locked
Device Not Functioning AND Mystery Disk Not Initialized in Disk Management RRS feed

  • Question

  • I have WHS PP2.  I am getting multiple messages about problems.  I had to rebuild the backups, which lost all my backups of each computer, but at least now I am back to nightly backups again.  I also have numerous Yellow warnings stating "A device attached to the system is not functioning," which is causing file conflicts in virtually all of my folders!  But I have NO issues with any of my hard drives listed--ALL are "HEALTHY".  I have a pretty big system, with 5 internal hard drives, 5 drives in an external SATA box plugged into an eSATA port, and 5 more USB drives.  One of the USB drives is not in the "pool," but is used instead for server backups.  

    Looking at WHS Console, all of the drives show as "Healthy."  I have the Disk Management Add-in installed (installed it after getting these problems) and it shows only one disk "in the red" at 98% full.  Of the other disks, I have 9 greens and 5 yellows.  

    The very strange thing is, I went on the Server itself and entered Computer Management, Disk Management.  It asked me if I wanted to initialize a disk.  Both Skipping and Trying to initialize have resulted in the same result: no change.  Looking at Disk Management, it shows the 15 Disks that I have, all listed as healthy.  Then there's a mysterious "Disk 10" in the middle of the list, showing up as "Unknown, Not Initialized."  Looking at the Properties on this "Disk" it says it is called "Config Disk SCSI Disk Device" and it says it's working properly, although under the "Volumes" tab, it says it is not initialized.  I cannot for the life of me figure out what this is, since all my Hard drives are present and accounted for, but I'm guessing this is the source of my failures.  

    Please any help is greatly appreciated!
    Monday, May 18, 2009 10:20 PM

Answers

  • I looked in the Registry, under Storage Manager\Disks, there are 14 disks listed, out of my 15 attached.  I'm guessing the disk that is not a part of the pool is not included in the list.  Correct me if I'm wrong.  Under Volumes, 15 are listed.  It's not the simplest of exercises to determine which attributes go with which disks, though.  When I have more time, I'll look at it further.  In the meantime, should all 15 be showing up under Disks?  or is that just for the Disks that are specifically assigned to the Pool?  I suppose I could try removing the backup USB drive and see if that makes a difference.  

    Your guess is correct, and it's indeed not simple. Take your time to find the logic, and as long as you don't understand don't mess with it.

    Before changing anything you could export the HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows Home Server key from the registry. If you run into problems you can the simply delete the whole key from the registry, then double click the previously exported key (.reg file) to merge it with the registry again.

    Basically you should:
     
    1. Check for each of the volume guid keys if the mountpoint key is pointing to an existing mountpoint (check in windows explorer)
    2. Check if each of the volume guid keys contains a subkey with the name of a valid disk guid

    Your event logs tell me that your client backup database is possibly corrupted, or more likely that one or more of the backup db files can no longer be found, which could be symptom of your basic problem (WHS thinks one or more disks are missing). If this problem still exists when (you think) the volume and disk keys in the registry are OK you should run chkdsk on each of your partitions:

    ogon to your server desktop, open command shell ("DOS" window), then run the following commands.

    1. chkdsk D: /x /f /r 
    2. for /d %1 in (c:\fs\*) do chkdsk /x /f /r %1 
    3. chkdsk C: /f /r

    System may ask you to run chkdsk at next boot (for C partition), Answer Y. Please note chkdsk may take a long time to run, so you may need to be patient. Results for chkdsk on C can be found in windows application eventlog , source winlogon.

    If chkdsk finds no more errors the second time around watch your eventlog log closely for at least a week or so if any more of these bad block errors appear. If yes, your disk could be about to fail, and it's probably best to replace it.

    If the disk seems to be OK you could try and repair your backup database from the WHS console, Settings Menu, Backup. If that doesn't fix it you are probably facing a server reinstall.


    • Marked as answer by ryrad Sunday, May 31, 2009 9:23 PM
    Wednesday, May 20, 2009 3:22 AM
    Moderator
  • I checked the 2 items in registry...
    1. I validated that the keys in Volume (15 total keys) were each pointing to mounts.  2 of the mounts were C:\ and D:\  the rest were C:\n\ where n was a letter or number.
    2. Checking the Subkeys, I found that 2 of them both pointed to the same disk (There are only 14 Disk Keys listed in the registry).  The Disk that had 2 volumes pointed to it was labeled as "System" under Friendly Name and the Volumes were the ones listed as C: and D:  So I'm assuming this is ok, right?
    1. That's OK
    2. That's what you expect

    If these check out, I'll proceed to the Chkdsk commands.  A question on those.  Do I literally type exactly what you have written (Aside from the 1., 2., 3.)?  Will I be waiting for hours between each of the 3 commands?  After #1 runs, if it finds/repairs problems, should I just stop there and reboot and see if the File Conflicts go away?  Or should I do all 3 commands regardless of the output?


