Through these interviews a few key points came to light:
· The widespread adoption of cloud services has significantly impacted the way companies are meeting their data centre infrastructure requirements.
· These complexities will be compounded by an anticipated groundswell of new distributed IT driven by the Internet of Things (IoT) and emerging edge computing workloads.
· The shift towards a hybrid data centre environment that consists of a mix of ‘off-premises’ services, public cloud and colocation, and privately owned ‘on-premises’ data centres is challenging traditional infrastructure management and oversight methodologies.
As cloud services are deployed, there are ripple effects across the organization. There is a significant shift in business models, while greater demand is placed on connectivity and workload management. To realize the full value of a hybrid approach, the management of a combination of data centre environments has become one of the most complex issues for modern enterprise leaders, forcing them to re-consider strategy and common practice.
The study also reveals that while the experiences, strategies and innovative technologies used varied greatly, there were clear common themes:
· Determining the best execution venue - Choosing the location of data centre capacity greatly affected risk, cost and IT service performance, on a workload-by-workload basis.
In one use case, a U.K.-based retailer found that when determining the best venue, security, performance and latency requirements needed to also be factored into the total cost analysis, along with data transit and application license costs. These vectors combined helped to determine the right mix of colocation and public cloud computing infrastructure to support connectivity needs while also controlling costs.
· Driving visibility across hybrid environments - Driving visibility across multiple sites is a key challenge. The process is time consuming and expensive, with monitoring hindered by siloed software capabilities and availability of consistent data.
As one study participant noted, “If I had my dream, we would have the visibility to predict failures before they occur.”
· Managing costs without sacrificing performance - Without proper research and planning, hybrid data centre costs can be high, but with the right plan, it is possible to reduce data centre capital and operating costs while maintaining or improving high levels of availability across hybrid environments.
“The biggest challenge of all is cost control, and management and financial applicability of capitalization,” noted one U.S. retailer.
· Unifying operational procedures across hybrid IT environments - Developing strict operational processes and ensuring that these requirements are part of SLAs and operations-level agreements (OLAs) with colocation and cloud service providers is crucial.
As one study participant advised, “You have to know [off-premises] is not going to be cheaper…and you lose a huge amount of control. To gain some of that control back, all you can rely on are your SLAs and OLAs.”
