I may be posting this in the wrong place, if so, please direct me to the proper forum.
I have an ESXi server with four guest machines. Two of the four shut down on Monday or Tuesday, sometime between 5 AM and 10 AM. The guest machines are running Windows server. The logs say:
"The kernel power manager has initiated a shutdown transition.
Shutdown Reason: Kernel API"
I suspect that ESXi is shutting the servers down, but not sure why or how. Would love some help in trying to root this out.
Please advise.
are you sure that windows machines have been activated currectly?
There are a lot of information gaps here. Is this a standalone ESXi host or is it managed by vCenter. Do you have the option of moving the VMs to another host to see if the problem follows the VM(s)?
It would help to know which version of ESXi you're running? What version(s) of Windows are the VMs affected? and what OS are the VMs that are unaffected?
As asked above by @miladmeh8, is Windows properly activated, as an expired license can cause such behaviour. Its a good question and it needs confirming the license status shows as "activated"
HTH
Yea, if the windows is downloaded from microsoft website directly for a test or trial version without activating that defenately after some hours of working it will be powered off.
Hi,
do you have running HA and DRS on your system and do you have enough resources on the host to run in it?
When do you have enabled HA and DRS and you don't have enough resources on your hosts then the system will disable (shutdown) some hosts for that isn't enough resources. I had the same problem.
Look at it.
Hello @cz1138 ,
VMs Powering OFF automatically can be caused by many reasons.
1. Windows is not Liscenced.
2. We have allocated more CPU and Memory to VMs and Esxi host is unable to satisfy the allocation.
3. The HA condition is not met hence it is Powering OFF the VMs.
Please select the Cluster on which you have the hosts and the VMs and then select Monitor - Tasks and see what is the task or event that is triggerred while Powering OFF the VM.
Share that Task/Event and we will have the answer.
Thank you for the input.
Looking at the UI Host > Monitor > Events, I only see entries for today. "Tasks" are the same. Doesn't look like there's a way to look at previous dates.
I'm a bit new to this side of VMWare, I'm not sure what "HA" and "DRS" mean in this context.
Someone else mentioned the possibility of a resource overrun creating an issue. I do wonder about that.
There are two processors on the host machine. They are Intel Xeon Silver 4309Y CPUs with 2.80GHz clocks. They have 8 cores each.
The first server that shuts down appears to have the following configuration:
For CPU the first value is 16, Cores per Socket is 8. So this VM was configured for a system that has 16 CPUs with 8 cores per socket, correct? Looking at the second VM I see the same configuration.
I reckon these may be our smoking gun.
Hello @cz1138 ,
The HA and DRS are resource allocation tools used to protect such events. For this issue we need to get on zoom session for 20 mins and need to check the config so the cause of the issue can be identified.
You have only one host cluster?
HA - high availability
enable that if one host is down that other will take its jobs
DRS - Distributed Resource Scheduler
distribute resources across all host that all host are equal usability or how you have it set in the Cluster -Configure - Services settings.
I thing I have find your problem. You have the added to the VM's more resources as you have and your additional settings of the host (ESXi or vCenter) then shut down to VM's while the host resources are overloaded.
You need to follow some rules by configuration: