Hello

We have no errors in hardware testing and no explicit events in Open Manage. Do you have any idea what caused this problem ?

If there are no errors or warnings then look at what happened just before the system shut down. Check the hardware log at the time of the shutdown. The hardware log should state what initiated the shutdown. If there is nothing in the hardware log that states what initiated the shutdown then this is a hardware issue.

Thanks

Hi Daniel

Thank you for your update

When you said "Check the hardware log at the time of the shutdown", how I can verify the hardware log ? By Open Mange Essential, iDRAC,...?

Thank you in advance for your answer

Best Regards

When the reboot occurs, i have only this hardware logs:

OEM software event.

C: boot completed.

So that means there's a hardware problem ?

Best Regards

It is under the log section of the iDRAC. It is called the System Event Log in the iDRAC. It is not the same as the operating system's System Event Log. In OpenManage it is listed as the Hardware Log.

On iDRAC, in Lifecycle Logs i have this event before the reboot

Just an idea but set the power options in the bios to Max proformance

Does it hapen when the when the GPUs/ systems are underloading?

Hi

Today I made the changes in the BIOS configuration because the servers had Watt Performance Optimization Profile Settings (DAPC) as the profile settings.

Now the configuration of each node is in custom mode with maximum performance and disables C1E and C-state options.

I hope that will solve the problem.

Best Regards

When the reboot occurs, i have only this hardware logs: OEM software event. C: boot completed. So that means there's a hardware problem ?

No, those are normal messages that occur during system shutdown and startup. You need to review all of the software and hardware logs and cross-reference them at the time the events occur. Until you find something in the logs or diagnostics to indicate what is happening it is just guess work.

Make sure you turn off automatic recovery on failure in the operating system. If the OS is faulting it automatically restarts the system by default.

Thanks

Assumeing the nodes are windows Vm's? if so, set those  to high power mode in the OS AND host OS, are they on the latest firmware?

Also look at the Idrac power/graph see if the systems are useing too much power.

Same issue with a single T430 / Windows Server 2016. No hardware errors, sometime the server reboots two or three time within 5 minutes, sometimes it is ok for ours. For the moment , i just installed the OS, no users, no activity !!!

DELL support asked for hardware test, no problem, so no other answer.