Question #176255 on Graphite changed:
https://answers.launchpad.net/graphite/+question/176255
Status: Open => Answered
Serge Danzanvilliers proposed the following answer:
I'm currently pushing 220K+ metrics per minute on a 8 eight core box (2x
quadcore) with 18GB ram and a ssd drive to store the whisper files. I'm
using the following setup:
1 carbon-relay -> 7 carbon-aggregators -> 4 carbon-cache (all using consistent
hashing)
The webapp is hosted on the same machine, using a 1GB memcache
MAX_CACHE_SIZE is set to inf
MAX_CREATES_PER_MINUTE is set to 250
MAX_UPDATE_PER_SECONDS is set to 1000
I use aggregators to do rewriting stuff and (obviously) aggregations.
Using multiple of them + multiple cache on the same bow allows for
better scalability as the number of input metrics will go higher. I'm
not IO bound (thanks to ssd) and the cpu load is equally balanced on all
cores except for the one running the relay wich is a bit higher.
--
You received this question notification because you are a member of
graphite-dev, which is an answer contact for Graphite.
_______________________________________________
Mailing list: https://launchpad.net/~graphite-dev
Post to : [email protected]
Unsubscribe : https://launchpad.net/~graphite-dev
More help : https://help.launchpad.net/ListHelp
