This simply contains a bunch of common example configs for certain backends and inputs
The basic mode for cadent is a consistent hashing relay which.
cadent --config=configs/graphite/graphite-relay.toml
prereg is the accumulators/writers/api section of cadent.
nameing is as follows .. each "input" type is put into a folder (statsd, carbon2, graphite)
{input}-{writer}-{type}.toml
where
{input} is "graphite", "statsd", "carbon2"
{writer} is "mysql", "cassandra", "elasticsearch", "kafka", "whisper", "relay"
the "relay" is simply using cadent as a consistent hash ring fowarder.
{type} is "series" or "flat" (or in the case of pure relay, the output format "graphite", etc)
To run
cadent --config={input}-config.toml --prereg={input}-{writer}-{type}.toml
The "apionly" directory has a config for just operating the API w/o all the writer stuff
Please note this API Only mode is in progress, It will easily read from the backend DBs, but if using the Series type for your metrics on different writers, the mechanism to pull from the relevent write-back caches is not yet complete.
cadent --api=configs/apionly/api.toml