We have a new vSphere 7 setup consisting of 1 x Vcsa , 3 x HPE DL360 Gen10 servers (ESXi v7), 2 x Mellanox sfp switches and a HPE Nimble storage (hybrid: flash & spinning disk) array.
We've just setup a Windows 2019 RDS collection (6 x RDSH, 1 x RD Broker, 1 x RDWeb & RDGateway - using FSLogix user & office containers). We performed some initial testing with 30 or so users and the feedback was positive. Everything was quick! This morning, we disabled access to our old Windows 2016 RDS collection (6 x physical RDSH servers) and had staff (roughly 200 users) connect to the new Win2019 RDS collection. Almost straight away, we receiving CPU alerts for the 6 RDSH vm's. We initially gave the new RDSH vm's 8 x cpu's (ie: 4 sockets x 2 cores). We then increased the vm's cpu's to 12 each but they're still running at 100% CPU. The old physical RDSH servers only had older dual Xeon quad core cpu's. Looking at task manager, on the vms, there's no single process which is hogging the cpu. Other than the fact the new RDSH servers are vm's, running Windows 2019 and FSLogix, they're pretty much setup the same as our old physical RDSH servers (ie: same apps and users).
Anyone else came across this situation before? The new RDSH vm's effectively have quicker and more CPUs.
