[rrd-users] SUM, in gprint, when DS type is GAUGE

[Simon Hobson]

Danny Kjærgaard wrote:

>However i have some problems with the normalisation that you speak of.
>I loose about 3 when i calculate the total of 3-4000 this might
>not be that much. But if there is a way to make this better i would
>love to know how. I enter the data to the rrd file as precise that i
>can and i cast all data to int before inserted.
>
>Is there a way to make the 3 mails apeare agine ?

If absolute precision is required then rrd is 
probably not the right tool - rrd was designed to 
do a particular task and only works with rates, 
not absolute numbers.