READ the docs on rrds at the rrdtool home page.

What you're seeing is a RATE 1004.465... units per second * 300 (or
whatever) seconds per interval = actual count

-----Burton

-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of The
Jetman
Sent: Monday, June 09, 2003 7:56 PM
To: NTOP Mail List
Subject: [Ntop] Is There a Trick to Understanding RRD ?




    I have a Python script, that reaches into a RRD file and extracts
data samples, based on whatever file, date/time, etc, I specify.  The
example below [is intended to show] the average amt of data my own
workstation rcvd from today starting at noon and the subsequent three
hours:

RRD Report for Mon Jun  9 12:00:00 2003:
os.system( 'rrdtool fetch
/var/db/ntop/rrd/interfaces/rl0/hosts/192.168.0.50/bytesRcvd.rrd \
AVERAGE -r 3600 -s 1055178000 -e 1055188800')
===============
                  counter

1055178000: 1.0044655778e+03
1055181600: 1.1332594444e+03
1055185200: 3.7978324444e+02
1055188800: 2.6146522222e+02

    Looking at the Host Traffic Stats for my workstation, I see something
completely different for the Total Traffic Rcvd (2.3MB, 3.9MB, 1.1MB,
862.9KB.)  I understand that the average amt of data received during the
desired period may be slightly different than the actual amt of data
received, but these results aren't even close.

    Hence, the title of this e-mail.  Truth is, every file I look at, I try
and compare the results I get from my script and I have yet to see RRD data
that corresponds to what's reported in web pages.

    Later....Jet


===============  From the desk of Jethro Wright, III  ================
+      Nothing causes self-delusion quite so readily as power.       =
===  [EMAIL PROTECTED]  =========================  Liu Binyan  ===
_______________________________________________
Ntop mailing list
[EMAIL PROTECTED]
http://listgateway.unipi.it/mailman/listinfo/ntop

_______________________________________________
Ntop mailing list
[EMAIL PROTECTED]
http://listgateway.unipi.it/mailman/listinfo/ntop

Reply via email to