[Top] [All Lists]

Re: Strange behavior on the 2.4.18 XFS tree?

To: Michael Sinz <msinz@xxxxxxxxx>
Subject: Re: Strange behavior on the 2.4.18 XFS tree?
From: Charles Shannon Hendrix <shannon@xxxxxxxxxxxxx>
Date: Wed, 15 May 2002 15:38:27 -0400
Cc: Steve Lord <lord@xxxxxxx>, Keith Owens <kaos@xxxxxxx>, linux-xfs@xxxxxxxxxxx
In-reply-to: <3CE2A4BE.7060508@wgate.com>
References: <22203.1021466116@ocs3.intra.ocs.com.au> <1021469940.19994.388.camel@jen.americas.sgi.com> <3CE2A4BE.7060508@wgate.com>
Sender: owner-linux-xfs@xxxxxxxxxxx
User-agent: Mutt/
On Wed, May 15, 2002 at 02:11:10PM -0400, Michael Sinz wrote:

> >For the record, there is no use of the BKL in xfs at all. Michael's
> >problem sounds more like memory getting chewed up by the dcache and
> >icache.
> Ahh, I have seen this happen in the past - I have not understood why
> so much memory was used at times in the system.  The machine is running
> headless (no local X display) and thus it should have lots of resources.

Can you elaborate on this?

Are you saying that a report from the free program does not account for all

I'm asking because a constant problem for me with Linux 2.4.x kernels is that
I seem to gradually lose memory.

Booting up my system takes under 20MB of RAM. After running a few days with X
on my XFS 1.1 system, going back to console shows 50-100MB unaccounted for in
the memory stats.

I mean, after subtracting buffers and everything, memory is missing.

A lot of wierd filesystem behavior also started happening on my Linux systems
after I started noticing memory seems to be going away.

It might be nothing but /proc lying about memory use, as someone suggested
some time back in a kernel discussion I read. However, it bothers me a bit
that bad behavior, espeically with XFS, seems to have accompanied this missing
memory problem.

Of course, if that isn't what you meant, maybe I'm the only one... :)


<Prev in Thread] Current Thread [Next in Thread>