How to add a host to OpenNebula

OpenNebula hosts are compute nodes that run virtual machines for the cloud. Adding a host registers a prepared KVM node with the front-end so monitoring can collect CPU, memory, and local datastore data before the scheduler places workloads there.

The CLI path uses onehost from the front-end, normally from the oneadmin account. The node should already have the OpenNebula KVM node package, matching network and storage preparation, and passwordless SSH from the front-end because OpenNebula uses that path to run host probes.

Register one node at a time and wait for the host state to move from initial monitoring to on. A host that lands in err usually needs an SSH, name-resolution, package, or remote probe fix before it can provide scheduler capacity.

Steps to add a KVM host to OpenNebula:

Open a front-end shell as oneadmin.
```
$ sudo -iu oneadmin
```
Test passwordless SSH from the front-end to the prepared node.
```
$ ssh host01 hostname -f
host01.example.net
```
Accept a new SSH host key only after comparing it with a trusted console, inventory, or rotation record. OpenNebula uses the same oneadmin SSH path for monitoring and remote operations.
Tool: SSH Key Fingerprint Checker
Register the KVM host.
```
$ onehost create host01 --im kvm --vm kvm
ID: 2
```
Use driver names that match the prepared node type. For an LXC node, replace both kvm values with lxc after the LXC node packages and storage path are prepared.

List the host pool while OpenNebula starts monitoring the new host.

$ onehost list
  ID NAME            CLUSTER     RVM   TCPU   FCPU   ACPU   TMEM   FMEM   AMEM STAT
   2 host01          default       0      -      -      -      -      -      - init

The initial row can show init or update until the first monitor probe returns.

List the host pool after the first monitoring cycle.
```
$ onehost list
  ID NAME            CLUSTER     RVM   TCPU   FCPU   ACPU   TMEM   FMEM   AMEM STAT
   2 host01          default       0    800    800    800  31.1G  31.1G  31.1G   on
```
STAT should become on before the host is trusted for normal VM placement. If it shows err, check /var/log/one/oned.log for SSH, name-resolution, missing remote script, or virtualization probe errors before retrying monitoring with onehost forceupdate host01.

Inspect the host record.

$ onehost show host01
HOST 2 INFORMATION
ID                    : 2
NAME                  : host01
CLUSTER               : default
STATE                 : MONITORED
IM_MAD                : kvm
VM_MAD                : kvm
LAST MONITORING TIME  : 06/25 10:14:42

HOST SHARES
RUNNING VMS           : 0
MEMORY
  TOTAL               : 31.1G
  USED (REAL)         : 0K
  USED (ALLOCATED)    : 0K
CPU
  TOTAL               : 800
  USED (REAL)         : 0
  USED (ALLOCATED)    : 0

LOCAL SYSTEM DATASTORE #0 CAPACITY
TOTAL:                : 467.7G
USED:                 : 113.5G
FREE:                 : 354.2G

The record should show the expected drivers, monitored CPU and memory, and local datastore data when the transfer mode exposes host-local capacity.