Balancing choice and standardization, utilizing DevOps automation and properly securing private clouds are among the necessary steps, according to members of Protocol's Braintrust.
CEO at HashiCorp
Hybrid cloud infrastructure by definition means embracing choice — using different infrastructure platforms because each provides specific benefits for the applications that they host. But that does imply that you are introducing additional complexity and so it is critical to first take a step back and answer the question: As we manage that hybrid fleet, what are those elements where we must have consistency and where can we have choice? Implicitly this is a balance — giving developers access to the richness of each platform so that they can deliver applications as quickly as they must, while empowering operations and security teams to maintain standardization.
Our experience shows that in practice most organizations standardize on a common approach to the core infrastructure elements while allowing for choice at the application layer. This means standardizing on the core infrastructure workflows: a consistent way to provision infrastructure safely across the fleet; a consistent way to enforce identity-based security regardless of application; and a consistent approach to connectivity and networking across the fleet. We see this described as a "cloud foundation" that is the basis of most successful cloud programs — which by definition are "hybrid" for any organization of scale.
CTO at Puppet
You can't do scalable hybrid infrastructure without automation, full stop — and I'm not just saying that because I'm the CTO at Puppet. Hybrid estate management is increasing in complexity, especially within the enterprise where teams need to manage workloads at significant scale. The value of infrastructure automation cannot be overstated when your workload is dispersed across multiple infrastructure types; this is magnified as workloads start to move to the edge. Automation is a requirement and the critical step in creating hybrid environments that can scale.
Security and compliance requirements do not go away when you move to a public cloud. I can't overstate how important it is to move from a reactive to a proactive approach to security and compliance in order to avoid spikes of activity and surprises at audit time, or deal with vulnerabilities that could have been addressed. Teams should ensure they are able to enforce continuous compliance, support self-service incident remediation and offer incident response support across a hybrid environment.
A company that truly embraces DevOps will have a much easier time with these critical steps. Why? Because cross-functional teams are aligned to a single purpose and in constant communication, prioritizing security and productivity from the get-go. They're all in it together, making the path forward more collaborative, secure and scalable.
COO at VMware
Companies are defined by both their cloud strategy and the speed at which they execute it. Decisions that are made to satisfy the urgent and ever-changing business environment will leave a lasting impact for years to come — we've just seen that with the dramatic change of pace in enterprise digital transformation caused by the pandemic.
Hybrid cloud has emerged as a proven model for unifying private cloud, public cloud and edge due to the consistent infrastructure and operations to gain the flexibility to work across a seamless pool of resources for all applications. The freedom to choose the best cloud offering to meet application needs is what will enable companies to preserve their existing investment in tools, training and operational processes while giving them the agility, economics, and scale of a modern public cloud. And that is also the most critical step in creating scalable infrastructure – using that freedom to set a clear strategy, make the right choices about clouds, and then execute without hesitation.
The digital revolution is placing an incredible demand on cloud solutions to be more open, flexible and scalable to help address a wider range of customer needs. Smart use of scalable, hybrid cloud infrastructure will help companies capitalize on new business opportunities and tap into the potential of emerging technologies.
General Manager, GreenLake Cloud Services at Hewlett Packard Enterprise
When building a hybrid cloud infrastructure, you are adding complexity to your current environment, which increases cost and adds risk to your operation. You now have to manage accessing resources via public cloud, on-premises and potentially in the edge. These environments have different points of entry, management requirements, levels of observability and security requirements.
In order to build a hybrid cloud infrastructure that can scale, you must first address how to automate and standardize as much of these disparate capabilities as possible. Without automation, you will find yourself hiring and spending significant amounts on resources to manually deal with these issues.
To achieve this:
- Find the right tooling that provides a single pane of glass from a monitoring and observability perspective. Avoid fragmented performance reports; you need a holistic picture of your IT resources and their overall performance.
- Consider an AI Ops system that handles routine failures. Your operations should have self-healing capabilities so most basic failures resolve themselves. If an automated resolution is not available, you should still have the ability to trace down the root cause of the issue.
- Bring in a trusted advisor and partner to help you de-risk your plan and give you the most optimal solution. Your partner should help with the design, architecture and assessment of which workloads are best served in which environment.
CIO at Nutanix
A flexible and scalable hybrid cloud is the ideal operating mode for enterprise IT, according to this year's Enterprise Cloud Index. Unified hybrid cloud architecture, built on a flexible substrate of hyper-converged infrastructure, enables IT to run both on-premises and public cloud infrastructure as code. A flexible and scalable hybrid cloud empowers enterprise IT, and the companies we support, to reap the benefits of a software-defined ecosystem.
However, not all hybrid clouds are equal. The most important step in building a hybrid cloud that scales to meet the vagaries and demands of the enterprise is to build that cloud on a common foundation of a single operating system and hypervisor that run the exact same code in the exact same way no matter on what vendor's hardware or in which vendor's cloud that code runs. Having a single common hyper-converged substrate enables enterprise IT to write code once, and scale it flexibly and securely to run anywhere.
Senior Vice President and General Manager of Cloud at Qumulo
A unified hybrid cloud environment can provide businesses with the ability to very quickly scale workload capacity. Hybrid clouds should offer end users a consistent experience in usability, administration, performance, visibility and support as you scale up. But maintaining scalability is easier said than done — because you can't always use the same resources and processes in the cloud as you do on-premises. So the most important step is to establish a powerful private cloud. Make sure your on-premises architecture is robust enough to retain control over your data, including file data and security services. Only then can an organization maintain interoperability between clouds as they continue to scale up.
When you're ready to transition to a hybrid cloud, you need to take advantage of the native scaling elements of each environment. In the data center, this means building software platforms that have scaling with the underlying number of servers applied to the solution as a core part of the architecture. In the public cloud, that means building on services that can use the core compute and storage infrastructure of any public cloud environment in ways that overcome the scaling limits of those individual building blocks.
It's never too early to build a strong foundation for a scalable hybrid cloud infrastructure — but it can be too late. To prepare for future workloads, IT teams should be able to maximize their existing infrastructure investments in the private cloud while expanding with the unlimited potential of the public cloud.
More from Braintrust