For Chapter 2 – OpenShift Data Foundation deployed using local storage devices, a few suggestions could help make the procedure clearer and easier to follow.

1. Add to the Prerequisites section, before removing a healthy node:

• Make sure the monitor quorum is healthy
• Confirm that the remaining OSDs and placement groups (PGs) are in a healthy state

This helps ensure the Ceph cluster is in a good state before removing any node.

2. Add a note before Step 3 (cordon):

As part of the documented process, the monitor pod on the node being removed is manually scaled to 0. Once that happens, the Rook operator starts a 10-minute timer, after which it will trigger a monitor failover creating a new canary and monitor pod on another node, but if there's no suitable nodes are available (e.g. only 2 ODF nodes remain), the new monitor pod will stay in Pending until a valid node becomes available.

This helps to understand why a mon pod could be in Pending status during the node removal.