Converged infrastructure has emerged from the nexus of several technologies:
- Increasingly capable hardware that can be packed in very high-density packages
- Virtualization software that can parcel physical compute, storage and network resources into arbitrarily sized logical units
- Distributed storage software that can glue physical volumes comprised of dozens of disks on multiple machines into logical object, file and block resources
- Management software that can consolidate and programmatically automate administration for racks full of nodes into a single UI
The sales growth of converged systems has been fueled by a variety of business factors, notably digital business initiatives that place a premium on the rapid development and deployment of new digital assets and the ability of IT and application development organizations to rapidly adapt to changing business conditions. The hyperconverged sub-category is an outgrowth of all of these factors, but also strongly influenced by the production environments of mega-cloud service providers and online services that use massive arrays of mostly identical nodes that contain both compute and storage and are logically consolidated using sophistical virtualization, containerization, distributed file system and management software.
Scale up CI systems are this era’s mainframes (in a positive sense): highly reliable hardware with OSs and runtime environments that can accommodate a wide variety of traditional enterprise applications. In contrast, from its roots in storage appliances for VM workloads, HCI is evolving into the building blocks for private cloud environments that share the key attributes of their public cloud counterparts: on-demand, self-service provisioning, the delivery of infrastructure and application services, not just VMs and disk volumes, rapid, scale out expandability and node-level system redundancy.
Both types of hardware have important roles in next-generation enterprise data centers and are particularly appealing to organizations that have moved IT into more strategic, higher-value roles involving digital business innovation and away from low-level system integration and quotidian operations. Over time, we expect vendors to evolve and differentiate their products by integrating additional hardware and software components into platforms that target and are optimized for narrower business and application segments like deep learning, engineering simulations/HPC, data warehouses or secondary file storage and archive. Together, the combination of pre-integrated hardware and task-customized software will further relieve IT from the chores of hardware/software assembly, freeing up resources for more strategic business initiatives.
Additional Growth and Differentiation in CI and HCI Platforms Resources