End users complaining of slow logon issues.
Identifying Root Cause
When trying to identify the root cause of a Citrix end user experience issue, the main display to view all of the Citrix user sessions is the XenApp & XenDesktop session display.The display is key because it includes user session data (both past and present), and allows you to track the complete user experience through the environment, from the login at the endpoint, all the way through the environment back to the underlying infrastructure, and present these data points over the course of the session so you can troubleshoot any issue that takes place during a user’s session.
To access this display, click View then XenApp & XenDesktop. This page is divided into three areas: App Servers and Published App & Desktops (for Citrix XenApp environments) and Virtual Desktops (for XenDesktop environments). You’ll want to navigate to the applicable section for your environment (Published App & Desktop or Virtual Desktops) to troubleshoot further.
To view the sessions with high logons sort the page by the Logon column or use the Search feature at the bottom of the page. Some of the key metrics of the page include Machine Name, Username, Group Name and Start Time which help confirm whether or not any patterns can be identified outside of high logon duration for the end users. For example, when widespread issues are occurring it can be good to note the client address to identify if all of the users are connecting from the same locations.
To analyze the 33+ stages of the users logon process, you can click on any user session to drill into the session metrics. Once you've drilled down, navigate to the logon duration tab. This tab is divided into three sections. The top section provides the logon summary showing the high-level stages of the overall logon process. The next two sections break down the details of what is happening in the brokering stages and the launch of the session on the App Server or VDI.
The brokering stages in the middle of the dialog enable you to identify slowness due to Citrix Delivery Controller brokering issues including Authentication, Logon Script, Profile Load time, delays in Windows session creation on the session host. On the client side, you can identify delays in establishing the micro-VPN connection, receiver launch, ICA file download, and more.
At the bottom of the dialog each stage of the logon process at the session host process is displayed including authentication at the domain controller, determining how long it took to copy over the policy files to the session host, logon script execution, drive mapping, printer mapping, and folder redirection. Each policy and script which is executed is included in this stage.
Figure 1 - Session dialog - Logon Duration tab
When you have many users in the environment complaining about the same issues, in this case logon duration, finding patterns can be key to identifying root cause.
Using the Logon Duration details, one user was able to determine that the widespread delays in logon duration ultimately stemmed from drive mapping (logon) scripts. This was discovered by tracking high GPO times in the top level of the logon duration breakdown, down to the brokering stage of PLSD (The time spent on the server loading the user's profile), down to the bottom section of the page, specifically Group Policy Drive Maps Extension.
Figure 1: Real-world Example - Logon Duration tab
In the example scenario described above, reviewing the logon duration metric trends resulted in identifying high group policy drive mapping. Since specific policy(s) that were contributing to the drive mapping extension were displayed, further investigation of the policy that was mapping the drives could take place. Ultimately, this user discovered that the scripts were mapping 30+ errant drives to the user’s profile. The engineers were able to reduce the number of logon scripts down to a single script, and removed the problematic, as well as incorrectly linked, GPOs. By determining the root cause and applying the aforementioned actions in their environment, the system engineers were able to reduce logon times down to an average of 45 seconds.