Drive Fails During "remove From Pool"-Procedure

Hi everyone,

 

I’m just an enthusiastic amateur and have been running a homebuilt Windows Home Server 2003 for about four years, as the central storage and backup place – for private use only. This is my first post and “call for help”. Please excuse my long post – and bear with my knowledge level and way of describing/phrasing.

 

 

 

A few weeks ago a 1.5 TB drive began to display "failing" in WHS Console. Online I found forum discussions indicating it was best to then run a "remove" rather than a "repair". To get enough space I temporarily mounted more disks that I had at home, both IDE and external USB.

 

I started the "remove" process, WHS checked space and I got the confirmation message that the disk could be removed without consequences – and removal was initiated.

 

When I seven hours later, via a local client, came back to check progress, the WHS Console is in "offline" mode (gray icon in taskbar). A pop-up message (the regular balloon) however still sits on top, saying something like "HD (serial #) failed. Data could not be written to \ \ Share \ Photos ... \. The files are permanently lost ".

 

Panic. The server is still running but the WHS Console cannot login to the server from the client. I switch on the server’s local monitor, on top of the login screen there are error messages (red x) with similar info as in the balloon, "HD failed ... files are permanently lost". There are approximately five of these dialogs atop one another, with different folder paths, which I need to OK my way past to get to the login dialog. In my shocked state, I only thing I can think of is that these paths hold critical files and I just click through them to be able to login.

 

Once logged in locally, WHS Console does not start here either and there aren’t any shares displayed in Explorer. In Windows Disk Management all disks are represented anyway, shown as "healthy". As a side note I also have Crash Plan installed since a few weeks, but it has only managed to upload a couple of GB. However, I note that it still uploads from the very share mentioned in the error messages.

 

I reboot WHS and log in locally. On the login screen I get two more error messages about the "HD failed, files permanently lost". I didn’t have presence of mind to write down any error messages / file paths nor take screenshots at any stage.

 

I start WHS Console locally on the server; the Disk Manager Add-on shows all 10 disks, now with the status "Healthy". The disc on which the "remove" began is now approx. half full and disks that were empty before, now have data on them. Seemingly the source disk was during the process emptied by half and these data then distributed to various disks, which then filled up. But then something obviously went wrong and the process interrupted.

 

In Explorer all shares are now shown. I open one of the paths (as I remembered it from the error message) down to leaf node, so that I can see the files, but I don’t open any file up so I don’t know if they were readable. I then decide to shut down the server and have not started it since.

 

Obviously something went wrong during the process, and it appears that it was a destination HD that "failed". According to the ID# I recall from the error message, one of the two IDE hard drives that was recently added to the pool.

 

 

1. Can I find all these different error messages in a log?
   I would like to check the paths to see if the entire folders are missing, or if just certain files are affected.
    
2. Am I correct in my understanding of the fundamentals of WHS Drive Pool – for all Shares with “duplication” active, WHS makes sure all files are in two copies on two different physical drives?
    
3. More specifically regarding a "remove" of a HD from drive pool. Does WHS make sure “duplication” is maintained – i.e. files exist on two physical disks at all times – so that if one of them "fails", data is maintained on the other?
    
4. To cause the least possible amount of damage – what is the smartest to do now?
   
   Alt. 1:
   Before booting up the server again, take out each HD and copy all the data to a backup destination?

​- From what I understand data on any drive from the Pool would be available as regular folders/files that can be read as usual if the drive is plugged into another computer?

- Will mounting the drive in another computer in any way affect/reset it so that WHS, when booting back up again, may perceive it as a drive “not added to the pool”?
- This is of course a cumbersome procedure – not a problem obviously – but is it the safest?

Alt. 2:
Start up the server and let WHS drive pool "do its thing" with balancing / Duplication 

- Perhaps insert a new large drive before?
- Monitor if any HD is indicated as "failing" – if so, do what?

Alt. 3:
Other - what would you have done in this situation?