Editor’s note: This post about proactively troubleshooting and fixing Citrix XenApp/XenDesktop end user logon issues is written by Ed Duncan, a former Citrite and one of the IT industry’s foremost experts on Citrix end user logon optimization. Ed’s post is written to be a resource, so please share this content with your peers and colleagues.
A few years ago I co-authored a white paper for Citrix called Logon Optimization for XenDesktop and XenApp. In it I discussed various ways to improve the logon experience when launching XenApp or XenDesktop sessions. I am proud that it has been one of the more popular whitepapers from Citrix and that it is still relevant with the current versions of XenApp and XenDesktop.
The whitepaper did not really delve into what administrators can do after following the best practices mentioned in the document, but are still experiencing logon issues. This is an important topic because slow logons give users the perception that the problem is with XenApp or XenDesktop, when the root cause could be due to something else. There are several components involved in the logon process, and trying to identify where the problem is occurring is not always an easy task.
Think about it, you have Citrix Receiver initiating the connection from the end client device, NetScaler Gateway or a similar appliance providing secure access into the datacenter, StoreFront or Web Interface servers handling authentication, and the Broker Service on the Controller determining which desktop and/or applications the user will be given access to all happening relatively quickly when the environment is configured optimally.
In addition to those components there is the supporting infrastructure to consider with Active Directory, DNS, group policies, SQL Services, Citrix Licensing, and optionally Provisioning Services. There is also the hardware to consider from the hypervisor hosts, to the underlying network, to the storage sub-system where the virtual machines reside. With so many components involved in delivering a XenApp or XenDesktop session how does one go about accurately finding the root cause when issues arise?
Most administrators will have a well-defined methodology for analyzing each step in the logon process using an assortment of tools, logs, and reports however this process can be painstakingly difficult for a few reasons:
- IT components may be siloed. This is particularly true in larger organizations where services like Active Directory and SQL, or infrastructure components like networking and storage are managed by different IT teams. As a Citrix consultant, many companies I’ve worked with had great interaction between the Citrix admins and the other IT teams working together to resolve issues. However I have seen organizations, where enlisting assistance from other teams is a formal procedure involving submitting support tickets and waiting for a resource from that team to respond. While efficient it can slow the troubleshooting process.
- Multiple monitoring tools are required. It is not uncommon to find administrators using multiple monitoring solutions within the datacenter for different services. There are times when it is even necessary to run multiple monitoring solutions for different versions of the same product. For example, I would often see organizations running multiple versions of XenApp usually because of the need to support a critical legacy business application that won’t run on Windows Server 2012 R2. Therefore EdgeSight is used to monitor the older XenApp farms and Citrix Director is used for the newer XenApp 7.x sites. This creates additional complexity when troubleshooting because both tools may be required.
- Data may not accurately reflect the issue. The data gathered by some tools may not be indicative of the real issue. For example, a report may show that logon performance is slower during the morning hours, which is expected due to the increased number of users starting sessions around the same time. However just assuming that the increased load is the problem and not digging further into the details may cause you to miss the real issue like a faulty domain controller or non-responsive DNS server. During periods of light user logons a bad domain controller or DNS server may be masked by a healthy domain controller or DNS server that is handling all of the logon requests.
These are just a few examples of what I’ve seen Citrix administrators having to grapple with when troubleshooting session logon issues. An ideal tool would be able to capture the entire logon process from initiation to launch, and provide in-depth details that can be analyzed quickly and easily.
I have recently been evaluating Goliath Technologies’ Logon Simulator. It is a product that is capable of simulating the full user logon experience, and through the use of agents it can report on what is occurring at every step of the logon process. The product can also send alerts based on failed or slow logons.I have been looking at this product for several reasons:
- There are not many tools available that can detect and report issues across the different components that make up the Citrix environment as well as the supporting infrastructure. If I am experiencing slow logons and can show that the problem is with a domain controller, or that network latency is high, or read/writes to storage is taking longer than usual, then the teams managing those services will know where to focus their investigation, which speeds up the troubleshooting process.
- Logon Simulator collects data from every component in the Citrix environment and provides an overview from one console, which is more efficient than having to pull logs or run reports from multiple sources and monitoring tools.
- A unique feature of Logon Simulator is being able to drill down into the details of the individual ICA channels, which makes it easier to identify the activities of the end user when performance issues occurred. This is useful for troubleshooting slow sessions in general and not just the logon performance.
The product provides many built-in reports. I won’t cover each one here but some important ones to note are the End-to-End Connection Report which gives details about the specific components that were touched during the logon process, from the type of device the end user is working from to the server the session was brokered to and how long it took to get there. The ICA Latency Report (previously mentioned and shown pictured above), and the End User Activity report which gives details on the number of sessions and the length the sessions ran.
At the time of writing the Logon Optimization whitepaper there wasn’t a technology that could preemptively logon with synthetic users to identify potential performance issues before real users attempt to logon. I can see great value in this technology to fundamentally change how Citrix Administrators manage and troubleshoot logon issues.
To see a live demo of the Logon Simulator, please watch our pre-recorded webinar, Introducing the Goliath Logon Simulator for Citrix. You can also get a fully supported free 30-day trial or personal demo.
If you have any questions about the Goliath Logon Simulator or you need help troubleshooting a Citrix end user logon issue please leave a comment. Thank you.