[rrd-users] Re: discrepancy with RRD consolidation

Alex van den Bogaerdt alex at ergens.op.het.net
Fri Jan 20 03:24:00 MET 2006


On Thu, Jan 19, 2006 at 05:29:45PM -0800, apex (sent by Nabble.com) wrote:

> I have a discrepancy with updating RRD. I update 9 PDP's with a value of 1, and the 10th PDP with a value of 10. I do these repeatedly, and I notice RRD retains an ever-decreasing value for the 10th PDP. Why wouldn't the value be consistent, if not simply "10"? Can someone explain this behavior? Is there a way to configure a GAUGE to be exactly the number sent during an update?

...snip...
> 1137718080: 1.0000000000e+00 0.0000000000e+00
> 1137718140: 1.0000000000e+00 0.0000000000e+00
> 1137718200: 8.7295979500e+00 4.2942210833e+00
> 1137718260: 2.2779059500e+00 7.0994775000e-01
> 1137718320: 1.0000000000e+00 0.0000000000e+00
> 1137718380: 1.0000000000e+00 0.0000000000e+00
...snip...
> 1 0 1137718088 Thu Jan 19 19:48:08 EST 2006
> 1 0 1137718148 Thu Jan 19 19:49:08 EST 2006
> 10 5 1137718208 Thu Jan 19 19:50:08 EST 2006
> 1 0 1137718268 Thu Jan 19 19:51:08 EST 2006
> 1 0 1137718328 Thu Jan 19 19:52:08 EST 2006
...snip...

No, it is not a discrepancy.  It is the way RRDtool works.
Why? See my site, and the archives of this list, and the
tutorials.

Your last question: yes, under the condition that you have
exactly one PDP per CDP, and update at exactly "n" times
your {step size} seconds past januari 1st, midnight UTC.
1137718200 is such a timestamp, 1137718208 is not.

You can make any number of seconds fit this test, by setting
your step size to 1.  Or you can massage your updates so that
they are at a whole multiple of your step size.

Or you can use the LAST consolidation function.  Then you get
what you ask, but probably not what you want.

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

--
Unsubscribe mailto:rrd-users-request at list.ee.ethz.ch?subject=unsubscribe
Help        mailto:rrd-users-request at list.ee.ethz.ch?subject=help
Archive     http://lists.ee.ethz.ch/rrd-users
WebAdmin    http://lists.ee.ethz.ch/lsg2.cgi



More information about the rrd-users mailing list