A few weeks ago, I introduced you to vKernel’s Capacity View, a free tool that provides you with an aggregate view of the status of your VMware environment. vKernel’s Capacity View is not intended to provide you with an in-depth analysis of your infrastructure; the tool is designed to do two things:

  • Provide you with an understanding of general areas that might need attention, such as storage performance or RAM capacity;
  • Entice you to want more information and make you willing to purchase vKernel’s full-featured Capacity Analyzer software.

Capacity Analyzer

Capacity Analyzer truly takes vKernel’s monitoring software to the next level by providing detailed analytics for every aspect of your virtual environment, from the host down to the individual components that make up a virtual machine.

The Capacity Analyzer main Overview page (Figure A) provides you with an at-a-glance look at current infrastructure bottlenecks along with the constraining resource. For example, if a virtual machine is getting low on RAM, Capacity Analyzer will indicate that RAM is the constraining resource.
Figure A

vKernel Capacity Analyzer Overview page. (Click the image to enlarge.)

On the Current Bottlenecks page, you’ll get an expanded overview of where you’re suffering resource constraints. In Figure B, note that a number of resources are outlined, including CPU usage, RAM utilization, and storage, as well as storage I/O. In this example, the Storage I/O column is not populated because these stats are not being captured yet. Also note that you can mouse over the various icons to get an expanded look at the status of a particular resource. I moused over the red triangle in the CPU box to get a better understanding of what condition created the red triangle.
Figure B

Current resource bottlenecks

When it comes to bottlenecks, it’s important to understand if the condition is a one-off or if it is a recurring condition. To get a better understanding, Capacity Analyzer provides you with a historical view of resources, including the RAM history you see in Figure C. As you can see from this graph for this particular virtual machine, RAM usage has stayed consistently in the yellow range this week, although it dipped briefly into green.
Figure C

Historical view of resources. (Click the image to enlarge.)

Although this historical look is great information, what steps do you need to take to contain the problem? In the case of the server named MGMT2, the yellow condition is simply a warning, but left unaddressed, it could create a problem. Capacity Analyzer takes the analytics a step further by providing proscriptive measures that you can take to correct the problem. This remediation information is based on best practices data. In Figure D, here’s what vKernel has to say about the RAM condition you saw in Figure C.
Figure D

Remediation recommendation, (Click the image to enlarge.)

RAM is certainly an important resource, but equally important is disk utilization inside a virtual machine. Capacity Analyzer also covers this resource as you see in Figure E, which is a list of where each individual disk for each virtual machine currently stands with regard to utilization. Throughout Capacity Analyzer, you will see the little arrows next to each resource. The direction of this arrow shows you in what direction a particular resource usage is trending. In the case of the server named MGMT2, both virtual disks are at 99% capacity, but the growth line is flat. Obviously, 99% full isn’t a great place to be, so steps will be taken to bring this capacity level down a bit.
Figure E

Where are you experiencing storage capacity issues? (Click the image to enlarge.)

Capacity Analyzer performs another critical function: It predicts how many more virtual machines can be added to each host and notes what resource is inhibiting the addition of even more virtual machines. In Figure F, you’ll see that we can add three more hosts to our server cluster with the constraining resource being memory. We actually have a few virtual machines that have too much RAM provisioned, so we’ll be able to take steps to free up some of that RAM; this will help us fully optimize our virtual environment.
Figure F

How many more virtual machines can be added to this environment? (Click the image to enlarge.)

Each virtual machine also enjoys its own full overview, which includes current bottlenecks and future bottlenecks. Future bottlenecks are based on historical growth trends and help you to take measures to avoid performance problems. In Figure G, note that the virtual machine named Angel_Vm has RAM and disk capacity alerts. In the Future Bottlenecks section, you’ll see a list of resources related to Angel_Vm that are predicted to face bottlenecks and at what point it’s expected that the resources will face these bottlenecks.
Figure G

Future bottlenecks. (Click the image to enlarge.)

An alternative view of the future is shown in Figure H, which outlines the RAM-based bottleneck that is expected to be suffered by the cluster in 23 days as a result of RAM growth on a number of virtual machines eventually outstripping the host’s resources. Thirteen remediation recommendations are made, starting with adjusting the RAM on the virtual machine named WebFS.
Figure H

An alternative look at the future. (Click the image to enlarge.)

The bottom line

At a list price of $299 per socket, Capacity Analyzer can easily pay for itself through cost avoidance. That is, when used right, Capacity Analyzer will help you to optimize existing resources and help you to avoid the need to purchase additional hardware to meet growing infrastructure needs. vKernel offers a fully functional 14 day trial version of Capacity Analyzer if you want to try out.

Want to keep up with Scott Lowe’s posts on TechRepublic?