[rrd-users] Re: Unexpected NaN values

jeffrey.j.petter at verizon.com jeffrey.j.petter at verizon.com
Sat Nov 4 01:05:04 MET 2006

Alex wrote: "At least you know that your script got values.  Try
duplicating the
update command, prepend with echo and log that too.

No special data source names that could be treated differently? I know
some version accept names like "3com" and others don't (or a similar
problem; can't remember exactly).

Look at the updates carefully.  Are the timestamps further than 600
seconds apart?

Show a couple of consecutive loglines, perhaps something obvious
is visible."

As you suggested, I modified the update script to echo the update command
to the log file. A few consecutive loglines are pasted here:
blh-ebiz-int-fw-1: 2006-11-03 15:00:01
1162584001:.RRDs::update ("/usr/local/rrdtool/rrd/blhint1.rrd",

blh-ebiz-int-fw-1: 2006-11-03 15:05:01
1162584301:.RRDs::update ("/usr/local/rrdtool/rrd/blhint1.rrd",

blh-ebiz-int-fw-1: 2006-11-03 15:10:00
1162584600:.RRDs::update ("/usr/local/rrdtool/rrd/blhint1.rrd",


Nothing really jumped out at me, so after a bit too long, I tried to enter
the command directly from the command line, with the values from random log
entries. Each time, I received an error that stated "<number> is not a
simple integer", and that *number* was always from the sixth field, and in
this case, has always been negative. I don't understand how a COUNTER could
be negative, but this one is. Since that field does not contain data that I
am presently graphing, I just modified the code to write the value of the
fifth field into the sixth field for the update statement, and it has been
working since. This was just a quick fix to try to isolate the problem. I
was thinking perhaps a better long-term solution would be to add a CDEF
that tested the value for less than 0, and if true, to multiply by a
negative 1. Does that sound like a viable approach, or could that lead to
unforeseen problems, which would then have me chase down why the counter is
returning negative numbers.


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