graphite-dev team mailing list archive

[Andrew Crowell <question171056@xxxxxxxxxxxxxxxxxxxxx>]

New question #171056 on Graphite:
https://answers.launchpad.net/graphite/+question/171056

Hi there,

Until recently, my company used graphite as a graphing system for aggregated server statistics. The version of graphite we use has webapp/graphite/readers.py and  webapp/graphite/finders.py, which it uses to collect data for metrics from our stats server, and it works. 

Recently we decided there were possible reasons to consider updating to the trunk version in Bazaar, to take advantage of some new features in graphite. But it seems that this Reader/Finder interface has disappeared in favor of an Carbon interface that requires continually pushing data to it. This presents us with some difficulties, however.

This means that we can longer do a "load on demand" request to our statistics server, but rather push data from our stats server into graphite's Carbon server. This would have to happen constantly, because we want to see up-to-date data plotsm, and not just a cached version. And there are lots of different metrics that would be getting updated all the time. The server traffic would be too much to handle.

Not to mention, Carbon has its own data store which is redundant, at least, for our purposes. Our statistics server already has its own storage system, specially catered to our needs.

If possible, we would like to cut out the middle man, and just use graphite for its graphing of metrics, without having to go through Carbon. Is there an interface to define other data sources than carbon, like there used to be? It would be very useful.

Thanks.

-- Andrew Crowell

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