Hi, On Fri, Feb 06, 2026 at 03:07:47PM +0000, Chris Green wrote: > Looking at rdtool I don't really see how it works for all these > different time periods. Maybe I'm misunderstanding it completely.
I don't understand why you think rrdtool can't do this. You decide what granularity you want at the time you create the rrd. If you want to keep 5 minute samples for 10 years that is fine, just accept that the rrd will be sized for the ~1.05 million entries this would entail. If instead you tell it to keep 5 minute samples for 1 year and 30 minute averages for the 9 years prior to that, that will be ~105k samples for the previous year and ~157k 30 minute averages for the 9 years prior to that. And so on. When you tell it to make the graph rrdtool will choose the data set with the most accuracy for the timespan being shown. Thanks, Andy -- https://bitfolk.com/ -- No-nonsense VPS hosting
