Peter Thurner <[email protected]> wrote: > I was wondering how the data in the RRD Databases is saved.
It is normalised and consolidated, except under certain vary precise conditions it doesn't sotre the values you put in. See http://www.vandenbogaerdt.nl/rrdtool/ - in particular the one on Rates, normalizing and consolidating. > Also what I don't understand is that rrdtool fetch always gives me a > seconds timestamp one hour after the time I tried to commit to the RRD: > > last line of rrdtool fetch output in my script: > 947808000: -nan > > blue@aptenodytes:~/test$ date -d "Jan 14 2000" +%s > 947804400 > blue@aptenodytes:~/test$ date --date @947804400 > Fri Jan 14 00:00:00 CET 2000 > > # The timestamp I get from fetch: > blue@aptenodytes:~/test$ date --date @947808000 > Fri Jan 14 01:00:00 CET 2000 Fetch does tend to include the end time of the currently accumulating bucket - which is normally unknown and so reports as NaN. Also, be aware that RRD always works in UTC, so if you have (say) a consolidation of 86400 seconds/CDP (Consolidated Data Point) then you will only see timestamps for CDPs that are at 1AM CET - or 2AM when you are on summer time. _______________________________________________ rrd-users mailing list [email protected] https://lists.oetiker.ch/cgi-bin/listinfo/rrd-users
