Red Hat global preferences

Red Hat Blog

15 Infrastructure Predictions for 2014

Editor’s note – This is part two of our week-long series of predictions for the new year. Yesterday, Red Hat CTO Brian Stevens kicked off the week with five predictions. Today, we’re taking our shot at infrastructure-related predictions for 2014. Enjoy!



The cloud will be hybrid.

“Traditional” infrastructure deployments as we know them are losing ground to hybrid deployments that include virtual and cloud infrastructure. When it comes to new applications, this trend will only accelerate in 2014. During development, companies will increasingly seek to use Platform-as-a-Service (PaaS) and Infrastructure-as-a-Service (IaaS) to rapidly build, deploy, and scale applications.

- Mark Coggin, senior director, Product Marketing, Platform, Red Hat



Open source offerings will continue to cut into proprietary market share.

In 2014 there will be more credible enterprise options for open and interoperable solutions. We will also see the continued growth of KVM (Kernel-based virtual machine) and other open source virtualization technologies. We'll also see technologies like OpenStack, which are more cloud-focused, continue to take off.

- Chuck Dubuque, director, Product Marketing, Virtualization and OpenStack, Red Hat



Open source storage will continue to mature and cut into proprietary storage market share.

- Joe Brockmeier, Open Source and Standards team, Red Hat



IaaS adoption deepens among enterprise customers.

There will be several factors driving adoption of IaaS in the enterprise in 2014, including the need to make the most of existing virtualization investments. Enterprises have invested a lot in virtualization, but this investment is threatened by newer technologies that may not fall under the same envelope as the solutions they've implemented. As a result, organizations will look for ways to manage sprawl and consolidate their infrastructure. This will cause them to turn to IaaS solutions, which allow them to pull disparate virtualization solutions together in a hybrid and open way.

- Chuck Dubuque, director, Product Marketing, Virtualization and OpenStack, Red Hat



Convergence will lead us to the Business-Driven Cloud.

We will see the integration of business rules, cloud management and cloud infrastructure platforms enabling the Business-Driven Cloud (BDC). This will enable decisions about cost, service levels and resource allocation to move up the stack to the business owners. The BDC is a convergence of three different trends that have been occurring discretely over the past several years - IT becoming a cloud service provider and broker, business demands for more self-control and agility expanding beyond self-service provisioning, and the increased use of business rules management systems to define and describe guidance for making business decisions.  In today’s clouds many of the policies and rules are embodied in scripts, workflows, tools and manual processes that are created and managed by the infrastructure and operations teams. This limits business agility by limiting the direct, real-time control and allocation of cloud resources to business needs. As we’ve seen with self-service provisioning and DevOps, IT will again need to “let go” and move more control from the IT and infrastructure teams to the business. Directly connecting the business to the cloud will finally deliver what enterprises have been hoping for since the advent of cloud computing - the Business-Driven Cloud.

- Joe Fitzgerald, senior director, cloud strategy, Red Hat



Compute and storage convergence will force the overhaul of IT operations.

With private cloud and SaaS environments proliferating in enterprises worldwide, the integration of compute, networking and storage infrastructure is changing the way technology services are defined, enabled and delivered. This will necessitate re-thinking of core IT tools, principles and staffing to support the converged infrastructure. IT specialist roles and accompanying tools focused on silos of hardware-based networking, compute and storage services will revert to the broad systems approach and IT will focus on delivering and managing user services across the complete technology environment however and from whomever they are sourced.

- Ranga Rangachari, vice president and general manager, Storage, Red Hat



Open source cloud management will disrupt the market; Monitoring, management, and resource enforcement become even more critical.

We believe that open source cloud management is going to disrupt and displace traditional management vendors and startups in this space. When we acquired ManageIQ (now part of Red Hat CloudForms), we committed to open sourcing the ManageIQ code. We’re still committed to doing that, and believe that huge value can be unlocked in this management space in the form of an ecosystem, such as a partner ecosystem for CloudForms, similar to Red Hat’s existing ecosystem of OpenStack partners.

- Bryan Che, general manager, Red Hat CloudForms



As virtualization and cloud architecture deployments continue to grow through 2014, monitoring, management and resource enforcement will become even more critical. Before, an application would run on a single piece of hardware, effectively containing the application within its "box." If it didn’t run well enough, the application could then be deployed on a bigger box and there was never any worry about it affecting other systems or applications because of this box. Moving forward, however, applications running in the cloud, on virtual machines or in containers aren't confined to one box. Instead, they share resources with other applications and affect far more than just their “box.”



