>Usually, only one time range is affected at any given time, and graphs >for any other time range look good. Usually, it just goes away after >a minute or a few minutes. But then, perhaps another time range >starts looking blocky.
I believe the RRA rollups happen at specific time offsets from UTC. So it is possible that you are looking at the graphs just before and after consolidation, when different data is available. Another issue is that in gmetad.conf you can set your polling interval. I have extended mine to as long as 60 seconds on very huge clusters to reduce the CPU/network consumption of that rather busy gmetad. But, this then also becomes the step interval in newly created RRDs. So now the RRAs cover a different range of time. This causes some of my graphs to be chunky while others are detailed, for the same time range. I think this is a bug. RRD does not require that the step be equal to the update interval: it interpolates the value at the step time. Since the number of rows are fixed and assume a specific step, that step should also be fixed, and not dependent on the update interval. I would like to study this further, and see if the default RRD settings could be improved. -twitham ------------------------------------------------------------------------- This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/ _______________________________________________ Ganglia-general mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/ganglia-general
