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

Zero Page Reclaim

by Hu Yoshida on Mar 19, 2009

It has often been difficult to find IT shops who accurately measure the improvements that they make in their shops due to the pressure of other demands. While this may make them seem more productive, it hampers their ability to quantify the improvements they have made and get credit for the monetary impact they have on the bottom line of their businesses.

With the economic down turn, we are beginning to find more IT shops willing to invest the extra effort to do this quantification and the results can be significant.

Recently one of our global financial customers completed a study on the use of Zero Page Reclaim, a feature of Hitachi Dynamic Provisioning (HDP). This feature which is available on the USP V and VM storage systems can reclaim allocated but unused capacity on existing open systems volumes without disruption to the application. Simply by using the virtualization capability of the USP V/VM, we can move a “fat” volume into an HDP pool, page by page, and after the move, reclaim the pages that are filled with zeros, so that the “fat” volume is reduced to a “thin provisioned” volume.

The net results are shown here in this table:
average-capacity-savings

The value benefits that they were able to show were:

  • Reclamation of about 185TB valued in excess of $2M USD
  • Maintenance Avoidance of an additional $2M USD
  • 40 to 60% reduction in physical storage
  • 40% reduction in environmental requirements
  • 40% reduction in real estate
  • Improvement in operational efficiencies due to the ease of provisioning

The improvement in operational efficiencies could be further quantified through the ability to do “thin” copies, moves, replication, migration, and tiering of only the used capacity, the ability to quickly provision new volumes and expansion of volumes with pre-formatted pages, and the performance improvement of pages which are striped across all the disks in the HDP pool.

Zero Page Reclaim on the USP V/VM is one of the most effective tools for enhancing your current assets and providing significant cost savings that go straight to your bottom line.

Related Posts Plugin for WordPress, Blogger...

Comments (10 )

RJP on 20 Mar 2009 at 4:49 am

Good post, Hu.

[...] pages of storage dynamically from a preformatted pool of storage, do thin moves and copies, zero page reclaims,  and improve performance by striping pages across the width of the HDP storage pool. With storage [...]

Kulvinder on 23 Mar 2009 at 7:45 am

This is very good post

[...] allocated but unused space. This service uses a feature of our Dynamic Provisioning product called, Zero Page Reclaim. Zero Page Reclaim is used after a migration or restore of a normal volume into a Dynamic [...]

[...] format, we can reclaim that page and return it to the pool for other allocations. This is known as Zero Page Reclaim. With other types of thin provisioning where the page or chunklet is a subset of a larger chunk, [...]

[...] pages of storage dynamically from a preformatted pool of storage, do thin moves and copies, zero page reclaims,  and improve performance by striping pages across the width of the HDP storage pool. With storage [...]

[...] not actually save you any money on disk.  This is where HDS’s Zero Page Reclaim can help.  Hitachi’s Dynamic Provisioning (with ZPR) can scan a LUN for sections where all the bytes are zero and reclaim that space for [...]

[...] along with its Thin Provisioning and Zero Page Reclaim capabilities, can actually reclaim the unused storage by virtualizing the application’s volume [...]

Tony Boucher on 26 Sep 2011 at 2:25 pm

Hello, I have a question about zero page reclaim.

We are supporting 400 Virtual LPARS running AIX 6.1 that the storage is being presented to the LPAR’s through Virtual I/O servers that are direct attached to HDS storage. We are being instructed that when we decommission storage from LPAR’s that we need to write “0″s to the LUN so that the Zero Page reclaim will work.

IE: dd if=/dev/zero of=/dev/hdisk10

We have tried this with mixed results on the AIX side. Most of our LUN’s are
100-500GB. It takes hours to try and do this. 10% of the time it never completes.

Is this the correct procedure ??

P.S> We present all the LUNS to 10 Virtual I/O servers so we can move LPAR’s between the frames. We also manage approximately 4000 SAN luns on these Virtual I/O servers.

Thanks,
Tony

Hu Yoshida on 10 Oct 2011 at 12:07 pm

Tony,

I am not certain about the requirement to use Zero Page Reclaim (ZPR) after a storage decommission. ZPR is used to reclaim pool capacity from LUNs that are still deployed but have locked away unnecessarily pool capacity.
Needing to reclaim capacity implies that you are trying to maintain some amount or oversubscription. If you are not interested in oversubscription then you do not need to use ZPR or any other steps to reclaim pool capacity. However if you are trying to have a larger total capacity defined to LUNs versus the capacity of the supporting pool then various steps can be used to ‘stay thin’.
Since it sounds like you are withdrawing (redeploying) the LUNs, there is a better alternative than ZPR. The first is to simply delete the LUNs (DPVOLs in our case) and define a brand new LUN. The delete will release all the capacity. If you are using a VSP then we now support VSP format to a thin LUN which will simply release the LUN’s mapped pool capacity

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