<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>Reclaim on Matt Allford</title><link>https://www.mattallford.com/tags/reclaim/</link><description>Recent content in Reclaim on Matt Allford</description><generator>Hugo</generator><language>en</language><lastBuildDate>Mon, 13 Jun 2016 11:28:44 +0000</lastBuildDate><atom:link href="https://www.mattallford.com/tags/reclaim/index.xml" rel="self" type="application/rss+xml"/><item><title>ESXi and Nimble Storage Unmap SCSI Blocks</title><link>https://www.mattallford.com/esxi-and-nimble-storage-unmap-scsi-blocks/</link><pubDate>Mon, 13 Jun 2016 11:28:44 +0000</pubDate><guid>https://www.mattallford.com/esxi-and-nimble-storage-unmap-scsi-blocks/</guid><description>&lt;h1 id="overview"&gt;OverView&lt;/h1&gt;
&lt;p&gt;Back in vSphere 5 a new VAAI primitive called unmap was released. When using thin provisioned volumes or LUNs for VMFS datastores, actions such as storage vMotion, consolidating snapshots and deleting data (such as VMDK files) do not automatically reclaim the space on the underlying datastore. For example, if you have a thin provisioned 1TB datastore that has 2 x virtual machines using 400GB each, and you delete one of these virtual machines (using the &amp;lsquo;delete from disk&amp;rsquo; action), the used space on the datastore will still report as 800GB.&lt;/p&gt;</description></item></channel></rss>