vCenter Operations Manager Dashboards

8. vCloud Integration Examples : 8.1 vCenter Operations Manager : 8.1.5 vCenter Operations Manager Dashboards

8.1.5 vCenter Operations Manager Dashboards

To get the full value from vCenter Operations Manager to monitor a vCloud environment, it is recommended that you configure vCenter Operations Manager dashboards. These dashboards provide a view of the health of the various vCloud constructs.

Dashboards can be shared between Admin groups. Examples of this are the disk and network dashboards. The storage administrators can have a dashboard that is related only to storage metrics from the cluster. This dashboard can include metrics such as cluster disk I/Os or read/write latency. At the same time, the network administrators can have a dashboard that is related to the cluster networking metric. These metrics can include physical switches that are connected to the vSphere cluster to give the network administrator visibility into statistics from the virtual and physical environment on one dashboard.

For this example, a dashboard has been created to display the following statistics on the resource cluster:

Capacity remaining.

Alerts and events.

Network.

Storage.

Memory.

CPU.

A second dashboard gives statistics on the management cluster. The dashboard has been configured with the following metrics:

vCenter SQL database transactions/database size.

vCloud Director SQL database transactions/database size.

SQL Server operating system drive space remaining.

vCenter Server operating system drive space remaining.

vCloud Director mount point space remaining.

8.1.5.1. vCenter Operations Manager Widget Configuration

The custom UI of vCenter Operations Manager uses widgets to display information about objects that are being monitored. Some widgets display only data, but others can be configured to display data and set thresholds that display different colors when thresholds are exceeded. The following are two examples of how to configure widgets.

8.1.5.2. Generic Scoreboard

The widget configuration settings are shown in the following table.

Setting	Value
Widget Title	Displayed title of widget
Self Provider	On
Refresh Widget Content	On
Widget Refresh Interval	300 (seconds)

After the widget settings have been set, the relevant objects and metrics must be selected. This example uses a filter for cluster. Then, the cluster for which we want to display statistics and metrics is selected. After all of the desired metrics are listed in Selected Metrics, the thresholds can be configured.

The following table shows the thresholds that can be set.

Threshold	Range
Green	Up to 10
Yellow	10–20
Orange	20–30
Red	30 and higher

The following figure shows the completed widget.

Figure 41. Generic Cluster CPU Scoreboard

8.1.5.3. Heat Map

The Heat Map widget can be used when a few objects (for example, datastores, cluster physical CPU cores) must be displayed in comparison with each other. For example, physical CPU cores can be displayed for all hosts in a cluster, and those over a certain threshold can be displayed as red to indicate hot spots.

The widget configuration settings are shown in the following table.

Setting	Value
Widget Title	Displayed title of widget
Self Provider	On
Refresh Widget Content	On
Widget Refresh Interval	300 (data collection is every 300 seconds)
Group By	Select object to group by
Resource Kind	Heat map data displayed for which type of resource

After the resource kind has been selected, choose the reported metric by selecting the attribute kind. Then select the tag from which the data will be reported.

After the selection of metric and tags, save the configuration. Click the green plus sign to give the configuration a name and save it. Multiple configurations can be saved and then selected to be displayed on the same widget.

The completed heat map widget is shown in the following figure.

Figure 42. Management Cluster CPU Core Utilization Heat Map

8.1.5.4. Physical CPU Resource Monitoring of Resource Clusters

The CPU Dashboard can be used to monitor the physical CPU performance of the resource cluster. Figure 43 shows a sample dashboard that was built using the following widgets.

Generic Scoreboard (Figure 44).

Heat Map (Figure 45).

Metric Graph (Figure 46).

Health Status (Figure 47).

Figure 43. CPU Resource Cluster Dashboard

The Generic Scoreboard widget is at the bottom right of the dashboard in Figure 43. The selection is based on the resource cluster metric which is a sum of all host metrics in the cluster.

Table 17. Generic Scoreboard Widget Configuration

Metric	Unit	Green Range	Yellow Range	Orange Range	Red Range
Capacity Usage %	%	50	50–75	75–85	85
Demand	%	50	50–75	75085	85
CPU Reserved Capacity	MHz	100000	100000–150000	150000–175000	175000
Wait	ms	100	100–200	200–300	300

NoteThese threshold/ranges are only examples. These values should be based on the cluster design threshold values and the customer requirements.

Figure 44. Resource Cluster CPU Scoreboard Widget

The Heat Map widget shows the resource cluster physical CPU core utilization. This displays all the ESXi physical CPUs and cores in the resource cluster to identify hot spots on the physical CPUs. This configuration uses the core utilization metric and resource cluster tag to display the heat map data.

Figure 45. Resource Cluster Physical CPU Core Heat Map Widget

The Metric Graph widget is used to give a graph view of some of the CPU metrics. The view is customizable to display from last hour to last year. The graph can also display the dynamic thresholds for certain metrics. As an example, this widget displays the cluster CPU usage and cluster CPU wait.

