This message was deleted.

What issues are you having Cory?

like what? i had issues too, but it had do with my existing gpo for printers

Share man, we'll help. It should honestly be as simple as out it in and leave it. Lol as long as you have enough storage.

Nope. Been busting our heads against this since implementation. We're on shared desktops (2016) though.

hence why i'm still on upm

What are you issues guys? I've rolled more than a few FSLogix implementations on windows 10 and server 2016.

Random freezing

Interesting. Have you captured any metrics on your fileservers etc... I'd be happy to take a look with you.

Of the desktops, they will run great for 10-12 hours users disconnects and it freezes. Happens to me too, working then the task at goes away but I can still use ie and other apps . It’s like a slow death

Interesting. Are you using cloud cache?

I have disabled ofdc and downgraded

Okay, but are you using cloud cache for storage or just standard vhd locations?

Sorry reg file share

Weird, I'd love to see your logs and some perf monitoring from your file servers. I'm sure we could nail this down.

Good idea I have removed the app for the users for now and using upm will go back to More testing.

Definitely worth a bit more testing if you're seeing issues. Let me know if you need a hand.

Im getting to the point where fslogix is to worth it anymore.

It shouldn’t be this complex.

@Ray Davis What are you seeing?

For me it's anything that maintenance that is performance on the user vhdx profile files. Then logins are destroyed. User shell goes to 80plus seconds. It when u redo the profile it's back to 8 seconds

I had the same experience as well where logon times were very high in random situations

Interesting. I'd love to try and reproduce this. I've never seen this before. @James Rankin have you?

Leee, what ur average login times? What is is your profiles on? Do you exclusde things?

Nope never seen that happen, although we do see the occasional bit of aberrant behaviour. if you spin up a totally vanilla lab machine, no GPOs, no apps, nothing that reflects your production environment, and test it - do you get the same odd behaviour? I'm aware some environments do exhibit problems - just trying to narrow it down

Hi Ray, depends on a lot of factors to be honest. In the main though I'd say average of around 30-40 seconds in a mature environment. Much quicker on a greenfield site. Exclusions dont really save a lot of time as its just the VHD attach. My advice as per James comment, setup a new OU. Move a box to that, apply FSLogix settings only. See what happens.

Ok will do guys. Thank you for your time.

Let us know how you get on Ray.

I figured it out. Man what a week . So I don't know how it happened. Lots of peocmons and I found that windows shell was 50 %cpu everylong. So I would log in and force it to load, and it would just kill my session. Procmon was showing alot of explore.exe reg keys loading. So I googled and found this https://support.microsoft.com/en-us/help/3161390/slow-logon-with-a-blank-screen-in-windows-7-or-windows-8-1 But I wasn't to sure. Then I found this But the this articlea https://support.citrix.com/article/CTX212905 And https://xenappblog.com/2016/optimize-logon-times-os/ So I creat a little cmd files to export it out and automated it with control up to export it and delete it. Test it on 10 users and logins we're back to 8-15 seconds. 8 on published apps 15 on desktops. I have no idea how this got into everyusers hive. Unless something I did on the default hkcu. Anyways, I was wrong about FSlogix being the issue. However their support is not really mature enough to understand the process and I didn't ha e time to teach them about the whole login process Procmon is a tuff tool to help you as well.

Nice work though Ray! Great find! It's hardly ever FSLogix in my experience (unless you use CloudCache)

Thank you. I would love to use cloud cache. But I am hearing that it's problematic. I only have 1 profile server now. I replicate it to our DR. But I need two of them in sync. James gave me some good pointers, but I need to figure that out next. DFS-R has file lock issues. I need something that goes back and fourth between the 2 servers and keeps the data in sync not matter what smb share you hit.

There's not really a 100% solid solution without looking at storage level replication right now. Personally for production I like a cluster with continuous availability. Then dependent on the backup product I schedule a restore to DR (I like Veeam for that) the alternative is the block level copy tool, always very good.

I like MSCS but I don't like it in a VMware setup. But again it's be a while since I did what they called it a while ago. Cluster in a box or cluster across a box. Rdm was was a pain point back then. But I'm not up to date on continuous replication. I assume your running this now?

VDI in a box 😉