Hardware failures are a fact of life. The larger your cluster grows, the more frequently you'll see failures. Fortunately Ceph goes to great lengths to ensure the durability and availability of your precious data.
With proper deployment and consideration of fault domains your Ceph cluster will cruise right through common hardware failures. It is however essential to integrate Ceph into your organization's monitoring framework so that you can address failed components before they pile up. Earlier in this chapter we introduced Ceph's logging strategies and showed some examples; in the next chapter we'll focus on monitoring. At the very least you will want to frequently consult your MON or admin node for cluster status, and ...
