Problems with 'Unknown' disk

[ace]

Hi.  I am currently having a problem with my pool, running on WHS2011.   Hope someone can help.

 

WHS started to complain about 1 of the disks being missing.   I checked the dashboard, and sure enough, my 3TB drive was no longer listed.   So i RD into the server, and fired up the Windows disk management tool.   The drive was there, but listed as 'unknown'.   I searched the internet (including, and especially, this site), and the best suggesting was to initialize the disk.  Many posts suggested the data would be safe ... so I did that.   My 3TB drive reappeared, and the data seems to be ok.

 

However, I have a bit of a hangover issue, whereby, the data from that drive seems to be included in the pool, and, the system is writing new data to the 3TB drive as if it is in the pool, however, the drive is listed under 'Not in Pool / Drives that can be added to the pool', and I can neither add it, nor remove it from the pool.

 

I would like to add a new 4TB drive (to replace a different 2TB drive), but would like to fix the problem with the 3TB drive before I do that.

 

Thanks!

[Christopher (Drashna)]

I'm sorry to hear about the disk issue. That's never a pleasant experience

 

 

As for the issues you're seeing... well re-initializing the disk is probably the issue.

 

If the data is actually still there, then the solution may be "simple", relatively speaking.

If it is, then do the following:

1. Stop the "StableBit DrivePool Service"
2. Enable "Show Hidden Files" on the system
3. Find the hidden "PoolPart.xxxx" folder on the disk in question
4. Move the contents of the folder into the "root" of the disk (basically, out of that Poolpart folder)
5. Delete the now empty PoolPart.xxxx folder
6. Start the StableBit DrivePool Service
   It should complain about the missing disk, and should not be able to properly identify it.
7. Remove the "missing" disk from the pool
8. Add the 3TB disk to the pool.
9. Move the contents of the disk into the newly created PoolPart folder on the drive
10. Reset DrivePool's settings 

Unfortunately, it's entirely possible that it will still identify the problem volume in step 6. It shouldn't, but ....

If that's the case, then you need to move the contents of the disk off of that disk and format it (and maybe re-partition it)

[ace]

Thanks for the response drashna.  I decided to move the data off that drive, drop the volume, recreate it, and then format it.   I am at that stage now, and am running a few tests on the drive with the Seatools app.   I really don't think anything is physically wrong with the drive (chkdsk has come back clean several times).  But I want to make sure.  

 

A couple or few questions if I may ....

 

1) do you have any idea why a disk would suddenly change to 'unknown' status in whs?

2) if re-initializing wasn't the correct way to fix an unknown disk, then what is a better solution?  (keep in mind, the drive was not accessible from explorer, or any other way I could think of).

3) once I finish testing the disk (and assuming all is well), then I have to continue your steps above at step 6.  do i then move the contents back into my virtual pool drive (in my case, my pool is K:), or back into the "PoolPart..." folders on the 3TB disk once it is readded to the pool?

4) how do I avoid extra duplicates in the recovery scenerio?

 

My own theory for #'s 3 & 4 is make sure real-time duplication is turned off, and then copy the files onto the new drive.  Then turn duplication back on.  I believe drivepool is smart enough to look for duplicate files, and won't duplicate them again (making them triplicates).   But as I said ... this is a theory, and I don't have enough info to feel comfortable risking my data (some of which is unduplicated, and was on that 3TB drive).

[Christopher (Drashna)]

That works too, and is probably a good idea, to boot.

 

1. Lots of reasons. Mostly damage to the MFT/Partition table on the disk. A loose cable, or bad connection, or an issue with the controller, or the like could cause this. The best way to find out may be to check the event viewer, and look for disk, IO or controller errors at about the time it occurred.
2. There isn't a "good" way. It's always bad when this happens. But sometimes, a power cycle can fix the issue. And there are tools out there that can repair the partition (or if you're crazy/skilled, it can be done "by hand"). But as I said, it's never a good thing, and whichever way you do it, you risk data loss.
3. Move it back into the Pool directly. The Pool will then handle the placement of the files (in fact, you should probably do that now, if you have the space). 
   We generally recommend against accessing the "PoolPart" folder directly, except in very specific circumstances. 
   And since you already have the data off the disk.... there isn't any benefit to using the PoolPart folder for this.
4. See #3. That's why I recommended copying the files to the Pool directly. DrivePool will take care of duplicates and such. Chances are, it may have already duplicated the necessary files, actually.  Otherwise, if you get any "this file already exists", you can overwrite, if you want.

As for Real Time Duplication, if you follow above, (and don't turn it off), then the Pool handles that for you. Period. Simple. I've done it myself a couple of times, and it's much less work.

 

But yes, DrivePool is smart enough to not reduplicate the files. It checks to see how many copies it finds. in fact, if it finds "too many", it will delete a copy.  Also, if the dates don't match, or the file sizes are different, it will actually prompt you about that. And it uses the same process to make additional copies.

 

As for not trusting the disk, open the Balancer settings in DrivePool. Then open the "Balancers" tab, and find the "File Placement Limiter" balancer. From there, you can actually disable duplicated or unduplicated storage per disk. Uncheck "Unduplicated" for that disk, and it will ONLY store duplicated data on that drive. It sounds like a good idea. At least for now.