[mrtg] Re: mrtg for home users: more precision than '5 minuteaverage'?

[Constantine A. Murenin]

On 30/03/06, PAUL WILLIAMSON <pwilliamson at mandtbank.com> wrote:
> >>> "Constantine A. Murenin" <mureninc at gmail.com> 03/30/06 2:25 PM >>>
> > Thanks for suggestion, but that is all too complicated, plus I don't
> > feel comfortable with installing some monster CGIs and putting them
> on
> > the web. Did you see the security warning on routers2 site? That's
> not
> > for me. :)
> >
> > I want a simple and elegant solution, which would use plain mrtg
> with
> > 'Interval: 1', the performance hit would be trivial anyways, so why
> is
> > this feature still not implemented?
> >
> > Constantine.
>
> Because the rate of "normalization" would be greater, thus, ironically
> enough,
> reducing the accuracy.  If you insist on using a 1 minute poll, you can
> use
> RRDTool and your own polling solution.  That will work perfectly fine.

All accuracy I care about is that if the dsl line is used full-speed
for 1 minute, it's reported as full speed in the graph; and if the
line is not used, reported not used.

Obviously, I don't have time now to write my own polling solution, and
I don't see why this cannot be done with mrtg.


> Just because you run routers2.cgi, doesn't mean you need a web browser
>
> exposed to the internet.  Just shut off port 80 for all internet,
> non-internal
> based addresses and there is no security risk.

I don't want to do that. I trust my chroot'ed httpd, and I trust that
mrtg should not stop at providing 5-minute accuracy in regular mode as
opposed to RRDTool mode.

I don't want to install any third-party CGIs, I'm quite happy with
what mrtg offers, except that 'Interval: 1' doesn't work. :(

Cheers,
Constantine.

--
Unsubscribe mailto:mrtg-request at list.ee.ethz.ch?subject=unsubscribe
Archive     http://lists.ee.ethz.ch/mrtg
FAQ         http://faq.mrtg.org    Homepage     http://www.mrtg.org
WebAdmin    http://lists.ee.ethz.ch/lsg2.cgi