force it to a hot spare to get the array rebuilding again. Didn't lose any data but definitely gave me a scare and took me about 4 hours to figure out how to fix on my own and I will soon be going to raid6!

jbrukardt said:

Yeah, i can tell you how to fix this. Its a notorious problem with arecas, you move one drive, and it breaks the entire array. Heres how to fix

In the repair section of the web interface do the following sequence of commands, to the T and your array will show back up

#1) RESCUE

reboot like it tells you

#2) SIGNAT
#3) LeVeL2ReScUe (caps is important)

reboot like it tells you

#4) Array should be back now, do a SIGNAT again and itll stick

That should fix it, its a pain that none of this is documented, but as i said, known problem, i ve had it a ahalf dozen times.

Click to expand...

haha wtf is that a joke

do 3ware cards have insane problems like that, if not so ill take them with the lower performance plzzzz

" my 12x1TB raid5 " that is too many, your chance of failure is so high

Well, in all fairness, my problem was apparently caused by user error.
It appears that the cause had something to do with unplugging drives from the controller while the system was powered off, but not electrically disconnected (meaning plugged in but turned off even though the powersupply still had power and was giving power to the components as well, whatever that type of powered on is called).

The Areca tech support had the following to say about LeVeL2ReScUe, posting it for reference in case someone else comes across this thread:

Areca Tech Support said:

the level2rescue works with some specify condition only.
an incorrect level2rescue action may corrupted the configurations in drives
and increase the risk of data lost.
so if you got abnormaly array problem, please contact with us before take
these actions to avoid the risk of data lost.

Click to expand...

bAMtan2 said:

haha wtf is that a joke

do 3ware cards have insane problems like that, if not so ill take them with the lower performance plzzzz

" my 12x1TB raid5 " that is too many, your chance of failure is so high

Click to expand...

Indeed, which is why I am going to 20x1TB (raid6)