<?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"
	>
<channel>
	<title>Comments on: Wishes for new  &#8220;Pure PHP&#8221; MySQL  driver</title>
	<atom:link href="http://www.mysqlperformanceblog.com/2006/10/28/wishes-for-new-pure-php-mysql-driver/feed/" rel="self" type="application/rss+xml" />
	<link>http://www.mysqlperformanceblog.com/2006/10/28/wishes-for-new-pure-php-mysql-driver/</link>
	<description>Everything about MySQL Performance</description>
	<pubDate>Mon, 13 Oct 2008 16:57:28 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.5.1</generator>
		<item>
		<title>By: .: El Blog de Inwe :. &#187; Importante mejora para php 5.3</title>
		<link>http://www.mysqlperformanceblog.com/2006/10/28/wishes-for-new-pure-php-mysql-driver/#comment-197857</link>
		<dc:creator>.: El Blog de Inwe :. &#187; Importante mejora para php 5.3</dc:creator>
		<pubDate>Mon, 19 Nov 2007 11:18:50 +0000</pubDate>
		<guid isPermaLink="false">http://www.mysqlperformanceblog.com/2006/10/28/wishes-for-new-pure-php-mysql-driver/#comment-197857</guid>
		<description>[...] del cliente y un monón de posibilidades de cara a futuras funcionalidades (se pueden comprobar en PHP driver whish list) como caché en las sentencias preparadas, balanceado de carga automático, etc.. Tags: mysqlnd, [...]</description>
		<content:encoded><![CDATA[<p>[...] del cliente y un monón de posibilidades de cara a futuras funcionalidades (se pueden comprobar en PHP driver whish list) como caché en las sentencias preparadas, balanceado de carga automático, etc.. Tags: mysqlnd, [...]</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: What is new in PHP 5.3 - part 3: mysqlnd &#124; Gergely Hodicska</title>
		<link>http://www.mysqlperformanceblog.com/2006/10/28/wishes-for-new-pure-php-mysql-driver/#comment-196250</link>
		<dc:creator>What is new in PHP 5.3 - part 3: mysqlnd &#124; Gergely Hodicska</dc:creator>
		<pubDate>Sat, 17 Nov 2007 20:50:14 +0000</pubDate>
		<guid isPermaLink="false">http://www.mysqlperformanceblog.com/2006/10/28/wishes-for-new-pure-php-mysql-driver/#comment-196250</guid>
		<description>[...] load balancing, built-in profiling. More foretastes for gourmands: the wishlist of Lukas, and Peter Zaitsev (MySQL Performance [...]</description>
		<content:encoded><![CDATA[<p>[...] load balancing, built-in profiling. More foretastes for gourmands: the wishlist of Lukas, and Peter Zaitsev (MySQL Performance [...]</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Georg</title>
		<link>http://www.mysqlperformanceblog.com/2006/10/28/wishes-for-new-pure-php-mysql-driver/#comment-131724</link>
		<dc:creator>Georg</dc:creator>
		<pubDate>Wed, 30 May 2007 13:36:33 +0000</pubDate>
		<guid isPermaLink="false">http://www.mysqlperformanceblog.com/2006/10/28/wishes-for-new-pure-php-mysql-driver/#comment-131724</guid>
		<description>Rolf,

currently there is no way to stop the running queries from client side. This could be an issue for the proxy (http://jan.kneschke.de/projects/mysql/mysql-proxy)</description>
		<content:encoded><![CDATA[<p>Rolf,</p>
<p>currently there is no way to stop the running queries from client side. This could be an issue for the proxy (http://jan.kneschke.de/projects/mysql/mysql-proxy)</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: rolf</title>
		<link>http://www.mysqlperformanceblog.com/2006/10/28/wishes-for-new-pure-php-mysql-driver/#comment-26393</link>
		<dc:creator>rolf</dc:creator>
		<pubDate>Thu, 04 Jan 2007 17:24:18 +0000</pubDate>
		<guid isPermaLink="false">http://www.mysqlperformanceblog.com/2006/10/28/wishes-for-new-pure-php-mysql-driver/#comment-26393</guid>
		<description>I don't know whether this is the best venue for this question, but here it is.

