In this final part i will explain how to create dashboards and widgets to display the data of the Metric. A dashboard is a single window inside the vCOps UI and can contain multiple widgets that display the data in various ways. For instance Graphs, Heatmaps or Badges are possible, but you also can show your data on a geo map to visualize locations. Some of the widgets can interact with others. I will show that for a Heatmap that feeds a metric graph.
Table of Contents
Select the Dashboards from the menu. It will show you all your defined dashboards and also the shared dashboards from other users. Right next the last dashboards you find the small Plus button. With that button you create a new dashboard.
Define “Group by Adapter Instance” to sort your vCenter by name in the wigdet view. Next choose the resource kind vCenter Server and Size by “Number of Running VMs”. For the color choose the created Super Metric. In the top color bar define 5 as the minimum and 10 as maximum. This configuration will a square for every vCenter resource. The size is defined by the number of running VMs in the vCenter and the color goes from green to red if the average %READY is more then 5% for all VMs in that vCenter. Normally you should choose zero and five as min and max, because of more then 5% ready times lead to a poor performance inside a VM. But i have choosen higher values to demonstrate the work of the colors. Next press the small Plus button on the top right to create a new configuration. Give it a name and save it.
The dashboards now displays a heatmap with a square for every vCenter. One of the bigger vCenter of the left side seems to have a problem with high READY times. That demonstrates the power of a heatmap. With one view you see in which part of your environment you have potential problems. If you hover the mouse over a square you see the values for both metrics (Number of running VMs and the avg. ready percentage). This one is a quite large vCenter a above the normal ready percentage.
The drawback of heatmaps is that they only show the current snapshot of the situation. Thats why i have added the metric graph below. It can display the history for READY times of a vCenter. It is still empty, but with Interactions on top of the widget we can change that. Select the heatmap as the providing widget for the metric graph and save it.
If you click on a square it will not only show the metric details as before, but also show you the history of the metric in the graph below. You can even select multiple vCenter to compare the graphs.
As you can see Super Metrics can be very handy to answers questions that require to combine several existing values. Another good example for that use case is the calculation of the provisioned space of a datastore. That metric is also not accessible within the vSphere SDK, but can be calculated. Maybe this is something for another article…
Many thanks to David from VMware for his great vCenter Operations Workshop and explaining me how the Super Metrics work!