To visually see the overall topology of a Ceph cluster, run ceph osd tree. This will show us at once the hierarchy of CRUSH buckets, including the name of each bucket, the weight, whether it is marked up or down, a weight adjustment, and an advanced attribute of primary affinity. This cluster was provisioned initially with 3 racks each housing 4 hosts for a total of 12 OSD nodes. Each OSD node (also known as host, also known as server) in turn houses 24 OSD drives.
# ceph osd tree ID WEIGHT TYPE NAME UP/DOWN REWEIGHT PRIMARY-AFFINITY -1 974.89661 root default -14 330.76886 rack r1 -2 83.56099 host data001 0 3.48199 osd.0 up 1.00000 1.00000 ... 23 3.48199 osd.23 up 1.00000 1.00000 -3 80.08588 host data002 24 3.48199 ...