ADSM-L

Re: AIX and async I/O--additional info

2003-04-24 16:39:46
Subject: Re: AIX and async I/O--additional info
From: "PINNI, BALANAND (SBCSI)" <bp3965 AT SBC DOT COM>
To: ADSM-L AT VM.MARIST DOT EDU
Date: Thu, 24 Apr 2003 15:39:12 -0500
Paul I am not from AUS ,rather from USA as shown in below message .

Paul as u are DBA did u try any time working with disk_asyncio in init.ora .
I am not talking about aio that is used while running rootpre.sh that
creates /etc/pwsyscall .

Thanks
Balanand Pinni


-----Original Message-----
From: Paul Ripke [mailto:stixpjr AT BIGPOND.NET DOT AU]
Sent: Thursday, April 24, 2003 3:23 PM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: Re: AIX and async I/O--additional info

On Friday, Apr 25, 2003, at 01:18 Australia/Sydney, PINNI, BALANAND
(SBCSI) wrote:

> Direct I/O does not enhance raw I/O performance. Raw I/O performance is

I did say almost...

> still slightly faster than direct I/O, but if you want the benefits of
> a JFS
> and enhanced performance, direct I/O is highly recommended

Which is why it's enabled by default.

> Summary
> Direct I/O requires substantially fewer CPU cycles than regular I/O.
> I/O-intensive applications that do not get much benefit from the
> caching
> provided by regular I/O can enhance performance by using direct I/O.
> The
> benefits of direct I/O will grow in the future as increases in CPU
> speeds
> continue to outpace increases in memory speeds.

Exactly - raw I/O and direct I/O bypass the buffer cache, hence the
pager
thread doesn't need to spend anywhere near as much time looking for
victim
pages.

> What types of programs are good candidates for direct I/O? Programs
> that are
> typically CPU-limited and perform lots of disk I/O. "Technical" codes
> that
> have large sequential I/Os are good candidates. Applications that do
> numerous small I/Os will typically see less performance benefit,
> because
> direct I/O cannot do read ahead or write behind. Applications that have
> benefited from striping are also good candidates.

I'd add: Any application which does it's own "smarter" cache management
(TSM bufferpool for DB, Oracle block cache in SGA, etc). The OS cache is
pretty much LRU, whereas most RDBMs implement LRU/MRU/??? based on
usage.

Cheers,
--
Paul Ripke
Unix/OpenVMS/TSM/DBA
101 reasons why you can't find your Sysadmin:
68: It's 9AM. He/She is not working that late.
-- Koos van den Hout

<Prev in Thread] Current Thread [Next in Thread>