soft lockup - CPU## stuck for ##s!

[trogne]

On my vCenter Server Appliance VM, I get "soft lockup - CPU## stuck for ##s!" message.

What can I do to prevent this, or at least, make them appear less often ?

See screen attached.

When this is happening, most of the time I have to restart my vCSA VM.

[trogne]

I think it was a memory issue.

In vmware workstation, in "edit - preferences - memory", I select "Allow most virtual machine memory to be swapped".

Now, more memory can virtually be used (swapped in the VMs).

[trogne]

Even with "Allow most virtual machine memory to be swapped",

I still have lots of lockup, like "NMI watchdog: BUG:  soft lockup - CPU#0 stuck for 21s!"

What is that?

Now, vcsa is started, and I wait and wait and wait for page to respond.

Since 30 minutes, I have this :

503 Service Unavailable (Failed to connect to endpoint: [N7Vmacore4Http20NamedPipeServiceSpecE:0x000055611f20c300] _serverNamespace = / action = Allow _pipeName =/var/run/vmware/vpxd-webserver-pipe)

What's going on ?

[trogne]

All works fine now.

What I did :

In vmware workstation, Preferences, Memory : reducing the amount of memory to 10 gb (keeping the "Allow most virtual memory..." checked).

I think that's enough. My host pc was going to 100% memory. Now it doesn't goes beyond 95%.  The VMs does swap for the remaining.

I also moved my vcsa machine to my SSD disk. If swap memory needed, it's much better on ssd.

I've tried back the vcsa on the hdd. No good.

So, to me, the vcsa absolutely needs to run on ssd. Bar none.