VMware Horizon Management User Guide
Overview
This DEX Pack shows performance and usage data for VMware Horizon virtual machines. Use this DEX Pack tassess provisioning needs, system issues, and user concurrency. This data can be organized by desktop pools, virtual machines, and users.
Machine Rightsizing
Assess provisioning of virtual machines. Determine which specific metrics are over-provisioned or under-provisioned.
Pooled User Health and Concurrency
View usage and performance issues for desktop pools.
Protocol Performance
View data metrics for VMware protocols. Identify sessions with high data transfer.
Sensor Details
Identify and troubleshoot specific system issues affecting VMware hosts.
Machine Rightsizing
Global Filters
By default, this Dashboard shows data for both virtual and physical systems. Use the System Type drop-down to show data for only one type of system. View virtual systems to determine provisioning needs for those systems. View physical systems to determine which systems could be virtualized.
Machine Rightsizing Summary
If a system is under-provisioned in any metric, it is labeled as under-provisioned. This is true even if that system is over-provisioned in other metrics. A system is only listed as right sized if it is not under-provisioned or over-provisioned in any metric.
ESX Server Sizing
This table suggests the total CPU, RAM, and Disk capacity needed to achieve right sized provisioning for all systems. Filter this data using the drop-downs in the Global Filters pane.
Use the Overhead % slider to adjust the extra provisioning for all metrics. For example, if Overhead % is set to 5, then this calculation assumes 5% additional CPU, RAM, and Disk capacity for every system. Every system will be slightly over-provisioned. This helps prevent performance issues if your users begin using slightly more resources than they previously have.
Machine Rightsizing Details
This grid shows detailed performance metrics for every system. To filter this data, use the Sizing Type drop-down as well as the drop-downs under Global Filters.
For virtual systems, you may need to alter provisioning for under-provisioned or over-provisioned systems. The Suggested Memory/CPU/Disk columns show suggested changes highlighted in yellow. Use C Drive Free GB and Graphical Intensity Usage to assess storage and GPU provisioning, respectively.
For physical systems, you may be able to virtualize over-provisioned systems. This may free up resources and be more cost effective for your enterprise. For under-provisioned physical systems, you may need to troubleshoot causes behind the performance issues, or the users may require an upgraded system.
Double-click any online system to more closely investigate it in Resolve.
Rightsizing Calculations
| When Max Memory Usage is between 80% and 100% of Memory Capacity | Right Sized |
| Greater than 80% | Under-Provisioned |
| Less than 80% | Over-Provisioned |
| When Max CPU multiplied by 1.25 is 80% to 100% | Right Sized |
| Greater than 100% | Under-Provisioned |
| Less than 80% | Over-Provisioned |
| Virtual Machines | |
|---|---|
| When Max CPU multiplied by 1.25 is greater than 100% of capacity | Remove Cores |
| When Max CPU multiplied by 1.25 is less than 80% of capacity | Add Cores |
| For Physical Machines | When Max CPU multiplied by 1.25 is less than 80% of CPU capacity |
| Remove Cores | When C Drive Free GB is between 10% and 75% of the total C Drive storage |
| Right Sized | Less than 10% |
| Under-Provisioned | Over 75% |
| Over-Provisioned | When Graphical Intensity Usage is over 75% |
| Heavy | Between 10% and 75% |
| Moderate | Less than 10% |
| Light |
Memory Rightsizing Calculations
-
Recommended memory size is calculated based on Max Memory Usage multiplied by 1.25
-
If the recommendation matches the existing memory capacity of the system, we say No Change (Right Sized)
-
The sizing is split into buckets. We recommend the possible memory values listed below. There is a different set of values for physical and virtual systems
If Max Memory Usage multiplied by 1.25 is between 0 and 4 GB, we recommend 4 GB
| Physical | |
|---|---|
| 4 to 8 | 8 GB |
| 8 to 12 | 12 GB |
| 12 to 16 | 16 GB |
| 16 to 20 | 20 GB |
| 20 to 24 | 24 GB |
| 24 to 28 | 28 GB |
| 28 to 32 | 32 GB |
| 32+ | No Change (Right Sized) |
| Virtual | |
|---|---|
| 0 to 2 | 2 GB |
| 2 to 4 | 4 GB |
| 4 to 6 | 6 GB |
| 6 to 8 | 8 GB |
| 8 to 10 | 10 GB |
| 10 to 12 | 12 GB |
| 12 to 14 | 14 GB |
| 14 to 16 | 16 GB |
| 16 to 18 | 18 GB |
| 18 to 20 | 20 GB |
| 20 to 22 | 22 GB |
| 22 to 24 | 24 GB |
| 24 to 26 | 26 GB |
| 26 to 28 | 28 GB |
| 28 to 30 | 30 GB |
| 30 to 32 | 32 GB |
| 32+ | No Change (Right Sized) |
Overall Provisioning Logic
-
If the Max Memory Usage, Max CPU, and C Drive Free GB all say No Change, then the Sizing Type for this system is labeled as Right Sized
-
If any of the following is true, we label the system as Under-Provisioned
-
CPU recommendation is Add Cores or Over-utilized
-
Memory recommendation is to increase memory size
-
Disk recommendation is Low Space
-
-
Any systems that do not meet the conditions listed above are labeled as Over-Provisioned
Pooled User Health and Concurrency
Desktop Pools Usage and Health
This table shows session and health score data for VMware desktop pools. For more infon creating and configuring these desktop pools, refer to VMware’s documentation.
