Hybrid cloud challenges

Customers trying to naively implement a hybrid cloud strategy encounter challenges in the five pillars of operational inconsistencies, different skill sets and tools, disparate management tools and security controls, inconsistent application service-level agreements (SLAs), and incompatible machine formats. Without making proper adjustments to those pillars, customers may encounter decreased agility and an increase in cost and risk.

The following figure summarizes those pillars:

Figure 1.1 – Five challenges of implementing a hybrid cloud strategy

Now, let’s explore those challenges in further detail in the following section.

Describing the challenges of the hybrid cloud

Cloud infrastructures have become more attractive to organizations driven by business transformation initiatives. The cloud improves agility with faster testing and development cycles and reduces costs and risks. Organizations are migrating to the cloud for those reasons.

While providing positive business values, many challenges arise when moving from on-premises to the public cloud. Many customers don’t realize the changes they need to go through to properly take advantage of the public cloud’s benefits. A cloud strategy that addresses the hybrid cloud challenges needs to consider people, processes, and technology.

Operational inconsistency

The tools and procedures that operation teams are leveraging to manage the life cycle of their applications and workloads on-premises are different from the public cloud.

For example, application and infrastructure monitoring and observability tools, automation, management, and CI/CD tools for deploying applications need to be repurposed from vSphere-based APIs/SDKs to AWS APIs and native monitoring services such as CloudTrail, CloudWatch, and adopting infrastructure as code with tools such as HashiCorp’s Terraform.

Disparate security controls

Expanding on operational inconsistency, customers achieve security and compliance through existing security procedures and tools. Adaptation ranges from how users consume authentication, identity access management, network security controls – such as firewalls, intrusion prevention systems (IPSs), and web application firewalls (WAFs) – and application-level protection, monitoring, and logging for Security Operation Center (SOC) environments.

Skill sets and certifications

IT personnel managing VMware-based infrastructure require an investment in recertification and retraining to operate workloads in the public cloud. Skilled IT and DevOps personnel are in short supply in the market.

Inconsistent application SLAs

Migrating workloads in a high-availability architecture while providing production-grade SLAs requires application-level architecture adjustments to enjoy the resiliency of public cloud services. For instance, migrating a virtual machine to an EC2 service in the cloud doesn’t make it highly available. On-premises resiliency mechanisms such as vSphere High Availability (HA) and Distributed Resource Scheduler (DRS) are unavailable on an EC2 service without making architecture adjustments.

Incompatible machine formats

Migration requires a manual conversion for each virtual machine, which includes the hypervisor format, operating system disks, and networking IP address configurations. This process takes into account unsupported configurations in the cloud, especially for legacy end-of-life and 32-bit operating systems. Additionally, the format conversion problem creates a vendor lock-in challenge.

Customers not considering those challenges in advance may experience a decrease in the developer’s agility instead of an increase, an increase in the risk of the project instead of a decrease, and an increase in costs instead of a decrease.

VMware Cloud on AWS was designed to address all of those challenges of the hybrid cloud deployment model.