From: Helge Hafting <helgehaf@aitel.hist.no>
To: m.knoblauch@teraport.de, linux-kernel@vger.kernel.org
Subject: Re: Possible Idea with filesystem buffering.
Date: Wed, 23 Jan 2002 12:52:27 +0100 [thread overview]
Message-ID: <3C4EA3FB.93A98AB9@aitel.hist.no> (raw)
In-Reply-To: <3C4E85BB.FBC5C2E4@TeraPort.de>
Martin Knoblauch wrote:
>
> > Re: Possible Idea with filesystem buffering.
> >
> > From: Richard B. Johnson (root@chaos.analogic.com)
> > Date: Tue Jan 22 2002 - 17:10:27 EST
> >
> >
> > We need a free-RAM target, possibly based upon a percentage of
> > available RAM. The lack of such a target is what causes the
> > out-of-RAM condition we have been experiencing. Somebody thought
> > that "free RAM is wasted RAM" and the VM has been based upon
> > that theory. That theory has been proven incorrect. You need
>
As far as I know, there is a free target. The kernel will try to get
rid of old pages (swapout program memory, toss cache pages)
when there's too little free memory around. This keeps memory
around so future allocations and IO request may start
immediately. Maybe the current target is too small, but it is there.
Without it, _every_ allocation or file operation would block
waiting for a swapout/cache flush in order to get free pages. Linux
isn't nearly _that_ bad.
> Now, I think the theory itself is OK. The problem is that the stuff in
> buffer/caches is to sticky. It does not go away when "more important"
> uses for memory come up. Or at least it does not go away fast enough.
>
Then we need a larger free target to cope with the slow cache freeing.
> > free RAM, just like you need "excess horsepower" to make
> > automobiles drivable. That free RAM is the needed "rubber-band"
> > to absorb the dynamics of real-world systems.
>
> Question: what about just setting a maximum limit to the cache/buffer
> size. Either absolute, or as a fraction of total available memory? Sure,
> it maybe a waste of memory in most situations, but sometimes the
> administrator/user of a system simply "knows better" than the FVM (F ==
> Fine ? :-)
[...]
> I know, the tuning-knob approach is frowned upon. But sometimes there
> are workloads where even the best VM may not know how to react
> correctly.
Wasting memory "in most situations" isn't really an option. But I
see nothing wrong with "knobs" as long as they are automatic by
default. Those who want to optimize for a corner case can
go and turn off the autopilot.
Helge Hafting
next prev parent reply other threads:[~2002-01-23 11:53 UTC|newest]
Thread overview: 87+ messages / expand[flat|nested] mbox.gz Atom feed top
2002-01-23 9:43 Martin Knoblauch
2002-01-23 11:52 ` Helge Hafting [this message]
2002-01-23 12:02 ` Rik van Riel
2002-01-23 12:11 ` Martin Knoblauch
[not found] <Pine.LNX.4.33.0201231301560.24338-100000@coffee.psychology.mcmaster.ca>
[not found] ` <3C4FC478.BCC44CDF@TeraPort.de>
[not found] ` <3C4FDB80.C9F83EBB@aitel.hist.no>
2002-01-24 13:59 ` Martin Knoblauch
