United States
Site Map Contacts Hitachi Global Community
Hu's Blog - Data Storage and Virtualization Thought Leader Hitachi - Inspire the Next

Hu Yoshida's Blog - Vice President | Chief Technology Officer

Home > Corporate > HDS Blogs > HDS Bloggers > Hu's Blog
Products, Solutions and more

Hu's Blog

Correcting PCSE Waste

by Hu Yoshida on Aug 5, 2007

The Enterprise Strategy Group published a new white paper on Power, Cooling, Space Efficient Storage, in which they introduced the acronym, PCSE, for Power Cooling, Space Efficiency, and identified 11 storage technologies that could raise the PCSE for storage. These technologies included:

Thin Provisioning

Dynamic Volume Management

Snapshots

Writeable Snapshots

Data Compression

Data De-Duplication

Internal Storage Virtualization – Wide Stripe Groups

Storage System Consolidation – Multi-protocol Support

Intelligent Internal Tiering

MAID

Storage System Architectures

It is worth a read since it is vendor neutral. While there is a lot of marketing hype about Green IT and Green Storage features, this paper removes the hype and focuses on ways that storage can be used efficiently. ESG avoids the use of the “Green” term and talks about the “two major aspects of storage that lead to Power Cooling and Space Efficiency waste: 1) Buying more storage than you need; and 2) Underutilizing what you have.” They caution users “When analyzing any storage system it is important to understand how they address both these problems.

Since this paper is the property of ESG, I leave it to you to down load it from their site. I agree with the effectiveness of all their 11 technologies in reducing PCSE waste in storage. However I do want to comment on one technology which may at first seem to be an omission. That technology is storage virtualization.

Shortly after the introduction of Storage Area Networks, storage virtualization was introduced as the next major new technology to increase utilization of storage. Leveraging the connectivity of the SAN, it was introduced as a way to move the function of volume management from the host servers to a single point of control in the network. This SAN based volume manager could create a virtual volume from slices of extents from a “liquid pool of physical storage.” Sounds like a great solution for utilization of storage. So why didn’t ESG include this technology?

The answer can be found in an ESG Lab Validation Report on the USP V, in which ESG points out that:

“Aside from being on the same network, these storage systems (attached to a SAN) do not, in any way, interact with each other. They are discrete and isolated devices that are networked to host systems, but not to each other. In reality, network storage systems are not fully networked. They do not work in concert, but rather separately, individually, discretely and therefore, inefficiently.”

As a consequence, adding virtualization of storage onto a SAN platform is not effective in reducing PCSE waste. Your users still ask for an over allocation of storage capacity and use an only a fraction of the capacity that the allocation has tied up. SAN based volume virtualization does not address that waste of allocated but unused space. That allocated space also gets copied many times over for backup clones, replication, development test, data mining, data distribution, etc. SAN based volume virtualization does nothing to reduce the waste of redundancies.

The technologies that are listed above to address the waste of PCSE are all technologies which are done in storage controllers where the information and the resources exist to support these technologies. They don’t exist in SAN based appliances or intelligent switch blades.

Also you do not see all these technologies available from one vendor. Most of these technologies appear as islands of niche storage products running on midrange or low cost, non enterprise class storage systems.

Hitachi comes the closest to providing all of these technologies, as services on their, enterprise class, Universal Storage V platform. The only technology on that list that we do not provide today is MAID, Massive Array of idle disks, due to concerns over the inability to monitoring the health of the disks during idle periods. By providing storage virtualization in the USP V platform we can extend these technologies as services to heterogeneous storage products that attach behind the USP V and to heterogeneous application servers that use the USP V for storage. Unlike SAN based virtualization, the USP V storage virtualization can leverage nearly all of the new technologies for reducing PCSE waste.

Using the USP V services platform, we can also expand the ESG list of technologies by adding file and content services. With our HNAS files services we can add to efficiencies for file systems consolidation, tiering, and deduplication. With HCAP we add consolidation of content or object data, keeping one immutable copy of data, deleting and shredding data based on policy so that copies don’t get lost and linger around when they are no longer needed.

Take a look at the ESG paper and learn about the technologies that you can use today to reduce PCSE waste and visit our website to learn how USP V with HNAS and HCAP can further add to reducing this waste.

Related Posts Plugin for WordPress, Blogger...

Comments (2 )

Open systems guy on 06 Aug 2007 at 5:13 am

“Aside from being on the same network, these storage systems (attached to a SAN) do not, in any way, interact with each other. They are discrete and isolated devices that are networked to host systems, but not to each other. In reality, network storage systems are not fully networked. They do not work in concert, but rather separately, individually, discretely and therefore, inefficiently.”

-I was under the impression that storage virtualization technology spread a workload across multiple storage devices and by doing so, connected those devices to each other. I also thought that it was mainly used to increase the overall used resources of the disks being virtualized… if every LUN in a virtualized environment is striped across several boxes and each box contributes space and spindles toward the workload, how is that less efficient than non-virtualized storage?

Tony Asaro on 24 Aug 2007 at 1:26 pm

To Open Systems Guy – I think you may be taking that paragraph I wrote out of context. I am saying that storage networks today are inefficient because they are siloed and don’t interact with one another. Further, in that report and in others I state that external storage virtualization networks these storage systems and enables efficiency. So I am actually saying that external storage virtualization does make things more efficient.

What I think Hu is saying is that external storage virtualization doesn’t further optimize PCSE as I have outlined it in the ESG report he is referring to. However, that was not my thesis. The reason I did not bring up external storage virtualization in that specific report was that it is about optimizing a storage system through various technologies for PCSE. The report was about internal storage and not external storage virtualization.

External storage virtualization actually can also help to drive PCSE by getting better use out of what you already have. If you can get to greater utilization of your existing assets then you many not need to buy more capacity or a new storage system, which results in more effective PCSE. Further, if an external storage virtualization solution had one or more of the 11 items I suggested for PCSE and could extend those capabilities to external storage – then it certainly would have an impact on PCSE.

Tony

Hu Yoshida - Storage Virtualization Thought LeaderMust-read IT Blog

Hu Yoshida
Vice President and Chief Technology Officer

Connect with Us

     

Switch to our mobile site