[rrd-users] Modifying results for GPRINT output

Alex van den Bogaerdt alex at ergens.op.het.net
Tue Sep 4 09:53:43 CEST 2007


On Sun, Sep 02, 2007 at 01:41:43PM +0100, Piers Kittel wrote:
> Alex,
> 
> On 2 Sep 2007, at 13:12, Alex van den Bogaerdt wrote:
> 
> >The way you set it up means that the database stores ticks per second,
> >not kWh or similar.
> 
> Yes.

It's getting complicated, it's early and I'm still on my first
cup of coffee. In other words: check my calculations and reasoning!


1 Wh is 1 W during 3600 seconds is 1 Joule per second during 3600
seconds is 3600 Joule.

Each 800 ticks is 1 kWh so each tick is 1.25 Wh is 4500 Joule.

Earlier I talked about watt per second but this is wrong. That was
because you wrote kW instead of kWh.

In your database you have 1/4500 Joule per second, aka 1/4500 W.

If you want to see kWh you need to:

- multiply by 4500 to get W
- multiply by 3600 to get Wh
- divide by 1000 to get kWh

If you don't mind seeing mWh, kWh, hWh, MWh and such, you can
skip dividing by 1000 and let GPRINT handle this:
GPRINT:x:%6.2lf%sWh.

If you don't use %s, you will see numbers like 0.000003 kWh:
GPRINT:x:%0.6lfkWh.

If you want to see W instead of Wh, don't multiply by 3600.

Showing W may be a way to check your setup. Power up any device
with a known power consumption (a vacuum cleaner, a microwave oven)
and notice the graph go up by (e.g.) 650 W.  At 650 W, you should
see 1 tick every 4500/650=7(rounded) seconds.


About your interface:  instead of counting pulses and periodically
calling RRDtool (e.g. every 5 minutes) it may be better to let those
same pulses trigger an RRDtool update.  This will improve accuracy
dramatically or so I think.

cheers,
-- 
Alex van den Bogaerdt
http://www.vandenbogaerdt.nl/rrdtool/



More information about the rrd-users mailing list