none
User Profile Disks Same Disk Identifiers RRS feed

  • Question

  • Hello.

    Running a test RDS2019 server with UserProfile disks hosted on an RDSbroker share on 2016. 

    Everything was fine but since the weekend, logging new users onto this server caused a black screen. The system became completely unresponsive. Upon investigation, we're noticing the following errors..

    1) Disk 4 has the same disk identifiers as one or more disks connected to the system. Go to Microsoft's support website (http://support.microsoft.com) and search for KB2983588 to resolve the issue.

    2) An error was detected on device \Device\Harddisk4\DR24 during a paging operation.

    4) The system failed to flush data to the transaction log. Corruption may occur in VolumeId: \\?\Volume{7483659d-0000-0000-0000-100000000000}, DeviceName: \Device\HarddiskVolume27.
    (A device which does not exist was specified.)

    5) The IO operation at logical block address 0x24ea0 for Disk 4 (PDO name: \Device\00000131) was retried.

    For issue 1 i've looked at the KB and checked several other posts, using diskpart to look at the uniqueid, they are all different. 

    2) Related? not sure

    3, 4 and 5 I could do with some advice please.


    Monday, August 19, 2019 9:30 AM

All replies

  • Hi,

     

    Please help to confirm more details for us to understand your issue.

     

    1. Did this issue occur for all the user accounts(admin/non-admin) or only the newly created user account?

    2. Have you tried any methods to recover the server like rebooting or check task manager and kill pending program?

    3. Is there any event logs rather than UPD related errors?

    4. What kind of web browser has been installed on session host server?

    5. When did the black screen occur? Was the user ever managed to see desktop of session host?

    Looking forward to your response.

     

    Thanks,

    Jenny


    Please remember to mark the replies as answers if they help.
    If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.com.

    Tuesday, August 20, 2019 6:57 AM
  • Hi,

    Is there any update? Please feel free to let us know if more assistance needed.

     

    Have a nice day.

     

    Thanks,

    Jenny


    Please remember to mark the replies as answers if they help.
    If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.com.

    Wednesday, August 21, 2019 1:13 AM
  • Hi,

     

    Please help to confirm more details for us to understand your issue.

     

    1. Did this issue occur for all the user accounts(admin/non-admin) or only the newly created user account?

    2. Have you tried any methods to recover the server like rebooting or check task manager and kill pending program?

    3. Is there any event logs rather than UPD related errors?

    4. What kind of web browser has been installed on session host server?

    5. When did the black screen occur? Was the user ever managed to see desktop of session host?

    Looking forward to your response.

     

    Thanks,

    Jenny


    Please remember to mark the replies as answers if they help.
    If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.com.

    Hi Jenny, 

    I backed out of this piece of work as we needed to do some other updates. Back on it now.

    to answer your questions

    1. Admin accounts are effected and user accounts, this is when the profile disk mounts on use login

    2. Server has been rebooted

    3. No - only event 158 - same disk identifier

    4. Chrome 

    5. Occurred between logons. All users affected. Seems ok now though after a restart. 


    Monday, September 9, 2019 8:30 AM
  • Doing some further testing.

    All users logging onto the server, get assigned a profile disk. Every user who logs on generates event 158

    Does anyone have any advice please? It's a 2019 server fully patched.


    Thanks.

    Monday, September 9, 2019 9:08 AM
  • here's some info that may help

    DISKPART> list disk

      Disk ###  Status         Size     Free     Dyn  Gpt
      --------  -------------  -------  -------  ---  ---
      Disk 0    Online          100 GB      0 B
      Disk 1    Online          100 GB  1024 KB        *
      Disk 2    Online         5120 MB      0 B
    * Disk 3    Online         5120 MB      0 B
      Disk 4    Online         5120 MB      0 B
      Disk 5    Online         5120 MB      0 B
      Disk 6    Online         5120 MB      0 B

    DISKPART> select disk 3

    Disk 3 is now the selected disk.

    DISKPART> uniqueid disk

    Disk ID: E5E39FE1

    DISKPART> select disk 4

    Disk 4 is now the selected disk.

    DISKPART> uniqueid disk

    Disk ID: 6DBCAB89

    DISKPART> select disk 5

    Disk 5 is now the selected disk.

    DISKPART> uniqueid disk

    Disk ID: 416D0C7A

    DISKPART> select disk 6

    Disk 6 is now the selected disk.

    DISKPART> uniqueid disk

    Disk ID: 1C6790B6

    DISKPART>

    I get event id 158 for all these disks.

    Monday, September 9, 2019 9:26 AM
  • exact same issue with RDS, we also have blank screens and the event 158

    server gets totally frozen and requires a reboot

    maybe some profile disks are hitting their size limits.  I know they never reclaim white space so really hard to troubleshoot in a large environment.  I think I'm going to increase the size and then delete everyones profile disks so everythings clean

    Tuesday, September 17, 2019 7:41 AM
  • We have the same issue on our Windows Server 2019 RDHS collection. At some point the user "desktop" goes black and we have to reboot the server to fix the issue, which end up that that all user lost their unsave work!

    We take a look at the VHDX for the user and they indead have all the same disk identifier (GPT) event if disk part show a MBR uniqueid who differ.

