Tuesday, June 17, 2014

vCenter Operations Manager - Capacity Planning

Most companies have a high reliance on the operations infrastructure that they have created. Outages of phone systems, e-mail, customer facing applications, and ERP systems can cripple a company.

Capacity management includes establishing and maintaining a safe and reliable amount of resources to meet the business demand. Demand management is an important component of providing reliable computing services. It requires a variety of non-IT oriented skills and knowledge, and is therefore an often-neglected area. However, it is becoming significantly more important as virtual machine growth weighs down IT staffing ratios and infrastructure budgets. For large organizations, most capacity management is done with spreadsheets; fortunately vCenter Operations Manager provides information to better account for the needed resources and capacity.

Enhancing data for capacity in vCenter Operations Manager involves finding opportunities for resource optimization and cost savings. On the Planning tab of VMware vCenter Operations Manager, you can monitor the use or resources and the available capacity in your virtual environment, and plan for capacity upgrading or optimization. On the Reports tab, you can create a report to capture the details relates to current or predicted resource needs and export that report to a file.

The Summary view is context sensitive to the object you have selected in the inventory tree. In the image above, I have selected the World object and in the Objects panel it is showing the capacity remaining in my environment for clusters, hosts, and virtual machines. This is typically going to show a downward trajectory, whereas if you select Deployed from the Perspective you should an upward trajectory. The Extended Forecast is going to give you the prediction for the amount of capacity remaining and the time remaining.

For the next month, I have enough capacity based on the existing trend; however next quarter I will need to add resource capacity to meet the virtual machine growth demand. The time remaining for the current capacity for virtual machines is 57 days.

When we switch to the Deployed perspective, we can see what is driving the estimates. The average growth trend is a few virtual machines each week, which running on my Macbook Pro is a substantial amount of future capacity.

The Resources pane provides a detailed view into the capacity of the data center.

Different settings such as Source (Virtual or Physical), Aggregation, Perspective, and Unite give a significant amount of visibility into the environment. Views of various time frames are available; these time settings provide additional information like how much time each resource has available before it runs out of capacity.

For capacity planning to be effective, we need to know what the constraint is in the environment that we need to address and the time remaining before we are out of capacity. The Resources panel shows us that my resource constraint is host memory, I have 57 days remaining before I run out of memory resources based on the current utilization trend.

The aggregation settings allow you to set filters for:
  • Sum: add all the values
  • Average by Cluster: shows the average value per cluster
  • Average by Host: shows the average value based on the average host
  • Average by VM: shows the average value for one average virtual machine
Values get smaller at each aggregation layer. While the Sum is the overall total, a cluster is larger than a host or a virtual machine. The aggregation level depends on the task at hand. For example, if you were considering expanding your hosts to identify whether you need to upgrade existing physical servers or add a new physical server, you would look at data for the host. In contrast, if you want to add more virtual machines, you would need to get more information about your average virtual machine.

The Perspective filter is used to display:
  • How much is used, consumed, or deployed
  • How much resource is left remaining
  • Total capacity (used and remaining)
You can display the results in the Actual Value or a Percentage using the Unit setting. 

Next, we are going to take a look at the Views tab. Views are organized by badges; there are six categories, one for each minor badge under Risk and Efficiency (Time Remaining - Capacity - Stress - Waste - Density), and All Views.

If we look at the above diagram, we can see that Memory is the resource constraint for my Gold Cluster. I have 56 days remaining, with an estimated capacity of an additional 13 VMs on the cluster. You will notice that I am using the default policy configuration. Clicking on the Change Default Policy Configuration hyperlink can modify this.

I have modified my Usable Capacity Rules to be a little more conservative; I like having 85% overhead to ensure there is no business impact during a host outage. After the modification, you will see that it shows up in the View information under Capacity Buffer Limits.

The Average Virtual Machine Capacity gives you the trend line for the average deployed virtual machines, the average virtual machine capacity remaining, and the average powered-on virtual machines. The interval for this setting can be extended out to monthly, quarterly, or yearly.

An important part of understanding capacity trending is being able to forecast changes to the environment. That is where What-If scenarios come into play. When you launch the What-If scenario wizard, you are able to model adding virtual machines, hosts, or data stores. For instance, if you wanted to see the trend for adding 10 more virtual instances with a specific profile, vCenter Operations Manager will determine the average deployed virtual machines, the average virtual machine capacity remaining, and the average powered-on virtual machines after they have been deployed. They will show up as dotted lines on the chart.

An important part of operations capacity management is helping IT leadership understand the trade-off between the business demand and cost. For instance, being able to trend the impact of adding those 10 additional virtual machines in our Gold Cluster may result in having to purchase additional server resources or an additional server host. Coupled with a charge-back model and support SLAs, business leaders can then decide if the new application or project should go into our Gold Cluster or if it should go into a lower tier. They may decide the project is going to be on hold until the next budget cycle.

vCenter Operations Management can provide valuable insights into capacity planning and trending into your data center environment, and help ensure you have the required resources to meet the business demand.
News: Top vBlog 2016 Trending: DRS Advanced Settings