Comments on: Dynamic Provisioning Delivered Through SOSS http://blogs.hds.com/hu/2007/11/dynamic_provisioning_delivered_through_soss.html Hu Yoshida, VP and CTO of Hitachi Data Systems, provides his insight into industry issues, discusses in his own words storage best practices, and provides realistic solutions to real storage problems of current and next generation storage environments. Fri, 29 Aug 2008 19:36:55 +0000 http://wordpress.org/?v=2.0.4 by: Carl http://blogs.hds.com/hu/2007/11/dynamic_provisioning_delivered_through_soss.html#comment-70577 Tue, 13 May 2008 06:04:29 +0000 http://blogs.hds.com/hu/2007/11/dynamic_provisioning_delivered_through_soss.html#comment-70577 I would like to know if HDP will ever be available on the USP/NSC platform. We have 2 NSC55s which we purchased with the expectation of virtualization and thin provisioning, and we cannot just replace them with USP VMs to get thin provisioning. I would like to know if HDP will ever be available on the USP/NSC platform. We have 2 NSC55s which we purchased with the expectation of virtualization and thin provisioning, and we cannot just replace them with USP VMs to get thin provisioning.

]]> by: Ahlert Hysing http://blogs.hds.com/hu/2007/11/dynamic_provisioning_delivered_through_soss.html#comment-42435 Fri, 09 Nov 2007 10:45:05 +0000 http://blogs.hds.com/hu/2007/11/dynamic_provisioning_delivered_through_soss.html#comment-42435 I lack a lot of arguments for why. The basic argument is the bad utilization of the the sentralized storage. If it can be documented less then 50 percent, dynamic provisioning is a must. If utilization is above 70 percent, dynamic provisioning is a yawn, a function "Nice to Have". One ot the most interesting functions of virtualization og storage must be the ability to pretend the storage is much bigger than it realy is. This argument was the basic argument for virtualization of memory in computers at the time memory was based on magnetic core. 4 KB pages is still a foundation of memory management. The other argument should be a discussion about sizes of LUNs. Change of size often requires work at night because of the necessety of IPL. The argument that applications are not affected is a consequence. Used as an primary argument it disguises the challenge. The result is only seen as extra overtime for the professionals doing the changes when others sleep. Ahlert MSEE Senior Editor Computerworld Norway I lack a lot of arguments for why. The basic argument is the bad utilization of the the sentralized storage. If it can be documented less then 50 percent, dynamic provisioning is a must. If utilization is above 70 percent, dynamic provisioning is a yawn, a function “Nice to Have”.
One ot the most interesting functions of virtualization og storage must be the ability to pretend the storage is much bigger than it realy is.
This argument was the basic argument for virtualization of memory in computers at the time memory was based on magnetic core. 4 KB pages is still a foundation of memory management.
The other argument should be a discussion about sizes of LUNs. Change of size often requires work at night because of the necessety of IPL.
The argument that applications are not affected is a consequence. Used as an primary argument it disguises the challenge. The result is only seen as extra overtime for the professionals doing the changes when others sleep.

Ahlert
MSEE

Senior Editor Computerworld Norway

]]>