Hey everyone! We're feeling the strain after upgrading several Windows servers for our clients from 2016/2019 to 2022 and 2025. As an outsourced IT provider, we've hit some bumps when it comes to RDS servers, especially those using FSLogix. We're seeing crashes and freezes quite often, and it seems like extreme RAM usage is the main culprit – we're talking 90-95% usage! On the 2025 servers, everything gets unresponsive, while the 2022 ones have FSLogix services stopping totally and only restart after a reboot, often leading to frozen sessions for users.
We've checked the Event Viewer but nothing odd stands out. It appears that individual user sessions are consuming a ton of RAM—some are around 700MB, others peak at 4-5GB! Our current RAM allocation is set with 6GB for the OS and 2GB per user, giving us about 26GB for 10 users. But now I'm questioning whether that's the right approach or if we have hardware issues or something misconfigured in VMware contributing to this mess.
I'm curious if anyone else has faced similar issues with FSLogix on these server versions? How do you calculate your RAM needs per user? Any troubleshooting tips or insights would be a lifesaver. Thanks a ton!
1 Answer
It really depends on what's being run on your farm. Adding something like Teams can stretch your resources way too thin. From our experience, 2016 and even 2019 could handle things better, but 2022 seems to be a resource hog. We ended up having to cut our user expectations in half because of it.
That's good to know! Do you have any sources or examples about how Teams impacts performance? I need something solid to show my supervisors, especially since Teams is typically on every RDS.