[rrd-users] Re: Help needed on rrdtool create and rrdtool upd ate

[Chang-Ping Hsiao]

> > Meanwhile, I am not clear about whether this includes the 
> end points of the
> > interval if they are primary data points in an RRD.
>
> There are no primary data points in an RRD, only intervals.  

	Well, according to the document of "RRD CREATE", there really is
"something" called primary data point as in the paragraph

"When data is entered into an RRD, it is first fit into time slots of the
length defined with the -s option becoming a primary data point."

although I still don't understand what it is.  :-)

> The interval
> labeled as 99417600 (if heartbeat is 3600) is actually the time period
> from 99417600-99421199.  Any data that comes in during that 
> time is considered as representing the time from the last sample to the
time of 
> the sample.
> If there is an interval of heartbeat seconds that does not 
> contain at least two samples, that interval is treated as NaN (unknown).

	Why should there be "two" data in an interval?  If The samples only
present once an hour, there will be only one datum in 3600 seconds.

> > Serge was right on the data I tried to updated, 7200 is 
> greater than 3600,
> > but I would think RRDTool can be smart enough to determine 
> 99417900 is
> > unknown while 994176000 is 3 and 994183200 is 1, instead of thinking
> > everything in the interval including the end points are unknown.
> > 
> Nope.  RRDTool is smarter than that, and it thinks that it 
> doesn't have data for an interval if it doesn't have more than one sample
in 
> the interval.

	This is not smart to me.  Why can it not just take what it's fed
with?

> Why not use a small step and a large heartbeat?  The only 
> thing the heartbeat controls is how much time can elapse between samples
without 
> declaring the entire period unknown.  Literally, heartbeat should
represent 
> your MAXIMUM acceptable time between samples, possibly plus some margin.

	The thing is, the MAXIMUM is unknown!!  Even if I make the heartbeat
hugely big, the timestamps that have no data would come up as something
inaccurate with the smaller step.

Chang-Ping

__________________________________________ 
NetZero Platinum
No Banner Ads and Unlimited Access
Sign Up Today - Only $9.95 per month!
http://www.netzero.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