Hybrid cloud implementations and considerations are rising, according to our work with cloud practitioners. The reasons are quite direct: Cloud technologists believe hybrid cloud may be a better path toward operational efficiency, better security, faster applications development, improved business insights, and better resiliency – among other reasons.
First, let’s define the meaning of hybrid cloud. In a hybrid cloud deployment, organizations look to leverage compute, networking, and storage resources – wherever they may reside. This might include a combination of resources in both public cloud and private infrastructure.
One of the major advantages of using hybrid resources is that organizations don’t feel “locked in” to one specific public cloud provider.
Economics Over Speed
A hybrid cloud deployment might come with a classic tradeoff you might be familiar with in many industries: It can be more complicated, because it involves more careful planning. But that also means you can gain more control over customization, as well as possibly better economics.
The public cloud revolution came about for two primary reasons: economies of scale as well as speed of deployment. Cloud leaders such as Amazon, Microsoft, and Google had all the resources, and when they made those resources available to the public, many found it faster to use that infrastructure rather than build it themselves.
But fast forward a few years. It’s now becoming easier to build hybrid systems, thanks to cloud orchestration technologies such as Kubernetes. There’s no reason for you to limit your cloud spend to one infrastructure provider.
In addition to the threat of lock-in, another major gripe about public cloud resources is that they are expensive, according to research and interviews with end users. Many CIOs are speaking out publicly, saying the compute costs in public cloud services are too high. In one example, McDermott International CIO Vagesh Dave told CIO.com that he shifted to private resources because they were cheaper.
Depending on the size and the type of organization, hybrid makes more sense. It of course depends on the type of the organization, the infrastructure, and which applications they are running. But if you are an organization with a large amount of private infrastructure and edge data applications, it’s likely that hybrid cloud is for you.
For example, if you are retailer Target — with 4,000 engineers, over 1,950 stores in the U.S. alone, and more than $107 billion in annual revenue – it makes sense to build out your own hybrid infrastructure, as Target has done.
Hari Govind, Senior Vice President of Infrastucture for Target, had this to say about their move to hybrid cloud in a public blog:
“Our private cloud footprint between enterprise data centers and the edge is extensive. We run our private cloud on an elaborate network of enterprise data centers, combined with a near edge infrastructure in our regional distribution centers (DCs) and a far edge infrastructure in our stores. Enterprise DCs still serve all our data-sensitive needs such as our unified guest data platform. A massively scaled edge footprint enables low latency workloads in stores and distribution centers.”
While some organizations may continue to outsource as much of their infrastructure and development to the public cloud as possible, others are leery of the loss of control, escalating costs, and potential “lock-in” associated with putting too many of their resources in public cloud. Instead, organizations may want to leverage cheaper, traditional resources (such as compute nodes in private datacenters) by building a hybrid architecture that can blend all the infrastructure as one.
Hybrid Cloud for Innovation and Control
Futuriom recently studied 37 hybrid cloud implementations through public records. In citing the chief benefits in public statements, the top three drivers were operational efficiency, security, and faster application development – named in more than half of the examined use cases.
In another example, AstraZeneca said it built a hybrid cloud using tools from VMware and cloud management provider Morpheus, to automate workflows across multicloud and hybrid cloud environments, including public cloud services such as AWS, Azure, and Google Cloud. The bio-pharmaceutical company said that streamlining the flow of resources to developers reduced labor costs, eliminated delays in provisioning, and speeded deployment times.
In another example, Walmart has built its own cloud-native infrastructure to put more compute resources closer to its stores and customers. And the TruGreen lawn-care company stores historical data in a private cloud while using Oracle Cloud Infrastructure in support of enterprise business applications.
Implications for Technology Vendors
As interest in hybrid cloud technology grows, it has big implications for the cloud infrastructure technology providers.
On the compute side, Kubernetes has solidified its role in helping implement hybrid environments, strengthening the position of Kubernetes management platforms such as IBM Red Hat’s RedShift or VMware’s Tanzu. Independent cloud management vendors such as CloudBolt, Morpheus, UBiqube, and Virtana are helping organizations get a handle on hybrid cloud infrastructure, providing integration, compliance, and operations tools. And cloud cost management (CCM) vendors such as CAST AI, Yotascale, and Zesty focus on the FinOps aspect of cloud management.
To respond to this trend, the major cloud providers are investing and promoting more of their hybrid infrastructure, such as Google Anthos or the Microsoft Azure Edge Stack for on-premises implementations. Some companies, such as HPE, have staked their future as hybrid facilitators, positioning all of their cloud wares “as-a-service” in either public or private infrastructure, with hybrid integration services offered as well.
Networking will become an essential hybrid tool to connect cloud services wherever they reside, which requires a more programmable, software-driven approach. This is an area being targeted by the many multicloud networking (MCN) specialists such as Aviatrix, F5 Networks, Graphiant, Prosimo, Versa Networks, and others.
In addition, orchestration and network automation companies such as Itential and UBiqube are making it easier to automate network connectivity across diverse platforms and environments, regardless of which vendors dominate the installed base. Observability and monitoring platforms such as Datadog and Kentik collect and route data for better automation. By tying together observability, automation, and orchestration, we get the NetDevOps movement, which aims to deliver network automation to facilitate hybrid cloud infrastructure connectivity.
Multicloud and hybrid cloud tools are in the early days of development, and this process is likely to unfold over the next 2-5 years. Larger vendors will look for strategic acquisitions. The size of the opportunity is large, and the technology providers that focus on end users’ ease-of-use, flexibility, and cost of operations could have lots of success developing automation for these new hybrid environments.
Read the full article here
