[rrd-users] How to make exact data

[Steve Shipway]

> I have been using rrdtool for years and being very pleased with the
> tool, except for one problem I never get to solve.
> 
> I have rrd to count occurences of viruses arriving to my mailbox, but
> I never found the right type of data and always end-up with half a
> virus arriving into 2 consecudtive heartbeat,
> 
> It would look much better to have one virus arriving at time t rather
> than having half virus a time t and half a virus at time t-1.

You are referring to the frequently-cursed Data Normalisation.

The way to avoid this is to always submit the data on the time window boundary.

Thus: for the data arriving at time t (in UNIX time format) into an RRD file with data step s, submit it to RRDTool at time (t-(t mod s)), IE at the last time step boundary prior to t, rather than using 'N'.

EG, if s=100, then t=1234567890 will be submitted for time 1234567800 and so on.

This will render the data normalisation function a null operation.

Steve

Steve Shipway
ITS Unix Services Design Lead
University of Auckland, New Zealand
Floor 1, 58 Symonds Street, Auckland
Phone: +64 (0)9 3737599 ext 86487
DDI: +64 (0)9 924 6487
Mobile: +64 (0)21 753 189
Email: s.shipway at auckland.ac.nz
 Please consider the environment before printing this e-mail : 打印本邮件，将减少一棵树存活的机会

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 5900 bytes
Desc: not available
Url : http://lists.oetiker.ch/pipermail/rrd-users/attachments/20121119/aaaa30b2/attachment.bin