<?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: What&#8217;s missing in MySQL Replication?</title>
	<atom:link href="http://blog.wl0.org/2009/03/whats-missing-in-mysql-replication/feed/" rel="self" type="application/rss+xml" />
	<link>http://blog.wl0.org/2009/03/whats-missing-in-mysql-replication/</link>
	<description>Random thoughts on different topics</description>
	<lastBuildDate>Thu, 22 Jul 2010 06:31:53 +0000</lastBuildDate>
	<generator>http://wordpress.org/?v=2.9.1</generator>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
		<item>
		<title>By: SQL Tutorials</title>
		<link>http://blog.wl0.org/2009/03/whats-missing-in-mysql-replication/comment-page-1/#comment-108</link>
		<dc:creator>SQL Tutorials</dc:creator>
		<pubDate>Fri, 01 May 2009 02:18:05 +0000</pubDate>
		<guid isPermaLink="false">http://blog.wl0.org/?p=67#comment-108</guid>
		<description>You know, the thing about SQL is, that there is virtually nothing that can replace it. 

Does anyone know if a substitute exists for sql? I mean besides MS SQL and Oracle and all that jazz. Thanks.</description>
		<content:encoded><![CDATA[<p>You know, the thing about SQL is, that there is virtually nothing that can replace it. </p>
<p>Does anyone know if a substitute exists for sql? I mean besides MS SQL and Oracle and all that jazz. Thanks.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: sjmudd</title>
		<link>http://blog.wl0.org/2009/03/whats-missing-in-mysql-replication/comment-page-1/#comment-87</link>
		<dc:creator>sjmudd</dc:creator>
		<pubDate>Tue, 31 Mar 2009 18:41:39 +0000</pubDate>
		<guid isPermaLink="false">http://blog.wl0.org/?p=67#comment-87</guid>
		<description>Hi Robert. I&#039;ve read about Tungsten before and it does sound a lot more complete. I don&#039;t know of anyone who uses it, nor how well it scales. What put me off from trying it when I looked several months ago was the use of Java. Probably unjustified. The functionality your product documentation says it provides does look nice and perhaps it would fit our requirements better. It also looks quite similar to how Sybase Replication server works. I&#039;ve not used other replication products so perhaps they do the same thing too. Fundamentally the problem of any replication product is the same so the solutions are likely to be similar.

I&#039;ll have to see if I can at least make some time to look at Tungsten and see how well it behaves and how well it scales. Thanks for the pointer.</description>
		<content:encoded><![CDATA[<p>Hi Robert. I&#8217;ve read about Tungsten before and it does sound a lot more complete. I don&#8217;t know of anyone who uses it, nor how well it scales. What put me off from trying it when I looked several months ago was the use of Java. Probably unjustified. The functionality your product documentation says it provides does look nice and perhaps it would fit our requirements better. It also looks quite similar to how Sybase Replication server works. I&#8217;ve not used other replication products so perhaps they do the same thing too. Fundamentally the problem of any replication product is the same so the solutions are likely to be similar.</p>
<p>I&#8217;ll have to see if I can at least make some time to look at Tungsten and see how well it behaves and how well it scales. Thanks for the pointer.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: sjmudd</title>
		<link>http://blog.wl0.org/2009/03/whats-missing-in-mysql-replication/comment-page-1/#comment-86</link>
		<dc:creator>sjmudd</dc:creator>
		<pubDate>Tue, 31 Mar 2009 18:36:37 +0000</pubDate>
		<guid isPermaLink="false">http://blog.wl0.org/?p=67#comment-86</guid>
		<description>Hi Tom.

it&#039;s true that to turn the bin log on or off is easy (on the source server), but that affects all slaves.

I&#039;ve worked in a location where I was able to selectively filter out DELETE statements for a specific table so that these statements weren&#039;t applied on one slave while they might have been applied on another one. The same is true of UPDATE, or TRUNCATE TABLE commands or even filtering on commands like DROP TABLE.

As far as I&#039;m aware none of this is possible with MySQL replication.

Common usage of this is: a topology consisting of a master and one or more &quot;normal slaves&quot; and a &quot;log slave&quot;.
The log slave has DELETEs disabled/filtered out yet the normal slaves and the master don&#039;t.

So with this it&#039;s possible to run a daily job on the master to &quot;clean up&quot; (delete old data). This gets replicated in such a way that the normal slaves have the same content but the &quot;log slave&quot; keeps the entire history. So historic reporting can be done on this box and the main master/normal slaves just have the upto date &quot;current state&quot;. 

