Wednesday, May 22, 2013

Showcasing Virtual Datacenter Capacity using vCenter Operations Manager Dashboards.

A few months back, I wrote about the capabilities of vCenter Operations Manager around Capacity Management in this article - vCenter Operations Manager - Solving Performance, Capacity and Configuration Problems!!. With this article I wanted to take a step further and showcase how you can use vCOps in a real world scenario to monitor the capacity of your Virtual Datacenter in an automated manner.

If you read my last 2 articles on vCenter Operations Manager, you would notice that custom dashboards in vCenter Operations is an amazing way to showcase a lot of important data which is collected,analysed & computed by vCOps Patented Analytics and Capacity engine. In this article I will share a Custom Dashboard which in a single pane can showcase multiple facets of Capacity & Consolidation in your Virtual Datacenter. 

To begin with let's have a look at the dashboard and then we will break down the discussion on how to create it using the custom dashboard widgets.

Executive Capacity Dashboard

Well that is colorful, isn't it. As complex as it looks, if your approach towards creating this dashboard is correct, this should be a matter of few minutes. I will not go into step by step of how I created it, I would rather ask you to get your hands dirty and learn how to create such dashboards.

Here are a few pointers on this dashboard, which will help you create one on your own:-

a) The over all dashboard is a 2 column and 8 widget dashboard.
b) Each widget used is a Score Board Widget.
c) No super-metric used here. Each widget uses metrics which are already populated by default vCOps collection.

Let's look at each part horizontally.

Part 1 - Datacenter Capacity and Consolidation Ratios

Here the widget on the left uses a scoreboard widget to show the capacity of 2 data-centers  DC 1 and DC2.  I have used the Datacenter as the Resource tags here shown the "Summary" of each Datacenter such as number of Clusters, Datastores, ESXi Hosts and Virtual machines in a DC.

The widget on the right uses the scoreboard widget to show the consolidation ratios of both the virtual data-centers by using the Datacenter as the resource tag and  "Density" ratios to populate the following vCPU to pCPU, VMs to Hosts and vMem to pMem. This can give you the density of your Datacenter in a snap.



Part 2 - Capacity & Time Remaining in Each DC/Cluster 
This takes the capacity information to the next level. This section talks about, how many VMs can be deployed, how many VMs are deployed currently and the number of days remaining for VMs to be deployed (from left to right). I have done this for one cluster in each Datacenter as this is what I have in my lab. I used Score Board Widgets, Datacenter/Cluster as Resource Tags and "Capacity Remaining", "Time Remaining" and "Summary" as the metric providers for these scoreboards.

Note:- The number of days are a weird looking number since there has been no deployment in this cluster after installing vCOps. vCOps does not have a deployment pattern in this case, using which it can suggest you the approximate amount of days remaining.


Part 3 - Resource Wise Capacity Remaining

This part also uses 2 scoreboard widgets for each Datacenter/Cluster and this breaks down the part 2 of this dashboard further down to tell which which is the most constraining resource due to which you have given number of virtual machines which you can deploy in your Datacenter/Cluster. Again I have selected Datacenter as the "Resource Tag" and used "Capacity Remaining" resource wise such as CPU, Disk Space, Disk I/O, Memory and Network and showcased, ho many VMs you can deploy as per each individual resource.



Part 4 - Resource Wise Time Remaining







Coming to the last part of the dashboard. This again is made by using the Scoreboard widget and tells you how many days would a resource last with your deployment pattern of creating new Virtual Machines in each Datacenter/Cluster. Again I have selected Datacenter as the "Resource Tag" and used "Time Remaining" resource wise such as CPU, Disk Space, Disk I/O, Memory and Network and showcased, how many VMs you can deploy as per each individual resource.

Note:- The number of days are a weird looking number since there has been no deployment in this cluster after installing vCOps. vCOps does not have a deployment pattern in this case, using which it can suggest you the approximate amount of days remaining.

Having said this before, it is your creativity & need which can take you to heights with vCOps customization. I prefer this way of dash-boarding rather than seeing what is available out of the box in a lot of other products, as this gives the power to create in my hands, rather that being dependent on the manufacturer of the product.

I will close this article now, feel free to reach out if you have questions around the discussed topic. 


Don't forget to Share and Spread the Knowledge. 


Wednesday, May 1, 2013

Part 2 - Using vCOps Super Metrics & Custom Dashboards to Display Datastore Usage % !!

This post speaks about creating a Custom Dashboard in vCOps using Super Metrics. Please read  the Part 1 of this series to ensure that you know the context.

Here is the link - 

Part 1 - Using vCOps Super Metrics & Custom Dashboards to Display Datastore Usage % !!