One of my biggest complaints about http+php+mysql is that when a browser stops any impending queries will run to completion on the mysqld side. Will mysqlnd allow these queries to be halted on script termination?</description>
		<content:encoded><![CDATA[<p>I don&#8217;t know whether this is the best venue for this question, but here it is.</p>
<p>One of my biggest complaints about http+php+mysql is that when a browser stops any impending queries will run to completion on the mysqld side. Will mysqlnd allow these queries to be halted on script termination?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: peter</title>
		<link>http://www.mysqlperformanceblog.com/2006/10/28/wishes-for-new-pure-php-mysql-driver/#comment-8259</link>
		<dc:creator>peter</dc:creator>
		<pubDate>Mon, 13 Nov 2006 03:10:03 +0000</pubDate>
		<guid isPermaLink="false">http://www.mysqlperformanceblog.com/2006/10/28/wishes-for-new-pure-php-mysql-driver/#comment-8259</guid>
		<description>Rene,

This is kind of mistake - I'm not going to Google HQ for MySQL Camp - I did not have visa so I could not go :(
The presentation you're is from our discussion in Frankfut at International PHP Conference.</description>
		<content:encoded><![CDATA[<p>Rene,</p>
<p>This is kind of mistake - I&#8217;m not going to Google HQ for MySQL Camp - I did not have visa so I could not go <img src='http://www.mysqlperformanceblog.com/wp-includes/images/smilies/icon_sad.gif' alt=':(' class='wp-smiley' /><br />
The presentation you&#8217;re is from our discussion in Frankfut at International PHP Conference.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Rene Leonhardt</title>
		<link>http://www.mysqlperformanceblog.com/2006/10/28/wishes-for-new-pure-php-mysql-driver/#comment-7766</link>
		<dc:creator>Rene Leonhardt</dc:creator>
		<pubDate>Sat, 11 Nov 2006 07:24:19 +0000</pubDate>
		<guid isPermaLink="false">http://www.mysqlperformanceblog.com/2006/10/28/wishes-for-new-pure-php-mysql-driver/#comment-7766</guid>
		<description>Peter,

I wish you a good time at the Google HQ, for a developer like me it would be impressive ;)
http://mysqlcamp.org/Session_PHP_mysqlnd

I hope you will post the presentation slides here afterwards.</description>
		<content:encoded><![CDATA[<p>Peter,</p>
<p>I wish you a good time at the Google HQ, for a developer like me it would be impressive <img src='http://www.mysqlperformanceblog.com/wp-includes/images/smilies/icon_wink.gif' alt=';)' class='wp-smiley' /><br />
<a href="http://mysqlcamp.org/Session_PHP_mysqlnd" rel="nofollow">http://mysqlcamp.org/Session_PHP_mysqlnd</a></p>
<p>I hope you will post the presentation slides here afterwards.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Mark Matthews</title>
		<link>http://www.mysqlperformanceblog.com/2006/10/28/wishes-for-new-pure-php-mysql-driver/#comment-5572</link>
		<dc:creator>Mark Matthews</dc:creator>
		<pubDate>Tue, 31 Oct 2006 22:30:45 +0000</pubDate>
		<guid isPermaLink="false">http://www.mysqlperformanceblog.com/2006/10/28/wishes-for-new-pure-php-mysql-driver/#comment-5572</guid>
		<description>Elder,

It would be rather straightforward to implement the interface to MySQL's XA implementation in PHP as the interface to XA in MySQL is standard SQL (unlike many other databases, which require native code or protocol-level calls to do the same thing):

http://dev.mysql.com/doc/refman/5.0/en/xa-statements.html

The issue is that PHP doesn't have a transaction monitor, that I know of, so that would be the first thing someone would have to write, or "wrap" (many opensource transaction monitors exist in the Java world).

You can "fake" the protocol (and we do in the JDBC unit tests), but that doesn't give you any recoverability if the application itself crashes without an actual logging transaction monitor, which then makes XA pretty much useless in real-world scenarios.

