1

I’ve got a weird issue with a customer where their profile load times are high. They use Citrix User Profile Management (yes, I know it’s outdated, but it’s unfortunately not up to me to tell them to change. I just support what they already have in place and try to optimize it).

They first started to come to me about long logon times for Citrix about a month ago, so I’ve been digging into finding the culprit for this for a while now. It’s like a needle in a haystack!

Looking at the Citrix Cloud dashboard and trends, I note that the Profile Load time is what seems very high. Looking at profile sizes, they’re not very big at all. Usually, no bigger than 500mb to 1.6 GB but nothing bigger. Excluding all the recommended directories, that Carl Stalhood notates in this blog .

Nothing has worked so far…except renaming the NTUSER.DAT file for the respective profile. When we do that…the profile speeds up and acts normally again, for about a week. Over the span of that week, we note that the profile load time creeps up until it’s back to “start session and grab a cup of coffee, it’s going to be five minutes”.

I’ve got a Citrix Support case rolling already, but thought I’d ask around in the Spiceworks community too while at it. Anyone else had Citrix Profile Load time issues and managed to emerge victorious?

5 Spice ups
2

What group policies are applying to the user - any that might be waiting for a mapped drive and timing out?

What about AV, are exclusions in place?

Is it the same for all users, some or one?

Are all profiles in one place, is that system under resourced?

These are not Citrix specific, but can relate to any VDI/RDS type setup.

3

Thanks for the reply, Rod!

What group policies are applying to the user - any that might be waiting for a mapped drive and timing out? Just some typical group policies that apply to all users for Citrix Profile Management, and then a GPO that sets Citrix VDA server session settings:

  • Citrix Components/Profile Management -
    • Active Write Back - Disabled
    • Enable Profile Management - Enabled
    • Path to user Store - Enabled
  • Citrix Components/Profile Management/Advanced Settings -
    • Customer Experience Improvement Program - Disabled
    • Process Internet Cookies on logoff - Enabled
  • Citrix Components/Profile Management/File system -
    • Enable Default Exclusion List - directories - Enable
    • Exclusion List - Directories - Enable
      • AppData\Local\Microsoft\Windows\INetCache
      • AppData\Local\Microsoft\Windows\WER
  • Citrix Components/Profile Management/Profile Handling -
    • Delete locally cached profiles on logoff - Enabled
    • Local profile conflict handling - Enabled
  • Citrix Components/Profile Management/Registry -
    • Software\Microsoft\AppV\Client\Integration - Enabled
    • Software\Microsoft\AppV\Client\Publishing - Enabled
    • Software\Microsoft\Speech_OneCore - Enabled
    • NTUSER.DAT backup - Enabled
  • Citrix Components/Profile Management/Streamed User Profiles -
  • Profile streaming - Disabled- (Scoped against the VDA hosts) Windows Components/Remote Desktop Services/Remote Desktop Session Host/Session Time Limits
    • Set Time limit for active but idle Remote Desktop Services sessions - Enabled (set to 3 hours)
    • Set time limit for disconnected sessions - Enabled (set to 1 minute)
    • Set time limit for logoff of RemoteApp sessions - Enabled (set to “immediately”)

Nothing waiting for mapped drives, though I did see the following error in the Application logs, just before RSOP starts calculating:

"Failed to read the VHDX files associated with the following policies: Profile container (error code 0x3), Search index roaming for Outlook (error code 0x3), Accelerate folder mirroring (error code 0x3). Cause: Could not access the VHDX storage path “\<storage_path>”. Action: Verify that the storage path “\<storage_path>” is accessible.

Source: Citrix Profile Management

Error ID: 4000

Level: Information

What about AV, are exclusions in place? We checked their AV, can definitely confirm exclusions are set.

Is it the same for all users, some or one? Only some. Others are just fine and average 1-10 seconds profile load time

Are all profiles in one place, is that system under resourced? One place. Storage, CPU, RAM all seem fine, nothing being pounded hard. Network is a busy hospital network, and we haven’t ruled that out, but profile sizes aren’t that large (less than 1GB, some even just a few 100mb), and nothing else is generally that slow for them from a network performance perspective.

These are not Citrix specific, but can relate to any VDI/RDS type setup.

@rod-it

4

I’d start here.

If every user has their own VHDx, can you recreate it and see how it performs, it’s been a long time since I used Citrix as a backend, but VMware VDI with UEM/Fslogix have similar features and use containers for profiles.

Confirm the clients (or the users sessions) are not having DNS issues and are using good and valid DNs settings.

If the VHDx has a specific purpose - it mentions policies for outlook, one assumes this is an outlook cache location - if true, for one of the problem users, can you disable outlook caching, if this resolves login times, focus again on that container, the OST cache could be corrupt or the container its sat inside.