[mrtg] Advice: monitoring GPS based clock

McDonald, Dan Dan.McDonald at austinenergy.com
Thu Sep 8 14:47:09 CEST 2011


> From: Steve Shipway [mailto:s.shipway at auckland.ac.nz]
> Sent: Wednesday, September 07, 2011 6:07 PM
> To: McDonald, Dan; mrtg
> Subject: RE: [mrtg] Advice: monitoring GPS based clock
> 
> > routers.cgi*UpperLimit[$target_name]: -120
> > routers.cgi*LowerLimit[$target_name]: -200
> ...
> > routers.cgi*Options[$target_name]: noo nototal nopercent fixunit
> > routers.cgi*Graph[$target_name]: $graphtarget_name noo $color
> ...
> > Do I need to set routers.cgi*UpperLimit[$graphtarget_name]: -120 to
> make
> > this work right?  

> The UpperLimit and LowerLimit options are not inherited by a
> userdefined graph; so, you should have
> 
> routers.cgi*UpperLimit[$graphtarget_name]: -120
> routers.cgi*LowerLimit[$graphtarget_name]: -200
> 
> to define the options for the userdefined graph as well.  If you also
> use
> 
> routers.cgi*Options[$graphtarget_name]: rigid
> 
> then this will force the userdefined graph to be between these Y
> values, regardless of the data values.

It's honoring LowerLimit now, but not UpperLimit (still drawing a graph all the way up to 1, which I think is the smallest MaxBytes[] value...)
 

 
> Since the limits, and the data, are all negative this can cause some
> odd behaviour in some situations.  You might like to upgrade to v2.22
> of routers2.cgi from the website as this fixed a minor bug in this
> regard?

Ok, no change.  Then again, I'm still running rrdtool 1.2.15, so no telling there....  I can probably upgrade that a bit.  One of these days I'll get around to rebuilding this machine completely....



 Daniel J McDonald, CCIE # 2495, CISSP # 78281
Austin Energy


 



More information about the mrtg mailing list