The Summary tab is the user experience dashboard that provides information and metrics about the user's specific VMware Horizon application or desktop session. This tab is essential to quickly diagnosing user session performance issues. Combining both a high-level view as well as detailed metrics, this should be the first place your help desk should look to determine a user’s performance issue.
Being able to see the user’s RTT, Jitter, and estimated bandwidth on the same screen that you have virtual machine and host performance metrics will eliminate unnecessary time troubleshooting. Why bother looking into a user’s connection when you can see that the problem lies with the user’s virtual machine resources? Have multiple users complaining about performance issues? Take note of the hosts they are connected to and be able to escalate to the proper team that manages the host infrastructure.
Figure 1: Summary tab
User Experience Summary and Supporting Infrastructure
The middle section of the summary tab includes metrics about the session, virtual machine, and host the session is running on.
User Session
This section provides key metrics that allows the viewer to come up with quick answers about the performance of the Citrix Session and if slowness is caused by the network or the resources.
- RTT (Round Trip Time): Indication of the performance of the session. If above 450ms, the end user is experiencing slowness in their session.
- Session Jitter: Network Jitter for uplink. Jitter represents a time delay, in milliseconds, when sending data packets over your network connection. Delays can be caused by network congestion or by route changes. Delays of about 30 ms and higher will begin to impact session performance and will be noticeable to the end user.
- Session Estimated Bandwidth: The estimated bandwidth for an uplink signal. Bandwidth allocated for sessions can be reserved via policy within your Horizon environment. The more bandwidth available to the session, the better the user experience. Refer to VMware guidelines for allocating bandwidth and ensuring there is enough available for your user's sessions.
|
How to Troubleshoot: If RTT is high, take a look at Jitter, if this is also high and similar to RTT then we can rule out the resources as the cause of slowness issues. If Jitter is low then take a look at the Virtual Machine Resources and Host Resources Sections. |
Virtual Machine Resources
This section provides insight into the resource utilization of the session host machine from the VM and OS level for the timeframe of the user’s session.
- CPU: Current and historical CPU load for the virtual machine.
- CPU Ready: The amount of time a virtual machine is ready to use CPU but was unable to schedule time because all CPU resources are busy. CPU ready above should be a warning and above 10% should be a major concern.
- Memory: Current and historical memory usage for the virtual machine.
- B/W: Network bandwidth consumed by the virtual machine.
|
How to Troubleshoot: The Virtual Machine Resource section will assist you with either finding troublesome applications causing your virtual machines to struggle or outline an issue with under sized virtual machines. You can compare this section to RTT graph to determine if spikes in RTT correlate with spikes in resource utilization most likely caused by a resource issue. |
Host Resources
This section provides insight into the resource utilization of the Hypervisor host supporting the Virtual machine session host for the timeframe of the user’s session.
- CPU: Current and historical CPU load for the virtual machine.
- CPU Ready: The amount of time a hypervisor host is ready to use CPU but was unable to schedule time because all CPU resources are busy. CPU ready above should be a warning and above 10% should be a major concern in a virtualization environment.
- Memory: Current and historical memory usage for the virtual machine.
- Storage Latency: Current and historical latency between the hypervisor host and the datastore the virtual machine is running from.
|
How to Troubleshoot: This section helps you quickly determine if your hosts or storage are oversubscribed. High CPU, Memory and Storage latency will affect all virtual machine sessions on the displayed host or storage. Make sure your user’s virtual machines are balanced across all available hosts. If necessary, set policies to power off virtual machines not in use to reduce the load on your hosts. |
Viewing Additional Data
Each graph on the summary tab can be expanded to view additional data points for the specific metric.
Figure 3: Virtual Machine Resources CPU graph expanded
In addition, you can use the mouse wheel to zoom in/out of each graph or selection specific areas of a graph to view more detailed data.
You can also click the header for any section on the summary chart to jump to the relevant tab within the session dialog.
Application Resource Usage
A detailed breakdown of application usage can also be essential to determining the cause of performance issues in a user’s virtual machine. Note that applications and processes are displayed for the current session host or virtual desktop as indicated by the timestamp included in the section.
Figure 4: Top 5 Applications by Resource Usage
|
How to Troubleshoot: A single application may be the cause of a user’s poor session performance. Ending the offending application may take care of the problem, but if you notice this same trend on other virtual machine sessions you may be able to narrow down a larger issue. |