We currently use Xen on SLES11SP2. Some want us to switch to VMWare ESX. I have a VCPU question in regards to how each platform handles CPU scheduling. I did some searching, but rather trying to read a bunch of web pages I'm hoping someone here can give me an answer. I have a Xen server (HP DL360 G8p), Dual Xeon with 24 VCPUs. I'm currently running 26 VMs with the following CPU assignments: 14 VMs have 4 vcpus 12 VMs have 2 vcpus 1 VM has 1 vcpu Obviously I'm over subscribing the vcpus and count on the scheduler to handle it. A VMWare guy is telling me you have to be very careful with VMWare to properly size maximum VM to host ratio or you'll end up with VMs waiting for CPU. If you assign 4 CPUS to a VM, all four CPUs must be available or the VM will wait. Maybe that's the same in Xen and there's enough idle time to handle it (on my server)? Maybe my Xen VMs are waiting on CPUs, I'm not sure, but I've never had a complaint about the performance on this host for any of the VMs. Can anyone with both Xen and VMWare experience comment? Much appreciated. Thanks James