EVE-NG Professional - bare metal - disabled perfctr wrmsr

[juror897]

I have a new deployed bare metal instance of EVE-NG professional:

Specifications: Intel(R) Xeon(R) CPU E3-1270 v5 @ 3.60GHz - 64 Gb RAM

I started a lab with one qemu node and connected it to the Cloud0 bridge interface.

As soon as I start (or stop) the image, I receive the vcpu0 error "disabled perfctr wrmsr"

The management IP for EVE-NG becomes unresponsive for just about 5 minutes and then recovers.

I have seen another post where Uldis has stated the CPU cannot keep up with the lab nodes. Does getting this error mean that my CPU is not built to handle this type of virtualization?

I know that I can ignore the errors as Uldis has listed in the FAQ page but having the EVE server become unresponsive for 5 minutes each time I start or stop an instance makes labbing difficult.

As a side note, prior to deploying EVE-NG as bare metal, it was running on this same server over ESXi and I did not experience these interruptions.

[Uldis (UD)]

but you must check must BIOS virtualization settings if it is ON!
https://www.intel.com/content/www/us/en ... tions.html

EPT virtualization is is supported see above

[juror897]

I have looked through the BIOS options and see that virtualization is enabled. This is the only mention of virtualization I could find in the BIOS.

2024-02-20_07-33-06_BIOS.png

[rusty725]

you're most likely overloading your Eve 4 cores 4 threads. If you want me to check come to our chat and give me anydesk.

[juror897]

Thank you for the information. I have re-hosted EVE on an older server with multiple sockets. I am experiencing no adverse effects with this new setup.

With virtualization, it's clearly better to have multiple threads/cores over a single (and newer) CPU.

Thank you very much for your responses.