Average Session Health shows the percentage of time where productivity was not negatively impacted by system issues. System issues can include a variety of factors such as latency, CPU utilization, and disk utilization. In general, a lower Health Score indicates that a system is affected by system issues more often. This score is an average for every session in this desktop pool.
For example, consider a desktop pool with an Average Session Health of 85. On average, these sessions were negatively impacted by system issues for 15% of their active time. If a session was active for 10 hours, then it was likely impacted by system issues for about 1.5 hours.
The color coding for Average Session Health is as follows:
| Red | Less than 70 |
| Yellow | Between 70 and 90 |
| Green | Greater than 90 |
Click a desktop pool to populate the data in User Concurrency, Top 5 Health Concerns, and Top 25 Users by Session Time
User Concurrency (Selected Desktop Pool)
This graph shows the number of active sessions for the desktop pool selected in Desktop Pools Usage and Health. This data is shown every hour.
Top 5 Health Concerns (Selected Desktop Pool)
This graph shows the top 5 types of system issues impacting the desktop pool selected in Desktop Pools Usage and Health. Troubleshoot these issues to improve the user experience for users in this desktop pool.
Users by Session Time (Selected Desktop Pool)
This grid shows the users with the longest cumulative session time for the desktop pool selected in Desktop Pools Usage and Health.
Click a user to populate the data in Health Trendand Health Impact Breakdown.
Health Trend (Selected User)
This graph shows the health trend for the user selected in Top25 Users by Session Time.
Health Impact Breakdown (Selected User)
This grid shows how much each type of system issue impacted the user selected in Top 25 Users by Session Time. For each system issue, this data shows the average number of minutes affected by that issue during every session. Troubleshoot prominent system issues to improve the user experience for this user.
Protocol Performance
Performance Summary
This grid shows data metrics for VMware protocols. Each metric is averaged from every session using that protocol over the past 30 days. Each metric is related to data sent or received via the corresponding VMware protocol. For more information on these metrics, refer to VMware’s documentation.
Click a metric to populate the data in Performance Trend and Session Details.
Performance Trend (Selected Metric)
This graph shows data for the metric selected in Performance Summary. Specifically, it shows the daily average across all VMware sessions using that protocol for the past 30 days.
Click any point on this graph to populate the table in Session Details.
Session Details (Selected Metric and Date)
This grid shows the worst performing sessions for the metric and date selected inPerformance Trend. By default, the worst 25% of sessions are shown. Use the slider to adjust this value.
Double-click a system to view the performance trend for this specific system, metric, and date in Resolve. High rates of data transfer may increase latency or other performance issues on this system. Use the Health page in Resolve to investigate performance issues for the selected system, and troubleshoot those issues as needed.
Sensor Details
Triggered Sensors
This grid lists all sensors relevant to VMware hosts that triggered in the past 30 days. It also shows which sensors are currently active on VMware hosts. If any VMware-specific sensors are triggered, they are shown at the top of the grid.
Click a sensor to populate the other panes in this Dashboard. Double-click a sensor to view additional sensor info in Prevent.
Sensor Description (Selected Sensor)
This pane shows the description for the sensor selected in Triggered Sensors.
Active Sensor Trend (Selected Sensor)
This graph shows the number of VMware hosts that triggered the sensor selected in Triggered Sensors. This data is shown every day for the past 30 days.
System Details (Selected Sensor)
This table shows all the VMware hosts that triggered the sensor selected in Triggered Sensors. This data is from the past 30 days. Double-click an online system to investigate it more closely in Resolve.
On This Page