HELP: 7TB data lost in my RAID 5?

Max-Powers

Gawd
Joined
Jan 9, 2005
Messages
558
I have had a Sans Digital TowerRAID 5bay for over a year now. Something similar to this. Don't know if it is the exact model I have 5 1.5 TB samsung drives in raid 5
It has worked well, only encountering a couple instances where my computer would stop seeing it and would require a shut down and restart of the computer. No big deal. Well now it has stopped working entirely. The status lights for all but the bottom HDD go out, my computer freezes completely for about a full minute. Then the computer responds by asking me to format drive E: the status lights of all the HHD come back on, and "my computer" shows the new drive as just a unformated drive E. No longer having the name that I gave it. Then within about a minute, the status lights go out, the computer freezes, and it just keeps doing this. So I turn the RAID tower off. Shut down my computer, reset the hardrives, check connections, and try again. It does the same thing, even freezing the windows load screen.
I have a huge problem here. I thought I was safe with RAID 5. I dont know if it is a hard ware problem or software. What should I do? Call tech support? Get a new system of the same brand/model and see if it works; recover my data, and then return the new tower?

it is a RSV-s5
 
As an Amazon Associate, HardForum may earn from qualifying purchases.
I recently recovered a large raid 5 array with rstudio. As long as you input the correct raid parameters, it should pick up the data.

The other program that might help is ZAR. At first I was unsure of the raid parameters of my array, and it was the only program able to properly work it out for me (although it didnt get the actual data back, I just input the parameters it found into r-studio).
 
What are "RAID parameters"? I will look into downloading ZAR but if it is required to run and look at the RAID array it will have problems because of the 1 min frozen 1 min free that my computer goes through when this thing is turned on.

And yes I had problems with it 11 months ago like I mentioned. But I managed to find a work around. I planned on buying a new system but I just didn't have the money to buy a server+hard drives to transfer the data.
It looks like rosewell tech support has gotten back to me, I will let you know how it goes

Dear Valued Customer,

Thank you for contacting Rosewill Support Services.
Please accept our humbled apologies for this inconvenient situation.
It sounds like you have lost the backplane connection to a short or defect.
Unfortunately, we do not carry independent parts for a replacement and will need to process an RMA to have your Server case exchanged for a brand new retail unit.
Please provide the following information and we can have a replacement adapter sent to you ASAP:
<contact info>

Hopefully they dont tell me it has been too long
 
Raid parameters are block size, order of the hard drives, and parity placement.

If the raid controller is inaccessible or causing freezing, you may want to label the drives (so you remember what order they were in, and then just connect them to the sata ports on your motherboard.
 
Kind of sounds like the power supply maybe weak in that external storage box..


But If you need to rstudio saved my bacon on a very lard raid-5 array. The controller lost it's config and the tech working on it did not keep track of the drive order when he removed them.

Found a cable with 6 connectors, 6 hot plug adaptors and connect all the drive to a bench pc and fired up rstudio. It first complained about the drive order and told me the correct order for the array (this was amazing!). It thave created the virtual array and recoverd the whole thing with 100% data recovered.
 
I would be buying the exact external case again, and slap those drives in the new enclosure. I wonder if the Raid would be seen..
 
What raid controller are you using? I know some of those external SansDigital just have an esata link. The tower might have a bad power supply or backplane depending on what model you have. If I were you I would immediately take the drives out mark the order. Get an RMA or check below....

All raid controllers software and hardware will both write whats called meta data to each disc.

This meta data contains the layout of the raid array, parity, stripe size, etc....

If you build this raid array on an ICH10 chipset for example, then you can place the disc in the correct order on anybody's mobo with that same chipset and it will pick up the array. Also do a smart scan on each hard drive. You may need to plug each disk up to an esata port on another computer to run smart scans and find out if you have a bad disk or a bad raid chip or both.

Good luck.

And for the future it is wise to backup the contents of your array to another external disk. Software like Acronis or others can break images down to certain sizes.

If you have 7TB of data you can buy 4 2 TB drives and have software make a complete image of all files and break it into 2tb chunks sort of like a massive zip file. This way you do not lose everything. I am going to be doing this to my Adaptec powered NAS in the next few days again just to be sure.

Not directed at you but a large amount of people are jumping head first into raid technology not knowing how to properly utilize it and or prevent catastrophe. Raid is not designed to preserve data, it is designed to preserve the array of disk and the bonus is your data might be retained in a disk failure. Redundant array of inexpensive disk, not data. You are still responsible for backing up all data. Raid prevents only one point of failure but you must prepare for others. For instance if the raid controller shits on you and wipes the meta data or parity stripe gets screwed then what good is all those disk for anyways.
 
