Hi
We are also getting this issue on our RDS servers running windows server 2019, after a few days the vm fully locks up and the only fix there on out is to turn it off in hyper v and turn it back on. Disabling Network threat protection fixes that issue, we have had no reoccurances since then
Hi Ben,
I've split your comment into a new forum post as the symptoms appear slightly different in your situation.
When the system locks up, would it be possible for you to create a crash dump? The following article explains how to force a system crash from the keyboard. I suggest opening a case with our support team once you have this log file.- Forcing a System Crash from the Keyboard
After power-on, are there any signs on the system that the issue is worsening, such as increased memory usage on certain components?
After the restart, are there any errors in the Windows Event log, System I guess is most likley, prior to the lockup?
We have been seeing a similar issues on RDS 2016 Virtual Machines
https://www.reddit.com/r/sysadmin/comments/wxehc0/rds_2016_session_hosts_on_different_esxi_hosts/?utm_source=share&utm_medium=android_app&utm_name=androidcss&utm_term=1&utm_content=share_button
Only fix is to hard power off VM.
We notice NTP Service stuck in starting state after boot.
Ben.
Thanks for adding to the discussion!
If you can replicate this issue, I suggest opening a support case so our team can take a closer look.
I suggest using the article linked in my previous comment to force a system crash if possible so that our team can see exactly what is loaded into the system at the time of lock-up.
Qoosh said:- Forcing a System Crash from the Keyboard
If you can also provide an SDU from one of the devices where this issue has occurred previously, this will help us to gather some initial information. Please provide me with the case ID either here or via PM so I can follow up on this.
Same issue here. SDU has been provided. Case ID: 05698133
Multiple SDUs have been provided. Case ID: #05632664
The NTP Service being stuck in a starting state is interesting, I wonder if this is related?
I've seen this with a bug in a Redhat network driver and another NIC driver affecting the registering of WFP filters. So the SophosNTPService.exe process gets stuck in a WFP call.
It is only registering with WFP to enable the WFP filters for web protection/Control. With those features disabled it shouldn't perform this task.
Are you able to test if you get the issue with "Web Control" AND "Scan downloads in progress" AND "Block access to malicious websites", I.e. in the Threat Protection policy applied to the server disabled these 2:
Then in the web control policy applied to the server, if enabled disable this:
Check at the client that SophosNetFilter.exe has stopped. This is evidence of the features being disabled. All 3 need to be disabled for SophosNetFilter.exe to stop running.
Then restart the Sophos Network Threat Protection service. This will ensure all WFP callout registrations are removed.
Does the problem occur from this state? If so, it suggest to me an issue with this feature/WFP.
any updates on this case?
Let me know if the following response in a neighbouring thread helps. - RE: Windows Server 2019: Network Setup Service constantly restarting. disabling Network Threat Protection fixes it
If you don't have a more recent version of Core Agent at this time, it is also possible to open a support case to request that your site be moved to an earlier release group so you receive the update sooner.