RAID failure?

[khayman80]

I have an MSI K8N Neo Platinum motherboard with an Athlon 64 2800+, 1 GB of PC3200 Crucial DDR RAM, an ATI x800 Pro video card, and two Maxtor SATA 160GB drives plugged into SATA 1 and 2 respectively. I have had these drives configured as a RAID 1 array ever since I built the computer.

I built this system 15 months ago, then 10 months ago I experienced a severe RAID failure (i.e. I lost all my data). Setting up the system wasn't difficult once I figured out which order the drivers had to be installed in, but one day (about 5 months after I built the computer) the RAID array *SPLIT* into two different RAID arrays. Each array read as "degraded", which I assume is because each array only had a single drive in it. I was too busy with school to fix it immediately, but 1 month later the whole damn thing crashed- hard. Wouldn't boot up at all.

I assumed that it might be caused by a cheap power supply, lack of adequate cooling, or some damage to the hard drives during a move that occurred several months prior to the crash. So I replaced the power supply with a very nice 400W model, improved case cooling and added hard drive cooling units, and replaced both hard drives under warranty. After rebuilding the system I had 10 months of peace.

Until today. Today I got a message from NVRaid that said "searching for spare hard drive...". Curious, I opened NVRaid and noticed that my array was "degraded", so I clicked on "rebuild array". This caused a BSOD! I *never* get BSOD's on this box- it's very stable. Anyway, when it rebooted I felt the fear of death wash over me- the BIOS was reporting that I had TWO RAID arrays again- just like last time! Each array is composed of a single hard drive, and each array is "degraded" according to BIOS and NVRaid.

In addition, I now see each hard drive separately in Windows explorer.

Does anyone have any advice? I really don't know what to do here- I've tried googling for NVRaid and the only thing I found was this forum. Nvidia apparently doesn't support NVRaid, and I wonder if MSI will be any help. Basically, I just want my array working again, hopefully without having to reinstall windows. I think I have several options:

1. Rebuild the array somehow. I believe this can be done in the BIOS, but I'm terrified of making a mistake and losing all my data forever. I'm not sure what command will "consolidate" the two RAID arrays into my original single array. Also, I can't tell which drive is which in the BIOS- they're not identical anymore that one of them is my system drive and the other is just sitting there.

2. Take both hard drives out, put two new hard drives in, set them up as a RAID 1 array and then somehow copy the data from the most recently updated old drive to the new array. That is, I want to take my current installation of windows with all its settings and simply copy that entire hard drive bit-by-bit to the new array so that I don't have to spend an entire day configuring windows. Can this be done? If so, how?

I'm sorry if these are redundant or trivial questions, but I'm worried sick about this computer and I'm not sure what to do. Anyone have any suggestions?

Thanks...

-Khayman

[wonkanoby]

they're not identical anymore that one of them is my system drive and the other is just sitting there.

then i suggest the one just sitting there has gone bad

[khayman80]

then i suggest the one just sitting there has gone bad

That's what I thought at first. However, two things (that I didn't mention) seem to contradict that conclusion:

1. The drive that's "just sitting there" is still working. (Yeah, I was vague about that.) What I meant was that it's not my "C:\" drive anymore, and therefore doesn't get updated automatically whenever I use a program.

2. When the crash occurred, I was in the middle of copying some files from another computer to my RAID array. The drive that is currently my C: drive did NOT finish copying, whereas the drive that is currently my E: drive DID finish copying. Weird.

I wish that one of the drives WAS bad. I can deal with drives going bad- that's not a problem. That is, in fact, why I set up the RAID array in the first place. What's confusing me is the fact that the array seems to SPLIT into several arrays whenever something goes wrong. Is this normal?

Anyway, I forgot to mention that I'm running Windows XP SP2, for what it's worth.

[khayman80]

I've been tinkering with the system a little, and I'm still confused. First thing I did was disable RAID in the bios, then use Powermax to diagnose both the hard drives. Powermax said that one of the hard drives had "errors" and I allowed it to "fix" them. It gave both hard drives a clean bill of health, so to speak, after this was done.

Then I attempted to return my system to the point it was at before the problem. In other words, I have two hard drives currently not RAIDed at the moment, but I'd like to use the good hard drive as a "template" and start up the RAID 1 array again, overwriting the drive that hasn't been used lately (and therefore isn't up to date).