Figure 46. Resource Cluster CPU Metric Graph Widget

The Health widget is used to display overall Resource Cluster Health. The widget can be configured to display data from the last hour to last month.

Figure 47. Cluster Health Widget

8.1.5.5. Memory Dashboard

The Memory Dashboard can be used to monitor the cluster memory usage and demand. Figure 48 shows a sample dashboard that was built using the following widgets.

Generic Scoreboard (Figure 49).

Heat Map (Figure 50).

Metric Graph (Figure 51).

Health Status (Figure 47).

Figure 48. Cluster Memory Dashboard

The Generic Scoreboard widget was configured with these metrics defined at the cluster level. The resource cluster is used for the memory metric selections.

Table 18. Generic Scoreboard Configuration

Metric	Unit	Green Range	Yellow Range	Orange Range	Red Range
Memory Granted	KB	8000	8000–12000	12000–13000	13000
Usage	%	50	50–75	75–80	80
Usage/Usable	%	50	50–75	75–80	80
Contention	%	50	50–75	75–80	80
Swap In Rate	KBps	300	300–400	400–500	500
Swap Out Rate	KBps	5	5–10	10–20	20

NoteThese threshold/ranges are only examples. These values should be based on the cluster design threshold values and the customer requirements.

Figure 49. Cluster Memory Scoreboard Widget

The Heat Map is configured with three metrics, created from the cluster resource, as follows:

Swap Out Rate (KBps).

Swap In Rate (KBps).

Mem Swap Used (Kb).

The following are settings for the Heat Map widget.

Table 19. Settings for the Heat Map Widget

Metric	Value
Attribute Kinds	Memory
Tags to Filter	Cluster compute resource (select the correct cluster)
Minimum Value	(Custom setting to environment)
Maximum Value	(Custom setting to environment)

Figure 50. Cluster Memory Heat Map Widget

The Metric Graph Widget can display line graphs of historical memory statistics. This example selects the following memory metrics to be displayed over a 24-hour period. Note that the historical time can be changed to up to one year and can also indicate the DT (dynamic threshold) values.

Table 20. Cluster Memory Metric Graph Settings

Metric

Memory | Contention (%)

Memory | Guest Active (KB)

Memory | Granted (KB)

Memory | Usage (%)

Figure 51. Cluster Memory Metric Graph Widget

8.1.5.6. Storage Dashboard

The Storage Dashboard widget can help with troubleshooting at the cluster or host level and can be used by the virtual infrastructure admins and the storage administrators. Figure 52 shows this dashboard, which is built using four widgets:

Disk Scoreboard (Figure 53).

Disk Capacity Scoreboard (Figure 54).

Cluster Storage Metric Graph (Figure 55).

Health Status (Figure 47).

This dashboard focuses on cluster statistics and datastore hot spots. Hot spots in this example are related to latency to the datastore as detected by the ESXi host. Heat maps can also be used to identify datastores with low capacity remaining, which is done using a super metric.

Figure 52. Storage Dashboard

This dashboard uses two Scoreboard widgets. The first widget displays statistics related to the cluster I/O usage. The second widget displays information about virtual memory snapshot space and virtual disk usage.

Table 21. Cluster Storage Usage Widget Settings

Metric	Unit	Green Range	Yellow Range	Orange Range	Red Range
Disk CMD Per Sec	#	200	200–300	300–400	400
Disk CMD Latency	ms	20	20–30	30–40	40
Disk Read Latency	ms	20	20–30	30–40	40
Disk Write Latency	ms	20	20–30	30–40	40
Disk Usage	KBps	10000	1000–2000	2000–3000	3000
Disk Workload	%	50	50–60	60–70	70

NoteThese threshold/ranges are only examples. These values should be based on the cluster design threshold values and the customer requirements.

Figure 53. Cluster Disk Scoreboard Widget

The second generic scoreboard widget displays virtual machine disk space information. This example shows how much snapshot space is used and the total amount of virtual disk space used.

Table 22. Cluster Disk Capacity Scoreboard Settings

Metric	Unit	Green Range	Yellow Range	Orange Range	Red Range
Virtual Machine Snapshot Space	GB	1	1–10	10–20	20
Virtual Disk Usage	GB	90	90–200	200–500	500

NoteThese threshold/ranges are only examples. These values should be based on the cluster design threshold values and the customer requirements.

Figure 54. Cluster Disk Capacity Scoreboard Widget

The Metric Graph widget displays historical disk statistics that include the DT values. The DT values are also displayed for the next number of hours, depending on the Date Control selection). Figure 55 shows the Anomalies view enabled (all of the yellow spikes). For the example we are displaying historical stats for the following metrics.

Table 23. Example Metrics

Metric	Unit
Disk	Commands per second
Disk	Disk commands latency (ms)
Disk	Usage rate (KBps)
Disk	Disk write latency (ms)
Disk	I/O usage capacity

Figure 55. Cluster Storage Metric Graph Widget