I just updated my nagios installation so that I can get cpu-core temperatures with lm_sensors. Works fine. I have pnp4nagios which give good time series trends of temperatures as well. But the problem is if I want a snapshot of CPU temperatures right now for my whole server room hardware (or at any other past instance of time) pnp4nagios is not useful. It gives a time series but not a spatial (trend over servers at a point in time) graph. If I wanted to hack together my own graphing tool what's the best place to pull out data from? Digging in I noticed the /nagios/share/perfdata/serverxxx/ directory which does have the xml and rrd files. Is this a good spot to pull data out from for each monitored server? Unfortunately that would only be for the latest time instance , right? But since pnp4nagios can plot over a time range it must have access to historic temperatures too. So, where are these stored? I suspect internally as a binary produced my rrdtool? In which case is there another way for me to access historic_perf_quantity at a given point in time for all servers? Or perhaps it is possible to generate something like this using custom rrdtool queries? Any sugesstions? -- Rahul