<?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: MySQL InnoDb inserts are slow, really slow!</title>
	<atom:link href="http://blog.cnizz.com/2010/06/12/mysql-innodb-inserts-are-slow-really-slow/feed/" rel="self" type="application/rss+xml" />
	<link>http://blog.cnizz.com/2010/06/12/mysql-innodb-inserts-are-slow-really-slow/</link>
	<description>Salt Lake City, Utah Developer / Human / Blogger</description>
	<lastBuildDate>Wed, 01 May 2013 14:55:16 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.org/?v=3.5.1</generator>
	<item>
		<title>By: Ajay Divakaran</title>
		<link>http://blog.cnizz.com/2010/06/12/mysql-innodb-inserts-are-slow-really-slow/comment-page-1/#comment-189653</link>
		<dc:creator>Ajay Divakaran</dc:creator>
		<pubDate>Wed, 13 Mar 2013 12:10:55 +0000</pubDate>
		<guid isPermaLink="false">http://blog.cnizz.com/?p=445#comment-189653</guid>
		<description><![CDATA[Chris,
InnoDB is a pretty good engine. However, it highly relies on being &#039;tuned&#039;. One thing is that if your inserts are not in the order of increasing primary keys, innoDB can take a bit longer than MyISAM. This can easily be overcome by setting a higher innodb_buffer_pool_size. My suggestion is to set it at 60-70% of your total RAM.
I am running 4 such servers in production now, inserting about 3.5 million rows a minute. They already have close to 3 Terabytes. InnoDB it had to be , because of the highly concurrent inserts. 
There are further ways to speed up inserts. And I&#039;ve benchmarked some.]]></description>
		<content:encoded><![CDATA[<p>Chris,<br />
InnoDB is a pretty good engine. However, it highly relies on being &#8216;tuned&#8217;. One thing is that if your inserts are not in the order of increasing primary keys, innoDB can take a bit longer than MyISAM. This can easily be overcome by setting a higher innodb_buffer_pool_size. My suggestion is to set it at 60-70% of your total RAM.<br />
I am running 4 such servers in production now, inserting about 3.5 million rows a minute. They already have close to 3 Terabytes. InnoDB it had to be , because of the highly concurrent inserts.<br />
There are further ways to speed up inserts. And I&#8217;ve benchmarked some.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: chris</title>
		<link>http://blog.cnizz.com/2010/06/12/mysql-innodb-inserts-are-slow-really-slow/comment-page-1/#comment-72415</link>
		<dc:creator>chris</dc:creator>
		<pubDate>Wed, 18 Aug 2010 19:52:18 +0000</pubDate>
		<guid isPermaLink="false">http://blog.cnizz.com/?p=445#comment-72415</guid>
		<description><![CDATA[Good to know Nick, thanks for the contribution.  Makes sense after reading &lt;a href=&quot;http://dev.mysql.com/doc/refman/5.0/en/commit.html&quot; rel=&quot;nofollow&quot;&gt;http://dev.mysql.com/doc/refman/5.0/en/commit.html&lt;/a&gt;.]]></description>
		<content:encoded><![CDATA[<p>Good to know Nick, thanks for the contribution.  Makes sense after reading <a href="http://dev.mysql.com/doc/refman/5.0/en/commit.html" rel="nofollow">http://dev.mysql.com/doc/refman/5.0/en/commit.html</a>.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Nick</title>
		<link>http://blog.cnizz.com/2010/06/12/mysql-innodb-inserts-are-slow-really-slow/comment-page-1/#comment-72413</link>
		<dc:creator>Nick</dc:creator>
		<pubDate>Wed, 18 Aug 2010 18:45:12 +0000</pubDate>
		<guid isPermaLink="false">http://blog.cnizz.com/?p=445#comment-72413</guid>
		<description><![CDATA[I found that if I set AutoCommit=0 before the inserts, then commit after Inserts i can cut off a lot of time. still not as fast but around 2 seconds.
doesn&#039;t work if the call is to a stored procedure(which does the insert).

On my development machine i&#039;m not experiencing this at all.  (i dont need to add the autocommit = 0 and i still see &lt; 1 second performance)

the production machine is a VPS, so im suspecting that disk usage may be the primary culprit and someone else has a high disk usage application.

