<?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: In-memory databases</title>
	<atom:link href="http://blog.geniedb.com/2010/05/27/in-memory-databases/feed/" rel="self" type="application/rss+xml" />
	<link>http://blog.geniedb.com/2010/05/27/in-memory-databases/</link>
	<description>Next generation distributed database technology</description>
	<lastBuildDate>Wed, 06 Oct 2010 17:37:23 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.org/?v=3.1.2</generator>
	<item>
		<title>By: Alaric Snell-Pym</title>
		<link>http://blog.geniedb.com/2010/05/27/in-memory-databases/comment-page-1/#comment-594</link>
		<dc:creator>Alaric Snell-Pym</dc:creator>
		<pubDate>Wed, 06 Oct 2010 13:25:53 +0000</pubDate>
		<guid isPermaLink="false">http://blog.geniedb.com/?p=119#comment-594</guid>
		<description>Sure, in that not having to commit writes to disk means you don&#039;t need to be limited by disk bandwidth and the time taken to sync.

But my point is that it&#039;s silly to make an &lt;em&gt;entire database&lt;/em&gt; be &quot;in-memory&quot; or not. The real issue is whether some of your data (eg, a table) needs to be persistent. Turn off persistence, and writes can go faster. Turn on persistence, and writes are limited by disks, but are persistent. Databases like Redis also let you fine-tune that a little, choosing degrees of persistence. And it shouldn&#039;t affect read performance at all, once the cache has warmed up.

Why can&#039;t we have a single database, with some tables (or even just some &lt;strong&gt;records&lt;/strong&gt;?) being persistent-but-slow-to-write, and some being fast-to-write-but-lossy, as a tweaking option? Why do we need an entire separate database?</description>
		<content:encoded><![CDATA[<p>Sure, in that not having to commit writes to disk means you don&#8217;t need to be limited by disk bandwidth and the time taken to sync.</p>
<p>But my point is that it&#8217;s silly to make an <em>entire database</em> be &#8220;in-memory&#8221; or not. The real issue is whether some of your data (eg, a table) needs to be persistent. Turn off persistence, and writes can go faster. Turn on persistence, and writes are limited by disks, but are persistent. Databases like Redis also let you fine-tune that a little, choosing degrees of persistence. And it shouldn&#8217;t affect read performance at all, once the cache has warmed up.</p>
<p>Why can&#8217;t we have a single database, with some tables (or even just some <strong>records</strong>?) being persistent-but-slow-to-write, and some being fast-to-write-but-lossy, as a tweaking option? Why do we need an entire separate database?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Seun Osewa</title>
		<link>http://blog.geniedb.com/2010/05/27/in-memory-databases/comment-page-1/#comment-567</link>
		<dc:creator>Seun Osewa</dc:creator>
		<pubDate>Thu, 12 Aug 2010 09:07:06 +0000</pubDate>
		<guid isPermaLink="false">http://blog.geniedb.com/?p=119#comment-567</guid>
		<description>In-memory databases are much faster in most cases, though.</description>
		<content:encoded><![CDATA[<p>In-memory databases are much faster in most cases, though.</p>
]]></content:encoded>
	</item>
</channel>
</rss>