    3. Yes you should type the commands literally (or copy and paste)
    4. It may indeed take a long time to for each of the chkdsks to finish, depends on the size of the disks, amount of data and the problems encountered
    5. I would run chkdsk on all disks regardless of the output
    6. Additionally you could issue the following two commands before running chkdsk to temporarily stop back backups from running:
       a. net stop pdl
       b. net stop whsbackup

    Incidentally, I still have file conflicts that have not gone away, but I have not had any more backup failures.  I have tried a few of the files that are listed in the conflicts and I have access to the ones I've checked. 

    7. Can you provide more detail on the file conflict errors?
    8. Do you have access to files through the WHS network shares folder (\\yourservername\shares), or do you access them in another way?
    • Marked as answer by ryrad Sunday, May 31, 2009 9:24 PM
    Monday, May 25, 2009 9:07 AM
    Moderator

All replies

  • I have WHS PP2.  I am getting multiple messages about problems.  I had to rebuild the backups, which lost all my backups of each computer, but at least now I am back to nightly backups again.  I also have numerous Yellow warnings stating "A device attached to the system is not functioning," which is causing file conflicts in virtually all of my folders!  But I have NO issues with any of my hard drives listed--ALL are "HEALTHY".  I have a pretty big system, with 5 internal hard drives, 5 drives in an external SATA box plugged into an eSATA port, and 5 more USB drives.  One of the USB drives is not in the "pool," but is used instead for server backups.  

    Looking at WHS Console, all of the drives show as "Healthy."  I have the Disk Management Add-in installed (installed it after getting these problems) and it shows only one disk "in the red" at 98% full.  Of the other disks, I have 9 greens and 5 yellows.  

    The very strange thing is, I went on the Server itself and entered Computer Management, Disk Management.  It asked me if I wanted to initialize a disk.  Both Skipping and Trying to initialize have resulted in the same result: no change.  Looking at Disk Management, it shows the 15 Disks that I have, all listed as healthy.  Then there's a mysterious "Disk 10" in the middle of the list, showing up as "Unknown, Not Initialized."  Looking at the Properties on this "Disk" it says it is called "Config Disk SCSI Disk Device" and it says it's working properly, although under the "Volumes" tab, it says it is not initialized.  I cannot for the life of me figure out what this is, since all my Hard drives are present and accounted for, but I'm guessing this is the source of my failures.  

    Please any help is greatly appreciated!

    Did you try rebooting the server?  Do you have any extra devices plugged in (USB flash drive) that might be "freaking it out"?  Did you install any add-ins/apps around the time the problem started?
    Monday, May 18, 2009 11:25 PM
    Moderator
  • Thanks for the reply.  I have rebooted multiple times.  It all may have precipitated after a power outtage that shut down the server a couple weeks ago, since before then things were fine.  There had been no software/hardware changes in between the outtage and the ocurrence of problems.  As I was looking for a solution, I did notice in Device Manager that 2 devices had "?" and I reinstalled them--one was the Real Audio driver and the other was the uGuru chip for the ABIT IP35 PRO motherboard that the WHS is on--doing this took the ? away, but didn't cure my problems.  Otherwise, all other devices are recognized.  I also tried running Update and it said everything was up to date.
    Tuesday, May 19, 2009 1:00 AM
  • I have WHS PP2.  I am getting multiple messages about problems.  I had to rebuild the backups, which lost all my backups of each computer, but at least now I am back to nightly backups again.  I also have numerous Yellow warnings stating "A device attached to the system is not functioning," which is causing file conflicts in virtually all of my folders!  But I have NO issues with any of my hard drives listed--ALL are "HEALTHY".  I have a pretty big system, with 5 internal hard drives, 5 drives in an external SATA box plugged into an eSATA port, and 5 more USB drives.  One of the USB drives is not in the "pool," but is used instead for server backups.  

    Looking at WHS Console, all of the drives show as "Healthy."  I have the Disk Management Add-in installed (installed it after getting these problems) and it shows only one disk "in the red" at 98% full.  Of the other disks, I have 9 greens and 5 yellows.  