To keep these applications and their associated operations running, monitoring and performance tools will be required in the coming year. These tools ensure that over- and under-utilization do not occur, preventing downtime and IT overspending, and that applications only receive the resources that they need, stopping runaway applications in their tracks. To deliver these necessary tools, a standard, open platform can  provide enterprises with built-in enforcement and monitoring and to connect management consoles to what they are actually intended to manage. In short, system and application management complexity will increase before it improves.

- Mark Coggin, senior director, Product Marketing, Platform, Red Hat



2014: Year of the Linux Container.

While building buzz now, 2014 will likely be the Year of the Linux Container. These lightweight and portable application containers include only the critical libraries, runtimes and operating system dependencies that they need to do their jobs. But effectively doing so requires built-in security models, resource management and the creation of defined namespaces for the container that technologies like SELinux, cgroups and Linux containers (like Docker) provide. Building containers at scale, which enterprises are likely to experiment with in 2014, will require advanced capabilities to build the container, demarcate the necessary components of the operating system to include in the container, and an understanding of which of these pieces must be maintained/updated and at what cadence. The majority of enterprise platforms will struggle to provide these capabilities, leading forward-looking enterprises on a Tolkien-like quest to find one platform to rule them all.

- Mark Coggin, senior director, Product Marketing, Platform, Red Hat



First, containers will become first-class citizens. Virtual machines are not necessarily the right separation model for SaaS, but containers could be; they will have a second life in the cloud.

- Gordon Haff, cloud evangelist, Red Hat



Virtualization will explode (in a good way).

Different types of virtualization will continue to explode in 2014. We'll see the growth of compute virtualization, especially for already virtualized workloads. Network virtualization and NFV (Network Function Virtualization) will also become more popular as organizations look to better handle scale-out data demands. And demand for storage virtualization will continue as organizations manage costs and data.

- Chuck Dubuque, director, Product Marketing, Virtualization and OpenStack, Red Hat



Virtualized power is something that has not yet gotten a lot of attention, but next year it will likely become much more prominent. Like OpenStack, which allows for the computation of networking resources, virtualized power enables organizations to determine where power needs to go, and when. This type of solution will gain increasing traction as it helps organizations be more efficient.

- Bill Mason, director, Partner Strategy & Programs, Red Hat



The Year of Software-Defined Everything.

Another favorite buzzword of 2013 that should take off in 2014 is software-defined networking or, as we will likely see in the coming months, software-defined everything. Software-defined networking (SDN) has come about due to the rise of virtualization and cloud computing, with networks existing more between virtual machines with less emphasis on physical switches and wires. The advantage is that SDN, at least hypothetically, makes networking easy, allowing IT generalists to handle what was formerly the domain of a specialist.



But in 2014, enterprises will be asking the question: Why stop with SDN? Why not abstract EVERYTHING? This means that we will begin to see the software-defined world in full, from storage to more specialized components of networking, like NFV. Of special interest to the telecommunications industry, NFV is designed to reduce the reliance on traditional proprietary hardware-based appliances for the deployment of network services, and instead, provide the same types of services as virtual machines running on industry standard servers, switches and storage.

- Mark Coggin, senior director, Product Marketing, Platform, Red Hat



ISVs will define a standard operating environment for both virtualization and cloud deployments.

These ever-expanding shifts towards virtualization and cloud means that companies need to find a constant across both of these new deployment models in 2014, before a nightmare scenario of incompatible technology sprawl develops. Without a standard operating environment across deployments, ISVs will find themselves either striving desperately to build and support software for a multitude of platforms or choosing a single, potentially proprietary platform that can easily put them at the mercy of one vendor. If this sounds familiar, it’s because it is: Software developers have seen this scenario with UNIX fragmentation, which required support for multiple platforms, and then the rise of Windows, which tied them to Microsoft. Remember, those who do not learn from the past are doomed to repeat it.

- Mark Coggin, senior director, Product Marketing, Platform, Red Hat



Government agencies will increase their focus on infrastructure management and orchestration.

In 2014, federal agencies will increase the rate of virtualization and movement to private and public clouds. As a result, they will allocate more resources towards new solutions for managing and orchestrating their infrastructures. While several federal agencies have already started the virtualization process, over the next year, many more will move to public clouds and adopt virtualized networking and storage solutions.

- Gunnar Hellekson, Chief Technology Strategist, U.S. Public Sector, Red Hat