Tuesday, April 8, 2014

Part 1 - Creating a Cluster Utilization Dashboard in vCOps Custom User Interface!

A few days back I wrote a post on creating a One-Click Cluster Capacity Dashboard which was liked by a number of people in the virtual world. In-fact it was so easy to download and replicate that within a few hours of rolling out that post, I could see a number of people coming back to me stating that they have deployed the Dashboard and have started looking at the Capacity numbers for their infrastructure. Many of them asked as to how can they create their own XMLs for creating similar dashboards. I wrote another post in a few days which spoke about Creating your own XMLs for Custom Dashboards and got some great feedback on that as well.

Doing this is so much fun, however more often it is difficult to think of the various problems you can solve with these Custom Dashboards. With my continuous engagements with the customers on the field, I get some fancy requests which keeps my brain working on digging out possibilities with vCenter Operations Manager. 

In my last engagement with a customer, I installed the Capacity Dashboard for him which I have already shared with you guys before. The immediate reaction of the customer after looking at the dashboard was to have something similar for his environment which could help him gauge the UTILIZATION of each cluster with a single click. This includes, CPU, Memory & Disk Utilization. The client was very clear with what he needs and that helped me create the following dashboard.

I will do this Dashboard in a 3 Part Series.

Part 1 - Introduction to the Cluster Utilization Dashboard (This post)
Part 2 - Creating the required Supermetrics for the Cluster Utilization Dashboard
Part 3 - Bringing it all together.

NOTE - This dashboard uses both Regular vCOps Derived Metrics and SuperMetrics. I have marked the Scorecard boxes which are using Supermetrics with the initials "SM <Number>". We would need to create these SuperMetrics before we could use the .XML file and import the Dashboard as we did previously. 

Let us first look at the features & use cases for this dashboard and then we would proceed with the steps on how to build the each component of this dashboard.


1 - Gives you a single page view of the Utilization of all your vSphere Clusters being monitored by vCenter Operations Manager. (One of Multiple vCenter Servers). Click on a cluster on the left pane for which you want to know the UTILIZATION data and on the right pane you will see the same.

2- The 1st row gives you an overview on how many VMs can be deployed in a cluster based on the vCOps average profile, how many are deployed and how many more can be deployed based on the deployment trends.

3- Then you have 3 rows of CPU , Memory & Storage Total Capacity,  Allocated Capacity, Usage, Usage in percentage & finally the number of VMs remaining keeping a resource in mind such as CPU, Memory & Storage.

4- It is important to note that vCOps uses local storage if available on the ESXi hosts, hence the storage stats might be a bit skewed. You can create groups within vCOps to isolate local storage from SAN cluster wise but then it will not allow you to have a dynamic dashboard such as this. You will have to then create a separate dashboard for each of your cluster & create a new one every time you add a cluster. But with THIS cluster utilization dashboard everything is automatic and dynamic ;-)

5- This is the most asked Dashboard by CxO's as they want a one click summary of what they have, how much they have used and what is remaining. This is the LARGEST use case hands down.

6- Their are last 30 days trend-lines as well for the specific metric which makes it a bit more informative.


I wrote another post on vCenter Operations Manager some time back where I have showcased how to create Supermetrics. In this post, I will show you how to create one of the Supermetric through screenshots and then give you the formula for rest of the super metrics which you can create yourself following the example. If you closely look at the screenshot of the dashboard, you will notice that this dashboard displays 20 metrics in total. 8 out of 20 are Supermetrics. The table below shows what metrics & Supermetrics are being displayed.

Need less to say that the one in RED are the Supermetrics and the others are derived metrics. Before we bring this all together, let me share the steps to create a Supermetric, let's say SM 1

I will close this part right now and in PART 2, I will explain as to how we need to create all these Supermetrics. That should be out by tomorrow so stay tuned :-)

***Share & Spread the Knowledge***

1 comment: