Have you ever tried to monitor your Ceph cluster? Have you ever tried it... with Riemann?
If your curious, go ahead and build it. We won't tell anyone. Running it is simple. Boot up a Riemann instance and run this from an account and machine with the ceph client available:
./ceph-reimann --delay 10
This will run ceph report, parse the results and ram them into Riemann with great speed. And it will do this every 10 seconds. Forever. Providing it's listening on 127.0.0.1:5555.
Is your glorious Riemann instance running on another server or port? No problem. Toss in the optional --server argument and behold: your metrics will flow to the whatever host you please.
./ceph-riemann --delay 10 --server 74.125.239.18:5555
That's it. There are no other options, flags or any other bullsh*t. Enjoy.