IT Consumption: A Closer Look at Storage Capacity Considerations

A pay-per-use model reduces the risk of overprovisioning resources in the on-premises data center.

Monitoring the network
iStock

IT and network leaders are continually challenged to manage their budgets for on-premises resources. Staffing costs aside, the next biggest line item is typically infrastructure.

Yet, budgeting for this equipment can be a shifting target. For example, most organizations over-provision for storage capacity, according to Futurum Research:

  • 67% over-invest in storage solutions
  • 33% have run out of capacity or experienced high utilization rates that impact performance, including downtime

It’s difficult to predict how much infrastructure will be needed in the future. And, with lengthy procurement cycles for capacity, it’s safer to err on the side of having too much rather than not enough. That’s why, Futurum predicts: “We expect to see increased adoption of pay-per-use models, both on-premises and in the public cloud, to better align storage capacity and costs with business needs.”

So, how does pay-per-use for on-prem infrastructure work?

Getting granular with storage capacity

Rather than procuring the actual infrastructure, the equipment vendor provides and owns the gear at your site. Then, you pay only for the capacity that you actually use. And with the right vendor, that usage is buffered to provide easy scalability based on business demands. If you need more capacity, just turn it on.

For a true pay-for-what-used model, it’s important to partner with an IT consumption provider that employs metering technology. For example, servers can be metered at the unit level down to individual blades or more granularly based on virtual machines. Usage can be tracked hourly and over time, providing visibility and helping managers better forecast future needs.

Yet, the benefits go beyond improved capacity planning:

  • Better workload balancing and hence, greater availability. Admins can immediately tackle cloud bursting to better manage application flow. In addition, the consumption model could be used for workloads that use significant network capacity — such as big data, backup, and edge computing.
  • Mix capacity resources according to needs, including flash, hybrid, file, block, and object storage technologies.
  • Improved planning for other network gear. The visibility into infrastructure usage also helps IT staff to better plan for and manage bandwidth, switches, routers, and other network gear.
  • Showback/chargeback. With granular metering, such as at the virtual machine level, the usage and cost of pay-per-use IT resources can be attributable — to the responsible business unit, department, application, etc. This enables better budgeting, provides the tools to motivate workload owners to optimize their consumption for lower costs, and helps put IT in the position of being a service provider to their internal customers

In addition, a consumption-based solution for storage infrastructure can be scaled to include maintenance, monitoring, and capacity management of equipment to help reduce strain on existing networking staff.

Taking control of the data center

By moving to an on-prem consumption model, IT and network leaders can eliminate costly upfront capital expenses. The solution also enables fast scalability for those organizations that must retain certain workloads in their data centers — such as for compliance or security reasons.

“Organizations that prefer to keep their storage on- premises for reasons of security, compliance, or performance can now gain the same advantages of the public cloud with pay-per-use storage offerings delivered in their own datacenter,” according to Futurum analysts. “This enables organizations to maintain control of their apps and data while gaining the benefits of pay-per-use, scalable storage.”

Learn more about the IT consumption model for on-prem infrastructure by visiting https://www.hpe.com/us/en/greenlake.html

Copyright © 2020 IDG Communications, Inc.