Capacity planning is defined as “the process of determining the production capacity needed by an organization to meet changing demands for its products. In the context of capacity planning, design capacity is the maximum amount of work that an organization is capable of completing in a given period.”  As it relates to IT organizations, this process has traditionally been performed for the server environment which is designed to anticipate processing and memory requirements to support the business applications needed for an enterprise.  Capacity planning for the storage environment is more complex and oftentimes less scientific because the information needed to make informed decisions is not in a readily usable form.


Capacity planning for the storage environment is a multi-dimensional challenge which is why most IT departments either over estimate their needs or make simplistic estimates.  This results in either over spending or worse – large unplanned purchases due to inaccurate forecasting.  There are certain aspects that need to be considered when capacity planning, such as:

  • Storage Growth
    • Types of storage needed to meet planned future requirements.
      • High performance storage
      • High capacity (lower cost) storage
  • Existing Organic Growth Trends
  • New Planned Application Needs
    • Tier 1 needs and Backup needs
  • Performance Capacity
    • I/O Workload Limitations
    • Response time requirements
  • Availability Considerations
    • High Availability Requirements
  • Additional Space Assumptions
    • Deduplication/Compression Usage
    • % of free space needed to meet performance expectations

These are just a few of the items to consider when developing a capacity plan for your storage environment.  Most of the time clients track storage usage and growth trends at a high level (typically at an array level) with no application level detail. This is because the application level of detail is unavailable or too difficult to gather.  Missing application data causes customers to not see high vs. low growth trends and can result in a significant capacity shortfall depending on specific growth patterns.

Today’s Typical Storage Capacity Plan

With todays typical storage plan using the array-based model and organic growth at the array level being the key metric for estimating future requirements, one element almost never considered is, “when has the array hit its performance limit.” The reason this is almost never considered is because identifying the performance capacity of the array requires some analysis by the customer to set thresholds – there are no values you can just use like there is for the other capacity planning resource categories.

The chart included here shows a single growth curve which communicates no storage upgrade is needed for this array in the foreseeable future.   However, this chart paints a very different picture for when performance capacity is exceeded vs. when storage space capacity is exceeded. Most customers only project storage usage and based on that projection no additional storage capacity is needed.  However, when you factor in performance capacity, you may have completely different growth curves which would result in a different set of budget decisions.

Storage Capacity Planning


Why does this matter and what is the impact of not having this information?  This matters a GREAT Deal.  Customers in this situation create several unexpected problems:

  1. They have to make “emergency purchases” with little to no planning.
    1. Typically, this results in high cost purchases because no negotiation process and no real alternatives can be considered.
  1. Performance upgrades cost more money than configuration upgrades.
    1. These purchases involve high priced components – like Flash drives vs, SAS or SATA Drives and other performance components like additional controllers, processors and memory – all of which are the higher priced components.
  1. Result in an unplanned purchase that was not accounted for in the IT budget.

So, the impact of incomplete storage capacity plans can result in budget shortfalls and overruns, as well as emergency staff time to install and configure the additional storage.  The sad part is, all the data to make more informed decisions is stored on storage arrays within the IT infrastructure – it is just not available or in the proper format to enable the business decision making process.


Visual Storage Intelligence® (VSI) is a tool which collects the data from any storage array(s) in a non-intrusive fashion and transforms the data into visual reports designed to facilitate the business decision making process.

Charts like the one shown here clearly depict the example outlined in the previous section showing both performance and storage usage capacity growth curves against their thresholds.  This chart infers that performance capacity will be exceeded in month 3. Having the ability to project this in advance allows customers to:

  1. Evaluate alternative solutions
  2. Develop a Plan and purchase additional resources prior to threshold values being exceeded.
  3. Reduce Budget Costs
  4. Eliminate Emergency Purchase events
Capacity Planning with VSI

VSI is a tool designed to collect and track storage data at all levels of the organization, so planning can be performed at of levels -including business unit or application, enterprise, data center, individual storage arrays, host clusters, individual servers.

Collecting performance and configuration data and summarizing at all levels of the organization allows for improved capacity plans while reducing overall budgets.  The chart below shows business application performance trends for all the key applications within an enterprise.  The data reflects is the IO workload, with the data for latency being included on the reverse side of the chart.  This trend data for IOPS is one of the 4 elements tracked and used to develop future forecast based on trends from previous months for each performance metric  (IOPS, Latency, MB/Sec, and Node utilization.

Storage Performance Trends

This next chart shows all 4 performance metrics in a single chart and how each measurement is tracking against its own percent of threshold.   In this example, Latency and node utilization are showing the highest percent of threshold, but node utilization has the highest growth factor based on starting at a lower percentage 12 months from the current measurement.

Performance Report

Being able to see how the array is performing against each metric allows for more targeted purchases vs. generic purchases.  Reducing node utilization may just involve purchasing additional controllers without any additional storage to improve utilization without incurring additional costs.

A graph depicting the "by application" tab. X-axis are dates in chronological years. Y-axis is forecasted capacity in millions of GB.

VSI includes multiple forecasting models that can be used to clarify capacity and growth assumptions.  These different models (array-based or a grid model) provide two different ways to see capacity growth.  These models allow customers to model and plan for:

  • Types of storage needed to meet planned future requirements
    • High Performance storage
    • High capacity (lower cost) storage
  • New Applications
  • Performance requirements by applications
  • Unique space considerations like deduplication & compression usage

Benefits of Using VSI for Capacity Planning

Reduce upgrade costs, minimize/eliminate unplanned budget events, improve IT staff productivity (no collection or analysis time – all data available and delivered through consumable reports).  No one makes this process as easy and effortless as VSI. The cloud -based nature of the VSI application means the data is available to all levels of management throughout the organization.  Organizing the data in an easy-to-use and accessible format for analysis allows customers at all levels to build, change and manage their storage capacity and performance from anywhere, at any time.