What is Capacity Management?
Capacity management is the practice of making sure IT resources meet business demands today and down the road—without over-provisioning.
But the role of capacity management has changed as IT environments have evolved. IT organizations everywhere have gone from physical systems to virtual systems and now into the cloud. And that impacts the way they approach capacity management.
That’s why Tom Huntington and I presented a webinar about the basics of capacity management, why you should do it, and how you’ll benefit from it.
Prefer reading to watching? Keep on reading for a full recap of the educational webinar below.
Capacity Management 101
Capacity management starts with a simple enough concept.
Over time, you’re going to have an increased demand for capacity at your organization. You’ll need some sort of sourcing mechanism to meet with capacity demands. In worst case scenarios, new capacity needs are discovered through incidents.
For some organizations, this means using simple rules of thumb to determine when to add capacity.
Since many companies don’t trust their ability to predict capacity, they deliberately overprovision. That can solve the capacity problem. But it gets very expensive.
So, the purpose of capacity management today is two-fold. You need to get sufficient capacity. And you need to avoid overspending.
How Do You Manage Capacity Today?
We polled the attendees of the webinar to find out how they’re managing capacity today. Most were using simple rules of thumb, like percentages of utilization. Some were doing capacity modeling based on demand. Others had no tools or processes in place. And a few were doing linear trending.
Where does your organization fall?
7 Drivers for Capacity Management
There are seven main drivers for capacity management today.
1. Business Critical Services
These are services you can’t afford to fail—or that you can’t afford to jeopardize the quality of. Capacity management can help you make sure your business critical services are always available.
2. Seasonality and Peaks in Business Activity
If your business has seasonal behavior—like retail—you need to manage those peaks in demand. Some examples of these just happened in the US—Black Friday and Cyber Monday.
On the flip side, you also need to know when those peaks wear off and adjust your provisioning accordingly.
Capacity management can help you ensure you have the right capacity available at the right times.
3. Business Growth
Do you have an expected period of business growth? Capacity management can help you make sure you have enough capacity available to meet the demand of your growing business.
4. Regulatory Requirements
Government regulations might require your business to provide a continuity plan for IT. Capacity management can help you meet those requirements and make sure your service stays up and running.
5. Provisioning Lead Times
Your operational model might cause you to have long provisioning lead times. Using capacity management can help you make sure you’ll be able to add capacity at the appropriate time—without worrying about missing the lead time.
6. Cost Optimization
You need to make the most of your budget and infrastructure. And capacity management can help you make sure you're finding cost savings through better planning and reclaiming unused resources.
Your IT department needs to be able to respond to opportunities or threats. By doing capacity management, you’ll have control over your current capacity situation.
Benefits of Capacity Management
There are three core benefits of capacity management.
Capacity management helps you improve efficiency by rightsizing IT resources to meet your needs.
Capacity management helps you avoid risks by pinpointing problems before they become problems.
Capacity management gives you a good handle on where you are now, so you’ll be better able to be flexible and exploit new opportunities.
How Do You Do Capacity Planning?
There are three main levels of capacity planning.
First, there’s the component level. You need to understand how your components are being used. And that means you need to be monitoring those components. This can give you quality improvement in terms of warnings and advice.
Next, there’s the service level. You need to be able to combine your components into services that make sense to the business. This will help you with SLA compliance—and it helps you increase your efficiency.
Finally, there’s the business level. This means business demands and making demand forecasts. Getting to this level is important to be able to cope with new things that happen in the market.
Where Does the Data Come From?
Capacity management requires data.
You probably already have monitoring set up and data collectors in place. Our capacity management solutions take a data agnostic approach. So, we don’t need to interrupt those.
We can take data from pretty much any data source—including public cloud service providers. And we can use that data to help you meet SLAs and keep infrastructure costs in check.
Here are some examples of data sources our solutions work with:
- AWS CloudWatch for enterprise capacity reporting
- Intermapper for network utilization and performance monitoring
- VMware vCenter and other virtualized systems
- Physical systems using Tivoli and SCOM
- ServiceNow for service level reporting
Doing Predictive and Prescriptive Analytics
The real power of capacity management is in predictive and prescriptive analytics.
This means answering questions like…
We expect increased demand in the third quarter due to a marketing campaign. How will that impact infrastructure supporting service XYZ?
Are there any lurking capacity issues in my data center that I should address?
I’m migrating a set of workloads from legacy systems to our new standard platform. How do I validate my strategy?
How to Do It
The traditional approach is taking a few metrics and focusing on utilization. You try to set appropriate thresholds for each metric type. And you use linear trends to forecast future behavior.
But this approach has flaws and limitations. In order for this to work, you’d need to evaluate each workload individually. That’s a manual task that typically creates bottlenecks. And you’ll wind up with false negatives using this approach. All that noise will make you disregard the important information.
A better approach is analytical modeling. It’s based on the same empirical monitoring data. But you take that data and run it through an analytical engine, create a queuing model, and use it to evaluate what-if scenarios.
This helps you evaluate the performance of a system in a controlled fashion. And it helps you take nonlinear information into consideration. You’ll be able to find out how far you actually are from a problem.
Capacity Planning Using Predictive Analytics
You have two different options for capacity planning with prescriptive analytics: scenario-oriented in-depth planning and enterprise-wide risk mitigation.
Scenario-oriented in-depth planning is good for predicting the impact of things like:
- Changes to transaction intensity or user concurrency
- Migration of VMs, workloads, and containers across platforms (including physical, virtual, container, and cloud)
- Reconfiguration of infrastructure or hypervisors
Enterprise-wide risk mitigation is good for assessing current health and predicting future risks. You’ll discover:
- Expected days to noncompliance
- Recommendations for how to improve health or mitigate risk
Common What-If Scenarios:
There are five common what-if scenarios to consider for capacity management:
- Plan for an increase in demand
- Add more virtual machines
- Balance loads across a tier of servers
- Deploy a new application
- Upgrade or replace existing hardware
When asked what scenarios they’d use capacity management for, our live webinar audience was equally split between:
- Planning for changes in customer demand
- Planning for moving workloads to new servers
- Planning for additional workloads on existing infrastructure
Capacity Management in the Cloud
There are some differences when it comes to doing capacity management in the cloud. Primarily, the difference is that the focus shifts to making sure you have the lowest possible cost, since capacity is infinite in the cloud.
A capacity management solution in the cloud should:
- Assist in the migration and deployment of workloads
- Monitor ongoing resource usage and charge
- Provide end-to- end resource visibility for hybrid IT
- Optimize and predict financial impact
- Do long-term planning to meet business demands
Learn more about doing capacity management in the cloud. Watch the webinar >
Getting Started with Capacity Management
For capacity management to be a success at your organization, it needs to be a strategic initiative. Implementing a tool will only get you so far.
Here are some of the results our customers achieve by making capacity management strategic:
- $10 million in reclaimed data center assets (telco)
- 94 percent decrease in application slowdowns in four months (healthcare)
- $12 million saved in hardware in 15 months (insurance)
- 730 transactions/second—without disruptions (finance)
- 25 percent reduction in IT budget in three years (insurance)
- Work volume quadrupled without adding staff (finance)
- 95 percent accuracy in predicting capacity needs (healthcare)