[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: ssh-vs-rsh benchmark result



On Tue, Aug 27, 2002 at 06:37:10PM +0900, Junichi Uekawa wrote:
> On Mon, 26 Aug 2002 09:12:48 -0500
> Rob Latham <rob@terizla.org> wrote:
> 
> > > The time is in seconds.
> > > I think it is rather significant.
> > > Starting a job using rsh for 60 nodes takes only 4 seconds,
> > > but running ssh for 60 nodes in sequence takes more than 20 
> > > seconds.
> > 
> > ... and when your 3 day comput job takes an extra 20 seconds to start
> > up, you can pat yourself on the back for finding a .00008 %
> > improvement.
> 
> ... and for a 0.1-second computing task of obtaining the load average
> of hosts, running through 1000 computer nodes, it will make a big
> difference.

But in that case why are you using ssh or rsh for that purpose, when
something like ganglia or pcp will provide you that information nearly
continuously with very little overhead.

	-Sean

-- 
_______________________________________________________________________
				
Sean Dague                sean@dague.net               http://dague.net

There is no silver bullet.  Plus, werewolves make better neighbors than
zombies, and they tend to keep the vampire population down.
_______________________________________________________________________

Attachment: pgpRN1E7bBknF.pgp
Description: PGP signature


Reply to: