<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	xmlns:sy="http://purl.org/rss/1.0/modules/syndication/"
		>
<channel>
	<title>Comments on: Virtualizing high performance SQL &#8211; VMFS or RDMs?</title>
	<atom:link href="http://vmetc.com/2008/12/08/virtualizing-high-performance-sql-vmfs-or-rdms/feed/" rel="self" type="application/rss+xml" />
	<link>http://vmetc.com/2008/12/08/virtualizing-high-performance-sql-vmfs-or-rdms/</link>
	<description>Go Green with Virtualization. Go UGLY Green with vmetc.com.</description>
	<lastBuildDate>Sat, 13 Mar 2010 09:25:14 +0000</lastBuildDate>
	<generator>http://wordpress.org/?v=2.9.1</generator>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
		<item>
		<title>By: Cameron Gocke</title>
		<link>http://vmetc.com/2008/12/08/virtualizing-high-performance-sql-vmfs-or-rdms/comment-page-1/#comment-3857</link>
		<dc:creator>Cameron Gocke</dc:creator>
		<pubDate>Wed, 17 Jun 2009 00:43:37 +0000</pubDate>
		<guid isPermaLink="false">http://vmetc.com/?p=2153#comment-3857</guid>
		<description>Yes indeed.  In fact at the vSpehere 4 launch event the presenter indicated that SRM support for NFS is due to come out in the second half of 2009.</description>
		<content:encoded><![CDATA[<p>Yes indeed.  In fact at the vSpehere 4 launch event the presenter indicated that SRM support for NFS is due to come out in the second half of 2009.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Cameron Gocke</title>
		<link>http://vmetc.com/2008/12/08/virtualizing-high-performance-sql-vmfs-or-rdms/comment-page-1/#comment-2655</link>
		<dc:creator>Cameron Gocke</dc:creator>
		<pubDate>Tue, 16 Jun 2009 18:43:37 +0000</pubDate>
		<guid isPermaLink="false">http://vmetc.com/?p=2153#comment-2655</guid>
		<description>Yes indeed.  In fact at the vSpehere 4 launch event the presenter indicated that SRM support for NFS is due to come out in the second half of 2009.</description>
		<content:encoded><![CDATA[<p>Yes indeed.  In fact at the vSpehere 4 launch event the presenter indicated that SRM support for NFS is due to come out in the second half of 2009.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: rbrambley</title>
		<link>http://vmetc.com/2008/12/08/virtualizing-high-performance-sql-vmfs-or-rdms/comment-page-1/#comment-2648</link>
		<dc:creator>rbrambley</dc:creator>
		<pubDate>Tue, 16 Jun 2009 00:22:50 +0000</pubDate>
		<guid isPermaLink="false">http://vmetc.com/?p=2153#comment-2648</guid>
		<description>Ed,&lt;br&gt;&lt;br&gt;Great point about the &quot;big picture&quot;. Things are changing though ....</description>
		<content:encoded><![CDATA[<p>Ed,</p>
<p>Great point about the &#8220;big picture&#8221;. Things are changing though &#8230;.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Ed Saipetch</title>
		<link>http://vmetc.com/2008/12/08/virtualizing-high-performance-sql-vmfs-or-rdms/comment-page-1/#comment-2647</link>
		<dc:creator>Ed Saipetch</dc:creator>
		<pubDate>Tue, 16 Jun 2009 00:11:42 +0000</pubDate>
		<guid isPermaLink="false">http://vmetc.com/?p=2153#comment-2647</guid>
		<description>One thing to remember with the storage decisions comes compatibility with things like Site Recovery Manager (SRM), Fault Tolerance and other things that people may consider using in their environment.  Not all features of VMware are supported on all storage types.  I believe SRM still isn&#039;t supported over NFS at this point in time.</description>
		<content:encoded><![CDATA[<p>One thing to remember with the storage decisions comes compatibility with things like Site Recovery Manager (SRM), Fault Tolerance and other things that people may consider using in their environment.  Not all features of VMware are supported on all storage types.  I believe SRM still isn&#39;t supported over NFS at this point in time.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: rbrambley</title>
		<link>http://vmetc.com/2008/12/08/virtualizing-high-performance-sql-vmfs-or-rdms/comment-page-1/#comment-2646</link>
		<dc:creator>rbrambley</dc:creator>
		<pubDate>Mon, 15 Jun 2009 23:20:47 +0000</pubDate>
