How to Design a Robust Multi, Hybrid-Cloud Strategy

This article was originally published in BW SMARTCITIES - Source Link
With IT infrastructure space turning into a commodity business, the cloud provider space has become increasingly cluttered. Cloud has seen unprecedented adoption in the last five years, and companies across industries and geographies continue to invest in the high-ROI technology to harness its scalability and cost benefits.

While more enterprises are looking to deploy cloud architectures, they are also inclined to continue using their on-premise data centres considering the legacy application landscape and data complexity. They may do so by leveraging multi-cloud and hybrid cloud architectures to increase business agility. As per Gartner, 75% of enterprises will have a hybrid cloud deployment by 2020.

According to Forrester’s estimates, around 62 percent of public cloud adopters are using two or more unique cloud environments or platforms, emphasizing on the need for selecting the right provider for the right offering and eliminating vendor lock-in; which was a common problem with traditional commercial off the shelf applications and infrastructure.

A hybrid approach fits for many organizations as it allows them to make the optimal use of on-premises infrastructure while utilizing the cloud for additional or burstable workloads of business agility or cost improvements. Enterprises are achieving the hybrid cloud strategy with cloud-native infrastructure to drive the adoption of emerging technologies and applications such as blockchain, chatbots, DevOps. That trend is seeing more traction with greater availability of cloud-native commercial off the shelf tools and platforms.

Recently, GCP and Azure released their hybrid cloud tools – Anthos and Azure Arc, respectively – and Azure Arc has played a significant role in Microsoft wining the USD 10 billion JDEI deal with the Pentagon. Some analytics platform providers like Cloudera are helping organizations optimize their data and migrate workloads to the cloud. Similarly, another open-source software provisioning, configuration management, and application-deployment tool – Ansible – delivers autoscaling with Infrastructure as Code (IaC) on the cloud, enabling the extension of DevOps frameworks.

Deploying the right workload at the right place

Experiencing the exponential cost and productivity benefits of cloud requires IT decision-makers to make the right choice while choosing the cloud provider. However, what happens when a workload is not suitable to be migrated to a cloud environment altogether? What happens to non-native applications? What happens to applications that cannot be migrated to a cloud environment hosted outside the country due to data laws? How can companies reduce time to market in a cloud environment? What are the cost benefits or business benefits of choosing a cloud strategy? There are real-life scenarios where the CIO must decide to simultaneously run and maintain both on-premise and cloud environments – commonly referred to as the ‘hybrid cloud’ deployment model.

That may be the most optimal choice in cases where:

  • Extreme workload fluctuations are observed, and the cloud is being considered as a possible solution to manage surges.
  • Legacy applications would require high cost, time and effort to make them ready for migration to the cloud.
  • Non-native applications experience performance lags when running on cloud infrastructure.
  • Geo-specific data security legislation does not allow personal data to be migrated outside the legislative boundaries.
  • Specific environments such as high-performance computing might require the data to be stored on-premise to ensure superior speed and minimize latency.
  • Leading-edge technologies such as IoT, chatbots, and serverless computing can be integrated with the existing IT landscape using a cloud-based, plug, and play model.

In all those cases, a hybrid cloud deployment can yield impressive outcomes in the form of workload, cost, and operational optimization. While the hybrid cloud is undoubtedly beneficial in the scenarios described above, managing it can prove to be expensive and cumbersome. Efficient and reliable deployments of the hybrid model require companies to embrace automation, which has been made possible with the proliferation of software-defined infrastructure. Using task-based templates and container libraries, developers can experience automatic provisioning, installation and testing of virtual machines, positively impacting the development lifecycle and accelerating time to market. Automation is also particularly useful while tackling the reduced utilization of cloud resources. By proactively monitoring utilization levels and automatically cutting off cloud servers based on pre-decided thresholds, digital enterprises can prevent wasteful provisioning. Moreover, savings can be achieved in terms of labor costs and response times as hybrid cloud orchestrators continuously monitor cloud deployments and shift and balance workloads optimally.

Role of User Experience and DevOps

The vision is for IT to be a one-stop-shop for all users, irrespective of the service delivery platform – public cloud or private cloud (data centers). For instance, visualize a process where all users would log into a central self-service portal, pick an item from the services catalog, and then get it delivered across any of the clouds orchestrated tightly with each other. Throughout the process, the user experience has to be at the core of the value proposition.

Kapil Uniyal
Kapil Uniyal
Former Vice President and SBU Head, Global Infrastructure and Cloud Technology Services, Birlasoft
Let's Connect