[rrd-developers] Accelerator Daemon

[Scott Brumbaugh]

Hi Tobias,

On Wed, Jul 02, 2008 at 05:54:12AM +0200, Tobias Oetiker wrote:
> Scott,
> 
> Yesterday Scott Brumbaugh wrote:
> 
> > Rrdtool allows for automated network monitoring through the rrd_xport
> > function.  You can readout the current rate estimates and use them to
> > update filters to do automated monitoring.  For this type of
> > application it is good to be able to read out the last few rate
> > estimates quickly after an update.
> 
> for monitoring, it is not very efficient to pull data back out of
> rrd files ... what you should do, is to use updatev (I wonder how
> the accelerator daemon deals with that one) and then take the
> output todo your monitoring on the fly ... this is MUCH more
> efficient.
> 
> cheers
> tobi

Thanks for the tip, I think you are absolutely right that reading
updatev output is more efficient than rrd_xport.  An advantage we get
from xport is the ability to use DEF and CDEF statments to form
aggregate rate estimates from a large number of rrd data sources.  We
do rely on most of the xport data being in filesystem cache for the
readout.

Also, our system is unusual in that we have our rrd databases
physically distributed across the internet and run our rrd_xport/graph
commands centrally on a different server than the rrd_updates run.
There was a patch posted to this list few weeks ago describing this
setup.  This also allows for horizontal scaling by letting us add
polling servers in parallel.

 http://thread.gmane.org/gmane.comp.db.rrdtool.devel/2216


Scott B