Windows Server 2019 sometimes crashes with the icinga2 agent

Icinga 2
,
1

Hi,

I have some virtual Windows Server 2016/2019 monitored by Icinga2 Windows Agent.

Sometimes Windows Server crashes with a blue screen. On a single server it happens once a month, more rarely several times.

I’m collecting memory dumps (MEMORY.DMP) and the common error is IRQL_NOT_LESS_OR_EQUAL (a)
but PROCESS_NAME is always an icinga process:

  • check_users.exe
  • check_load.exe
  • check_uptime.exe
  • check_procs.exe
  • check_service.exe
    or a script powershell
  • powershell.exe

The ram on the hypervisor has no errors.

My client

  • VM Windows 2019 run on varius phisical machine.
  • Windows 2019 all with the latest patch
  • Icinga2 Agent 2.13.7
  • No NSClient++
  • Several plugin custom via powershell.

I already reduced the frequency of checking and added the icinga2 bin folder exclusion to the antivirus real-time scan.

Note: With a vm Windows 2012 vm server I have no problems.

Anyone else having similar problems with windows server monitors?
Any suggestions on where to investigate?

Thank you
Michele

2

Obtaining a memory dump would be very helpful: Collecting User-Mode Dumps - Win32 apps | Microsoft Learn

These can then be analyzed with Visual Studio or WinDbg. Visual Studio has the less annoying interface, but I also had dumps where I could only get stack traces with WinDbg, so it might be worth trying both.

3

Please try to update your icinga client on the windows systems.

The icinga team has some fixes like: Retry file rename operations on Windows for some errors by julianbrost · Pull Request #8691 · Icinga/icinga2 · GitHub