RE: Testing Direct I/O

From: <oracle-l-bounce_at_freelists.org>
Date: Mon, 13 Feb 2006 20:52:33 -0800
Message-ID: <5D2570CAFC98974F9B6A759D1C74BAD03FDCB9@ex2.ms.polyserve.com>

 >>>"However, I'd like to test the results of actually using

>>>direct i/o. In order to not re-invent the wheel, I was
>>>wondering if anyone had performed similar testing?"

These tests have been performed by about eleventeen, or 42 thousand people. In the end you will likely do what you "believe". You are asking if it is better to cache in the SGA, or to cache in the SGA AND in the OS buffer pool.

I remember fondly re-linking the direct IO libs into 6.0.26 and festidiously performing such tests. And you know what? It is better to not double buffer! Imagine! I performed that test on a 30 CPU SMP about 5 years before 97.42% of all living DBAs even touched an SMP with more than 8 CPUs too.

--
http://www.freelists.org/webpage/oracle-l

Received on Mon Feb 13 2006 - 22:52:33 CST