Paradise by the (vROps) dashboard light

There seems to be a misapprehension that using and customizing vRealize Operations Manager is difficult and requires professional assistance. That’s not necessarily true. While a professional services engagement is great, especially for larger deployments, it’s not necessary to get value out of Operations Manager. vROps comes with a lot of awesome pre-built dashboards and reports to get you started plus it’s really easy to customize.

Monitoring IT infrastructure, in general, is a challenging thing to stand up. Reactive organizations may just use simple ICMP monitoring that triggers late night pages after something has gone wrong. Others may be trying to figure out what to monitor via trial and error.  Every IT organization wants to be proactive. It cuts down on the number of late night pages and makes for happier customers. But few are able to give monitoring the priority and time that it deserves to reach that state.

VMware has set out to change all of that with vRealize Operations. This is perhaps the most customizable monitoring and reporting solution I’ve ever used. vROps is capable of pulling in far more VMware metrics than any other solution, plus it can monitor your hardware and applications with management packs available in the VMware Solution Exchange and third parties like Blue Medora.

Out of the box, vRealize Operations Manager includes some very handy dashboards. Take for example the vSphere Datastores dashboard. Included in this dashboard is a heatmap which helps identify datastores with high latency. Heat maps are an awesome way to visualize this type of data, but what if you wanted to dive in deeper and see what VM’s are impacted with as few clicks as possible?

 

This isn’t a problem when your dashboard is nice and green like the one above, but unless you’re using vROps daily to identify and resolve issues, your dashboard may look more like this.

 

How do you prioritize? Normally you’d resolve issues impacting your most critical workloads first, but identifying them could be time-consuming. What if you could see it all in a single view? If you’re running the advanced or enterprise license you absolutely can!

The advanced and enterprise licenses of vROps allow you to create your own custom dashboards and reports. It’s pretty easy to do. Let me walk you through it.

Let’s build a vROps dashboard!!

To create your own custom vROps dashboard, click on Content in the menu on the left side of the home screen.

 

Select Dashboards and click Create Dashboard.

 

Let’s give this dashboard a clever name like “Datastore IOPS Heatmap.” That has a nice ring to it!

 

Click on Widget List on the left hand menu.

 

Drag the heatmap from the widget list to our new dashboard.

 

Let’s add a happy little object list that we can use to display the affected VM’s.

 

Widget configuration

That’s it for widgets at the moment. We just need to configure the heatmap and object to display what we want. Click the pencil icon on the top right corner of the heatmap to edit it.

 

Give the heatmap a name and description and group by vCenter Adapter –> Cluster Compute Resource. This will group your datastores by host clusters but you could group by datacenters or datastore clusters as well.

 

Object type is vCenter Adapter –> Datastore

 

Size by Capacity –> Total Capacity (GB)

 

Color by Datastore I/O –> Disk Command Latency (ms)

 

I chose to set a Max Value of 25ms. If you leave this blank, then it will autoscale which means your worst datastores will appear red even if there are no significant latency issues.

 

Click Save and move on to editing the object list.

Click the pencil above the object list.

 

Name the object list and set the mode to Parent. This tells the object list to show the datastore’s parent objects such as Virtual Machines.

 

 

Filter by Object Types –> Virtual Machine

 

So far we have told the object list to show the VM’s. But what if we wanted to see more information such as disk latency per VM? It’s really simple! Click on the Additional Column bar on the far right side of the screen.

Click Pick Metrics

 

Select vCenter Adapter as the adapter type.

 

Under Metric Picker select:

• Virtual Disk
• Then select Commands per Second
• Read Latency (ms)
• Write Latency (ms)

 

Click OK and then Save.

Configuring widget interactions

Because we want the heatmap to feed the Object list, we need to build the interaction between the two widgets.

Click on Widget Interactions on the left side of your screen.

 

Leave the first drop down menu empty because the heatmap isn’t being fed from any other widgets.

Select Datastore IOPS Heatmap in the second drop down menu. This tells vROps that we want whatever is selected in the heatmap to feed the object list.

Click Apply Interactions.

 

Click Save.

That’s it! You’ve now created a custom dashboard!

View your custom dashboard

Go Home and select your new dashboard from the dashboard list.

 

Here it is! Your custom dashboard!!! We can now see what VM’s are on any given datastore just by clicking on a square on the heatmap as well as each VM’s disk latency.