Now since we have the Super Metric from the previous post, we will go ahead and use it to create a Custom Dashboard which reflects this metric and helps you to monitor the Datastore Usage %. Here is how you do it:-

1- Login to vCOps Custom UI with a user with appropriate access rights.

2- Click on Dashboards -> Add









3 - Drag and Drop a Heat-map widget from the left pane to the right pane. (To know more about custom dashboards in vCOps, read this Guide.)

4- Give a Name to the Dashboard, Select the number of Columns as 1 and click on OK. (refer to screen shot below.)




















5- Now lets configure this widget. Re-size the widget as per your viewing requirement and then click on Edit widget to configure the heat map.















6- Here is how you will configure the Heatmap.






















7- And once you click on OK, you have you Heat Map which you can start monitoring to ensure you never run out of space on your datastores.




















You can hover over these objects and you will see 2 metrics for each object - Total Capacity & Usage in %age

Here you can clearly see that some of my datastores in my primary DC are actually reaching the 100% usage  and I have some work to do to ensure I have enough free space to keep them green. Since this heat map is auto refreshing, it will ensure that your datastores are continuously monitored for the thresholds you have set for utilizing space on datastores.

If you want to see me present other tips and tricks on vCOps at VMworld 2013, please vote the session which I have submitted. 









For more details refer to - VMworld 2013 - Call for Papers Voting is Now Open. Cast your vote to see your favorite sessions!!


Don't forget to Share and Spread the Knowledge. This will help others!!

Part 1 - Using vCOps Super Metrics & Custom Dashboards to Display Datastore Usage % !!

For the past few weeks, I have been working on vCenter Operations Manager as a part of a project. As always, I am here to share my experience around the work I have been doing and hopefully help the larger VMware community.

As we all know, one of the basic rule of thumb is to ensure that you have a certain percentage of space always available on your datastores to ensure that you cater to the need of all the virtual machines which are a part of these datastores. For example, you would always want your datastores to be around 10% free to ensure that you have enough space for snapshots or any other unplanned requirement. At the same time, you would want to keep an eye what percentage of space is available in your datastores so that you can provision VMs accordingly, of-course I am assuming that you are not using Storage DRS ;-)

Okay so let's quickly jump into action with our use case. We will use vCOps custom dashboard to create a heat-map which would change colors according to free space remaining and would help you to monitor and plan things accordingly in your Virtual Datacenter.

Since, vCOps pulls all the metrics from vCenter, it will be able to pull the Datastore Total Capacity and Used space. Using these 2 metrics, we would create a super metric which would give us the usage %. Using simple math:

Datastore Used % = (Used Capacity/Total Capacity)*100

So how do we create this super metric.

Note:- For someone new to super metric concept, I would recommend you read this white paper from VMware.

1- Login to vCOps Custom UI with a user with appropriate access rights.

2- Click on Environment -> Super Metrics










3- Click on the Add New icon on this window to create a new super metric.











4-  Configure your super metric as shown in figure below



Your super metric will now reflect in the list of available super metrics. Now we need to create a super metric package to attach the same with all the datastores, so that the system automatically can calculate the usage percentage for every datastore which exists or will be added to vCOps, in the future for monitoring. Here is how you create a package and attach it.

1- On the same screen click on -> Super Metric Package -> Add Super Metric Package.










2-  Give it a name and select the super metric which we create above. Refer to the screen shot below.

























3- Put a Check-mark against the given thresholds to make this super metric a Key Performance Indicator (KPI) - I just left the setting to default and clicked on OK.

Now we need to add this Super Metric to all the datastores. This will enable vCOps to calculate this new metric for all the datastores which exist and would be added in the future. This can be done by adding this Super Metric will the Resource Kinds - Datastore. Let's see how we will do it.

1- Click on Environment -> Configuration -> Resource Kind Defaults












2- On this Screen, select the Adapter Kind as VMware Adapter, Select Datastore as the Resource on the left pane and under super metric package select the super metric package which you created before.

























UPDATE 19th June 2013 - At this point if you want to collect this super-metric on the existing datastores as well, you need to Click on Environment -> Environment Overview -> Click to Expand Resource Kinds -> Select Datastores.

This will list all the Datastores in the right pane - Select All the Datastores -> Click on the small Pencil on the top to edit all these resources -> In the new window, put a check against Supermetric and then select the Datastore Usage% supermetric-> Click on Ok to save.

Well now you need to wait for the next collection and  performance calculation cycle which is usually 5 minutes. In another 5 minutes we will be able to use this super metric for creating custom dashboards.

We now create a custom dashboard to display this Metric in Part 2.


If you want to see me present other tips and tricks on vCOps at VMworld 2013, please vote the session which I have submitted. 









For more details refer to - VMworld 2013 - Call for Papers Voting is Now Open. Cast your vote to see your favorite sessions!!


Don't forget to Share and Spread the Knowledge. This will help others!!