I thought I knew how to do this. I re-enabled RAID, on both hard drives. I went into the RAID config utility right after the BIOS by hitting F10, and saw two arrays. I deleted the array that corresponded to the bad hard drive, and added it to the array with the new hard drive and hit "rebuild" then exited.

At this point, my Windows XP boot screen flashed for a SECOND, then my computer automatically restarted- right back to the BIOS screen. This continues in a seemingly infinite loop unless I do something- which I did. At the moment RAID is completely disabled, and my system boots up fine. Of course, this means I see two hard drives in windows explorer and I have absolutely no data redundancy at the moment, which is driving me insane.

I'm completely and utterly confused by what's going on here. I would be very grateful to anyone who could shed some light on this subject. What's going on? How to fix it? Etc...

Thanks in advance...

[wonkanoby]

if its giving a bsod them i suggest that drives still bad no matter what

powermax says

[aragorn_246]

Khayman,

I have exactly the same problem on my Asus K8N (NForce3) mobo.

Two Maxtor 160's in Raid 1 suddenly decide to split and become two one disk raids.

I think this is a firmware problem with the RAID controller.

Haven't fixed it yet this time, but I think I have to clear the secondary disk in the NVRaid bios util and then rebuild the array on to this disk. Going backup some data first tho!

Had no end of problems with this crap controller - I expected more from NVidia

[andy b]

I also have the identical problem.

It seemed to coincide with the deletion of a large file on the raid volume. (About 4.5 GB). Windows notified me of a raid volume problem (don't remember exactly what it said) - I rebooted and got the double 'degraded' BIOS red message, and the nvidia tool showed TWO mirrors, each degraded.

Both disks are fine. I can see them both in 'my computer'. I should only be seeing one... my raid 1 array!

Sigh.

I'm scared to do anything that might result in a total reformat, especially since if it's a firmware problem I don't fancy rebuilding my machine each month.

[mschoaf]

I'm going through this exact problem right now. I have an ASUS A8N-SLI Premium board using the NV raid channel and yesterday during the reboot, my raid array split into 2 and both were reported as degraded. I am only using RAID 1 with 2 SATA drives.

Fortunately, I was able to save all my data off in case I had to re-install the OS.

The first thing I tried was to delete the secondary array, add it back in to the primary array without destroying data and rebuild. This seemed to hang XP after booting up, logging in and trying to use it. I was watching the rebuild stat, but it never got to 2% before it would hang.

So, I then deleted the primary array, then created a new array with both drives, never destroying data and then rebuilding the secondary disk on the array. This time, before the welcome screen, XP ran through a buch of chkdsk repairs. After this, I was able to log in and I just let the rebuild go and XP did not hang. I haven't had a chance to use the system yet, so I'll followup my post later. I do know that it was up to about 86% rebuild when I had to leave for work.

[andy b]

Following a conversation with the company who assembled my PC (www.evesham.co.uk), they told me that they're approaching nvidia about it.

One of the most frustrating things is that it seems to be impossible to get through to nvidia ourselves. This forum seems to be the closest thing.

Given that many people will be using the onboard RAID to mirror the system disk specifically so we don't have to worry about reinstalling all our stuff if the system disk fails, I'd describe this as a rather critical problem.

I wish there was some way of knowing that:

a) Nvidia know the problem
b) They're on the case, with some sort of resolution in sight.

[mschoaf]

I don't think I have very good news for you all. Windows does SEEM to be running ok, but I have a bunch of little quirky problems. When I brought Outlook up for the first time, it said my mailbox was corrupted, so I pulled that from my backup. Word had a problem with the normal.dot. Norton said it's settings were corrupt and reset to the defaults. And who knows about the stuff I haven't seen yet.

So, I'm reluctantly coming to the conclusion that I need to reformat and re-install. I'm also thinking about pulling an old IDE drive from one of my spare parts computers to use as a backup drive and backing up my full system weekly and my data daily. The sad part is, that's why I bought a MB with RAID 1 capability, so I wouldn't HAVE to do this.

I guess I'll also open a ticket with ASUS and maybe they could start leaning on NVidia as well. One can only hope.