Guys, I have a customer that moved to Windows 11 (CVAD 2402 CU3, non-persistent PVS) around 6 months ago. We are seeing an average latency of around 40ms in Director (Disk Latency), the VMware cluster and PURE Storage shows zero latency or bottleknecks. I have noted that the vDiskdiff write cache grows to around 12 GB after around 30 mins when nothing is used/launched on the desktop. Windows update is disabled on the image but Windows is re-enabling it, I can see Windows Update thrashing for a while but doesnt appear to be installing any updates when I check. Whats the best way to disable while a safely stopping Windows from re-enabling it? I'm conscious that a load of disk IO and caching could be contributing to latency/performance issues, cheers

I usually will point provisioned machines to a non existing wsus update machine with gpo, and also disable the ability to connect to microsoft for windows updates. What are you using for profile management? The Appx packages also can fill that cache quite quickly.

Hey Jeff, FSLogix with Cloud Cache and redirections so the IO inherently is pretty heavy. I do see in FSLogix ADMX install AppX Packages is enabled, that could be causing it to fire, cheers

AV can also mess up the cache big time.

Cloudcache can inflate that as well, as it creates proxy and cache files on the c:\ that go into that writecache folder.

Thanks, the fsl cache is redirected to D: but I hear what you are saying. Crowdstrike is the AV solution, the customer was sure no exclusions were required but we have got them added to be sure. The only ones yet to go in are the fsl cache directories. It's managed by CS as a service so I’m hoping they get added today, cheers