    The very strange thing is, I went on the Server itself and entered Computer Management, Disk Management.  It asked me if I wanted to initialize a disk.  Both Skipping and Trying to initialize have resulted in the same result: no change.  Looking at Disk Management, it shows the 15 Disks that I have, all listed as healthy.  Then there's a mysterious "Disk 10" in the middle of the list, showing up as "Unknown, Not Initialized."  Looking at the Properties on this "Disk" it says it is called "Config Disk SCSI Disk Device" and it says it's working properly, although under the "Volumes" tab, it says it is not initialized.  I cannot for the life of me figure out what this is, since all my Hard drives are present and accounted for, but I'm guessing this is the source of my failures.  

    Please any help is greatly appreciated!

    Probably the ID of one the hard disks has changed after to the power outage, so even though all of the current disks appear as healthy WHS might still think one disk is missing. Possibly you could fix his by editing the appropriate entries in the WHS registry, however considering the number and variety of disks you have attached this may be very troublesome. I would therefore advise you to do a server reinstall. This should preserve your data and backups, however you will need to reconfigure users accounts and right, re-connect clients, re-install add-in's and all other modifications you have made after your last install. Please note if you need to provide storage drivers for the system to properly recognize all of your disk you will need to this twice, once during the initial phase of installation, then again at the beginning of text mode install (F6, provide drivers on diskette). Server reinstall is described in more detail in this FAQ.
     
    The config Disk SCSI device is ery likely caused be the eSATA enclosure and should be ignored.
    Tuesday, May 19, 2009 7:41 AM
    Moderator
  • I'm familiar with the Registry, having done a lot of tweaking over the years on MCE/VMC.  If you point me toward the right keys to look at, I could try that.  If not, I'll try the reinstall.
    what if I dismounted the drives in the eSata enclosure and then unplugged it?  Maybe WHS is confused by the Config Disk SCSI Device?
    Tuesday, May 19, 2009 11:30 AM
  • Relevant keys are

    1. HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows Home Server\Storage Manager\Disks 
    2. HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows Home Server\Storage Manager\Volumes
    3. Also search fo other occurences of the disk guid's  in the registry to make sure you don't miss an essential entry.

    There's a few threads about changing disk_id's in registry on this forum, perhaps you need to these first. Currently I don't have time to give you detailed guidance, also got no eSATA or USB disks attached to my WHS boxes, so can't really test

    I don't think Config Disk SCSI Device has anything to do with it, just Google perhaps together with something like "Silicon" or "Sil" or "RAID" to find out more.

    [EDIT] dismounting / mounting disk shouldn't harm , although AFAIK this comparable to reboot.

    Tuesday, May 19, 2009 11:43 AM
    Moderator
  • Did you check the event log on the server already for related errors?
    A problem with external drives was, that they did tend to jump out of the storage pool sometimes, but this would also be told as a health notification about a missing drive. Also with a larger amount of attached devices the USB port may become a problem.
    Do you have an USB thumbdrive attached to the server?
    Best greetings from Germany
    Olaf
    Tuesday, May 19, 2009 12:35 PM
    Moderator
  • I do Not have any thumb drives attached.  I have nothing attached other than the 5 USB drives, the one eSata cage, a monitor, keyboard and mouse.

    Also, checking the log, the errors did Not start immediately after the power outtage.  It was a few days after the power outtage.  The errors started 5 seconds after a backup of one of my PCs was kicked off.  Below is the last good information event and then the first few errors.  It MAY have been related to running out of space, as I had everything set on Duplicate and was down to about 900 GB of free space left on a 10 TB system.  I have since turned off duplicate on one of my folders, and now have 25% free.  But oddly, one disk stays at 98% filled (a 750GB USB drive), with everything else at or below 82%. 

    Event Type: Information
    Event Source: HomeServer
    Event Category: Backup
    Event ID: 260
    Date: 5/16/2009
    Time: 12:29:24 AM
    User: N/A
    Computer: SERVER
    Description:
    Backup of MOMKITCHEN C:\ has started.

    For more information, see Help and Support Center at http://go.microsoft.com/fwlink/events.asp.


    Event Type: Error
    Event Source: HomeServer
    Event Category: Backup
    Event ID: 272
    Date: 5/16/2009
    Time: 12:29:29 AM
    User: N/A
    Computer: SERVER
    Description:
    Unexpected error 0x570 from CreateFile on D:\folders\{00008086-058D-4C89-AB57-A7F909A47AB4}\MOM.C.FileRecordHash.4096.dat: The file or directory is corrupted and unreadable.

    For more information, see Help and Support Center at http://go.microsoft.com/fwlink/events.asp.

