[rrd-users] NaN Math options

[Philip Molter]

We've got an .. interesting scenario.

We're graphing a traffic graph for several interfaces where we're
summing the traffic from each interface to get a total (the total
is what we're graphing, not each individual interface).  If one of
our interfaces goes down and returns no data, we get unknowns in
our database for that time period.  When we put these unknowns into
the sum, the graph disappears, because the entire value becomes
unknown.

Simple enough fix, right?  You just do something like:

  DEF:int1=file1:octets:AVERAGE
  DEF:int2=file2:octets:AVERAGE
  CDEF:int1,UN,0,int1,IF,int2,UN,0,int2,IF,+

and those pesky unknowns become '0's and all is well.  Except now,
when you graph the graph, the graph dips at the very last second
to zero and if you have any GPRINTs, the value of LAST is '0',
rather than the actual last value (unless you happen to graph it
right near the interval moment).

What we really want is to start the graph from the interval closest
in time to the current time.  Our last value isn't /really/ unknown,
and it's not really zero.  Is there a way to make the graph start
from there rather than the absolute nearest time?

* Philip Molter
* DataFoundry.net
* http://www.datafoundry.net/
* philip at datafoundry.net

--
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