external rrd create and update

[noname]

>> so is it mean I increase the heartbeat will fix that problem?

Probably yes -- as far as you update rrd within the time of heartbeat.

FYI: You can modify heartbeat value by rrdtune without re-creating rrd file.

[chefchanyu]

no looking good.
because they taking last valid input to replace with "NaN".
is there possible way replace "NaN" with "0" rather than last valid input?

[noname]

because they taking last valid input to replace with "NaN".

Sorry I can't understand what this means.

Please indicate, "What value you updated (last and previous last)" and "When you updated that value (last and previos last)".

[chefchanyu]

Code: Select all

20110615030017 1
: 15 seconds
20110615030032 1

the 15 seconds has no input it mean = "0". when i change the heartbeat to 60, the result is like this.

Code: Select all

<!-- 2011-06-15 03:00:17 EST / 1308070817 --> <row><v> 1.0000000000e+00 </v></row>
                        <!-- 2011-06-15 03:00:18 EST / 1308070818 --> <row><v> 1.0000000000e+00 </v></row>
                        <!-- 2011-06-15 03:00:19 EST / 1308070819 --> <row><v> 1.0000000000e+00 </v></row>
                        <!-- 2011-06-15 03:00:20 EST / 1308070820 --> <row><v> 1.0000000000e+00 </v></row>
                        <!-- 2011-06-15 03:00:21 EST / 1308070821 --> <row><v> 1.0000000000e+00 </v></row>
                        <!-- 2011-06-15 03:00:22 EST / 1308070822 --> <row><v> 1.0000000000e+00 </v></row>
                        <!-- 2011-06-15 03:00:23 EST / 1308070823 --> <row><v> 1.0000000000e+00 </v></row>
                        <!-- 2011-06-15 03:00:24 EST / 1308070824 --> <row><v> 1.0000000000e+00 </v></row>
                        <!-- 2011-06-15 03:00:25 EST / 1308070825 --> <row><v> 1.0000000000e+00 </v></row>
                        <!-- 2011-06-15 03:00:26 EST / 1308070826 --> <row><v> 1.0000000000e+00 </v></row>
                        <!-- 2011-06-15 03:00:27 EST / 1308070827 --> <row><v> 1.0000000000e+00 </v></row>
                        <!-- 2011-06-15 03:00:28 EST / 1308070828 --> <row><v> 1.0000000000e+00 </v></row>
                        <!-- 2011-06-15 03:00:29 EST / 1308070829 --> <row><v> 1.0000000000e+00 </v></row>
                        <!-- 2011-06-15 03:00:30 EST / 1308070830 --> <row><v> 1.0000000000e+00 </v></row>
                        <!-- 2011-06-15 03:00:31 EST / 1308070831 --> <row><v> 1.0000000000e+00 </v></row>
                        <!-- 2011-06-15 03:00:32 EST / 1308070832 --> <row><v> 1.0000000000e+00 </v></row>

[noname]

Thanks, I've understood about your situations.
Unfortunately(?), it seems that this depends on RRDtool behavior.

http://oss.oetiker.ch/rrdtool/doc/rrdcreate.en.html

The HEARTBEAT and the STEP
...
The "heartbeat" can be short (unusual) or long (typical) relative to the "step" interval between PDPs.
A short "heartbeat" means you require multiple samples per PDP, and if you don't get them mark the PDP unknown.
A long heartbeat can span multiple "steps", which means it is acceptable to have multiple PDPs calculated from a single sample.
An extreme example of this might be a "step" of 5 minutes and a "heartbeat" of one day, in which case a single sample every day will result in all the PDPs for that entire day period being set to the same average rate.

Then, if you want to avoid storing valid value when no data comes, choose between the below two.
(Is there another choice?)

(A) Reduce heartbeat (2 or 3 or else). But you will lost more "valid" data. (as you faced)
(B) Use this heartbeat (60secs), and supplement by NaN (or 0) during every seconds which no data comes. (as you mentioned)

---

BTW, I misunderstood in the previous post.

>> don't know why start print out form 2008 to 2011.

Probably you specified that timestamp (3 years ago) by "--start" option when creating this rrd.

It seems that this relates "--start" option & rows (94608000) in RRA definitions. Sorry.
(I'm not sure in detail)

[chefchanyu]

Thanks noname. I had learn lot from you.
I think only way fix this problem is change the Perl script.