netdev - Re: [PATCH] net: implement emergency route cache rebulds when gc

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20081002131403.GA6261@hmsreliant.think-freely.org>
Date:	Thu, 2 Oct 2008 09:14:03 -0400
From:	Neil Horman <nhorman@...driver.com>
To:	Evgeniy Polyakov <johnpol@....mipt.ru>
Cc:	Eric Dumazet <dada1@...mosbay.com>,
	David Miller <davem@...emloft.net>, netdev@...r.kernel.org,
	kuznet@....inr.ac.ru, pekkas@...core.fi, jmorris@...ei.org,
	yoshfuji@...ux-ipv6.org, kaber@...sh.net
Subject: Re: [PATCH] net: implement emergency route cache rebulds when
	gc_elasticity is exceeded

On Thu, Oct 02, 2008 at 11:16:31AM +0400, Evgeniy Polyakov wrote:
> Hi.
> 
> On Tue, Sep 30, 2008 at 02:42:49PM -0400, Neil Horman (nhorman@...driver.com) wrote:
> > I think what you're looking for here is a simple standard deviation, isn't it?
> > Compute the mean chain legnth, sum the squares of the deviations of each chain
> > and take the square root.  Any individual chain longer than the mean chain
> > length + 1 standard deviation can be considered an 'outlier' and therefore
> > trigger a rebuild of the table for that net namespace.
> > 
> > I full well realize that thats easier said than done, but does that seem about
> > right?  If so, I can start working on trying to build something to accomplish
> > that.
> 
> You are right, but in kernel it may not that simple to get a square
> root... We can probably play with logarithms though.
> 

I don't in truth, need a quare root operation (at least I don't think I do).
I'm abel to compute the variance rather easily, with a walk of the hash table,
then I approximate the square root by use of a for loop in which I square the
iterator and compare it to the variance.  The iterator when squared that is
closest to the variance is taken as my standard deviation.  As mentioned
previously 4*sd should cover almost all of my sample universe, identifying any
remaining outliers as the trigger for a rebuild.

Neil


-- 
/****************************************************
 * Neil Horman <nhorman@...driver.com>
 * Software Engineer, Red Hat
 ****************************************************/
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html