		<guid isPermaLink="false">http://vmetc.com/?p=2153#comment-2646</guid>
		<description>Cameron,&lt;br&gt;&lt;br&gt;NFS performance is surprising isn&#039;t it? Thanks for posting your NFS/iSCSI/RDM test results here!</description>
		<content:encoded><![CDATA[<p>Cameron,</p>
<p>NFS performance is surprising isn&#39;t it? Thanks for posting your NFS/iSCSI/RDM test results here!</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Cameron Gocke</title>
		<link>http://vmetc.com/2008/12/08/virtualizing-high-performance-sql-vmfs-or-rdms/comment-page-1/#comment-2641</link>
		<dc:creator>Cameron Gocke</dc:creator>
		<pubDate>Mon, 15 Jun 2009 21:06:00 +0000</pubDate>
		<guid isPermaLink="false">http://vmetc.com/?p=2153#comment-2641</guid>
		<description>Well, I am finally getting around to running some performance benchmarks against my storage.  The results I&#039;m getting, if accurate, seem pretty crazy to me.  My main problem, however, is that I don&#039;t have a storage system that is not being used in production and so I worry that the production traffic is skewing my numbers even though all I really care about are the relative comparisons not the raw values.&lt;br&gt;&lt;br&gt;Basically, the test I am running is using SQLIO to test the storage performance of a VMDK file stored on an NFS datastore, vs. an RDM accessed via iSCSI, vs. the MS initiator run directly from the Guest OS accessing an iSCSI LUN.  On my NetApp all 3 are stored on the same aggregate of disks.  Because I do have production IO hitting those disks, the only good way I could think of to eliminate skewed data from workload changes throughout the day was to run all 3 IO tests simultaneously.&lt;br&gt;&lt;br&gt;The results that I am seeing are as follows:&lt;br&gt;The VMDK stored on the NFS datastore dramatically beats out the other 2 options hands down for 8KB sequential IO both for reads and writes.  The VMDK stored on NFS datastore also demonstrates much stronger performance on random writes 64KB and larger as well as on sequential reads of all sizes.  As you get larger sequential reads, however, the RDM and MS initiator start to close the gap.&lt;br&gt;&lt;br&gt;The MS initiator outperformed the other 2 storage volumes only on 8KB random write IO and the difference was substantial.  For some reason from 64KB random writes and larger the VMDK on NFS then performed much better.&lt;br&gt;&lt;br&gt;For random reads of all sizes the performance of all three volumes was close enough to be insignificant.&lt;br&gt;&lt;br&gt;The only category where the RDM sustained higher IO was for 8KB random reads, but only by a very small margin.&lt;br&gt;&lt;br&gt;The latency on the VMDK over NFS was always substantially lower than the other 2 with the RDM and MS initiator generally being pretty close to one another.&lt;br&gt;&lt;br&gt;The throughput for the VMDK over NFS was much better for random writes, but for all other IO the results were very close for all 3.&lt;br&gt;&lt;br&gt;I don&#039;t know if I&#039;d be willing to 100% stand behind these results, and I really wish I had a dedicated NetApp to run this test against, but I thought I&#039;d post the results here anyway.</description>
		<content:encoded><![CDATA[<p>Well, I am finally getting around to running some performance benchmarks against my storage.  The results I&#39;m getting, if accurate, seem pretty crazy to me.  My main problem, however, is that I don&#39;t have a storage system that is not being used in production and so I worry that the production traffic is skewing my numbers even though all I really care about are the relative comparisons not the raw values.</p>
<p>Basically, the test I am running is using SQLIO to test the storage performance of a VMDK file stored on an NFS datastore, vs. an RDM accessed via iSCSI, vs. the MS initiator run directly from the Guest OS accessing an iSCSI LUN.  On my NetApp all 3 are stored on the same aggregate of disks.  Because I do have production IO hitting those disks, the only good way I could think of to eliminate skewed data from workload changes throughout the day was to run all 3 IO tests simultaneously.</p>
