[rrd-users] Re: Urk. Help with a little nagging problem I'm having..? Prolly newbie mistake.

[S.D.]

On Friday 20 December 2002 11:32, Alex van den Bogaerdt wrote:
> On Fri, Dec 20, 2002 at 11:14:43AM -0800, S.D. wrote:
>
> > RRDs::create ("fxp0.rrd","--start=now","--step=4", 
> > 
"DS:inbytes:COUNTER:5:U:U","DS:outbytes:COUNTER:5:U:U","RRA:AVERAGE:0.5:4:3075840")))
> 
> This keeps > 569 days worth of information.  Not 6 months to a year.
> 
> You are keeping the data in a 16-second resolution.  That doesn't
> explain the high value but I just wanted to let you know this is
> not what you tell us you want.

But recording the data itself must happen every 4 seconds, correct? The 
average itself is what's created every 16 seconds? ah I see. So that value is 
quite a large RRA then, because it's room for all the entries that would make 
up 3075840 *average values* which is that value times 16 seconds for time, 
and 3075840 x 4 for recorded values (or does it just take four spots for the 
recorded values and not remember anything but the AVERAGE of past values?)

> > Argh! Notice the e+08? Impossible! :-)
> > 
> > Grateful thanks in advance for any assistance.
> 
> Make a small example script that uses fixed times (so: do not use
> "N" as the time) with fixed values and which demonstrates the problem.
> Do not keep a lengthy RRA, just a couple of lines should do.

Sure--I'll make a quick little dinky Perl script.

Here:

#!/usr/pkg/bin/perl -w

use RRDs;
$|=1;

if ( ! -f "fxp0.rrd" ) {
        print "Apparently fxp0.rrd doesn't exist..";
        if (!(RRDs::create ("fxp0.rrd","--start=1040408332","--step=4", 
"DS:inbytes:COUNTER:5:U:U","DS:outbytes:COUNTER:5:U:U","RRA:AVERAGE:0.5:4:50"))) 
{
                print "RRDs::create failed..?\n";
                exit 1;
        } else {
                print "RRDs::create successful!\n";
        }
}

RRDs::update ("fxp0.rrd", "1040408336:54006369:8317243");
RRDs::update ("fxp0.rrd", "1040408340:54008843:8317295");
RRDs::update ("fxp0.rrd", "1040408344:54009338:8317347");
RRDs::update ("fxp0.rrd", "1040408348:54011862:8317399");
RRDs::update ("fxp0.rrd", "1040408352:54013059:8317682");
RRDs::update ("fxp0.rrd", "1040408356:54013702:8318239");
RRDs::update ("fxp0.rrd", "1040408360:54016412:8335945");
RRDs::update ("fxp0.rrd", "1040408364:54017607:8336219");
RRDs::update ("fxp0.rrd", "1040408368:54017213:8336271");
RRDs::update ("fxp0.rrd", "1040408372:54017500:8336323");
RRDs::update ("fxp0.rrd", "1040408376:54019567:8337219");
RRDs::update ("fxp0.rrd", "1040408380:54021420:8337271");
RRDs::update ("fxp0.rrd", "1040408384:54051667:8342237");
RRDs::update ("fxp0.rrd", "1040408388:54053327:8342289");
RRDs::update ("fxp0.rrd", "1040408392:54055685:8342341");
RRDs::update ("fxp0.rrd", "1040408396:54057681:8342393");
RRDs::update ("fxp0.rrd", "1040408400:54059534:8342445");

exit 1;

...  and now retrieve the values with:

rrdtool fetch fxp0.rrd AVERAGE --start=1040408336 --end=1040408400

And here's what I see:

1040408352: 4.1812500000e+02 2.7437500000e+01
1040408368: 2.6843571562e+08 1.1618125000e+03
1040408384: 2.1533750000e+03 3.7287500000e+02
1040408400: 4.9168750000e+02 1.3000000000e+01

There's that e+08 again. D'oh!

> This way you make it easy for us to help you (we only have to run
> the script).

My apologies.

> Said script should
> - create a small database
> - insert some values into the database
> - fetch the problem area from the database

Does everything but the fetch part, which is a cut&paste.

> Also compute the numbers yourself and show what you think should
> happen.  A formula with an outcome.
> 
> Alex

Well, I was hoping that simply by inspection the data itself would show that 
the numbers don't grow (at any time) bigger than ..  oh..  20,000 units at a 
time. That's nowhere near the 268435715.62 value the AVERAGE claims it is.

Hey maybe the version I'm using is buggy: 

:draco:12:25:54 /v/ht/rrd$ pkg_info | grep -i rrd
rrdtool-1.0.33nb1   Data analysis tool generating graphical representations
:draco:12:29:10 /v/ht/rrd$ 

Hrm.. have to check out what's changed in the new version.

But I just have the feeling I'm probably just designing the RRD incorrectly 
somehow.

Thanks for your time..! I know I'm asking a lot.

--
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://www.ee.ethz.ch/~slist/rrd-users
WebAdmin    http://www.ee.ethz.ch/~slist/lsg2.cgi