    Event Type: Error
    Event Source: HomeServer
    Event Category: Backup
    Event ID: 267
    Date: 5/16/2009
    Time: 12:29:29 AM
    User: N/A
    Computer: SERVER
    Description:
    Client Backup server failed at d:\wssg_src\whs_pp2\qhs\src\backup\util\diskfile.cpp(145)

    For more information, see Help and Support Center at http://go.microsoft.com/fwlink/events.asp.


    Event Type: Error
    Event Source: HomeServer
    Event Category: Backup
    Event ID: 272
    Date: 5/16/2009
    Time: 12:30:07 AM
    User: N/A
    Computer: SERVER
    Description:
    Unexpected error 0x570 from CreateFile on D:\folders\{00008086-058D-4C89-AB57-A7F909A47AB4}\{97A8A80B-5917-4D14-95CF-84DD133398BF}.Machine.configdat: The file or directory is corrupted and unreadable.

    For more information, see Help and Support Center at http://go.microsoft.com/fwlink/events.asp.

    Event Type: Error
    Event Source: HomeServer
    Event Category: Backup
    Event ID: 267
    Date: 5/16/2009
    Time: 12:30:07 AM
    User: N/A
    Computer: SERVER
    Description:
    Client Backup server failed at d:\wssg_src\whs_pp2\qhs\src\backup\util\diskfile.cpp(145)

    For more information, see Help and Support Center at http://go.microsoft.com/fwlink/events.asp.

    Tuesday, May 19, 2009 2:07 PM
  • Just experimenting, I powered down the WHS, turned off the eSATA enclosure and restarted the WHS.  Within a few seconds of logging in, I got the BSOD.  Turned it off, reconnected eSATA again, and am back where I was before, with same errors about device not functioning.  
    Tuesday, May 19, 2009 2:15 PM
  • I looked in the Registry, under Storage Manager\Disks, there are 14 disks listed, out of my 15 attached.  I'm guessing the disk that is not a part of the pool is not included in the list.  Correct me if I'm wrong.  Under Volumes, 15 are listed.  It's not the simplest of exercises to determine which attributes go with which disks, though.  When I have more time, I'll look at it further.  In the meantime, should all 15 be showing up under Disks?  or is that just for the Disks that are specifically assigned to the Pool?  I suppose I could try removing the backup USB drive and see if that makes a difference.  
    Tuesday, May 19, 2009 3:38 PM
  • The registry key you mention contains only the disks, which are integrated in the pool (at least on my server).
    Best greetings from Germany
    Olaf
    Tuesday, May 19, 2009 7:14 PM
    Moderator
  • I see that there is a download for WHS restore Disk.  Should I use that to reinstall my system, versus the original Intall Disk?
    Wednesday, May 20, 2009 1:45 AM
  • I'm guessing that WHS restore disk is the client restore disk, not the actual WHS install disk.
    Use your original install disk to do a server reinstallation
    Wednesday, May 20, 2009 2:14 AM
  • I looked in the Registry, under Storage Manager\Disks, there are 14 disks listed, out of my 15 attached.  I'm guessing the disk that is not a part of the pool is not included in the list.  Correct me if I'm wrong.  Under Volumes, 15 are listed.  It's not the simplest of exercises to determine which attributes go with which disks, though.  When I have more time, I'll look at it further.  In the meantime, should all 15 be showing up under Disks?  or is that just for the Disks that are specifically assigned to the Pool?  I suppose I could try removing the backup USB drive and see if that makes a difference.  

    Your guess is correct, and it's indeed not simple. Take your time to find the logic, and as long as you don't understand don't mess with it.

    Before changing anything you could export the HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows Home Server key from the registry. If you run into problems you can the simply delete the whole key from the registry, then double click the previously exported key (.reg file) to merge it with the registry again.

    Basically you should:
     
    1. Check for each of the volume guid keys if the mountpoint key is pointing to an existing mountpoint (check in windows explorer)
    2. Check if each of the volume guid keys contains a subkey with the name of a valid disk guid

    Your event logs tell me that your client backup database is possibly corrupted, or more likely that one or more of the backup db files can no longer be found, which could be symptom of your basic problem (WHS thinks one or more disks are missing). If this problem still exists when (you think) the volume and disk keys in the registry are OK you should run chkdsk on each of your partitions:

    ogon to your server desktop, open command shell ("DOS" window), then run the following commands.

    1. chkdsk D: /x /f /r 
    2. for /d %1 in (c:\fs\*) do chkdsk /x /f /r %1 
    3. chkdsk C: /f /r

    System may ask you to run chkdsk at next boot (for C partition), Answer Y. Please note chkdsk may take a long time to run, so you may need to be patient. Results for chkdsk on C can be found in windows application eventlog , source winlogon.

