[hobbit] Folders RRD info

W.J.M. Nelis nelis at nlr.nl
Tue Sep 1 12:31:38 CEST 2009


> I am writing a new ext called “folders” that will provide a check to 
> see that folders do not exceed a certain size. For example, it will 
> check the folders that contain each user’s personal network drive, 
> list them all with their sizes, and indicate any that are over the 
> limit. I have this much worked out and think that I have the red, 
> green, yellow parts of the script working properly.
>
>  
>
> However, it would be even better if I could store the sizes of these 
> folders within Hobbit/Xymon as data in RRD and have  them graphed.
>
>  
>
> How do I send the data to Hobbit in a format that it will understand 
> as being RRD data and that will be tied to the “folders” test for this 
> host?
>
>
This sounds like a variable number of measurements (one per folder) with 
a fixed number of parameters per measurement (one (size) or two 
(size,max)). In that case you should use  the '--extra-test' option (see 
http://www.hswn.dk/hobbit/help/manpages/man8/hobbitd_rrd.8.html). You 
probably also need a script to build the header of each graph. If you 
like, I can send you sample scripts.

Within the body of the status message, you'll include the measured 
values in NCV format, which might be embedded in HTML comment. Note that 
you might need to change other lines in the status message in order to 
avoid them to be interpreted as NCV. One way of doing so is to replace 
each ":" by ":" and each "=" by "=". The script named in 
"--extra-test" writes those values to RRDs. O, include a line in the 
status message, before the measurements, stating the number of 
measurements, that is include a line with "<!-- linecount=N -->" with a 
suitable value for N.

Kind regards,
  Wim Nelis.



*******************************************************************************************************
The NLR disclaimer (http://www.nlr.nl/emaildisclaimer) is valid for NLR e-mail messages.
*******************************************************************************************************




More information about the Xymon mailing list