ADAPTING TO CHANGING INFRASTRUCTURE: A MULTI-CLOUD, DISTRIBUTED, AND INTERCONNECTED APPROACH TO INCREASINGLY COMPLEX AND HYBRID ENVIRONMENTS, WHERE ON-PREMISE AND AS-A-SERVICE MODELS JOIN FORCES
As business infrastructures adapt according to multi-cloud operating models, data centres must follow suit. In the digital-first economy, where data volume, advanced analytics, and resource distribution in hybrid environments are on the rise, it is crucial to modernise and rethink information exchange between cloud and traditional infrastructures. This involves reducing reliance on a single cloud provider and re-evaluating supplier relationships.
Projections1* indicate that enterprise-managed data centres will decrease in capacity by 6.6% over the next five years, while service provider-managed data centres will see a significant increase of 20.9%. In IDC’s report, “Moving from Datacenters to Centers of Data” – whose title highlights the growing preference for a technical and holistic system approach – IDC reports that 66% of decision-makers in IT companies plan to rely on a hybrid cloud2* architecture to achieve their business goals.
With a hybrid model, businesses can mitigate the cost uncertainties of a – potentially unpredictable – single-cloud approach and ensure compliance, security, and regulatory adherence by maintaining control over data location and characteristics.
In cloud-native workflows, seamless data exchange between cloud and on-premise applications is essential (e.g. financial systems, databases, calculation tools). This is why providers who deliver advanced services – using artificial intelligence, machine learning, and automation – will become increasingly competitive in the field.
1*IDC, Moving from Datacenters to Centers of Data, October 2022
2* IDC, Future Enterprise Resiliency and Spending Survey, March 2022
What are the implications for traditional data centre models?
According to IDC, the heterogeneous multi-cloud approach requires several changes:
- Companies need to embrace more agile and elastic IT deployment models;
- Companies need to increase proximity to customer clouds – physical adjacency to hyperscaler data centres will become critical;
- Performance limitations, bandwidth constraints, and latency issues must not be overlooked. They pose obstacles to digital transformation, particularly in modern distribution models where the seamless sharing of data and models across various locations (in-house, public, edge) is becoming a fundamental requirement.
An increasingly distributed application model requires multiple structures that can communicate with users and their applications. This shift is driven by the understanding that data holds the key to competitive differentiation in the digital era. Traditional data centres often incur high costs related to connectivity, power supply, cooling, and space, unlike interconnected, distributed, next-generation models.
In fact, this model delegates infrastructure management to suppliers and business partners, who oversee data centre operations, allowing companies to focus on their business objectives. Specialists can thus free themselves from the worries of day-to-day infrastructure maintenance and focus instead on more important goals.
By centralising and optimising data management within sustainable operating environments, entities can enhance their overall agility. As-a-Service models are flexible and allow infrastructure costs to be adapted to real needs, expanding as activities increase. The migration process can be costly, especially for smaller and unprepared businesses. As-a-Service models make it possible to modulate and optimise network costs while shifting workloads to higher-performance applications and systems.
State-of-the-art data storage, security measures, emphasis on localisation, and provisioning through As-a-Service models are vital components of an optimised approach. They enable improved network performance, overcome space and resource constraints, and provide secure simultaneous access to analytics and machine learning services.
Sustainability is also a key goal, a fact confirmed by Forrester. The most attentive colocation providers invest in technologies that can cool equipment without increasing carbon emissions. IT managers will select colocation providers based not only on trust, reputation, and global reach but also on their ability to forge strategic partnerships.
Green Energy, Data Gravity, Specialisation
In a 2021 study, Forrester3* revealed that an increasing number of operators are positioning themselves as business partners in digital transformation. These operators differentiate their services based on the unique needs of each customer and are attentive to energy-saving initiatives. Interestingly, despite a 650% increase in total data centre workloads between 2010 and 2019, overall energy consumption remained unchanged. This points to the ability of new efficient data centre models to operate on “green” energy.
3* Forrester, Data Centre And Colocation Market Trends, January 2021
The World Economic Forum predicts that by 2025, 463 exabytes of data will be generated per day. With the growth of IoT, machine learning, artificial intelligence, and edge computing, the production of data continues to surge. This trend shines a light on the crucial concept of Data Gravity (which refers to the capacity, quantity, and speed with which large volumes of data attract services, applications, and users).
Decision-makers in the IT sector will take Data Gravity into consideration when selecting data centres and defining the market for colocation providers. It is this concept that is driving the adoption of a hybrid model, where critical applications are hosted in data centres managed by specialised colocation providers, allowing direct access to the cloud through gateways within the same data centre.
In addition to the increasing workload, there is a growing demand for specialised services. The pandemic accelerated the need for sophisticated analysis systems, such as health data mining. Interconnected and reliable data centres are essential for several sectors, including financial analysis, petrochemical applications, and advancements in artificial intelligence. Strategic service providers with ultra-specialised skills are needed to meet specific requirements in these areas.
Infrastructure & Operations: key digital transformation predictions
The data provided by Gartner4* analysts indicate that the hybrid on-premise/As-a-Service model will grow in the next two to four years:
- By 2025, 70% of companies will embrace structured infrastructure automation to enhance flexibility and efficiency, a significant increase from the 20% reported in 2021.
According to the Gartner 2021 I&O Leaders Survey, 80% of respondents consider automation the best technology for optimising costs. However, many companies still feel unprepared to fully seize the opportunities presented by automation.
Infrastructure automation should be viewed as a continuous process aimed at increasing scalability and efficiency. Achieving cost optimisation is only possible when infrastructure automation reaches certain levels of maturity.
- By 2025, only 50% of companies will develop skills for infrastructure automation through hybrid and multicloud platforms, compared to less than 10% in 2021.
Implementing automation processes from a technical standpoint should not overshadow skills development. Designers and architects must prioritise efficiency and agility to ensure that automation is meaningful, well understood, and effectively utilised. In fact, a study from 20215*
identified a lack of specific skills as one of the main obstacles to creating and managing mature automation.
- By 2027, 35% of application workloads will not be optimal or ready for deployment in the cloud, down from 55% in 2022.
Currently, it is estimated that around 10% of all workloads are linked to cloud-native applications. In the near future, approximately one-third of workloads will run in the cloud. While some workloads may migrate or transition to the cloud with increased complexity, others will not be suitable for cloud deployment due to an unpreparedness for the complexities of new environments.
- By 2025, 40% of newly acquired local computing and storage resources will be consumed As-a-Service, compared to less than 10% in 2021.
With the widespread adoption of cloud computing, there will be a growing demand for on-demand solutions based on operational and consumption expenditure. Specialised infrastructure providers will focus on emerging models such as Bare Metal As-a-Service (BMaaS) and Edge as a Service (EaaS) to cater to the evolving needs of companies.
4*Gartner, 4 Predictions for I&O Leaders on the Path to Digital Infrastructure, January 2022
5* https://www.gartner.com/en/documents/4007208 Gartner, Predicts 2022: Driving Toward Digital Infrastructure Platforms, 2021