It’s no surprise that the cloud is highly utilized in the enterprise for everything from workloads to failover to DevOps. According to a 2018 State of the Cloud Survey, multi-cloud adoption has arrived —81 percent of respondents currently use a multi-cloud strategy, and organizations leverage almost five clouds on average. To stay ahead of the curve, CIOs need to implement a multi-cloud strategy sooner than later.
We can and should learn from our past mistakes. When public cloud first came into existence, everyone saw it as suitable only for small-scale enterprises with a multitude of issues ranging from cumbersome IT workflows and budgeting nightmares to widespread security concerns. Nowadays, it is safe to say that if you are not using public cloud in some form, you are years behind your competitors.
But as the name implies, a multi-cloud strategy involves dealing with multiple cloud vendors. As with every digital solution, cloud providers want to completely immerse you in their ecosystem, so that it’s easier to upsell you their value-added services. This introduces the IT headache of learning and managing cloud vendor-specific tools and techniques. Solutions like Red Hat OpenShift and Pivotal Cloud Foundry help you in avoiding these traps and provide a cloud management layer so that your IT teams are only responsible for building and running applications.
Problems to address
Even though adopting a multi-cloud strategy is a smart move for most organizations, there are still problems and growing pains that need to be addressed. And fast.
- Security: This is one of the biggest concerns when using any public cloud solution, and there is no magic bullet to solve it. Security provided by public cloud vendors is generally robust for cloud-only workloads, but a significant amount of planning is necessary to secure data that flows over the highly vulnerable public network from on-premise infrastructure to public clouds
- Cost: Controlling costs has been an issue since public cloud came into existence. It’s easy to run into budget bleeding if cloud consumption is not monitored. Multi-cloud adoption gives you an opportunity to commoditize cloud resources and transition between public cloud vendors based on your needs. This demands a significant shift in the way enterprise IT works because it requires moving from vendor-enforced workflows to vendor-agnostic infrastructure
- Lack of true cross-cloud solutions: Most solutions providing multi-cloud support use makeshift approaches that enable data and application movement from one cloud to another. There is a dearth of products that can provide true cross-cloud fabric so that end-users can visualize and utilize multiple clouds as a single platform instead of numerous isolated entities.
Multi-cloud deployments are still in the relatively nascent stage, limited to applications and workloads which are not business-critical or mission-critical. Some enterprises are forced to implement a multi-cloud strategy because they depend on a line-up of solutions that are not entirely supported by a single vendor. For instance, VMware support only exists for AWS and cloud support for Microsoft office applications is most cost-effective through Azure.
Making the most of multi-cloud
Looking ahead, there’s various strategic justifications for a multi-cloud deployment. The business environment is global and highly dynamic, with requirements and relationships changing constantly. Not to mention that the challenges of privacy and security threats plus related regulations are relentless. Data is multiplying exponentially, IoT is growing rapidly, and artificial intelligence and machine learning projects are sure to test the limits of cloud compute capabilities. It’s already true that traditional data storage systems cannot meet the dynamic demands of distributed applications and cloud deployments, and storage designed with distributed systems in mind makes it possible to quickly provision application specific, policy-based data services.
To respond with agility to emerging technology standards and challenges, multi-cloud adoption strategies must focus first on the standardization of cloud orchestration and commoditization of resources. As digital business models and enterprise systems grow in scale and complexity, automated orchestration capabilities will be essential to maintaining control and getting the most out of any cloud investments. Commoditization of hardware, managed services, cloud platforms, and even security solutions is making infrastructure components more turnkey and interoperable — but only if you have an overarching layer of control and visibility.
It’s especially imperative that enterprise-level multi-cloud management include automated orchestration tools for replicating data across multiple sites (data center, private cloud, and public cloud). Use cases like analytics, test and development, unstructured data and secondary storage demonstrate the necessity of finding ways to reliably and easily move workloads from cloud to cloud.
The promise of a cloud-agnostic infrastructure is to make data easier to access and more affordable to store long-term by putting different types of data into different clouds for their various benefits and cost structures. Multi-cloud deployments strengthen business continuity and resilience, empower DevOps development and cloud-native applications, and optimize regulatory compliance and service delivery for global organizations.
The trajectory and dynamism of cloud technology reflects the nature of the modern world — explosive growth and sudden contractions, the surge and retreat of markets, the fluidity of consumer trends, and the nonstop inventiveness of a young, diverse, and international citizenry. To match this energy and catch its currents, you’ll need the flexibility, control, and wide-open potential of multi-cloud. Up, up, and away we go.