From: unixsys@ssg.com (Rick Emerson) Subject: Re: [comp.os.linux]: Re: File system issues! Date: Tue, 14 Jul 1992 11:20:02 GMT
news-daemon@optical.bms.com writes:
> Newsgroups: comp.os.linux
> Path: synapse.bms.com!jvnc.net!yale.edu!qt.cs.utexas.edu!cs.utexas.edu!zaphod
> From: quale@saavik.cs.wisc.edu (Douglas E. Quale)
> Subject: Re: File system issues!
> Message-ID: <1992Jul9.082537.19043@daffy.cs.wisc.edu>
> Sender: news@daffy.cs.wisc.edu (The News)
> Organization: Undergraduate Projects Lab, UW Madison
> References: <710630507.F00090@remote.halcyon.com>
> Date: Thu, 9 Jul 1992 08:25:37 GMT
> Lines: 18
>
> In article <710630507.F00090@remote.halcyon.com> Rob.Levin@f217.n3802.z1.fido
> >
> >2) Caching. It should be possible to turn on and off the write-back
> >portion of caching, both as a default at system startup time, and
> >dynamically.
>
> I'm afraid that this is unacceptable. There is already such an OS that
> runs on more even hardware than Linux does. You guessed it -- its name is
> Mud. Err, I mean DOS. The first software that anyone gets when complaining
> of DOS's poor I/O performance is a disk cache "utility". I have never known
> anyone with DOS disk cache software to disable the write cache (if available)
> just because it would be safer -- everyone wants faster, safety be damned.
Well, now you know at least one MS-DOS user who routinely disables write back
caching to limit problems while running a mailer on a timer switch on
weekends.
> Seriously, losing caching on writes would hurt *a lot*. The real solution
> is clear documentation and a good fsck, run automatically by /etc/rc.
There are two issues here: 1) Is write-back caching useful? 2) Would a
switch that disables write-back caching cripple Linux. Discussions of
MS-DOS are just so much wasted sophistry.
1) Is write-back caching useful?
Of course. The increase in performance has been documented over and over.
Well, the arguments go, since disk caching is so hot, why turn it off?
There are several users who've identified situations where the computer
running Linux may lose power without warning. Disabling write-back caching
is an attempt to minimize the exposure to risks to the filesystem structure.
2) Would a switch that disables write-back caching cripple Linux?
Who knows? I have yet to see a reasoned review of the Linux kernel code (I
just don't know enough about the kernel to comment). Many people (the above
is only one sample) have expressed opinions and subjective comments about
this question but *no one* has stated a why such a switch is good or bad
in objective terms. I have yet to see anyone say something like "in xyz
module there are this and that pieces of code which fail when write-back
stops" or "implementing this switch will take an additional 100K" or "the
whole darn kernel will have to be re-written because..."
Those who just flame MS-DOS or people who ask for "fscache -write_back 0"
don't advance the discussion.
Rick
+--------------------------------------------------------------------------+
| Richard B. Emerson | Replies may be sent to: |
| System Support Group | rick@ssg.com |
| Post Office Box 180 |-------------------------------------------------+
| Lansdale, PA 19446 USA | "Answers are the easy part |
| Voice: 215.855.1607 | questions raise the doubt." - Jimmy Buffett |
+--------------------------------------------------------------------------+