Where I work now something like this sort of functionality would simplify our setup signficantly. It&#039;s not possible.</description>
		<content:encoded><![CDATA[<p>Hi Tom.</p>
<p>it&#8217;s true that to turn the bin log on or off is easy (on the source server), but that affects all slaves.</p>
<p>I&#8217;ve worked in a location where I was able to selectively filter out DELETE statements for a specific table so that these statements weren&#8217;t applied on one slave while they might have been applied on another one. The same is true of UPDATE, or TRUNCATE TABLE commands or even filtering on commands like DROP TABLE.</p>
<p>As far as I&#8217;m aware none of this is possible with MySQL replication.</p>
<p>Common usage of this is: a topology consisting of a master and one or more &#8220;normal slaves&#8221; and a &#8220;log slave&#8221;.<br />
The log slave has DELETEs disabled/filtered out yet the normal slaves and the master don&#8217;t.</p>
<p>So with this it&#8217;s possible to run a daily job on the master to &#8220;clean up&#8221; (delete old data). This gets replicated in such a way that the normal slaves have the same content but the &#8220;log slave&#8221; keeps the entire history. So historic reporting can be done on this box and the main master/normal slaves just have the upto date &#8220;current state&#8221;. </p>
<p>Where I work now something like this sort of functionality would simplify our setup signficantly. It&#8217;s not possible.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: tom hanlon</title>
		<link>http://blog.wl0.org/2009/03/whats-missing-in-mysql-replication/comment-page-1/#comment-85</link>
		<dc:creator>tom hanlon</dc:creator>
		<pubDate>Tue, 31 Mar 2009 17:46:28 +0000</pubDate>
		<guid isPermaLink="false">http://blog.wl0.org/?p=67#comment-85</guid>
		<description>It’s not easy to filter the replication commands ? 

12.6.1.3. SET sql_log_bin Syntax

SET sql_log_bin = {0&#124;1}

Disables or enables binary logging for the current connection (sql_log_bin is a session variable) if the client has the SUPER privilege. The statement is refused with an error if the client does not have that privilege.</description>
		<content:encoded><![CDATA[<p>It’s not easy to filter the replication commands ? </p>
<p>12.6.1.3. SET sql_log_bin Syntax</p>
<p>SET sql_log_bin = {0|1}</p>
<p>Disables or enables binary logging for the current connection (sql_log_bin is a session variable) if the client has the SUPER privilege. The statement is refused with an error if the client does not have that privilege.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Robert Hodges</title>
		<link>http://blog.wl0.org/2009/03/whats-missing-in-mysql-replication/comment-page-1/#comment-83</link>
		<dc:creator>Robert Hodges</dc:creator>
		<pubDate>Tue, 31 Mar 2009 00:55:09 +0000</pubDate>
		<guid isPermaLink="false">http://blog.wl0.org/?p=67#comment-83</guid>
		<description>Hi Simon!

Please have a look at Tungsten Replicator (http://www.continuent.com/community/tungsten-replicator). Tungsten Replicator is an open source replacement for MySQL Replication with flexible replication topologies, event filtering, etc. already covered.  Also, we have a lot more than just replication--we have components for proxying, SQL routing, and (very soon) cluster management.  

I look forward to your feedback and hope you will try it out.  Meanwhile, thanks for the article!

Cheers, Robert
CTO of Continuent, all disclaimers apply ;)</description>
		<content:encoded><![CDATA[<p>Hi Simon!</p>
<p>Please have a look at Tungsten Replicator (<a href="http://www.continuent.com/community/tungsten-replicator" rel="nofollow">http://www.continuent.com/community/tungsten-replicator</a>). Tungsten Replicator is an open source replacement for MySQL Replication with flexible replication topologies, event filtering, etc. already covered.  Also, we have a lot more than just replication&#8211;we have components for proxying, SQL routing, and (very soon) cluster management.  </p>
<p>I look forward to your feedback and hope you will try it out.  Meanwhile, thanks for the article!</p>
<p>Cheers, Robert<br />
CTO of Continuent, all disclaimers apply <img src='http://blog.wl0.org/wp-includes/images/smilies/icon_wink.gif' alt=';)' class='wp-smiley' /> </p>
]]></content:encoded>
	</item>
</channel>
</rss>
