EMC (NYSE: EMC) today unveiled a caching and data mobility appliance that the company hopes will ease the problem of latency in global storage networks.
The new VPLEX, announced today at EMC World, is designed to virtualize far-flung heterogeneous storage networks into one massive global storage network, speeding data access and I/O and limiting latency through caching and cache coherence.
Local and metro versions of VPLEX are already generally available, and EMC plans extended distance and global versions in the next year or two, offering what EMC vice president of marketing Barbara Robidoux said will be “transparent data mobility over unlimited distance.”
The VPLEX appliances, which start at $77,000 and are also available by subscription, come with 32 GB/s Fibre Channel directors, ports and interconnects and 64 GB of read cache, and are based on Intel multi-core CPUs.
Initial arrays supported include EMC Symmetrix V-Max and DMX, Clariion CX and CX4, HDS 9970/9980 and USP-V/VM, and IBM DS8000, with support planned for other IBM, Hitachi Data Systems, HP and NetApp (NASDAQ: NTAP) arrays.
VPLEX Local supports one to four Virtualization Engines per rack and up to 8,000 virtualized LUNs per system, while the Metro version offers twice that across data centers with synchronous distance support.
The Metro version includes support for VMware VMotion between ESX clusters up to 100km, with support planned for Microsoft Hyper-V, clustered file systems, and IP-based storage, including FCoE. Metro VPLEX also promises non-disruptive migrations and relocations greater than 100km.
Metro lets users link two VPLEX clusters within a data center or up to 100 km apart and to federate some or all of the storage volumes across both clusters, presenting the “stretched” volumes at each site as if they were local shared volumes. It promises 5ms round-trip maximum response time between sites.
The technology is based in part on intellectual property acquired from YottaYotta. A spokesman for the data storage giant called the technology “the culmination of a multi-year EMC development effort that combines EMC expertise and intellectual property with the intellectual property EMC acquired from YottaYotta. EMC has additionally invested in bringing new capabilities to VPLEX in multiple areas.”
One of those areas is VPLEX’s storage operating software (GeoSynchrony), which runs on
top of a specially built hardened Linux kernel with native optimizations to minimize I/O latency and maximize throughput. The extended distance and global versions (Geo and Global) will be based on asynchronous technology and will begin reading data before it fully arrives, and even anticipate what data will arrive next, according to EMC. EMC calls the new approach AccessAnywhere, and says it will eventually be added to other EMC storage solutions.
Patrick Gelsinger, the head of EMC’s Information Infrastructure Products division, said in a statement that VPLEX “is groundbreaking technology and that will change the way data centers are designed and managed. It’s a core component of EMC’s Virtual Storage vision and will do to storage what server virtualization has done to computing and will provide game-changing levels of efficiency and flexibility. The ability to share, move and access large amounts of data regardless of location is a key element of the journey to the private cloud. VPLEX technology will enable follow the sun computing, the relocating of workloads to low-cost energy regions or moving them out of the way of approaching storms — in general, IT environments will be more dynamic and flexible than ever before.”
NetApp claims to have beaten EMC to the punch. NetApp vice president of solutions and alliances marketing Patrick Rogers said the company “is already at the forefront of this evolution with our existing and proven FlexCache technology combined with our best-of-breed technology partner ecosystem. What sets NetApp apart is our ability to deliver these capabilities through a single, unified platform instead of forcing customers to leverage disparate architectures that require an entirely new set of processes, tools and training.”
Stay tuned for more coverage from EMC World.
Follow Enterprise Storage Forum on Twitter