Re: [SPAM?] Re: Asynchronous I/O Support

Zeugswetter Andreas ADI SD <zeugswettera@spardat.at>

From: "Zeugswetter Andreas ADI SD" <ZeugswetterA@spardat.at>

To: <mark@mark.mielke.cc>

Cc: "NikhilS" <nikkhils@gmail.com>, "Martijn van Oosterhout" <kleptog@svana.org>, "Mark Kirkwood" <markir@paradise.net.nz>, "Luke Lonergan" <llonergan@greenplum.com>, "Raja Agrawal" <raja.agrawal@gmail.com>, <pgsql-hackers@postgresql.org>

Date: 2006-10-23T07:59:30Z

Lists: pgsql-hackers

> > Yup, that would be the scenario where it helps (provided that you
have 
> > a smart disk or a disk array and an intelligent OS aio
implementation).
> > It would be used to fetch the data pages pointed at from an index 
> > leaf, or the next level index pages.
> > We measured the IO bandwidth difference on Windows with EMC as
beeing 
> > nearly proportional to parallel outstanding requests up to at least
> 
> Measured it using what? I was under the impression only one 
> proof-of-implementation existed, and that the scenarios and 
> configuration of the person who wrote it, did not show 
> significant improvement.

IIRC the configuration of that test was not suitable to show any
benefit.
Minimum requirements to show improvement are:
	- very few active sessions (typically less than number of disks)
	- a table that spans multiple disks (typically on a stripe set)
	   (or one intelligent scsi disk)
	- only random disk access plans
 
> You have PostgreSQL on Windows with EMC with async I/O 
> support to test with?

No, sorry. Was a MaxDB issue.

Andreas