<?xml version="1.0" encoding="UTF-8"?><!-- generator="wordpress/2.6.5" -->
<rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	>
<channel>
	<title>Comments on: ORA-01450 During Online Index Rebuild</title>
	<link>http://www.pythian.com/blogs/1111/ora-01450-during-online-index-rebuild</link>
	<description>News and views from Pythian DBAs</description>
	<pubDate>Thu,  4 Dec 2008 23:21:11 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.6.5</generator>
		<item>
		<title>By: Che fine ha fatto ORA-01450? &#171; Oracle and other</title>
		<link>http://www.pythian.com/blogs/1111/ora-01450-during-online-index-rebuild#comment-270605</link>
		<dc:creator>Che fine ha fatto ORA-01450? &#171; Oracle and other</dc:creator>
		<pubDate>Tue, 02 Sep 2008 15:45:55 +0000</pubDate>
		<guid>http://www.pythian.com/blogs/1111/ora-01450-during-online-index-rebuild#comment-270605</guid>
		<description>[...] fase successiva è la ricerca su Google, questi mi porta a un post di Alex Gorbachev su Pythian. Riporta un problema diverso, ma con lo stesso [...]</description>
		<content:encoded><![CDATA[<p>[&#8230;] fase successiva è la ricerca su Google, questi mi porta a un post di Alex Gorbachev su Pythian. Riporta un problema diverso, ma con lo stesso [&#8230;]</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Alex Gorbachev</title>
		<link>http://www.pythian.com/blogs/1111/ora-01450-during-online-index-rebuild#comment-247436</link>
		<dc:creator>Alex Gorbachev</dc:creator>
		<pubDate>Wed, 30 Jul 2008 17:03:52 +0000</pubDate>
		<guid>http://www.pythian.com/blogs/1111/ora-01450-during-online-index-rebuild#comment-247436</guid>
		<description>Yeah... Actually, I must be wrong regarding minimum two index key entries per branch block -- it doesn't need to be two entries. The beginning of a range for an n level branch block is inherited from an n-1 level parent branch block.</description>
		<content:encoded><![CDATA[<p>Yeah&#8230; Actually, I must be wrong regarding minimum two index key entries per branch block &#8212; it doesn&#8217;t need to be two entries. The beginning of a range for an n level branch block is inherited from an n-1 level parent branch block.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: YAP</title>
		<link>http://www.pythian.com/blogs/1111/ora-01450-during-online-index-rebuild#comment-247384</link>
		<dc:creator>YAP</dc:creator>
		<pubDate>Wed, 30 Jul 2008 15:02:56 +0000</pubDate>
		<guid>http://www.pythian.com/blogs/1111/ora-01450-during-online-index-rebuild#comment-247384</guid>
		<description>Oops! I see "Index Internals", sorry.</description>
		<content:encoded><![CDATA[<p>Oops! I see &#8220;Index Internals&#8221;, sorry.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Alex Gorbachev</title>
		<link>http://www.pythian.com/blogs/1111/ora-01450-during-online-index-rebuild#comment-247364</link>
		<dc:creator>Alex Gorbachev</dc:creator>
		<pubDate>Wed, 30 Jul 2008 14:25:21 +0000</pubDate>
		<guid>http://www.pythian.com/blogs/1111/ora-01450-during-online-index-rebuild#comment-247364</guid>
		<description>Not sure where you take it from... Page 11 of the IOT Internals presentation that I referenced says that &lt;i&gt;Maximum primary key lengths in Oracle 9.2&lt;/i&gt; for 16K block size is 3800 bytes, which actually contradicts to the Metalink Note 136158.1 (6498 bytes) but it might be the difference between and index and IOT table (thought, I don't see how it would be *significantly* different).

Now, I can't imagine why the requirements of having at least two keys per block would be lifted - it's impossible with the structure of a b-tree index and more specifically, it's about the *branch* blocks that should fit at least two index keys and not leaf blocks.</description>
		<content:encoded><![CDATA[<p>Not sure where you take it from&#8230; Page 11 of the IOT Internals presentation that I referenced says that <i>Maximum primary key lengths in Oracle 9.2</i> for 16K block size is 3800 bytes, which actually contradicts to the Metalink Note 136158.1 (6498 bytes) but it might be the difference between and index and IOT table (thought, I don&#8217;t see how it would be *significantly* different).</p>
<p>Now, I can&#8217;t imagine why the requirements of having at least two keys per block would be lifted - it&#8217;s impossible with the structure of a b-tree index and more specifically, it&#8217;s about the *branch* blocks that should fit at least two index keys and not leaf blocks.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: YAP</title>
		<link>http://www.pythian.com/blogs/1111/ora-01450-during-online-index-rebuild#comment-247319</link>
		<dc:creator>YAP</dc:creator>
		<pubDate>Wed, 30 Jul 2008 12:30:03 +0000</pubDate>
		<guid>http://www.pythian.com/blogs/1111/ora-01450-during-online-index-rebuild#comment-247319</guid>
		<description>Julian Dyke`s "IOT Internals" whitepaper, page 5, say that for Oracle version 9.X, in fact, it is NOT required that any index 
block must contain at least TWO index entries per block. For example, with block size 16K - maximum key lengths = 12958. Can you comment this?</description>
		<content:encoded><![CDATA[<p>Julian Dyke`s &#8220;IOT Internals&#8221; whitepaper, page 5, say that for Oracle version 9.X, in fact, it is NOT required that any index<br />
block must contain at least TWO index entries per block. For example, with block size 16K - maximum key lengths = 12958. Can you comment this?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Alex Gorbachev</title>
		<link>http://www.pythian.com/blogs/1111/ora-01450-during-online-index-rebuild#comment-231290</link>
		<dc:creator>Alex Gorbachev</dc:creator>
		<pubDate>Sun, 06 Jul 2008 13:25:05 +0000</pubDate>
		<guid>http://www.pythian.com/blogs/1111/ora-01450-during-online-index-rebuild#comment-231290</guid>
		<description>Yep - that's workaround. Another one would be to use a larger block size tablespace but there is a lot to think about before changing block size for an index.</description>
		<content:encoded><![CDATA[<p>Yep - that&#8217;s workaround. Another one would be to use a larger block size tablespace but there is a lot to think about before changing block size for an index.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Asif Momen</title>
		<link>http://www.pythian.com/blogs/1111/ora-01450-during-online-index-rebuild#comment-231173</link>
		<dc:creator>Asif Momen</dc:creator>
		<pubDate>Sun, 06 Jul 2008 07:32:22 +0000</pubDate>
		<guid>http://www.pythian.com/blogs/1111/ora-01450-during-online-index-rebuild#comment-231173</guid>
		<description>Hi Alex,

Probably, you ended up rebiulding index without ONLINE. 

Regards</description>
		<content:encoded><![CDATA[<p>Hi Alex,</p>
<p>Probably, you ended up rebiulding index without ONLINE. </p>
<p>Regards</p>
]]></content:encoded>
	</item>
</channel>
</rss>
