I'm giving a hand to a partner to evaluate Winflector with their application (we are still running the 2-user free version).

By now we are running it Windows Server 2022 on Azure.

From time to time, the wfserver.exe simply exits. No relevant entry in the Event Logs.

In tapplog it just shows the following:

[APP   v3990a|W32] 2022.02.07 00:51:03:  4c0: 01:SaoJoaoItaperiu:Error   18: Cannot read on TCPIP connection. Loc [ClsConnection::ReadNextInCmdBlock] Info [Connection closed by peer,10,10054] SysErr [0] Mod [C:FibraBombeiros_SaoJoaoDoItaperiuMenuAcessos.Exe]

(That is, seems like a connection drop but nothing relevant).

I've deployed Microsoft's Debug Tools "Monitoring Silent Process Exit" (https://docs.microsoft.com/en-us/windows-hardware/drivers/debugger/registry-entries-for-silent-process-exit) to see if it was another process killing the Winflector server.

To my surprise, last time the wfserver.exe process was down, this was the event logged:

The process 'C:Winflectorserverwfserver.exe' exited with exit code code 0. The creation time for the exiting process was 0x01d81b95a01d8b16.

So it seems the exit was spontaneous.

So my questions:

- I know Windows Server 2022 is not (yet) officialy supported. Any known issues regarding this OS?

- Is there any new limitation to the free 2-user version? Some time limit what the server will run?

- Any other known issue that I must be aware of?

Hi Mirek,

Unfortunately I don't have any other relevant details.

In summary:

- Winflector version: 3.9.9.0a (free 2-user version)

- OS Version: Windows Server 2022 Datacenter Azure Edition

- Winflector exits itself with exit code 0 (zero). No relevant error is logged.

- As I stated before, a dropped connection is logged, but I think this may be because Winflector is cancelling the thread before exiting? (If I force kill the wfserver.exe process while a connection is active, the same message is logged before it shuts down).

- I couldn't identify a recurrency pattern, but it happend +/- once each 2 days. Never twice in the same day. Last time (when I opened the ticket) Winflector lasted 3 days up.

- File mapping (WfMount) and printing are enabled.

- No server shutdowns / Windows Update disabled.

Now I downgraded the environment to Windows Server 2016 (Datacenter) - same OS version I run in another cluster with a paid license that seems to run fine. No drops so far (since Monday 2022-02-07). In the next days I'll drop updates here.

E-mail sent as requested.

Q: "Did you have any antivirus installed on the server?"

A: No, just the usual built-in Windows Defender with standard config.

Hi Mirek,

> It looks like the problem is in Winflector auto-start.

Sorry - I have forgotten this. Actually what I did was changing the built-in "on logon" scheduled task trigger and I've added also a cron trigger, so it could forcibly start a new instance in case the main one dropped. A did that after several wfserver.exe shutdowns, to soften the consequences (specially during weekends). So the logs I've sent you must contain lots of false positives.

(Un)Fortunately, since we downgraded the OS to Windows Server 2016 (same OS version we use a paid version), the process termination issue is gone. It's been 7 days now running straight.

> Maybe you run Winflector during RDP session? It is closed automatically by the system when you log out from RDP session.

That was my first thought as well, but I've looked close at the server and this was not the issue. I also configured the machine's Group Policy and took off the logoff actions (from start menu / Ctrl+Alt+Del menu).

> I can't send you any e-mails, because they are blocked by your spam filter

I've whitelisted the @winflector.com domain. It has blacklisted your e-mail because it contained no Subject.