[rrd-users] the min and max settings ...

[Tobias Oetiker]

Today you sent me mail regarding Re: [rrd-developers] new 1.0.6 fix patch ...:
*> On Mon, Aug 23, 1999 at 04:26:03PM +0200, Tobias Oetiker wrote:
*> 
*> > *> > The other problem is, why are there so large numers in your rrd ,,, 
*> > *> 
*> > *> There shouldn't be, I use 32 bit counters as input .. strange ...
*> > 
*> > this is strange indeed ... hmpf ... guess there may be another bug lurking
*> > somewhere ... 
*> 
*> I'm afraid there is, I had all kinds of strange core dumps with rrdtool,
*> now I installed 0.99.42, and restored my backup of the files prior to
*> installing 1.0.6, and now everything seems to work as expected ...
*> 
*> First I installed 0.99.42, but that core dumped on the files that was
*> modified by 1.0.6, so I expect 1.0.6 to misbehave in some way ...

Hi Jesper,

I had another look at the RRD you have send me ... and there is one very
special line in this file:


<!-- 1999-08-19 15:35:00 MEST --> <row><v> 1.8835254802e+04 </v><v> 3.4898033753e+03 </v><v> 3.5626959293e+01 </v><v> 2.9392968287e+01 </v><v> 0.0000000000e+00 </v><v> 0.0000000000e+00 </v></row>
<!-- 1999-08-19 15:36:00 MEST --> <row><v> 1.8835254802e+04 </v><v> 6.6134086027e+182 </v><v> 4.6393178266e+180 </v><v> 2.3537585903e-145 </v><v> -1.9742871170e+108 </v><v> 5.3263290373e-315 </v></row>
<!-- 1999-08-19 15:37:00 MEST --> <row><v> 0.0000000000e+00 </v><v> 3.4898027344e+03 </v><v> 3.5626959293e+01 </v><v> 2.9392968287e+01 </v><v> 0.0000000000e+00 </v><v> 0.0000000000e+00 </v></row>
<!-- 1999-08-19 15:38:00 MEST --> <row><v> 1.8835254802e+04 </v><v> 3.4898033753e+03 </v><v> 3.5626959293e+01 </v><v> 2.9392968287e+01 </v><v> 0.0000000000e+00 </v><v> 0.0000000000e+00 </v></row>



As you can see, it seems that there was a mayor problem with your data
aquisition on 1999-08-19 15:36:00 MEST I don't know how such an enormous
values into the rrd, but I do see that you have not set a MAX limit in any
of the data source definitions. Without this, can not detect such freak
values. OTOH, as you are using COUNTER type datasources I can not see how
such a large value can have been produced ... Memmory being overwritten?
Strange input breaking the generic diff routine ?

In any case, I am pretty sure that this value accounts for a large number of
your problems ...

It would be interesting to see whether such values occur in more of your
RRDs and if they show up at the same point in time ...

cheers
tobi
-- 
 ______    __   _
/_  __/_  / /  (_) Oetiker, Timelord & SysMgr @ EE-Dept ETH-Zurich
 / // _ \/ _ \/ / TEL: +41(0)1-6325286  FAX:...1517  ICQ: 10419518 
/_/ \.__/_.__/_/ oetiker at ee.ethz.ch http://ee-staff.ethz.ch/~oetiker

--
* To unsubscribe from the rrd-users mailing list, send a message with the
  subject: unsubscribe to rrd-users-request at list.ee.ethz.ch