Symptoms:
Hosts udpated manually, and VCF has not registered them as updated, as well as listing one host as in a failed status when it is not.
VMware Cloud Foundation
see workaround.
(1) fetch the id and hostname from the inventory. the following command will get that information.
curl localhost/inventory/hosts|jq -r '.[]|{id,hostName}'
(2) update the VCF inventory with the new version that was upgraded out of band. The <VMware ESX HOST ID> is the id corresponding to the hostname in step 1 response.
curl -X PATCH '<SDDC_Manager_FQDN >/inventory/entities/<VMware ESX HOST ID>' -d '{"version":"7.0.2-18426014", "type":"ESXI"}' -H 'Content-Type:application/json'
we need to create a new key and add it the known_hosts file. Commands would be as below:
ssh-keyscan -t ecdsa-sha2-nistp256 -p 22 <ESXHOST_FQDN> 2>/dev/null >> /etc/vmware/vcf/commonsvcs/known_hosts
ssh-keyscan -t ssh-ed25519 -p 22 <ESXHOST_FQDN> 2>/dev/null >> /etc/vmware/vcf/commonsvcs/known_hosts
Replace the ESXHOST_FQDN with the actual FQDN of the host.