<p>The results that I am seeing are as follows:<br />The VMDK stored on the NFS datastore dramatically beats out the other 2 options hands down for 8KB sequential IO both for reads and writes.  The VMDK stored on NFS datastore also demonstrates much stronger performance on random writes 64KB and larger as well as on sequential reads of all sizes.  As you get larger sequential reads, however, the RDM and MS initiator start to close the gap.</p>
<p>The MS initiator outperformed the other 2 storage volumes only on 8KB random write IO and the difference was substantial.  For some reason from 64KB random writes and larger the VMDK on NFS then performed much better.</p>
<p>For random reads of all sizes the performance of all three volumes was close enough to be insignificant.</p>
<p>The only category where the RDM sustained higher IO was for 8KB random reads, but only by a very small margin.</p>
<p>The latency on the VMDK over NFS was always substantially lower than the other 2 with the RDM and MS initiator generally being pretty close to one another.</p>
<p>The throughput for the VMDK over NFS was much better for random writes, but for all other IO the results were very close for all 3.</p>
<p>I don&#39;t know if I&#39;d be willing to 100% stand behind these results, and I really wish I had a dedicated NetApp to run this test against, but I thought I&#39;d post the results here anyway.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: rbrambley</title>
		<link>http://vmetc.com/2008/12/08/virtualizing-high-performance-sql-vmfs-or-rdms/comment-page-1/#comment-2153</link>
		<dc:creator>rbrambley</dc:creator>
		<pubDate>Tue, 10 Mar 2009 11:03:13 +0000</pubDate>
		<guid isPermaLink="false">http://vmetc.com/?p=2153#comment-2153</guid>
		<description>Cameron,&lt;br&gt;&lt;br&gt;True. NFS performance has been proven to be as equal as iSCSI in most&lt;br&gt;scenarios. So, without knowing your exact application mix, I would say&lt;br&gt;that the only benefits you are getting from RDMs is possible SAN&lt;br&gt;monitoring or the VMs native OS file system features. Probably not&lt;br&gt;significant performance benefits. Not sure if those benefits are worth&lt;br&gt;the complexity of RDMs?</description>
		<content:encoded><![CDATA[<p>Cameron,</p>
<p>True. NFS performance has been proven to be as equal as iSCSI in most<br />scenarios. So, without knowing your exact application mix, I would say<br />that the only benefits you are getting from RDMs is possible SAN<br />monitoring or the VMs native OS file system features. Probably not<br />significant performance benefits. Not sure if those benefits are worth<br />the complexity of RDMs?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Cameron Gocke</title>
		<link>http://vmetc.com/2008/12/08/virtualizing-high-performance-sql-vmfs-or-rdms/comment-page-1/#comment-2149</link>
		<dc:creator>Cameron Gocke</dc:creator>
		<pubDate>Mon, 09 Mar 2009 19:43:35 +0000</pubDate>
		<guid isPermaLink="false">http://vmetc.com/?p=2153#comment-2149</guid>
		<description>I wonder if there are any comparisons on running a similar environment using VMDK files over NFS datastores.  Our storage is a NetApp system and we use either iSCSI or NFS for our ESX environment.  Right now all of our OS/Program Files are on the NFS datastores and we use iSCSI RDMs for data volumes.  I&#039;ve often wondered if there was any real benefit for me even bothering with the RDMs there since the published papers comparing NFS to iSCSI on NetApp for VMDKs has often favored NFS.</description>
		<content:encoded><![CDATA[<p>I wonder if there are any comparisons on running a similar environment using VMDK files over NFS datastores.  Our storage is a NetApp system and we use either iSCSI or NFS for our ESX environment.  Right now all of our OS/Program Files are on the NFS datastores and we use iSCSI RDMs for data volumes.  I&#39;ve often wondered if there was any real benefit for me even bothering with the RDMs there since the published papers comparing NFS to iSCSI on NetApp for VMDKs has often favored NFS.</p>
]]></content:encoded>
	</item>
</channel>
</rss>
