Start decommissioning an old MinIO server pool to migrate its data off and retire the hardware

Documented steps

1. Back up cluster settings first: run `mc admin cluster bucket export`, `mc admin cluster iam export`, and `mc admin config export` to snapshot bucket, IAM, and cluster config.
2. Confirm the deployment has at least two pools and one remaining pool has sufficient free space to absorb the data (usable capacity per the erasure code calculator must exceed the decommissioned pool's total storage).
3. Get the pool TARGET: run `mc admin decommission status ALIAS` with no TARGET to list every pool with its ID/endpoint; for config.yaml deployments pools may appear as a `hash:<hex>` identifier.
4. Start decommissioning with `mc admin decommission start ALIAS TARGET` where TARGET is the full pool description (e.g. `https://minio-{01...04}.example.net:9000/mnt/disk{1...4}`) or the full `hash:<hex>` string.
5. The pool immediately enters read-only/draining status; MinIO automatically migrates objects to remaining pools weighted by each pool's free space, serially if multiple pools were queued.
6. Monitor progress with `mc admin decommission status ALIAS TARGET`; it shows migration progress (objects moved, data moved, remaining data, transfer speed, ETA) and next steps.
7. After status shows Complete, remove the drained pool's definition from the cluster config and reload (see the expand/decommission guides) to finish retiring the hardware.