IT and the datacenter have gone through several transitions over the last 25 years. First there was the mainframe to server transition, which allowed IT teams to better adapt and provision applications and services for business users. Then with virtualization, IT teams could consolidate and rapidly deploy new systems in days instead of months. Now cloud offers the ability to provision, scale and change direction with a level of agility never before seen.
Several years ago, colocation in shared data center facilities began gaining popularity as an alternative to on-premises datacenters, which required deep-pockets and all of the underlying facilities challenges they pose. Now, virtualization sets the foundation for moving out of colocation datacenters and into an infrastructure as a service (IaaS) in the cloud. Gartner predicts that by 2021, more than half of global enterprises already using cloud today will adopt an all-in cloud strategy.
The realized business benefits of cloud adoption, including reduced CAPEX in favor of ‘pay-for-what-you-use’ models, scalability on-demand, and agility are driving future IaaS adoption. IT budgets are not changing but the demands on IT to quickly adapt and innovate for the business are more strenuous than ever. Organizations are realizing that moving to the cloud allows their IT teams to focus on enhanced business value instead of worrying about keeping the lights on. However, with the growing dominance of hyperscale cloud providers, are all the benefits of IaaS being realized?
As cloud adoption increases, organizations need to focus less on adhoc cloud usage and more on strategic implementations to avoid the risks to realizing IaaS benefits. Some of these risks include cost, vendor lock-in, security and lost visibility.
Contain Cloud Costs, Avoid Sprawl and Understand the TCO of Cloud
The risk of cloud sprawl and hidden cloud costs is very real – particularly in the hyperscale clouds that offer limited billing visibility. A provisioning model based on flexible cloud resource configuration drives down costs. IaaS customers should be able to reserve a pool of resources covering CPU, storage, and RAM to allocate and reallocate to different VMs to meet changing workload needs. This approach differs from the instance based provisioning model of many hyperscale providers, which has pre-sized server instances and no flexibility to apply resources to other VMs, leading to over-provisioning and higher costs.
Another hidden cost is training and support of your cloud initiative. Hyperscale clouds differ from traditional on-premises infrastructures requiring IT teams to completely change how they manage, provision and support environments. Leveraging existing skillsets that IT teams have honed over years can help ease the transition to the cloud and refocus on innovating for the business instead of learning a new way to provision resources.
To avoid cloud cost surprises, future cloud resource requirements need to be considered and existing resources should constantly be reviewed. Having a reservation as well as burst pricing models helps customers adapt to changing capacity requirements, without having to pay for cloud resources needed only during peak times.
Avoid Cloud Vendor Lock-In
Outages, cyber-security risks and financial risks can justify a ‘multi-cloud’ provider strategy. After all, it’s hard to name a hyperscale cloud provider that hasn’t had an outage in the last 12 months. With hyperscale cloud providers, migration out of the cloud is more complex, especially when solutions have been configured to rely on proprietary services within those clouds. The prevailing advice is to pursue a diverse cloud vendor strategy by allocating the right workload to the right environment.
Make Cloud Security a Top Priority
With cybersecurity threats on the rise, compliance requirements grow exponentially every day. Don’t sacrifice security and compliance in the cloud. Make sure cloud security levels are on par with or better than on-premises IT security.
Many hyperscale cloud providers place the burden of cloud security on the customer; providing security services as add-ons. However, these services should come natively with an arsenal of proactive security features including vulnerability scanning, anti-virus/anti-malware, web reputation monitoring, intrusion detection and prevention, integrity monitoring and log inspection.
Avoid the IaaS Black Hole
Visibility and ease of management often determines how successful an organization is with their cloud journey. Most IT teams are used to managing every component of the IT stack. When adopting IaaS, especially with the hyperscale clouds, a great deal of visibility into infrastructure is lost. Some businesses have the luxury of employing a group of cloud administrators to monitor and manage their IaaS workloads. But many simply don’t have these resources and need help from their cloud provider. Customers should look for intuitive, self-service interfaces and even APIs that provide visibility into capacity, performance, billing, predictive and historical analytics and security reporting.
Author: Will Urban at iland, Director of Technical Marketing