[not found] <Pine.LNX.4.33L.0201222008280.32617-100000@imladris.surriel.com>
2002-01-22 23:31 ` Shawn Starr
2002-01-22 23:37 ` Rik van Riel
2002-01-23 5:26 ` Shawn Starr
-- strict thread matches above, loose matches on Subject: below --
2002-01-22 21:02 Rolf Lear
2002-01-20 9:04 Shawn
2002-01-20 11:31 ` Hans Reiser
2002-01-20 13:56 ` Rik van Riel
2002-01-20 14:21 ` Hans Reiser
2002-01-20 15:13 ` Rik van Riel
2002-01-20 21:15 ` Hans Reiser
2002-01-20 21:24 ` Rik van Riel
2002-01-20 21:30 ` Hans Reiser
2002-01-20 21:40 ` Rik van Riel
2002-01-20 21:49 ` Hans Reiser
2002-01-20 22:00 ` Rik van Riel
2002-01-21 0:10 ` Matt
2002-01-21 0:57 ` Hans Reiser
2002-01-21 1:28 ` Anton Altaparmakov
2002-01-21 2:29 ` Shawn Starr
2002-01-21 19:15 ` Shawn Starr
2002-01-22 22:02 ` Hans Reiser
2002-01-21 9:21 ` Horst von Brand
2002-01-21 9:13 ` Horst von Brand
2002-01-21 15:29 ` Eric W. Biederman
2002-01-20 17:51 ` Mark Hahn
2002-01-20 21:24 ` Hans Reiser
2002-01-20 21:32 ` Rik van Riel
2002-01-21 15:37 ` Eric W. Biederman
2002-01-20 22:45 ` Shawn Starr
2002-01-20 23:11 ` Rik van Riel
2002-01-20 23:40 ` Shawn Starr
2002-01-20 23:48 ` Rik van Riel
2002-01-21 0:44 ` Hans Reiser
2002-01-21 0:52 ` Rik van Riel
2002-01-21 1:08 ` Hans Reiser
2002-01-21 1:39 ` Rik van Riel
2002-01-21 11:10 ` Hans Reiser
2002-01-21 12:12 ` Rik van Riel
2002-01-21 13:42 ` Hans Reiser
2002-01-21 13:54 ` Rik van Riel
2002-01-21 14:07 ` Hans Reiser
2002-01-21 17:21 ` Chris Mason
2002-01-21 17:47 ` Hans Reiser
2002-01-21 19:44 ` Chris Mason
2002-01-21 20:41 ` Hans Reiser
2002-01-21 21:53 ` Chris Mason
2002-01-22 6:02 ` Andreas Dilger
2002-01-22 10:09 ` Tommi Kyntola
2002-01-22 11:39 ` Hans Reiser
2002-01-22 18:41 ` Andrew Morton
2002-01-22 19:03 ` Rik van Riel
2002-01-22 20:19 ` Hans Reiser
2002-01-22 20:50 ` Rik van Riel
2002-01-22 14:03 ` Chris Mason
2002-01-22 14:39 ` Rik van Riel
2002-01-22 18:46 ` Hans Reiser
2002-01-22 20:13 ` Steve Lord
2002-01-22 20:20 ` Rik van Riel
2002-01-22 22:31 ` Hans Reiser
2002-01-22 23:34 ` Rik van Riel
2002-01-23 17:15 ` Josh MacDonald
2002-01-22 19:19 ` Chris Mason
2002-01-22 20:32 ` Hans Reiser
2002-01-22 21:12 ` Rik van Riel
2002-01-22 21:28 ` Shawn Starr
2002-01-22 21:31 ` Rik van Riel
2002-01-22 21:08 ` Chris Mason
2002-01-22 22:05 ` Hans Reiser
2002-01-22 22:21 ` Rik van Riel
2002-01-23 0:16 ` Hans Reiser
2002-01-22 22:10 ` Richard B. Johnson
2002-01-23 1:14 ` Stuart Young
2002-01-23 17:16 ` Daniel Phillips
2002-01-22 21:22 ` Chris Mason
2002-01-21 0:28 ` Hans Reiser
2002-01-21 0:47 ` Rik van Riel
2002-01-21 1:01 ` Hans Reiser
2002-01-21 1:21 ` Rik van Riel
2002-01-21 1:26 ` Hans Reiser
2002-01-21 1:40 ` Rik van Riel
2002-01-20 15:49 ` Anton Altaparmakov
2002-01-20 21:21 ` Hans Reiser
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=3C4EA3FB.93A98AB9@aitel.hist.no \
--to=helgehaf@aitel.hist.no \
--cc=linux-kernel@vger.kernel.org \
--cc=m.knoblauch@teraport.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®