    If chkdsk finds no more errors the second time around watch your eventlog log closely for at least a week or so if any more of these bad block errors appear. If yes, your disk could be about to fail, and it's probably best to replace it.

    If the disk seems to be OK you could try and repair your backup database from the WHS console, Settings Menu, Backup. If that doesn't fix it you are probably facing a server reinstall.


    • Marked as answer by ryrad Sunday, May 31, 2009 9:23 PM
    Wednesday, May 20, 2009 3:22 AM
    Moderator
  • Thanks.  I am getting a "file corrupt on \DE, please run chkdsk", but I did not know how to do that on the disks.  I will give your method a shot.  Hopefully I dont do any damage to my data.  
    Wednesday, May 20, 2009 11:13 AM
  • If there is damage, it is already done.
    Best greetings from Germany
    Olaf
    Wednesday, May 20, 2009 1:28 PM
    Moderator
  • I checked the 2 items in registry...
    1. I validated that the keys in Volume (15 total keys) were each pointing to mounts.  2 of the mounts were C:\ and D:\  the rest were C:\n\ where n was a letter or number.
    2. Checking the Subkeys, I found that 2 of them both pointed to the same disk (There are only 14 Disk Keys listed in the registry).  The Disk that had 2 volumes pointed to it was labeled as "System" under Friendly Name and the Volumes were the ones listed as C: and D:  So I'm assuming this is ok, right?

    If these check out, I'll proceed to the Chkdsk commands.  A question on those.  Do I literally type exactly what you have written (Aside from the 1., 2., 3.)?  Will I be waiting for hours between each of the 3 commands?  After #1 runs, if it finds/repairs problems, should I just stop there and reboot and see if the File Conflicts go away?  Or should I do all 3 commands regardless of the output?

    Incidentally, I still have file conflicts that have not gone away, but I have not had any more backup failures.  I have tried a few of the files that are listed in the conflicts and I have access to the ones I've checked. 

    Thanks!
    Thursday, May 21, 2009 2:27 AM
  • I checked the 2 items in registry...
    1. I validated that the keys in Volume (15 total keys) were each pointing to mounts.  2 of the mounts were C:\ and D:\  the rest were C:\n\ where n was a letter or number.
    2. Checking the Subkeys, I found that 2 of them both pointed to the same disk (There are only 14 Disk Keys listed in the registry).  The Disk that had 2 volumes pointed to it was labeled as "System" under Friendly Name and the Volumes were the ones listed as C: and D:  So I'm assuming this is ok, right?
    1. That's OK
    2. That's what you expect

    If these check out, I'll proceed to the Chkdsk commands.  A question on those.  Do I literally type exactly what you have written (Aside from the 1., 2., 3.)?  Will I be waiting for hours between each of the 3 commands?  After #1 runs, if it finds/repairs problems, should I just stop there and reboot and see if the File Conflicts go away?  Or should I do all 3 commands regardless of the output?


    3. Yes you should type the commands literally (or copy and paste)
    4. It may indeed take a long time to for each of the chkdsks to finish, depends on the size of the disks, amount of data and the problems encountered
    5. I would run chkdsk on all disks regardless of the output
    6. Additionally you could issue the following two commands before running chkdsk to temporarily stop back backups from running:
       a. net stop pdl
       b. net stop whsbackup

    Incidentally, I still have file conflicts that have not gone away, but I have not had any more backup failures.  I have tried a few of the files that are listed in the conflicts and I have access to the ones I've checked. 

    7. Can you provide more detail on the file conflict errors?
    8. Do you have access to files through the WHS network shares folder (\\yourservername\shares), or do you access them in another way?
    • Marked as answer by ryrad Sunday, May 31, 2009 9:24 PM
    Monday, May 25, 2009 9:07 AM
    Moderator
  • Running the chkdsk steps allowed me to avoid having to reinstall WHS!  Thanks!  And I dont appear to have lost anything, despite the warning messages.  
    Step 1 took a while.  Step 2 ran for like 2 days straight (during which, I got error messages for the backup system not being able to run, so I probably could have benefitted from turning that off first, but doesnt seem like any lasting effects now.)  After rebooting after Step 2, the problems disappeared.  I ran step 3 anyway, and it finished very quickly.  
    Still don't know if there is a bad disk in the group, as they all indicate healthy.  I had been running close to capacity, with Duplication enabled on all folders.  I will not run at or near full capacity anymore.  
    That odd uninitialized disk still shows up, so it must be the eSata controller and was not related to my problems.  
    Thanks again for everyone's help.   
    Sunday, May 31, 2009 9:23 PM