← Back to team overview

graphite-dev team mailing list archive

Re: [Question #160966]: How do I configure retention aggregation for carbon?

 

Question #160966 on Graphite changed:
https://answers.launchpad.net/graphite/+question/160966

chrismd posted a new comment:
Totally different. The carbon-aggregator daemon is used to generate
aggregate metrics from an incoming stream of metrics based on some
configured rules. So if you've got "server1.cpu" and "server2.cpu" going
to carbon, you could setup carbon-aggregator to generate a
"all_servers.avg_cpu" metric from the former two.

The type of aggregation this thread is about is rollup aggregation,
which is when datapoints already stored on disk get aggregated to reduce
the number of datapoints and hence save space. Whisper only supports
averaging, which is rather limiting. With ceres you can specify
completely custom aggregation logic.

-- 
You received this question notification because you are a member of
graphite-dev, which is an answer contact for Graphite.