Virtualization-based Security - Performance Gain

[Diogo_BR]

Hello,

Based on this article: Virtualization-based Security

Is this correct to assume that by activating VBS in Windows 2016 virtual machines with SQL 2016, on the correct hardware, we would expect a good performance gain?

The article does not make it clear whether the performance gain can be attributed to performance improvements from version 6.7 over 6.5 or whether it can be attributed to VBS activation.

[sjesse]

This is my own interpretation, I don't work for VMware if you look at the line

"As shown in the following figure, these engineering efforts resulted in a 33% improvement in transactions per minute." And the graph references transactions with vbs turned on I believe it's only the difference between the two, not if it's on or off.

[scott28tt]

I conclude 3 things from the article:

1. 6.5 doesn’t support VBS

2. 6.7 does

3. If you can use 6.7 to get VBS support, you should see a performance gain compared to 6.5 where there is no VBS.

[Diogo_BR]

The question remains regarding a possible performance difference between VMs with and without VBS enabled running on ESXi 6.7

[scott28tt]

What difference would the answer make to you?

With such a big performance gain, if I was using 6.5 with no VBS and saw this it would make me want to shift to 6.7 to be able to take advantage of VBS support.

[Diogo_BR]

I think that this information is really important for the scenario of Windows 2016 VMs running SQL Servers at ESXi 6.7, but however they were  installed in BIOS mode instead of UEFI, as this prevents the activation of VBS. But if the performance gain is significative with the activation of VBS , it would be worth reinstalling the VM OS in UEFI mode, just to activate the VBS feature and take advantage of the performance gain.