<?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: Unused indexes by single query</title>
	<atom:link href="http://www.mysqlperformanceblog.com/2008/09/12/unused-indexes-by-single-query/feed/" rel="self" type="application/rss+xml" />
	<link>http://www.mysqlperformanceblog.com/2008/09/12/unused-indexes-by-single-query/</link>
	<description>Percona&#039;s MySQL &#38; InnoDB performance and scalability blog</description>
	<lastBuildDate>Sat, 11 Feb 2012 16:45:54 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.org/?v=3.2.1</generator>
	<item>
		<title>By: Normann</title>
		<link>http://www.mysqlperformanceblog.com/2008/09/12/unused-indexes-by-single-query/comment-page-1/#comment-802299</link>
		<dc:creator>Normann</dc:creator>
		<pubDate>Tue, 22 Mar 2011 21:32:32 +0000</pubDate>
		<guid isPermaLink="false">http://www.mysqlperformanceblog.com/?p=501#comment-802299</guid>
		<description>to enable stats:
percona &lt; 5.5:
set global userstat_running = 1;
5.5:
set global userstat = 1;</description>
		<content:encoded><![CDATA[<p>to enable stats:<br />
percona &lt; 5.5:<br />
set global userstat_running = 1;<br />
5.5:<br />
set global userstat = 1;</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Baron Schwartz</title>
		<link>http://www.mysqlperformanceblog.com/2008/09/12/unused-indexes-by-single-query/comment-page-1/#comment-782790</link>
		<dc:creator>Baron Schwartz</dc:creator>
		<pubDate>Sun, 14 Nov 2010 21:39:55 +0000</pubDate>
		<guid isPermaLink="false">http://www.mysqlperformanceblog.com/?p=501#comment-782790</guid>
		<description>Yes.  See mk-index-usage in Maatkit.  Also see http://www.mysqlperformanceblog.com/2010/11/11/advanced-index-analysis-with-mk-index-usage/ for example usage.</description>
		<content:encoded><![CDATA[<p>Yes.  See mk-index-usage in Maatkit.  Also see <a href="http://www.mysqlperformanceblog.com/2010/11/11/advanced-index-analysis-with-mk-index-usage/" rel="nofollow">http://www.mysqlperformanceblog.com/2010/11/11/advanced-index-analysis-with-mk-index-usage/</a> for example usage.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Webo</title>
		<link>http://www.mysqlperformanceblog.com/2008/09/12/unused-indexes-by-single-query/comment-page-1/#comment-782747</link>
		<dc:creator>Webo</dc:creator>
		<pubDate>Sun, 14 Nov 2010 12:45:52 +0000</pubDate>
		<guid isPermaLink="false">http://www.mysqlperformanceblog.com/?p=501#comment-782747</guid>
		<description>Is there any solution that does not require the patch?
Even if it requires some manual work.</description>
		<content:encoded><![CDATA[<p>Is there any solution that does not require the patch?<br />
Even if it requires some manual work.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Vadim</title>
		<link>http://www.mysqlperformanceblog.com/2008/09/12/unused-indexes-by-single-query/comment-page-1/#comment-510105</link>
		<dc:creator>Vadim</dc:creator>
		<pubDate>Tue, 17 Mar 2009 21:52:47 +0000</pubDate>
		<guid isPermaLink="false">http://www.mysqlperformanceblog.com/?p=501#comment-510105</guid>
		<description>xx

I fixed it</description>
		<content:encoded><![CDATA[<p>xx</p>
<p>I fixed it</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: xx</title>
		<link>http://www.mysqlperformanceblog.com/2008/09/12/unused-indexes-by-single-query/comment-page-1/#comment-504592</link>
		<dc:creator>xx</dc:creator>
		<pubDate>Fri, 13 Mar 2009 10:09:34 +0000</pubDate>
		<guid isPermaLink="false">http://www.mysqlperformanceblog.com/?p=501#comment-504592</guid>
		<description>ERROR 1064 (42000): Something is wrong in your syntax obok &#039;IS

&quot;IS&quot; is restricted keyword</description>
		<content:encoded><![CDATA[<p>ERROR 1064 (42000): Something is wrong in your syntax obok &#8216;IS</p>
<p>&#8220;IS&#8221; is restricted keyword</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Vadim</title>
		<link>http://www.mysqlperformanceblog.com/2008/09/12/unused-indexes-by-single-query/comment-page-1/#comment-357036</link>
		<dc:creator>Vadim</dc:creator>
		<pubDate>Sun, 21 Sep 2008 17:03:31 +0000</pubDate>
		<guid isPermaLink="false">http://www.mysqlperformanceblog.com/?p=501#comment-357036</guid>
		<description>Tamar,

