We have a customer with Windows Server 2022 VDA running on PVS. We see multiple performance issues and app crashes in this environment. The VDA are running on XenServer with local storage for the cache disks. The PVS servers are running on VMWare with fast DataCore storage. Finally I convinced the customer to use ControlUp to monitor the environment. Thanks to ControlUp we are now able to see what’s the real issue. The VDA have a lot of pages/s and page faults and the pagefile is heavily used. We configured the pagefile with a static size of 4 GB on the cache drive, after some hours in the morning the pagefile was at 100 %. We now increased the pagefile to 12 GB and it’s getting better. Here’s my question: Although the VDA only uses 40-50 % of the physical RAM (45 GB), the pagefile is always in use until it’s full, obviously no matter what size it has. How is that possible, what could be the reason, how can I find out what causes the massive pagefile usage?

There are performance counters that measure the Page File Bytes per process, this way you can find the culprit. In Performance Monitor you can setup a Data Collector Set to store the data for a longer period, or use ControlUp to explore the processes.

Thanks 🙏 I will give it a try.

Screenshot 2024-07-10 at 16.10.14.png

Also there is #CKL5UCNCX

PS: Don't assume the pagefile is only used when it runs out of physical RAM. While that is one of its functions, swapping out memory from RAM to disk if needed, some core Windows systems are always in the pagefile. 4GB is really low IMO, why save out on a couple of GB of disk space in 2024? Guidelines is 1.5-2x RAM size for a pagefile