everythingpossible - Fotolia
Better access to digital information has opened new revenue opportunities in nearly every industry. Whether it involves business intelligence and analytics, mobility or internet of things, it's clear the next level of business competiveness is being built upon a foundation of data availability. None of this is news, of course. And, if you're reading this column, you are likely already heavily involved in the ongoing battle to ensure that IT resources keep pace with the increasing demands placed on business applications and data.
The good news is IT infrastructure innovation has accelerated as well. Hardware, for example, continues to become quicker and more affordable. As a result, storage systems perform faster, scale higher and hold more capacity than ever. In theory, you'd think the two, (1) growing demands served by (2) ever-more capable infrastructure, would cancel each other out. But that's not how it works in the real world. While there are a number of reasons for this inconsistency, one that doesn't get discussed enough is the time to provision new storage capacity.
Time to provision
When storage vendors discuss time to provision, they tend to focus on how easy it is to set up and configure an array physically located on site in a rack with adequate power and cooling. Here, setup time is often a very small portion of the entire process. The true time to provision, however, encompasses everything that occurs from once you identify a storage resource need to the moment newly acquired resources are made available to applications. The full end-to-end process can take months, and the few minutes or hours it takes to set up the final storage array is only a small part of the overall pain. Meanwhile, application demands continue to increase while the infrastructure provisioning process happens.
Delays in provisioning infrastructure deployments not only slow down new IT initiatives. In this era where business competitiveness is often determined by data access, delays can negatively impact revenue opportunities and the bottom line as well. For years, you could address the time-to-provision challenge by simply deploying more storage capacity than immediately necessary, giving the environment room to support near-term growth during the sometimes lengthy process of new storage system procurement. While still considered a best practice by some, having excess infrastructure just sitting around doing nothing adds unnecessary cost, a nonstarter in this age of tighter budgets.
One obvious method for reducing the time to provision storage is using public cloud services. While this can provide near-immediate access to new capacity, performance and security, other business considerations often lead many firms to prudently retain a significant portion of data on premises. The trick here is to achieve cloud-like agility in storage provisioning while maintaining those on-premises capabilities required by many workloads. There are a number of options available to help improve, or at least mask, time-to-provision challenges for on-premises infrastructure, these include the following:
- Advanced storage analytics: Storage array management continues to evolve. A number of storage systems offer management tools that enable administrators to forecast capacity growth patterns months in advance, with some allowing architects to play out different what-if scenarios investigating the performance and capacity impacts of new workload deployments on existing infrastructure, helping to size what additions may be needed. This added level of intelligence does more to mask the time-to-provision challenge than reduce it, though. What it does do is enable storage administrators to recognize resource needs sooner in an attempt to start the infrastructure provisioning process earlier.
- Pay per usage for on-premises infrastructure: In response to the success of public cloud services, several on-premises storage providers have begun offering more flexible payment options for storage systems on site, where you can unlock additional capacity already on the system by paying extra. In addition, a few storage startups now focus on deploying on-site hosted cloud infrastructure using a pay-per-usage model. With this model, storage is owned and managed by the vendor, but the infrastructure is in your data center. Keep in mind, these offerings vary on capabilities, service commitments and what, if any, obligations are required from customers in terms of future growth. Nonetheless, this type of on-premises storage can ease the cost burden of on-site infrastructure provisioning while possibly speeding up the deployment of new capacity.
- The hybrid cloud: A wealth of hybrid cloud offerings has entered the marketplace recently. Some leverage the public cloud predominantly, while placing high-performance caching devices on site. Since the bulk of new capacity is located on the public cloud, they can deliver on-premises performance while deploying new capacity at the pace of the public cloud because of the cache. Other hybrid clouds allow you to deploy storage capacity on or off premises with the ability to move data back and forth between them. With these, you can add new capacity quickly to the public cloud side and data can be moved off on-site resources to make room locally as needed. These hybrid storage offerings greatly reduce the time to provision and improve workload agility.
IT demands change so rapidly that new resources are often needed immediately, not months down the road. Some look to the public cloud to solve these challenges, but these services alone aren't right for everyone or every workload. In response, on-premises vendors are offering greater intelligence and more flexibility in payment options to ease the burden of deploying new capacity on site. While there are benefits to this approach, it can still be a challenge to match the agility of the public cloud. For that, hybrid clouds have stepped in as an excellent option to deliver on-premises performance and security while integrating the agility of public cloud infrastructure.
Provisioning a software-defined data center
How automation and provisioning mesh with cloud computing
Develop a private cloud environment