VMware’s announcement of EVO:RAIL at VMworld 2014 set a new bar for converged infrastructure efficiency. EVO is architected to accelerate time to production and improve operational efficiencies, the holy grails for modernizing IT. VMware uses the term ‘hyper-converged’ to describe this class of converged infrastructure, defining it as ‘server hardware pre-integrated with virtualization software for compute, network, storage and management, providing a single point of entry for the entire SDDC lifecycle.’
With commodity servers and local vSAN storage, EVO:RAIL is ideally suited for test/dev, remote office, VDI and other workloads that can failover and restart. By contrast, Hitachi Unified Compute Platform is designed for workloads that serve as the enterprise backbone, typically requiring shared storage, multi 9’s availability, a high degree of redundancy and robustness, and comprehensive data protection.
Can the promise of EVO be extended to business critical converged? VMware describes traditional converged as made of ‘data center components (e.g. shared storage hardware, servers, switches) integrated and sold in a single chassis, along with software options.‘ Traditional converged is designed to achieve faster time to production and predictable performance. However, it lacks the operational efficiency, integrated lifecycle and ‘single point of entry’ promised by EVO and hyper-converged.
Fast Mean Time to Resolution (MTTR) is one indicator for operational efficiency. In enterprise class environments with hundreds or thousands of devices made by different vendors, many with different management tools, identifying and resolving performance and other issues is frustrating and time consuming. Determining whether the root cause of the problem is related to the server, storage, networking or other component is difficult. Moreover, on many occasions problem remediation require hand-offs between different organizations.
Hitachi Unified Compute Platform with UCP Director is the first enterprise-class converged system to address these set of challenges at the enterprise level. In line with VMware’s definition of hyper-converged, Hitachi UCP offers a federated entry to the converged system where the entry point is not a new tool, but simply VMware vCenter or vCenter Operations Manager (vC Ops), or in the case of Microsoft users, Microsoft Systems Center. Using REST API, UCP Director is integrated natively within either VMware or Microsoft, and provides detailed information about the physical infrastructure. Drill-down for problem detection is performed directly from within these environments, using a single set of tools across the entire converged stack. In the case of the VMware environment, vC Ops is used to analyze performance and to drill-down into every infrastructure element and component, be it storage, network switches, servers or chassis.
UCP is designed with Hitachi enterprise or modular class storage, Hitachi high density blade servers, and Cisco or Brocade networking gear. Combined with UCP Director, UCP offers customers hyper-converged class operational efficiency for business-critical environments. And, with the recent announcement that Hitachi will offer EVO:RAIL system, the benefits of UCP will be extended to workloads that can fail and restart, all under UCP Director as the unifying federated management framework. With this framework, customers will be able to monitor, provision and protect converged systems for any class of workloads, using a single tool.