Thanks for the link.]]></description>
		<content:encoded><![CDATA[<p>I found that if I set AutoCommit=0 before the inserts, then commit after Inserts i can cut off a lot of time. still not as fast but around 2 seconds.<br />
doesn&#8217;t work if the call is to a stored procedure(which does the insert).</p>
<p>On my development machine i&#8217;m not experiencing this at all.  (i dont need to add the autocommit = 0 and i still see &lt; 1 second performance)</p>
<p>the production machine is a VPS, so im suspecting that disk usage may be the primary culprit and someone else has a high disk usage application.</p>
<p>Thanks for the link.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: chris</title>
		<link>http://blog.cnizz.com/2010/06/12/mysql-innodb-inserts-are-slow-really-slow/comment-page-1/#comment-72408</link>
		<dc:creator>chris</dc:creator>
		<pubDate>Wed, 18 Aug 2010 17:26:07 +0000</pubDate>
		<guid isPermaLink="false">http://blog.cnizz.com/?p=445#comment-72408</guid>
		<description><![CDATA[Nick, thanks for the response.  Unfortunately I haven&#039;t looked into &quot;fixing&quot; this issue.  It is known that MyISAM is faster than InnoDb, but this is the trade-off you agree to when you want to gain referential integrity and normalized data at the database layer.  If this is a high availability production server I would look into some books on the subject.  You can also check out some of the posts over at &lt;a href=&quot;http://www.mysqlperformanceblog.com/&quot; rel=&quot;nofollow&quot;&gt;http://www.mysqlperformanceblog.com/&lt;/a&gt;.  The folks over there have lots of great stuff.]]></description>
		<content:encoded><![CDATA[<p>Nick, thanks for the response.  Unfortunately I haven&#8217;t looked into &#8220;fixing&#8221; this issue.  It is known that MyISAM is faster than InnoDb, but this is the trade-off you agree to when you want to gain referential integrity and normalized data at the database layer.  If this is a high availability production server I would look into some books on the subject.  You can also check out some of the posts over at <a href="http://www.mysqlperformanceblog.com/" rel="nofollow">http://www.mysqlperformanceblog.com/</a>.  The folks over there have lots of great stuff.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Nick</title>
		<link>http://blog.cnizz.com/2010/06/12/mysql-innodb-inserts-are-slow-really-slow/comment-page-1/#comment-72406</link>
		<dc:creator>Nick</dc:creator>
		<pubDate>Wed, 18 Aug 2010 17:20:49 +0000</pubDate>
		<guid isPermaLink="false">http://blog.cnizz.com/?p=445#comment-72406</guid>
		<description><![CDATA[I am experiencing the same symptoms,   
I have no indexes or foreign keys   one table, one column.   if i use InnoBb 1000 inserts takes around 10-20 seconds  MyIsam  &lt; 1 second
server is 4 core @ 2.0Ghz with 2818176KB free


any suggestions how to fix this?]]></description>
		<content:encoded><![CDATA[<p>I am experiencing the same symptoms,<br />
I have no indexes or foreign keys   one table, one column.   if i use InnoBb 1000 inserts takes around 10-20 seconds  MyIsam  &lt; 1 second<br />
server is 4 core @ 2.0Ghz with 2818176KB free</p>
<p>any suggestions how to fix this?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: chris</title>
		<link>http://blog.cnizz.com/2010/06/12/mysql-innodb-inserts-are-slow-really-slow/comment-page-1/#comment-69813</link>
		<dc:creator>chris</dc:creator>
		<pubDate>Thu, 01 Jul 2010 15:23:44 +0000</pubDate>
		<guid isPermaLink="false">http://blog.cnizz.com/?p=445#comment-69813</guid>
		<description><![CDATA[I was a little suprised myself.  This was an empty table actually with only 3 columns: a primary key set to auto increment, a varchar, and date_time.  Try it for yourself.]]></description>
		<content:encoded><![CDATA[<p>I was a little suprised myself.  This was an empty table actually with only 3 columns: a primary key set to auto increment, a varchar, and date_time.  Try it for yourself.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Hannes</title>
		<link>http://blog.cnizz.com/2010/06/12/mysql-innodb-inserts-are-slow-really-slow/comment-page-1/#comment-69792</link>
		<dc:creator>Hannes</dc:creator>
		<pubDate>Thu, 01 Jul 2010 07:11:37 +0000</pubDate>
		<guid isPermaLink="false">http://blog.cnizz.com/?p=445#comment-69792</guid>
		<description><![CDATA[Do you have tons of indexes or foreign keys in your tables? The difference seems to big..]]></description>
		<content:encoded><![CDATA[<p>Do you have tons of indexes or foreign keys in your tables? The difference seems to big..</p>
]]></content:encoded>
	</item>
</channel>
</rss>
<!-- This Quick Cache file was built for (  blog.cnizz.com/2010/06/12/mysql-innodb-inserts-are-slow-really-slow/feed/ ) in 2.02648 seconds, on May 14th, 2013 at 3:02 pm UTC. -->
<!-- This Quick Cache file will automatically expire ( and be re-built automatically ) on May 21st, 2013 at 3:02 pm UTC -->
<!-- +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ -->
<!-- Quick Cache Is Fully Functional :-) ... A Quick Cache file was just served for (  blog.cnizz.com/2010/06/12/mysql-innodb-inserts-are-slow-really-slow/feed/ ) in 0.07459 seconds, on May 18th, 2013 at 10:50 pm UTC. -->