Windows 2019 RDS sessionhost Filesystem error after some time

[PeterPodvodnik]

Dear community, 

We have a cliënt using an RDS farm with 6 sessionshosts (all virtual machines, running on Hyper-V with several more (application and service) VM's without any problem. 

After appr. 2 days of running the filesystem reports errors and a repair is needed. This eventually happens on all sessionhosts.  

When executing chkdsk or repair-volume (-scan), which by the way is working fine and reports no problems until ot goes wrong, we see the following message: 

A snapshot error occured while scanning this drive. Run an offline scan and fix. 

Followed by: 

Repair-volume : A snapshot error occurred while scanning this drive. Run an offline scan and fix.

Activity ID: {ffc7a4d1-d448-0005-c9b1-c7ff48d4d601}

At C:\Windows\Logs\checkfs.ps1:14 char:1

+ Repair-volume -driveletter C -scan -verbose  >> $Logfile

+ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

    + CategoryInfo          : NotSpecified: (StorageWMI:ROOT/Microsoft/...age/MSFT_Volume) [Repai 

   r-Volume], CimException

    + FullyQualifiedErrorId : StorageWMI 43012,Repair-Volume  

The only way to fix is to execute: 

vssadmin add shadowstorage /for=c: /on=c: /MaxSize=20% 

Then we are able to scan and spotfix. The filesystem is damaged. Rebooting does not work, we have to atach the disk to another server to do a repair. When fixed, again after about 2 days same issue.. 

Done so far: 

- Checked all Hypervisors for scheduled tasks etc. Nothing 

- Spreaded the RDS session hoste over physical servers, same result 

- Created a blank (new, fresh install) RDS session host, did a scan after each installed application, no problem. After 2 days, same issue Filesystem defective.. 

- Shadow copy is disabled 

Looking in the System eventlog we see severall messages of (automatic maintenance scheduling??) NTFS checks all with "no error" no action needed. Then suddenly 

 volume C:\ (device\harddiskvolume4) needs an online scan. 

At this moment we have no clue where to go. Any suggestions would be highly appreciated! Especially on the error message regarding the snapshot (!?) Apart from the defective filesystem we cannot explain how and why this error message appears.  

Thanks! 

Added information, in one occasion we noticed a message: 

"Low on registry space: "the system has reached the maximum size allowed for the system part of the registry"

[#2]

@PeterPodvodnik 

This is weird issue and need to be investigate and look into log files and reproduce this problem which is NOT possible through forum. I would recommend you either contact Microsoft Support (in case this is an urgent issue) or in case you have access to Windows 10 PC, copy log files and move it to Windows 10 PC and then open start and search for feedback and open Feedback Hub app and report this issue and under operating system select Windows Server (Feedback Hub is not available in Windows Server) so it would be investigated.

Are you running SSD or HDD on your physical disk?

[#3]

@Reza_Ameri-Archived A supportcase has been made. The RDS host are running as VM's and the datastores on a HPE MSA with various disktypes. This environment run's severall VM's without problems, only RDS has issues. I posted ther question to find out if anyone ran in to this issue previously. 

Kind regards, 

[#4]

@PeterPodvodnik did you find the root cause? how did you resolve the issue?

[#5]

@behrooz amiri No we did not. Even with extensive contact from support, no resolution. We are no longer in contact with this site so I don't know the current status.