With same system resources, VMware is much slower than physical machine due to technology nature.
We want to know more about impact on monitoring. Our current monitored object is BMC Remedy for user log in/out experience. It is browser based interface. The transaction testing includes
Open Change Console
Open Incident Console
Open Problem Console
Open Service Request
I am wonder that metrics values will be different with VMware comparing with physical machine. In other word, FTR in VMware may not reflect user experience who use physical machine.
FTR will run fine on VMWare. There are no special requirements to run FTR on VMWare machines, other than the standard documented requirements.
Having said that VMWare has limitations of its own, given VMWare machines and physical machine of the same hardware spec the VMWare machine will always run slower as it has to run though the OS it is running on top of.
If you are seeing the FTR image running more than 2X slower than the physical machine and the CP # speed and RAM are comparable then I would start looking into VMWare performance issues.
Many of our customers are running FTR on VMWare, in cases where we see performance problems they have all been traced back to overloaded VMWare hosts.
The short answer to your question is that ANY metrics involving timing will show different times between a Physical machine and a VMWare machine.
VMWare works by time slicing CPU time access the number of VMWare images running on a server.
Assuming you have a very powerful VMWare server running only one image you could see faster times then on a normal physical desktop.
But in most cases there is a base penalty you pay for running on VMWare to start, this varies based on the type of VMWare(workstation vs. ESX), the platform hardware and the number of images running on the platform.
In short you can't rely on times in VMware to reflect what a physical users experiences.
We have some customers who benchmark a physical browsers against scripts and then run the same scripts on VMWare and get a multiplier that they use for their SLA's, this means that they periodically need to rebaseline or are not allowed to add new scripts or images to their VMWare servers.