Debugging Login PI Workloads
Login PI offers two out of the box workloads to monitor your virtualized desktop environment from the perspective of an end-user. These workloads use the same applications as your regular users like native Windows and Microsoft Office apps, but it’s also possible to customize the workloads. When you create a new workload in Login PI you can run into several issues. In this blog I will provide some tips and tricks to debug your customized workloads.
There are usually three common issues with workload errors, from a simple typo, to focusing on the wrong window or using the wrong parameters. By default Login PI generates warnings to help you isolate the issues in the workload. These warnings can be found in the Windows Event Viewer, Login PI web interface but also on the console of the launcher itself.
1. Event Viewer
Typos can be easily located by opening the Event Viewer and clicking on “PI – Alerts”. In the example below I wrote “MSPaint.ex” instead of “MSPaint.exe”. This is just a small typo but it will definitely cause errors in your workload.
Alerts in the Event Viewer
2. Login PI Dashboard
In the Login PI Dashboard you can find alerts that indicate that the workload is focusing on the wrong window. In this case I have misconfigured the title of the Calculator. Instead of using Calculator I used “Calculating”. The timeout appears in the Alerts Panel of the dashboard.
Alerts in the Login PI Dashboard (logname/executable/title)
3. Console on the Launcher
An error like “Invalid parameter count” is visible in the console on the launcher.
Invalid parameter count in the console on the launcher
These three places will help you to locate the root cause of the most common workload issues, but sometimes these warnings require additional information. To gather more details about what Login PI is doing you can enable Engine and Workload debugging. This can be enabled in the database:
1. Open the MS SQL Management Studio and go to the Login PI database.
2. Edit the table dbo.WorkloadSettings
3. Change the EngineDebug and WorkloadDebug value to true
After enabling debugging, the Login PI engine will be displayed in red instead of green within the session. During the session Login PI will now write all actions into the debug folder. The debug folder can be found in the temp directory of the virtual user (%Temp%\PI\DebugLogging).
The default status bar and with debugging enabled
In case you are not able to locate a workload issue with debugging enabled, our support team is always willing to help you!