Oracle FAQ Your Portal to the Oracle Knowledge Grid

Home -> Community -> Mailing Lists -> Oracle-L -> RE: (Fwd) Re: (Fwd/Oracle) Does NT write to random locations on d

RE: (Fwd) Re: (Fwd/Oracle) Does NT write to random locations on d

From: Kevin Kostyszyn <>
Date: Mon, 12 Mar 2001 14:11:55 -0800
Message-ID: <>

don't know Ross, I didn't want to dig to deep into it.  The way I see it is as follows, "hey, that computer is a dual 550 with 10k rpm scsi drives and a gig of ram.  I betcha it would be faster than that piii600 with ide drives" 
    <FONT face=Arial
color=#0000ff size=2>I know it's simple minded, but it's kind of true:)
<FONT face=Arial color=#0000ff

<FONT face=Tahoma

  size=2>-----Original Message-----From:   []On Behalf Of Mohan, RossSent:   Monday, March 12, 2001 4:17 PMTo: Multiple recipients of list   ORACLE-LSubject: RE: (Fwd) Re: (Fwd/Oracle) Does NT write to random   locations on d
  I don't see the logic in the last post: "You can't have fast   and best."
  First, he doesn't define terms. "Fast"?  Is that peak   I/O? Streaming I/O? Single block read? Seek time?   Write time? Come on, trying to reduce this to an   undifferentiated "fast" or "slow" verges on the useless unless one   takes the effort to provide an EXPLICITLY CITED METRIC for   speed. And this fellow didn't.
  Second, it's confusing: why is "fast" set against "best" as   though the one is somehow the enemy of the other?   Huh?
  Third, it leaves out any discussion of the effect of on-disk   and on-controller cache. (Not to mention system-level   cache.)  As far as the application is concerned,   it does not see the "disk" sees controller and disk   cache and disk as an amalgam, performance-wise....)   

  Fourth, since WHEN did the choice become forced into "Do you   want a fast hard disk array with lots of fragments, or   a slow disk array with minimal fragments?" Geez, can I   have a slow disk array with lots of fragments?   The only statement I agree with, either logically or from   experience is the bit about OS vendors keeping a bit   secret from the world on their...well, "secret sauce". <FONT   size=2>Sure, you can keep a little bit secret, but come on, folks, it's not   like MS has any other/better/special MoJo than any   other vendor. What? when the aliens landed on the   Redmond campus and revealed their special VASTLY SUPERIOR alien OS technology,   no one else noticed?
  The fact is, data access through a system access   through a system. The *whole* system -- including   caches -- counts. And, logic will tell you that long-stride <FONT   size=2>streaming I/O ( think Oracle Video Server, e.g. ) will work FASTER and   therefore BETTER on a DEFRAGMENTED disk. (geez)   

  I guess this one needs someone who really cares enough to   actually test it.

  side note: clearly Oracle's internal NT development   groups(s) have considerable knowledge of NT's file   system, and they would presumably have a good idea of   how to approach this issue. Why that knowledge isn't   easily propagated out to/through Oracle Tech Support   seems mysterious/weird.
  regards, ep

Received on Mon Mar 12 2001 - 16:11:55 CST

Original text of this message