     -Mark</description>
		<content:encoded><![CDATA[<p>Elder,</p>
<p>It would be rather straightforward to implement the interface to MySQL&#8217;s XA implementation in PHP as the interface to XA in MySQL is standard SQL (unlike many other databases, which require native code or protocol-level calls to do the same thing):</p>
<p><a href="http://dev.mysql.com/doc/refman/5.0/en/xa-statements.html" rel="nofollow">http://dev.mysql.com/doc/refman/5.0/en/xa-statements.html</a></p>
<p>The issue is that PHP doesn&#8217;t have a transaction monitor, that I know of, so that would be the first thing someone would have to write, or &#8220;wrap&#8221; (many opensource transaction monitors exist in the Java world).</p>
<p>You can &#8220;fake&#8221; the protocol (and we do in the JDBC unit tests), but that doesn&#8217;t give you any recoverability if the application itself crashes without an actual logging transaction monitor, which then makes XA pretty much useless in real-world scenarios.</p>
<p>     -Mark</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: elder</title>
		<link>http://www.mysqlperformanceblog.com/2006/10/28/wishes-for-new-pure-php-mysql-driver/#comment-5533</link>
		<dc:creator>elder</dc:creator>
		<pubDate>Mon, 30 Oct 2006 20:13:48 +0000</pubDate>
		<guid isPermaLink="false">http://www.mysqlperformanceblog.com/2006/10/28/wishes-for-new-pure-php-mysql-driver/#comment-5533</guid>
		<description>Would be also nice ability to perform XA transactions as in JDBC</description>
		<content:encoded><![CDATA[<p>Would be also nice ability to perform XA transactions as in JDBC</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: peter</title>
		<link>http://www.mysqlperformanceblog.com/2006/10/28/wishes-for-new-pure-php-mysql-driver/#comment-5531</link>
		<dc:creator>peter</dc:creator>
		<pubDate>Mon, 30 Oct 2006 16:42:38 +0000</pubDate>
		<guid isPermaLink="false">http://www.mysqlperformanceblog.com/2006/10/28/wishes-for-new-pure-php-mysql-driver/#comment-5531</guid>
		<description>Mike, 

Persistent connections were never bad thing per say but they had number of problems.  Estabilishing new connection in many cases is not significantly more expensive than recycling connection (which also needs special command).  Plus on Gbit ethernet connection establishment did not add too much latency anyway.   On other hand  large number of etablished connections has some overhead in scheduling, memory usage etc.  

If pages are complicated there is very little win for persistant connections as connection creation will constitute very small proportion.  For  pages running single simple query which is frequently the case with AJAX it becomes significant. 

Now about limiting number of persistent connections - Apache would never act as connection pooling daemon as connections can't be shared between  processes.  Simply if there are too many persistent connection as connections are closed they would not be kept as persitent connections but closed at once.</description>
		<content:encoded><![CDATA[<p>Mike, </p>
<p>Persistent connections were never bad thing per say but they had number of problems.  Estabilishing new connection in many cases is not significantly more expensive than recycling connection (which also needs special command).  Plus on Gbit ethernet connection establishment did not add too much latency anyway.   On other hand  large number of etablished connections has some overhead in scheduling, memory usage etc.  </p>
<p>If pages are complicated there is very little win for persistant connections as connection creation will constitute very small proportion.  For  pages running single simple query which is frequently the case with AJAX it becomes significant. </p>
<p>Now about limiting number of persistent connections - Apache would never act as connection pooling daemon as connections can&#8217;t be shared between  processes.  Simply if there are too many persistent connection as connections are closed they would not be kept as persitent connections but closed at once.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Mike</title>
		<link>http://www.mysqlperformanceblog.com/2006/10/28/wishes-for-new-pure-php-mysql-driver/#comment-5530</link>
		<dc:creator>Mike</dc:creator>
		<pubDate>Mon, 30 Oct 2006 16:20:08 +0000</pubDate>
		<guid isPermaLink="false">http://www.mysqlperformanceblog.com/2006/10/28/wishes-for-new-pure-php-mysql-driver/#comment-5530</guid>
		<description>I'm curious about your statement that "Ajax applications with frequent case of one query per request make persistent connections relevant again." Persistent connections were irrelevant? I can see where they'd incur some serious overheard on heavily-loaded systems (all my experience is with pretty lightly-loaded applications) but I didn't know there was any consensus in the PHP/MySQL world that persistent connections were a bad thing. I would think with appropriate limits on the maximum number of persistent connections, Apache itself would essentially act as a connection-pooling daemon. Am I totally off base here?</description>
		<content:encoded><![CDATA[<p>I&#8217;m curious about your statement that &#8220;Ajax applications with frequent case of one query per request make persistent connections relevant again.&#8221; Persistent connections were irrelevant? I can see where they&#8217;d incur some serious overheard on heavily-loaded systems (all my experience is with pretty lightly-loaded applications) but I didn&#8217;t know there was any consensus in the PHP/MySQL world that persistent connections were a bad thing. I would think with appropriate limits on the maximum number of persistent connections, Apache itself would essentially act as a connection-pooling daemon. Am I totally off base here?</p>
]]></content:encoded>
	</item>
</channel>
</rss>
