Going Beyond Virtualization
by Hu Yoshida on May 14, 2007
Today, Hitachi and Hitachi Data Systems announced the new Hitachi Universal Storage Platform V, which will take us to new levels of virtualization, performance, scalability, and delivery of storage as a service.
The highlights of this announcement include:
Dynamic Provisioning which enables the ability to virtualize capacity within virtual volumes. This enables IT to eliminate the waste of allocated but unused space. While virtual volumes enabled non disruptive data movement for migration, replications, and tiered storage, it did not address the problem of allocating more space than is actually used in the primary volume as well as in all the replicas of that volume. Some storage systems can virtualize capacity with thin provisioning, but they don’t offer virtualization of the volume for non-disruptive data mobility. Hitachi’s Dynamic provisioning provides all this and much more. With Dynamic provisioning Hitachi also provides wide striping across large storage pools, where hundreds of disk drives can be employed to service an I/O request and provide self tuning.
Hitachi Data Systems also introduced a new trajectory in performance and scalability. When people hear the term “new trajectory” they may dismiss it as marketing hype, but I know of no other way to describe what Hitachi is doing in performance and scalability. We have increased the performance of external storage ports by 500% over our previous USP generation. We can not compare that to anything else on the market because, no one else does virtualization of storage by externally attaching heterogeneous storage to their control unit. We can manage 247 PB of storage, over 7 times the managed capacity of our previous product. Again there is no one else to compare this with in the industry since no one else has the virtual port connectivity to attach tens of thousands of FC ports nor the switched architecture to support this level of scalability.
With a new trajectory there are so many new facets for performance that there is no way to provide a comparison with other vendor’s products. The closest comparison that we can provide is in the maximum IOPs that a system can generate. We have measured 3.5 Million IOPs, which is up to 500% more than our competition. But this only measures the front end performance. It does not describe the improvements in the other parts of the USP V architecture like, single port performance, random write performance, random read performance, replication performance, back end internal disk performance, etc. Speaking of back end performance, the USP V has the industry’s first 4Gb/s switched backend for improved performance and availability.
With the enhancement in control unit virtualization through Dynamic Provisioning, the improvements in performance and scalability of the USP V, and the previous addition of content and file services, we can now provide storage as a set of common, block, file and object (content) services across heterogeneous data center assets. IT can now provide storage as a service to meet their business objectives, using a single platform that supports their current and future heterogeneous storage systems.
In some quarters, the debate still lingers as to whether storage virtualization should reside in SAN appliances and switches or in a storage controller. This announcement should put that debate to rest. SAN appliances and switches just do not have the architecture to meet the scalability, performance, functionality, and TCO demands of exploding storage growth and complexity. With the exception of volume pooling, the very best they can do is pass through the piece meal performance and functionality of their attached storage systems. They can not improve and revitalize the external storage systems with the common services of an enterprise class controller.
For more information on Hitachi Data System’s announcement of the Hitachi Universal Platform V. Please see the following links:
- Hitachi Introduces Services Oriented Storage Solutions, Enables Business and IT to Forge Stronger Partnership
- Hitachi Announces a New Dimension in Storage Virtualization, Introduces World’s First Enterprise-Class Implementation of Virtualization with Thin Provisioning
- Hitachi Announces a New Trajectory in Storage Performance and Scalability with Universal Storage Platform V—World’s Most Advanced Storage Services Platform
- IDC – Hitachi Data Systems: Thinner, Greener and at Your Services
Comments (3 )
My most sincere congratulations to everybody at Hitachi who have worked on developing Hitachi’s new trajectory in enterprise storage. From my days as a Hitachi employee, I believe that I know people in Odawara who have spent a great deal of time and have worked diligently developing this excellent technology.
It was not planned this way, but my company, EqualLogic, also introduced our version of thin provisioning today. Hitachi’s and EqualLogic’s new provisioning products give our respective customers much better control of the money they spend on storage.
Please be patient if you visit the EqualLogic web site (http://www.equallogic.com) to review our thin provisioning announcement – we are in the middle of a corporate relocation and are experiencing some web site performance problems.
Thanks you for the “new trajectory” concept.
I have long been thinking Information needs to be defined as “vector” or “scalar” based on a number of factors. Your “new trajectory” makes this at least possible with HDS.
In your statement:
“With a new trajectory there are so many new facets for performance that there is no way to provide a comparison with other vendor’s products. The closest comparison that we can provide is in the maximum IOPs that a system can generate.”, IOPs are scalar information as are all of the other “performance” facets.
I use “information” for scalar and “Information” for vector. The distinction is a definition in evolution.
I wait with great anticipation to hear the “vector” properties that give a “new trajectory”. These “vector” properties could finally enable the much needed SFO (Search, Find, Obtain) capability.
This should be the “Dawning of a new Age” for Storage.
Congratulations on the next generation USP-V. I have seen one of your customer presentations that shows the IOPs performance on the ‘external storage’ with a statement of 20K IOPs. I have a couple of questions regarding this :
1. Is that per port or in total – I hope per port…
2. With 296PB claimed external attach, by my calculation of 500GB 15K RPM FC disks, that would be 592K DDM’s. Lets assume 250 IOPs per DDM (generous) and JBOD – then that says you need to be able to drive 148 Million IOPs …………
Just wondering how the marketing figures actually scale up to real life application?
I work for IBM and these views / posts are my own thoughts and do not necessarily reflect the opinions of the IBM Corp.