Hybrid Cloud Implementation for Enterprises

As part of the digital transformation and technology modernization, many large enterprises are looking at embarking on a transformation roadmap to enable a hybrid cloud framework to address the following:

• Modernize legacy applications and coexist with cloud-native applications.
• Automate technology operations to enable agility and speed of delivery.
• Optimize TCO to address increase workloads due to business demand.
• Improve service quality.

This article focuses on key enablers that will be required for building a hybrid-cloud architecture and operating model to support the enterprise goals and transformation strategy, especially with the accelerated and increased adoption of digital services in a post COVID world.

Hybrid Cloud

Hybrid cloud is a hosting model that provides an ability to run your workloads on a combination of private and public cloud and provides 

• An integration layer that provides unified orchestration and management for workloads across clouds.
• Portability across various hosting models and clouds to minimize the risk of vendor lock-in.
• Enable cloudburst into the public cloud for unpredictable workloads.

Difference Between Hyperscalers and Hybrid Cloud Platforms

Hyperscalers like AWS, GCP, Azure, et al., provide public cloud platforms that host a number of technology services (FaaS, PaaS, IaaS) to build and run applications. They do not provide the ability to have a mix of these services and enable them to run across various providers to enable a hybrid model, thus minimizing portability of applications, increasing the risk of lock-in. 

Business Challenges That Hybrid Cloud Addresses 

Large Enterprises, especially financial institutions, are heavily governed by data privacy and technology risk regulations. These are critical challenges for running their entire infrastructure on the public cloud hyperscale platforms. 

As part of the business transformation to digital, these enterprises require a cloud foundation to support the:

• Agility of delivery.
• Speed of implementation.
• Unpredictable workloads that business use-cases can generate.

Thus, the need to provision and scale on demand. Building an on-prem solution to support this will be expensive as they have to buy the infrastructure required to support it.

The hybrid cloud model provides an ability to scale on-demand onto the public cloud for burstable compute workloads. This enables enterprises to keep their TCO flat and reduce over a period of two by leveraging more on-demand compute from hyperscalers. 

Hybrid cloud model also significantly minimizes technology risk of the following factors:

• Data privacy: Enterprises can leverage on cloud services with reduced risk of hosting sensitive information on public clouds, thus reducing the governance overhead required
• Portability: This enables enterprises to shield themselves from any lock-in or risk of any serious outages on public clouds. Hybrid clouds can be ported across the hosting model without having to change the underlying architecture, governance, and operating model.  

Key Enablers for a Hybrid Cloud Platform

Technology 

Container technology and orchestration platforms have a natural affinity towards enabling a hybrid cloud model. These platforms decouple functional requirements from non-functional requirements, thus enabling:

• Portability with standardization – Containers.
• Agility with automation – DevOps.
• Predictability of workloads and fault tolerance – Autoscaling, standard monitoring.

Cloud Governance

Cloud governance is an essential part of the implementation that enables the setting of goals and policies for the operating model and technology architecture aligning to address the functional and non-functional requirements that the cloud will deliver.  This is to simplify technology delivery and enable automation capabilities to fix governance and policy standards. Some of the key aspects of Cloud governance are the following:

• Target operating model: Define the target goals of the technology operating model that the hybrid cloud will help achieve.
• Technology architecture and blueprint: Define the hybrid cloud architecture, Cloud integration, tools and technologies, Container Architecture, Define NFR (monitoring and logging, capacity planning), and key metrics for operations management.
• Development: Align with an agile development model embedding the DevOps automation to simplify development and deployment. 
• IT Operations: Enable unified operations across the cloud and embed SRE-like capabilities and monitoring automation, including AI-Ops.
• IT Security: Establish standard security policies and Implicitly apply security policies and standards as part of DevSecOps automation.
• Risk and compliance: Identify key technology regulations and risks and enable automated capabilities to address them. Provide visibility of these standards and deviations through unfired dashboards.

The key to a successful governance model lies in enabling the culture change required and establish a lean governance model.

Cloud Integration

The hybrid cloud platform should have the ability to integrate through a single integrated management pane to:

• Support the coexistence of existing legacy applications and cloud-native applications using a unified operating model.
• Orchestrate workloads of containers across clouds.
• Manage the lifecycle of containers (deployment, patching, etc.).
• Support various versioning strategies and A_B deployment models.
• Provide zero downtime for applications during upgrades.
• Deliver automatic discovery and load balancing.
• Enable the platform to scale-out (workload containers and worker nodes).
• Support multi-tenanted architecture while addressing the organization policies required for isolation of workloads across LOBs.

Management Across Multiple Clouds

Since the workloads are running across multiple clouds (private and public), having a unified management console that enables

seamless operational view across clouds 

• Unified application/platform monitoring
• Ability to perform workload management (burst workloads, etc.) across clouds 
• Ability to continuously monitor organisational standards and policies and ensure adherence to the same. 

This management platform should have seamless visibility of applications running in different cloud providers and clusters under a unified technology operating model.