If your cluster is performing really slowly, to the point that it is barely servicing IO requests, then there is probably an underlying fault or configuration issue. These slow requests will likely be highlighted on the Ceph status display with a counter for how long the request has been blocked. There are a number of things to check in this case.
Extremely slow performance or no IO
Flapping OSDs
Check ceph.log on the monitors, and see whether it looks like any OSDs are flapping up and down. When an OSD joins a cluster, its PGs begin peering. During this peering process, IO is temporarily halted, so in the event of a number of OSD’s flapping, the client IO can be severely impacted. If there is evidence of flapping OSDs, the next step is to go through the logs...
