graphite-dev team mailing list archive

Thread
Date

[Question #213491]: In cluster mode, how do I do aggregation?

To: graphite-dev@xxxxxxxxxxxxxxxxxxx
From: DiegoV <question213491@xxxxxxxxxxxxxxxxxxxxx>
Date: Tue, 06 Nov 2012 20:31:04 -0000
Reply-to: question213491@xxxxxxxxxxxxxxxxxxxxx
Sender: bounces@xxxxxxxxxxxxx

New question #213491 on Graphite:
https://answers.launchpad.net/graphite/+question/213491

I am running graphite in cluster mode, where each node has one relay, several caches and one webapp. All the nodes communicate and work correctly with each other.

If I now want to add stat aggregation to the mix via the carbon-aggregator daemon, where should it be placed? Should stats coming from statsd go first to one of the node's aggregator (any at random), then to the local relay and then to the cache that the relay decides? What protocol does the aggregator use to communicate with the relay?

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