graphite-dev team mailing list archive

[Mark Seger <question183898@xxxxxxxxxxxxxxxxxxxxx>]

New question #183898 on Graphite:
https://answers.launchpad.net/graphite/+question/183898

For those not familiar with collectl you can read more about it here - collectl.sourceforge.net - but the short description is it's a very lightweight data collection utility that collects just about everything you'd ever want to see. Given its root in HPC monitoring, I believe it's one of the few utilities that can also collect data on InfiniBand, Lustre and even NUMA memory.  It's currently running on a number of multi-K node clusters.

One of the big wins with a tool like graphite is you can instruct collectl to log everything it collects locally (hundreds of data points) every 10 seconds and send a subset of that data to graphite if you want to keep the traffic down.  You can even send data at a  different monitoring interval to graphite.

I've written up more about it at http://collectl.sourceforge.net/Graphite.html, which is currently not linked to.

My plan is to release this in the next version of collectl, which will probably come out in a month or so, but I'd feel a lot better about it if some serious graphite users could take it for a test drive and tell me what they think.  It will also give me an opportunity to incorporate any enhancements or bug fixes people might identify.  Just shoot me an email OR reply to this posting if you're interested.

It's literally a matter of installing collectl from a tarball or rpm and running it - no configuration/setup required.

-mark



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