[Reposted using links instead of attachments] Hi,
I have done some further testing using the TOTAL function and I don't believe the TOTAL function is calculating the value correctly in some scenarios. To try and explain this easily I have put together two example recreates: http://netnix.org/correct.txt http://netnix.org/incorrect.txt Please bear with me on this one, but in the examples I have two RRAs - one with a 5 minute step and another with a 24 hour step. Thanks to the information that Alex has previously provided I calculated that my 24 hour RRA runs from midnight to midnight. I have manually calculated the total amount of bytes I have placed into my RRD as follows: ifOutOctets = 7.158 TB ifInOctets = 1.458 TB However, the results I get from rrdgraph (both graphs are covering the entire data period) are as follows: (1d - 5m step) ifOutOctets Total = 7.158 TB (1d - 5m step) ifInOctets Total = 1.458 TB (1y - 24h step) ifOutOctets Total = 12.126 TB (1y - 24h step) ifInOctets Total = 2.471 TB The 5 minute RRA is spot on, but the 24 hour RRA has over calculated the data. When I looked at my data I noticed I started to put entries into the RRD at some time after midnight and not at midnight so the beginning part of the 24 hour RRA would be missing. I then went and filled my RRD back to midnight duplicating the first update every 300 seconds - I haven't added or removed bytes. I then generated the graphs again and the data was spot on: (1d - 5m step) ifOutOctets Total = 7.158 TB (1d - 5m step) ifInOctets Total = 1.458 TB (1y - 24h step) ifOutOctets Total = 7.158 TB (1y - 24h step) ifInOctets Total = 1.458 TB I am assuming the data that is missing at the beginning of the RRA would be considered UNKNOWN and I would expect the TOTAL function to ignore it? Thanks, Chris _______________________________________________ rrd-users mailing list [email protected] https://lists.oetiker.ch/cgi-bin/listinfo/rrd-users
