This is Data Center Transformation
by Hu Yoshida on Sep 27, 2010
Today is a major step forward in the transformation of the data center. With a clear understanding of the transformation that is currently underway in the IT industry, Hitachi Data Systems is introducing a unique platform of both hardware and software that delivers unprecedented levels of scalability and performance for a truly end to end data center. Yesterday’s solutions of adding more storage capacity just don’t cut it anymore.
Customers require a data center designed for flexibility and efficiency, which is why we’ve introduced the new Hitachi Virtual Storage Platform (VSP) and the new Hitachi Command Suite (HCS), our comprehensive management software. Instead of taking a storage centric view of how we can provide more storage to contain compounding growth of data, we are taking an application and server centric view, and our new platform delivers integration of server and storage virtualization resources to enable data center transformation without major disruption to IT resources.
Applications and servers see file systems and volumes. They don’t see the storage infrastructure that sits behind the volume. They don’t know what other volumes are contending for the resources in that storage system. They also cannot predict with any accuracy how much capacity their volumes will require as their applications grow and where the hot spots in the volume will occur from day to day. As a result, application users want to have dedicated storage systems with more capacity than they will ever need with the highest performance available.
The cloud concept of providing shared infrastructure services on a pay as you grow basis was tried over 10 years ago during the dot com boom. This business model of providing services to many users by leveraging a shared set of infrastructure resources died because each user wanted their own dedicated resource and as a result we had the dot com bust in the early 2000s. The same will happen to the resurrection of this concept with cloud providers if we do not solve the problem of safe multi-tenancy and scalability.
Hitachi Virtual Storage Platform & Hitachi Command Suite
The Hitachi VSP and HCS are designed to answer the requirements for a shared infrastructure that will enable applications and servers to scale beyond their current limitations with the confidence that their data is secure. This will provide them with the capacity they need when they need it, and ensure that the data in a volume that needs performance will be on the highest performance tier, while the less active data in a volume will reside on lower cost tiers to reduce their TCO. All of this will be provided automatically to the application without the need to increase complexity with redefining LUNs, classifying application data, assigning volumes to tiers of storage, or redistributing workload across storage clusters.
The secret behind this unique capability is a 3D scaling architecture and a 3D management suite of tools. By providing 3D scaling in our storage and management platforms (not just capacity) we are enabling servers and applications to scale to new heights in consolidation, availability, agility, and ease of use. The Hitachi VSP is the only storage architecture that can dynamically and concurrently scale up, scale out and scale deep for performance, capacity and support of multivendor storage. The 3D architecture enables users to start with a diskless, single rack, 12 core system for use in a virtualized midrange application environment and scale up to a 6 x rack configuration with 96 cores, 1 TB of data cache, and 2048 x 2.5 inch, Serial Attached SCSI drives. This could scale up to support a large VMware cluster with thousands of virtual machines or it can scale out to support hundreds of application servers with safe multi-tenancy and QoS. It can also scale deep to include external storage as lower cost tiers of storage or to provide non disruptive migration for technology refresh or lease expiration.
Storage virtualization and Dynamic Provisioning enable the spin up of new storage allocations in a matter of minutes, eliminates the waste of over allocation with thin provisioning and zero page reclaim, and enhances application performance through automatic wide striping of data across all the spindles in a storage pool. In addition, a new enhancement of Hitachi Dynamic Tiering enables a volume to be assigned to a pool of storage that spans multiple tiers of storage, including external storage, and allows the Hitachi VSP to manage the placement of that volume’s data to the appropriate tier of storage. The application volume no longer needs to be classified and assigned to a tier of storage and moved to another tier when the performance characteristics change.
The new HCS is a major revision of our Hitachi Storage Command Suite. This Suite has been rewritten using Adobe Flex and has the look and feel of a web browser. It is multi-threaded and has consolidated management data bases so that it is simpler and more efficient to use with the ability to scale up to 5 million objects. It scales out with the ability to manage all data, file, content, and block with common management, data protection, and search. The HCS also scales deep, providing closer integration with application and server APIs and providers as well as application transparency into the virtualized infrastructure of the VSP. As a firm supporter of the SNIA SMI-S initiative for open management interfaces, HCS supports SMI-S providers from Clariion and EVA, and in turn, supports the Hitachi VSP and Device Manager providers for other vendors’ use.
What about migration? Integration?
The Hitachi VSP and HCS are backward compatible and provide an easy migration path when our customers are ready to migrate off their current Hitachi USP and USP-V systems or their current Hitachi Storage Command Suite. They integrate well with our AMS midrange products, as well as our HNAS, HCP, and HDDS products for file and content services.
Instead of taking a storage centric view, Hitachi Data Systems has taken an application and server view of what is required to enable servers and applications to scale beyond their current limitations. We don’t expect our customers to rip and replace all their current investments. With our scale deep approach, users can continue with their current investments without sacrificing the benefits of 3D scaling.
For more information, please take a look at our videos and white papers. Also, please read the blogs published by Michael Hay, Claus Mikkelsen, Pete Gerr, and David Merrill on why today’s announcements are so important to the industry.
I look forward to your comments and questions!
Comments (3 )
HDS has pulled it off then, sounds like this great product can now really join consolidation with virtualization, while allowing for scalability and enterprise class design and components. Now please make sure the software matches the HW and not only device manager but also the other tools from the suite, notably Tuning Manager, live up to the new product.
Best of luck
[...] can find more information about our announcements here and blog postings from Hu, Claus, Michael, David, Ken and [...]
[...] Hu noted in his blog, another important point is that we wanted our existing USP-V(M), AMS, USP, etc. customers to be [...]