Re: Raw Disk Partition vs. Unix File System (Oracle 6)

From: Andrew Finkenstadt <andy_at_vistachrome.com>
Date: Tue, 24 Aug 1993 05:08:43 GMT
Message-ID: <CC8zMK.8Cx_at_vistachrome.com>


lparsons_at_exlog.com (Lee Parsons) writes:
>In article <CC7IoC.FtA_at_vistachrome.com> andy_at_vistachrome.com writes:
>>cafe_at_cbnewse.cb.att.com (richard.dib) writes:
>>>- Does anybody have any experience with Oracle using a raw disk partitition
>>> instead of the Unix file system? Could you share your experiences?
>>
>>I have experience only with Unix fs-based database files. Some of our
>>australian friends believe highly in raw file systems. Some have seen
>>as much as 25% improvement in speed over cooked file systems, and sometimes
>>as much as 50% when used over a fragmented cooked file system.
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 

>I find that if my car is dirty and has a flat tire, then fixing the
>flat and washing it makes it go 95% faster. Does this mean that going
>to the car wash improves by road performance by 95%? :-}

<grin>

>Of course your going to get a huge increase going raw over a
>fragmented cooked file system. The question though is how much of that
>is the result of going raw vs unfragmenting the datafiles.
 

>Could it be that the vast majority of perf increase was do to the
>compression of the extents and that they could have gotten most of
>that increase by just rebuilding the File System and database?

Tis quite possible; just reporting the case studies. :)

>--
>Regards,
 

>Lee E. Parsons Baker Hughes Inteq, Inc
>Oracle Database Administrator lparsons_at_exlog.com

-- 
Andrew Finkenstadt  |  andy_at_{homes.com,vistachrome.com,genie.geis.com}
Systems Analyst     |  Vista-Chrome, Homes & Land Publishing Corporation
                    |  1600 Capital Circle SW, Tallahassee Florida 32310
+1 904-575-0189     |  GEnie Postmaster, Unix & Internet RoundTables Sysop
Received on Tue Aug 24 1993 - 07:08:43 CEST

Original text of this message