Subject: Re: vm.bufmem_hiwater not honored (found trigger)
To: Thor Lancelot Simon <tls@rek.tjls.com>
From: Arto Selonen <arto@selonen.org>
List: tech-kern
Date: 11/20/2004 12:32:13
Hi!

On Mon, 15 Nov 2004, Thor Lancelot Simon wrote:

> size, I don't understand how your system got into the situation it is
> in, at all.
>
> And I would very much like to.

The vm.bufmem growth is triggered by /etc/daily find_core routine.
I have three directory structures (as separate file systems) that
hold some amount of files/data:

	/squid		# disk cache of squid
	/cvs		# anoncvs sources (src,xsrc,pkgsrc)
	/obj		# build destination

As soon as find_core starts going through those three hierarchies,
vm.bufmem starts growing (seemingly without bounds). My current
bufmem_hiwater mark is at 24 MB, and even with just find_core limited
to 'find /cvs /obj ...', vm.bufmem jumps to ~40-50 MB. Then, should it
get to /squid, it will double again to about ~90-100 MB.

I think this should be fairly easy to reproduce (with /cvs and /obj type
setup), unles there are some hidden factors involved. Let me know if you
want me to look further into this, or possibly file a PR on it.


Artsi
-- 
#######======------  http://www.selonen.org/arto/  --------========########
Everstinkuja 5 B 35                               Don't mind doing it.
FIN-02600 Espoo        arto@selonen.org         Don't mind not doing it.
Finland              tel +358 50 560 4826     Don't know anything about it.