You can use the Container Cloud API to restart an inspection of a bare metal
host in MOSK clusters. For example, this procedure is useful
when hardware was changed. This works for bare metal hosts that were not
provisioned yet or were successfully deprovisioned.
The workflow of the reinspection procedure desribed above is as follows:
Ensure that the BareMetalHostInventory object is not bound to any
Machine object and it is in the available state.
Edit the BareMetalHostInventory object to initiate an inspection of the
bare metal server that hosts the node.
Note
Before update of the management cluster to Container Cloud 2.29.0
(Cluster release 16.4.0), instead of BareMetalHostInventory, use the
BareMetalHost object. For details, see BareMetalHost resource.
Caution
While the Cluster release of the management cluster is 16.4.0,
BareMetalHostInventory operations are allowed to
m:kaas@management-admin only. This limitation is lifted once the
management cluster is updated to the Cluster release 16.4.1 or later.
To restart the inspection of a bare metal host:
Using kubeconfig of the management cluster, access the Container Cloud
API and inspect the BareMetalHost object:
NAME STATE CONSUMER ONLINE ERROR AGE...managed-worker-a-storage-worker preparing managed-worker-a-storage-worker false 5d3hmanaged-worker-b-storage-worker preparing managed-worker-b-storage-worker false 5d3hmanaged-worker-c-storage-worker available false 5d3h
In the system response above, the managed-worker-c-storage-worker bare
metal host is in the available state and has no consumer (not bound
to any Machine). Therefore, you can reinspect it.
Open the required bare metal host object for editing:
Since the management cluster update to 16.4.0 (MCC 2.29.0)