[rrd-users] [unsure] Re: Things not adding up

Simon Hobson linux at thehobsons.co.uk
Tue Apr 5 22:59:12 CEST 2011


Jared Watkins wrote:

>Ah.. now I think I get it.  Put another way... the individual circuit
>maxes aren't occurring at the same moment so you can't simply add those
>max values up and expect to get the total from the stacked values.

Correct.

>This
>is because (in our case) the circuits are never going to be perfectly
>balanced in traffic volume. This also means the delta between the two
>values is a way of representing the amount of imbalance at the time of
>peak traffic.  I don't suppose there would be a way to calculate that
>value?

You would have to do it at data collection time, though I appreciate 
it might be difficult.

You can do with averaged values, but that may not give you what you 
want since the more you consolidate the data, the smoother it tends 
to get. My coarsest graphs show one pixel per day over a year, so the 
daily variation in traffic flows completely disappears.Where it makes 
sense, I have an option to show/hide a max line - and the more the 
data is consolidated, the more pronounced the difference between 
average and max.

So sticking with just two sources, you would have to modify your 
script to collect A and B, and store A+B and a measure of imbalance - 
perhaps |(A-B)|/(A+B) which would give the imbalance as a fraction of 
the total traffic (min 0 if A=B, max 1 when A=0 or B=0).

You can then consolidate A, B, sum, imbalance, using both average and 
max consolidations. You should then be able to graph a stack of A and 
B, plus a line for max(sum), and lines for max(imbalance) and 
average(imbalance) which I suspect is what you are looking for.

Obviously calculating imbalance gets harder as you add more data sources.

-- 
Simon Hobson

Visit http://www.magpiesnestpublishing.co.uk/ for books by acclaimed
author Gladys Hobson. Novels - poetry - short stories - ideal as
Christmas stocking fillers. Some available as e-books.



More information about the rrd-users mailing list