See previous post
http://www.mysqlperformanceblog.com/2008/09/12/googles-user_statistics-v2-port-and-changes/</description>
		<content:encoded><![CDATA[<p>Tamar,</p>
<p>See previous post<br />
<a href="http://www.mysqlperformanceblog.com/2008/09/12/googles-user_statistics-v2-port-and-changes/" rel="nofollow">http://www.mysqlperformanceblog.com/2008/09/12/googles-user_statistics-v2-port-and-changes/</a></p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Tamar Rosen</title>
		<link>http://www.mysqlperformanceblog.com/2008/09/12/unused-indexes-by-single-query/comment-page-1/#comment-357028</link>
		<dc:creator>Tamar Rosen</dc:creator>
		<pubDate>Sun, 21 Sep 2008 15:07:50 +0000</pubDate>
		<guid isPermaLink="false">http://www.mysqlperformanceblog.com/?p=501#comment-357028</guid>
		<description>Where can I get this patch, and for what version of MySQL?</description>
		<content:encoded><![CDATA[<p>Where can I get this patch, and for what version of MySQL?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Vladimir</title>
		<link>http://www.mysqlperformanceblog.com/2008/09/12/unused-indexes-by-single-query/comment-page-1/#comment-355543</link>
		<dc:creator>Vladimir</dc:creator>
		<pubDate>Tue, 16 Sep 2008 15:12:46 +0000</pubDate>
		<guid isPermaLink="false">http://www.mysqlperformanceblog.com/?p=501#comment-355543</guid>
		<description>First of all I agree that it&#039;s safer to keep indexes than to drop them in most cases. Next I&#039;m not sure when it&#039;s really ok to use this stats to correct your indexes - after a day of work, or a week, or maybe a month? In other words its highly application dependent question - when your data reaches some &quot;steady&quot; state in regard to how it&#039;s treated by optimizer. Now let&#039;s say you think for your dataflow 1 week makes sense. This means that you have to wait 1 week before you can make a decision. Too much for many applications I&#039;d say. What I&#039;d propose is to go for static analysis, and if possible - do that on the design stage.</description>
		<content:encoded><![CDATA[<p>First of all I agree that it&#8217;s safer to keep indexes than to drop them in most cases. Next I&#8217;m not sure when it&#8217;s really ok to use this stats to correct your indexes &#8211; after a day of work, or a week, or maybe a month? In other words its highly application dependent question &#8211; when your data reaches some &#8220;steady&#8221; state in regard to how it&#8217;s treated by optimizer. Now let&#8217;s say you think for your dataflow 1 week makes sense. This means that you have to wait 1 week before you can make a decision. Too much for many applications I&#8217;d say. What I&#8217;d propose is to go for static analysis, and if possible &#8211; do that on the design stage.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: peter</title>
		<link>http://www.mysqlperformanceblog.com/2008/09/12/unused-indexes-by-single-query/comment-page-1/#comment-355321</link>
		<dc:creator>peter</dc:creator>
		<pubDate>Mon, 15 Sep 2008 16:38:22 +0000</pubDate>
		<guid isPermaLink="false">http://www.mysqlperformanceblog.com/?p=501#comment-355321</guid>
		<description>I would add couple of points.

1) It is right you should START  using this patch on development as with any unproven code. However in general this patch is low impact and in particular previous versions of it are used on many high traffic web site.  So this is not development only toy.

2) It is important to run patch long enough so you catch all activities.  Sometimes there are activities which are done on weekly basics only in others (ie monthly billing processing) it can be happening once per month and you need to keep it covered.

3) If you have a choice (like Master-Master, setup or Multiple Slaves) drop indexes  in a way you can switch back to version with indexes quickly.  If you do mistake and drop needed indexes the system may effectively go down and adding index back may just take too much time.</description>
		<content:encoded><![CDATA[<p>I would add couple of points.</p>
<p>1) It is right you should START  using this patch on development as with any unproven code. However in general this patch is low impact and in particular previous versions of it are used on many high traffic web site.  So this is not development only toy.</p>
<p>2) It is important to run patch long enough so you catch all activities.  Sometimes there are activities which are done on weekly basics only in others (ie monthly billing processing) it can be happening once per month and you need to keep it covered.</p>
<p>3) If you have a choice (like Master-Master, setup or Multiple Slaves) drop indexes  in a way you can switch back to version with indexes quickly.  If you do mistake and drop needed indexes the system may effectively go down and adding index back may just take too much time.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Baron Schwartz</title>
		<link>http://www.mysqlperformanceblog.com/2008/09/12/unused-indexes-by-single-query/comment-page-1/#comment-355315</link>
		<dc:creator>Baron Schwartz</dc:creator>
		<pubDate>Mon, 15 Sep 2008 15:52:54 +0000</pubDate>
		<guid isPermaLink="false">http://www.mysqlperformanceblog.com/?p=501#comment-355315</guid>
		<description>Pat,

&quot;probably&quot; worth it is likely to be an overstatement.  &quot;May be worth it&quot; might be more accurate.  As Vadim said, the impact of extra indexes can be very large indeed.

Of course all the comments on this post are correct, and we really all agree: you have to think carefully about it.  Nobody is advocating to just drop all indexes without further thought.</description>
		<content:encoded><![CDATA[<p>Pat,</p>
<p>&#8220;probably&#8221; worth it is likely to be an overstatement.  &#8220;May be worth it&#8221; might be more accurate.  As Vadim said, the impact of extra indexes can be very large indeed.</p>
<p>Of course all the comments on this post are correct, and we really all agree: you have to think carefully about it.  Nobody is advocating to just drop all indexes without further thought.</p>
]]></content:encoded>
	</item>
</channel>
</rss>

