System + Drivepool drive crash

[locus123]

I have Drivepool running on WHS2011.  Pretty sure it's the latest 1.3 version. 

 

We had a power outage and my server died.  Now I get a bluescreen as Windows boots up.  I've tried a System Restore - same bluescreen.  I've tried chkdsk /f - same bluescreen.  

 

If all I had on that drive was the System Image, then I'd just restore it to a new drive.  Unfortunately, half of it was the System drive, and the other half was part of the pool.  What is the best way to attempt to recover the drive pool piece of this drive?

 

I'm thinking to System Restore to a new drive and attempt to boot.  If successful, then bring online the "broken" drive, plus the other 4 drives, and check the pool status.  Will that work?  Will the pool be smart enough to just piece it all together?

[Christopher (Drashna)]

First, I'm very sorry to hear about your power outage, and resulting corruption. That is never a pleasant experience.

 

As for the data, was that on a separate partition, or all on the same partition? 

I ask, because if it was on a separate partition from the system, you should be able to restore *just* the system partition, IIRC. That would be the best solution.

 

If you are unable to do that, then did you have duplication enable for the whole pool? If you did, then just reinstall, and wipe the data on that drive (disconnect all the other drives first, to make sure it installs to the correct drive).

 

If that's not an option (such as you didn't duplicate all the data), then you can definitely install/restore to a new drive. Then DrivePool will see the "old" drives, and recreate the pool.

 

And yes, DrivePool is generally "smart enough" to recreate the pool in a lot of different circumstances. In fact, a lot of time and effort was put into making sure that it will. It's not fool proof, but it does work very well.

[locus123]

Thanks for the empathy!  Not pleasant at all - especially the words from my wife.

 

Yes, the system and the data were on separate partitions.  However, not ALL the data was duplicated.  Only the stuff i deemed worthy (pics, music, docs, etc).  

 

I'm not sure why, but even a System Restore on the old drive doesn't work.  It still blue screens.  The drive itself might be toast.

 

OK, it sounds like my plan should work then.  I'm going to do a System Restore to a NEW drive and try to boot.  If that works, I'll hook up the "bad" drive, and the other 4 working drives, and hopefully the entire original pool is recognized.  If it is, then I'll ask Drivepool to remove the bad drive from the pool which should copy everything off of it, then format that drive and see if there was physical damage.  Thanks Christopher!

[Christopher (Drashna)]

Yeah, hardware failure will definitely drop the WAF.... especially as that's 8+ hours away from her, to fix everything, and no TV/Movies/Music from the server until it's fixed.

 

Yeah, it sounds like the drive is toast. In fact, I've had drives do similar.... even when not the boot drive. The power surge probably damaged the firmware on the drive.  

 

If you can get the contents of the hidden "PoolPart" on the bad drive, and copy that to another disk, you will have "recovered" the data. This is where all the contents are stored on the drive, by DrivePool.

[locus123]

The ugly:

- System Restore wouldn't work.  Even when I restored the last backup, and a backup from a week ago, to a new hard drive, I still got the blue screen.  It doesn't makes sense, but I didn't have time to figure it out.  I gave up and reinstalled WHS2011.

 

The good:

- After WHS2011 install, then hours of updates, and Drivepool installation, it recognized all my drives and the pool was back. (except for not taking my license, which I've opened a support case for)

 

The bad/unknown:

- There are two "folders" that have crazy long names.   e7ad45af09d7fc7db3283f942b724458 is one of them.  They're empty.  

- When everything first came back, it looked perfect.  Then Drivepool was "duplicating data" for a while.  All of a sudden I went from 800Gb free (which is what I had before) to 3.6Tb free.  I have no idea where the space came from, but I'm afraid something was lost.  Is there any way to tell what happened?  Logs or something?  I've looked and I can't tell if anything is missing.  There are hundreds of movies, and thousands of pictures (which is what I really care about).  

[Christopher (Drashna)]

Ah, okay, that's what happened here, then.

 

I'd recommend running a memory test here, just in case. And a full check on the system disk. 

Those are the possible culprits.

 

These folders are used as temp folders for some updates. I'm sure if you look at the times, you'll see they were created during a time when you were updating. If they're empty, they are safe to delete.

 

If you could post logs, we can take a look. But my guess is that the duplication flag got un-set, and it unduplicated a bunch of data.

That is very unusual. We store the duplication "flag" on the pooled disks, so they should never get "unset" unless somebody manually did that. In fact, there are a number of safeguards in the code to prevent that. So if you could get the logs, that would really appreciated:

http://wiki.covecube.com/StableBit_DrivePool_2.x_Log_Collection

(steps #6-8)

[locus123]

Those folders only showed up as "shares" after I ran the WSS Troubleshooter and recreated the shares.  Although they could have been there prior, I really don't know.  They're empty however, and I've deleted them.

 

I looked at the folders I had duplicated previously, and they won't account for the size mismatch (if it was their duplicated data that was lost).

 

As for the logs, I'm running 1.3.  It looks to me that unless I enable tracing, logs aren't captured.  Is that correct?  If so, any suggestions?

[Christopher (Drashna)]

Sorry, if you're using StableBit DrivePool 1.X, then these are the instructions you want:

http://wiki.covecube.com/StableBit_DrivePool_Log_Collection