none
"harddrive4 has a bad block" -- and which harddrive is that? RRS feed

  • Question

  • I'm getting freezing problems on my WHS since yesterday. I checked the event logs which say:
    "The device, \Device\Harddisk4, has a bad block."

    How do you determine which disk is which?
    Sunday, February 14, 2010 5:25 PM

Answers

  • Try running chkdsk on all the drives in your server .
    I'm not on the WHS team, I just post a lot. :)
    Sunday, February 14, 2010 6:09 PM
    Moderator
  • The link I provided will run chkdsk on all but your system drive simultaneously. On your system drive it will check the two partitions individually. Yes, chkdsk is slower, but seatools won't always fix the errors Windows detects (and vice versa). As for identifying a drive, I assume you have a self-built server? If so, probably Harddisk4 will be the drive attached to the 4th SATA port, or a drive attached to the secondary IDE cable. YOu may be able to use Device Manager on your server to identify the disk.

    It's safe to simply disconnect a drive, then boot your server, if a drive is showing up as unhealthy or missing in the console, by the way. In a self-built server, sometimes that's the only way to identify a particular drive with an issue...
    I'm not on the WHS team, I just post a lot. :)
    Sunday, February 14, 2010 6:52 PM
    Moderator

All replies

  • Try running chkdsk on all the drives in your server .
    I'm not on the WHS team, I just post a lot. :)
    Sunday, February 14, 2010 6:09 PM
    Moderator
  • I did try chkdsk but it runs one at a time and rather slowly, so I'm curious if there's a way to know which disk it refers to.

    SeaTools for Windows seems to work ok for checking all the drives at once for errors.
    Sunday, February 14, 2010 6:43 PM
  • The link I provided will run chkdsk on all but your system drive simultaneously. On your system drive it will check the two partitions individually. Yes, chkdsk is slower, but seatools won't always fix the errors Windows detects (and vice versa). As for identifying a drive, I assume you have a self-built server? If so, probably Harddisk4 will be the drive attached to the 4th SATA port, or a drive attached to the secondary IDE cable. YOu may be able to use Device Manager on your server to identify the disk.

    It's safe to simply disconnect a drive, then boot your server, if a drive is showing up as unhealthy or missing in the console, by the way. In a self-built server, sometimes that's the only way to identify a particular drive with an issue...
    I'm not on the WHS team, I just post a lot. :)
    Sunday, February 14, 2010 6:52 PM
    Moderator
  • That was the script I used. By simultaneously I assume you mean all disks will be checked eventually but not in parallel? I got a command window with chkdsk running for one of my disks, and I didn't see any other such windows. Chkdsk was still checking the same disk by the time I got some results in Seatools on the other disks. I have an Acer H340 server. I think harddrive4 refers to a drive in my USB enclosure. Seatools revealed one of the disks as producing errors, but it is the same model as another drive, and I can't get the serial number through the USB interface. If Disk Management uses the same numbering scheme as Event Viewer, which it might be, then it should be straight forward.

    Sunday, February 14, 2010 9:26 PM
  • The script has to run C: and D: alone because they're on the system disk, and require even more special handling than the secondary disks. Then it runs chkdsk on all your secondaries at once.
    I'm not on the WHS team, I just post a lot. :)
    Monday, February 15, 2010 12:32 AM
    Moderator
  • Thanks. I ran chkdsk again but it's been running for D:, partition two of the 1TB system drive, for the last 8 hours. It's at stage 4/5, 49% and I don't see progress at the moment. Does this sound normal?

    Update: the chkdsk file counter went up eventually so I guess it's just doing it very slowly. Still at 4/5 49%, but now file processed is 286,634, vs 171,634 about five hours ago.
    Monday, February 15, 2010 5:40 AM
  • I was having the same issue as you are experiencing now but in a different disk. See this thread for detail. SInce 1/17, I had fewer failure notice in system log. But it is still happening. I have run chkdsk and no error was reported.
    Tuesday, February 16, 2010 3:00 AM
  • Tuesday, February 16, 2010 3:14 AM
  • Strange, I manually removed the disk that got errors in Seatools, put it in a different PC to check, but now Seatools no longer reports any error on the drive. This is similar to the situation encountered in the thread above.

    Dcdyd: is your drive a few years old? Mine is. Perhaps our drives are dying but so far the bad blocks are being detected and skipped over so they no longer show up?

    Tuesday, February 16, 2010 10:59 AM
  • Strange, I manually removed the disk that got errors in Seatools, put it in a different PC to check, but now Seatools no longer reports any error on the drive. This is similar to the situation encountered in the thread above.

    Dcdyd: is your drive a few years old? Mine is. Perhaps our drives are dying but so far the bad blocks are being detected and skipped over so they no longer show up?

    In my case, the system always reports \device\harddevice1 even after I removed hard disk which was reported by Disk Management as \device\harddisk1. I removed three HDDs nad the result was the same. I ran chkdsk on different system and nothing was found. I also ran chkdsk on all disks in WHS system and nothing was found. I am starting wondering that this is due to something other than the hard disk.

    My HDDs are two years from being manufactured. I haven't run detailed analysis on any disk. May be that is the nest step I need to take.
    Tuesday, February 16, 2010 2:16 PM
  • The hard drive I assumed to be failing has now failed, despite passing Seatools' long generic test the second time around.

    • Proposed as answer by Dcdyd Friday, February 26, 2010 2:50 AM
    Monday, February 22, 2010 1:58 AM
  • Glad that you sort it out.
    Tuesday, February 23, 2010 3:03 AM