<?xml version="1.0" encoding="UTF-8" ?>
<?xml-stylesheet type="text/xsl" href="http://blogs.technet.com/utility/FeedStylesheets/rss.xsl" media="screen"?><rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:slash="http://purl.org/rss/1.0/modules/slash/" xmlns:wfw="http://wellformedweb.org/CommentAPI/"><channel><title>Microsoft SQL Server Performance Top Tip: Multi Processor (SMP) sudden death syndrome.</title><link>http://blogs.technet.com/mat_stephen/archive/2005/02/02/365370.aspx</link><description>You’ve been really proud of your nice new multi processor box, its got eight processors and some serious grunt. Everyone has been really pleased with the system’s lightening responsiveness and the big boss thinks you’re a super computer guru – looks like</description><dc:language>en-GB</dc:language><generator>CommunityServer 2.1 SP1 (Build: 61025.2)</generator><item><title>re: SQL Server Performance Top Tip: Multi Processor (SMP) sudden death syndrome.</title><link>http://blogs.technet.com/mat_stephen/archive/2005/02/02/365370.aspx#365396</link><pubDate>Wed, 02 Feb 2005 22:13:00 GMT</pubDate><guid isPermaLink="false">d5e57398-b9ef-4490-9955-07cbb4e4a80d:365396</guid><dc:creator>Mike Dimmick</dc:creator><description>I was going to ask what happens when the database runs out of file space, but then I suppose in a large system with SMP you would have one file per logical disk, which you'd make the size of the disk.&lt;br&gt;&lt;br&gt;Automatic growth is really only for development and test, I suppose. In production you need to monitor how full the files are and plan for additional disks when getting close to capacity. I'm not sure it's even possible to add an additional stripe to an existing striped set.&lt;br&gt;&lt;br&gt;At my employer, a small ISV/VAR, we don't see a lot of serious enterprise hardware. At least, the developers don't get to see it - some of our customers have it. The nearest we've got lately are some HP ProLiant DL380 G4 boxes set up with three disks in a RAID-5 configuration. It sounds like that customer might well run into the problem you describe, although the _first_ thing they'll need to do is put different databases on different spindles and the transaction logs on different spindles again.</description></item><item><title>Today is one of those I days I miss Take Outs.</title><link>http://blogs.technet.com/mat_stephen/archive/2005/02/02/365370.aspx#366080</link><pubDate>Thu, 03 Feb 2005 09:26:00 GMT</pubDate><guid isPermaLink="false">d5e57398-b9ef-4490-9955-07cbb4e4a80d:366080</guid><dc:creator>Enjoy Every Sandwich</dc:creator><description>Today is one of those I days I miss Take Outs.</description></item><item><title>re: SQL Server Performance Top Tip: Multi Processor (SMP) sudden death syndrome.</title><link>http://blogs.technet.com/mat_stephen/archive/2005/02/02/365370.aspx#366099</link><pubDate>Thu, 03 Feb 2005 15:14:00 GMT</pubDate><guid isPermaLink="false">d5e57398-b9ef-4490-9955-07cbb4e4a80d:366099</guid><dc:creator>Patrick Wellink</dc:creator><description>We had a situation like this.... and one processor would be up to 100% all the time....&lt;br&gt;&lt;br&gt;It became clear that paralel execution of the query caused the problem. It would give a TempDB error message and from then on the thread would push the processor to 100 % &lt;br&gt;&lt;br&gt;after disabeling paralel execution (setting query cost to 99999) plans this problem has gone away as wel as the tempDB errors.....</description></item><item><title>re: SQL Server Performance Top Tip: Multi Processor (SMP) sudden death syndrome.</title><link>http://blogs.technet.com/mat_stephen/archive/2005/02/02/365370.aspx#366914</link><pubDate>Fri, 04 Feb 2005 18:21:00 GMT</pubDate><guid isPermaLink="false">d5e57398-b9ef-4490-9955-07cbb4e4a80d:366914</guid><dc:creator>Matthew Stephen</dc:creator><description>Automatic growth is also good for low Maintenance databases - typically MSDE applications that only run with one processor</description></item><item><title>re: SQL Server Performance Top Tip: Multi Processor (SMP) sudden death syndrome.</title><link>http://blogs.technet.com/mat_stephen/archive/2005/02/02/365370.aspx#366915</link><pubDate>Fri, 04 Feb 2005 18:22:00 GMT</pubDate><guid isPermaLink="false">d5e57398-b9ef-4490-9955-07cbb4e4a80d:366915</guid><dc:creator>Matthew Stephen</dc:creator><description>Parallelism - I'll post a blog on this sometime soon</description></item><item><title>AutoGrow Apparently Means &amp;#8220;I Don&amp;#8217;t Care&amp;#8221; | SQLRockstar</title><link>http://blogs.technet.com/mat_stephen/archive/2005/02/02/365370.aspx#3246270</link><pubDate>Wed, 27 May 2009 20:13:51 GMT</pubDate><guid isPermaLink="false">d5e57398-b9ef-4490-9955-07cbb4e4a80d:3246270</guid><dc:creator>AutoGrow Apparently Means &amp;#8220;I Don&amp;#8217;t Care&amp;#8221; | SQLRockstar</dc:creator><description>&lt;p&gt;PingBack from &lt;a rel="nofollow" target="_new" href="http://thomaslarock.com/2009/05/autogrow-apparently-means-i-dont-care/"&gt;http://thomaslarock.com/2009/05/autogrow-apparently-means-i-dont-care/&lt;/a&gt;&lt;/p&gt;
</description></item></channel></rss>