Last edited:
So i got a new raid tower. Same model. I plug it in same problem. I change out the backplate. Same thing. Change the slot on the mobo, same thing. Mess around with it a bit and now it is staying online.
But it is not recognizing my RAID. At first computer management saw it as a 5.5 TB and 1.3 GB hard drives. However, ZAR saw it as a 1.3 GB hard drive. I tried running ZAR and it had an error. Now computer management can no longer see the other hard drive.
What do I do? It only sees one of the 1.3 GB hard drives. All the lights are on.

EDIT:
Okay, I turned everything off and reseated the hard drives and now the 5.5 TB drive is back. So computer management says I have a 5.5 TB drive (E) and a 1.3 TB drive (D) I have ZAR, Rstudio with the R project installed. This is running through the raid tower. What do I do?
 
Last edited:
I usually always figure out my computer problems using the internet. I am lost on this one. This is over my head. I do not want to admit defeat. What do I do then? Just trash all my data?

I am running windows 7 32bit. The raid tower is a RSV-S5. It runs a silicon image raid controller. Specifically "Silicon image SI 3132 SoftRaid 5 Controller" I have the updated bios, drivers, and utility. The utility says "Silicon image controllers not present" when it tries to start. It has done that in the past, but I have always had access to my data just fine. I think the utility is more of a GUI.
My computer sees 2 drives and asks me to format them. Computer management sees one 5.5 TB drive and one 1.3 TB drive. (even when one of the drives is unplugged it sees this)
I try to run ZAR and it hangs after I choose the 2 drives, specify raid 5, and click next. It also sees the two drives as 1.3 TB. It does not see the one that is 5.5
I do not know how, and do not understand how to run Rstudio. Is there a way I can find out my "parameters" from just knowing the raid controller?
Any information would be greatly appreciated. This is quite frustrating. If anyone thinks they can walk me through this please let me know and we can get on facebook chat or something.
 
Last edited:
I would plug all the disks directly into a PC and boot off a recent linux livecd/livedvd and see if the array is recognized.
 
I would plug all the disks directly into a PC and boot off a recent linux livecd/livedvd and see if the array is recognized.

I tried that. But my computer wasnt recognizing all my SATA ports or something.
I have a GA-MA790X-UD4P
 
I do not know how, and do not understand how to run Rstudio.

Have you tried reading the manual thoroughly? R-Studio has a very well written step-by-step guide in their manual on how to recover a RAID array.

I tried that. But my computer wasnt recognizing all my SATA ports or something.
I have a GA-MA790X-UD4P
Which Linux distro did you use?
 
I was able to get bios to see all my hard drives when plugged directly into the mobo. I do not know the order of the SATA ports tho. computer managment sees 3 drives that are healthy and RAW, and 2 drives that are unallocated. One of the "healthy" drives is 5.5TB, the rest of the drives are 1.3 TB. I am having the same problem with ZAR. I select the drives, and it hangs on RAID analysis.
How do I get ZAR to work and why is one of the drives 5.5 TB?
 
why is one of the drives 5.5 TB?

Probably because that is the drive with the partition table on it (from the raid) and it is being seen as a individual disk.
 
ok. Well ZAR sees 5 disks of 1.3TB. So i am guessing it is seeing all the drives. But it freezes and during RAID analysis. What do i do?
 
I can not help you with ZAR.

I would however use CrystalDiskInfo (or some other program that looks at teh SMART data) to examine each drive and report if any of them are bad perhaps that is the problem.
 
crystakdiskinfo says DiskC good
Disk 1 good, D Caution, F caution, E Bad, Disk 5 Caution
Drive E is the one that is 5.5 TB. the thing that is bad is attribute name Reallocated Sectors Count.
What does this mean
All my pictures and lots of files were on here. What are my options.
 
You may be able to pull the one marked "bad" and see if ZAR can assemble the array with that disk removed.
 
ok. Well ZAR sees 5 disks of 1.3TB. So i am guessing it is seeing all the drives. But it freezes and during RAID analysis. What do i do?

Enable the log when running ZAR. It will usually tell you the issue its having (mostly commonly bad sectors on one of the drives).

You can also email the log to the person who makes ZAR, and he will be willing to help solve the problem you having (ive done it myself)

Lastly, it is also possibly to find out from the ZAR log what raid parameters it thinks it has found (as long as its got say 10%+ the way through the analysis). From there, you can input those raid parameters into something like rstudio (this is what I did, and how I recovered 90%+ of my data).
 
Back
Top