No subject


Wed Mar 11 18:56:50 EDT 2009


a better match for a 2Ghz+ processor but that all depends on how many
servers you are running. If its a single server it shouldn't be an
issue.

    Steve / K
----- Original Message ----- 
From: "Anton Jansen" <gradius at fmf.nl>
To: <bf1942 at icculus.org>
Sent: 21 February 2003 15:50
Subject: Re: [bf1942] Proof of a memory leak?


On Fri, 21 Feb 2003, Killing wrote:

> That it may but 140MB is not a huge amount in the slightest.
That's true.

>If your
> machine is killing process @ this size its an issue on your machine
not
> the application. 
The only thing that is sure is that the size of the BF process is at
least 
140 mb when it's killed. It can be however that the size has grown 
exponential in the time after the 140 mb measurement.

> You might want to look @ what kernel you are running
> and if it has any know issue for ages Linux 2.4 kernel's / net drivers
> where buggy as hell and leaked like a sieve. 
2.4.19-rc1, I'll check that. Thanx for the tip.

> One check that seems to
> identify this is increasing allocations to buffer allocation under
> vmstat. 
Hmm, I used a ps aux | grep bf | cut ... approach to determine the
memory 
use.  I could off course add the total used memory of my system in the 
graph ... Nice idea!

>Also what's the spec on your machine and swap available?
> 
P4, 2.0 ghz 512 mb ram, 256 mb swap

Kind regards,

Anton Jansen
--------------------------+-------------------------





More information about the bf1942 mailing list