    Get-VHD -Path \\server\share\*.vhdx | Format-List -property Path,DiskIdentifier

    Path           : \\server\share\UVHD-S-{SID}-1006.vhdx

    DiskIdentifier : F69E260C-D4C3-4C49-83D4-DA035026FB2B
     
    Path           : \\server\share\UVHD-S-{SID}-1018.vhdx
    DiskIdentifier : F69E260C-D4C3-4C49-83D4-DA035026FB2B
     

    Path           : \\server\share\UVHD-S-{SID}-11597.vhdx
    DiskIdentifier : F69E260C-D4C3-4C49-83D4-DA035026FB2B

    John and Glen any solution on your side?

    We will ResetDiskIdentifier tonight on all disk via the following script and let`s you know.

    Set-VHD  \\server\share\UVHD-S-{SID}-11597.vhdx -ResetDiskIdentifier

    Thursday, October 3, 2019 4:36 PM
  • There is another megathread on technet regarding rds blank screens....I'll try and find it if I can

    I'm not sure the disk identifiers are the root cause though as I'm sure we had the same issue on RDS2016 and never had a single blank screen.  Now we are on 2019 with new server hardware, we used to get them multiple times a day.  They have significantly dropped, although we're not sure if we have 100% fixed them.

    So far we have done:

    Increased size of profile disks and given everyone new ones

    Disabled ipv4 and ipv6 offloading on the nics

    Made the microsoft recommended registry key to delete user created firwewall rules at logoff

    Disabled RDS timeouts across the board (everything set to never)

    We have also updated all the servers bios/network card firmware.....I'm going to wait a week to see if we get anymore.  If you find anything, please let me know.

    I did read up a lot about this issue, a lot of people get it when running firefox on rds which crashes the windows audio service.  This does not apply to us, but something to test next time you get one is if you can CTRL_ALT_END and get into task manager, then click run and a command to restart the audio service.  If the blank screens go and the server recovers you know you are on to something.

    Friday, October 4, 2019 7:47 AM
  • I found the other thread https://social.technet.microsoft.com/Forums/windowsserver/en-US/3fdfa58b-fe1b-4546-85d2-d43dac9bcc10/black-screen-on-all-new-connections-sessionhost-has-to-be-rebooted?forum=winserverTS

    Guid reset haven`t change anything, No one have reach the max size of their profile disk and we already have in place the registry key for the firewall.

    We have 4-5 crash this morning, most of our user are falling back to our deprecated 2008R2 RD farm. We just deploy that new infra this week to all user, we had 10-12 users on a single server for two week without issue. 

    I am updating the server to disabled the offloading.

    We are running on VMWare, I am wondering if the display driver could not be a issue there.

    I am completely clueless

    Friday, October 4, 2019 3:51 PM
  • We've just had a server fail with a black screen!!!

    Back to square one

    Tuesday, October 8, 2019 1:45 PM
  • We remove UPD and install FSlogic, we are just starting to put load on the new server, will let’s you know if we seem good or not.

    Tuesday, October 8, 2019 3:47 PM
  • We had another crash as soon we hit 18 users. 

    With event log error : {Delayed Write Failed} Windows was unable to save all the data for the file C:\Users\[UserName]\AppData\Local\Microsoft\Windows\Notifications\wpndatabase.db-shm. The data has been lost. This error may be caused by a failure of your computer hardware or network connection. Please try to save this file elsewhere.

    Our next thought is a network bottleneck, we use VmWare esxi, and the default network card is a 1Gig card (e1000), they have a 10Gig card(VMXNET3), we are updating them on the host and file server tonight as our entire esxi infrastructure run over a 10Gig physical network.

    https://www.lewan.com/blog/choosing-the-right-vmware-nic-should-you-replace-your-e1000-with-the-vmxnet3


    Tuesday, October 8, 2019 7:56 PM
  • FSLogix sounds like a brilliant idea, I'll look into this today.

    Just for info, our RDS SH servers are all physical with 2x10gb intel nics so I'm not sure it's a network issue.

    The fact Microsoft bought out FSLogix and are giving it away free speaks volumes about their own technololgy.  Profile disks is yet another b@stardised technology built for one purpose, yet used for another.  After all these years, they never got profiles or remote desktop working properly.  I read that RDS will ultimately come to an end in the next few years anyway as they move to this new Windows 10 session based stuff.

    Wednesday, October 9, 2019 7:27 AM
  • Yesterday afternoon we had 2 servers fail with black screens.  We have removed profile disks to see if we get anymore.  If we last a week, we know 100% it's those.

    Strangley, we noticed something this morning.  User calls up with a temp profile.  Checked and found a lot of .bak entries in the registry under profilelist.  I've made a small script to delete them from all our RDS servers.

    also, notice on one of our servers it's moved from the network domain profile, to guest/public.....maybe not related but we're checking it out anyway.

    Thursday, October 10, 2019 9:28 AM
  • Servers crashed on Friday, no profile disks in use.  Back to the drawing board.
    Monday, October 14, 2019 1:12 PM
  • On our side, we remove the User Profile Disk and use Folder redirection and FSLogix Office Container for Outlook and Teams.

    For a week now no crash, previously with both UPD and FSLogic full container as soon we hit some load (around 15-18 users) it was crashing with NTFS error in the event log and black screen.

    We are deploying to 50 users tomorrow will see what happens!

    @Glen but I am worries that with any of those technologies you still hit a black screen. 


    Tuesday, October 22, 2019 12:45 PM