Ceph: collect cluster statistics

In order to be able to troubleshoot ceph related issue we need to collect continuously the cluster health statistics.

The following commands are proposed to be run every 5 minutes on a control storing the output in logs (format: date, command name, output):

    ceph -f json status
    ceph -f json health detail
    ceph -f json osd dump
    ceph -f json osd perf
    ceph -f json osd pool stats
    ceph -f json df

Logs are rotated and saved for at least one week.

Log format:

    /var/log/ceph/ceph-stats.2015-06-09.log:

    2015-06-09 00:00:58 [df] {"stats":{"total_bytes":2852453203968,"total_...
    2015-06-09 00:00:59 [health] {"health":{"health_services":[{"mons":[{"...
    2015-06-09 00:00:59 [osddump] {"epoch":12,"fsid":"425fed93-8ac7-436d-8...
    2015-06-09 00:00:59 [osdperf] {"osd_perf_infos":[{"id":0,"perf_stats":...
    2015-06-09 00:00:59 [poolstats] [{"pool_name":"rbd","pool_id":0,"recov...
    2015-06-09 00:00:59 [status] {"health":{"health":{"health_services":[{...

These logs should be collected by fuel snapshot.