Hyper-converged infrastructure explained

Hyper-converged infrastructure explained further

The promise and appeal of HCI lies in overcoming many of the traditional impediments that have long-been present in traditional IT resource provisioning and management. Consider that, only a few years ago, a data center used one physical server for each workload, and admins had to configure each server -- often manually -- for the workload it was intended to host. Similarly, storage and networking gear was typically tied to specific workloads and managed with a high degree of manual interaction. Each resource relied on its own specialized IT team and used different tools for monitoring and management. Simply deploying a business application required days or even weeks of planning and coordination to prepare resources.

With HCI, traditional IT silos are essentially erased. A single IT team can provision compute, storage and network resources.

The adoption of virtualization has changed much of this, enabling admins to provision servers -- later storage and networks -- much faster and more flexibly and usually enabling far more efficient use of compute, storage and network resources. But even with the broad adoption of virtualization technology, impediments remained, such as IT silos, gear integration issues and common management limitations.

The emergence of converged infrastructure, and now HCI, really advances virtualization to address these higher-level IT problems. At its heart, hyper-converged infrastructure is explained as a comprehensive software layer that combines virtualization with automation, orchestration and management. This software layer virtualizes all of the compute, storage and network resources in its scope of operation. Those resources are then organized into pools and categorized into performance-based tiers. For example, admins can virtualize a group of Serial Advanced Technology Attachment hard disk drives into a low-performance pool for archival-type tasks and a group of solid-state drives into a high-performance pool for data stores.

With HCI, traditional IT silos are essentially erased. A single IT team can provision compute, storage and network resources. Automation capabilities enable many provisioning tasks to be handled immediately and usually within guidelines set by established policies. For example, rather than provisioning specific vCPUs and memory or creating a network segment, an admin can simply request a VM or container for a desired type of workload, and the provisioning can be accomplished dynamically to maintain security, compliance and resource efficiency. And finally, management capabilities provide a common view into all of the resources and underlying hardware within the HCI scope. Admins can see everything in the HCI environment.

HCI is increasingly implemented on white box hardware rather than preselected or purpose-built hardware. This enables a greater diversity of hardware, alleviates many of the problems normally associated with vendor lock-in and enables easier growth and expansion of the HCI environment.