2.1 Describe device virtualization technologies

Device virtualization lets one physical server run multiple logical devices. For ENCOR, the important pieces are the hypervisor, the virtual machine, and the virtual switch.

A hypervisor slices hardware into usable chunks. A virtual machine consumes those chunks as CPU, memory, disk, and NICs. A virtual switch connects the virtual NICs to other VMs and to the physical network.

That is the whole idea. Do not make it harder than it is.

The traffic path

Imagine a virtual router named R1 running on a server.

R1 virtual NIC
virtual switch / port group
physical host NIC
physical switchport
enterprise network

When a packet leaves R1, it does not magically appear on the wire. It passes through a virtual switch first. That virtual switch may apply a VLAN tag, a security policy, a port group setting, or a distributed switching policy before the packet leaves the host.

If a lab breaks, check the whole path.

1. VM NIC connected?
2. Correct port group?
3. Correct VLAN or trunk?
4. Host uplink connected?
5. Physical switchport allows the VLAN?
6. Gateway reachable?

What to know cold

Why network engineers care

Virtualization moves part of the network inside the server. That means a network problem may not be visible only from the physical switch.

A VLAN can be correct on the physical switch but wrong on the port group. A VM can have the right IP address but be connected to the wrong virtual network. A physical trunk can allow VLAN 20, but the virtual switch may present VLAN 10 to the VM.

You do not need to administer every hypervisor. You do need to understand where the packet could be stopped, tagged, bridged, or forwarded.

Mini-lab: map a VM packet

Goal

Explain the path from a VM to the physical network.

Topology

VM-A eth0 -> vSwitch / port group -> Host NIC -> SW1 Gi1/0/10 -> R1 gateway · VLAN 20

Physical switch example

interface GigabitEthernet1/0/10
 description Hypervisor-Host-01-Uplink
 switchport mode trunk
 switchport trunk allowed vlan 20,30
 spanning-tree portfast trunk

VM settings to verify

1. VM-A connected: yes
2. VM-A vNIC: eth0
3. Port group: USERS-VLAN20
4. Port group VLAN: 20
5. Host uplink: vmnic0
6. Physical switchport: Gi1/0/10
7. Allowed VLANs: 20,30

Test plan

From VM-A:

ping <default-gateway-on-vlan-20>
arp -n
traceroute <remote-ip>

From SW1:

show interfaces trunk
show mac address-table interface gi1/0/10
show spanning-tree interface gi1/0/10 detail

Pass condition

You can explain whether a failure is likely inside the VM, inside the virtual switch, on the host uplink, or in the physical network.

Troubleshooting approach

Use layers, not guesses.

1. Is the VM powered on and the NIC connected?
2. Is the VM on the correct virtual switch or port group?
3. Is the VLAN tag handled in the right place?
4. Is the physical uplink up?
5. Is the physical switchport access or trunk as intended?
6. Is the gateway reachable?
7. Does the MAC address appear where you expect?

Exam traps

• A virtual switch is still a switch. It forwards frames between virtual NICs and uplinks.
• Type 1 vs Type 2 is about where the hypervisor runs, not whether the VM is “better.”
• A VM is not a container. A VM runs a guest OS with virtual hardware.
• Port group VLAN settings and physical switch trunk settings must agree.
• Network engineers may need the hypervisor team to check the virtual side of the path.

Quick check

1. Where does a Type 1 hypervisor run?
2. What connects a VM to a virtual switch?
3. Why can a VM fail to reach the gateway even when the physical trunk allows the VLAN?
4. What command on a physical switch helps confirm that a VM MAC address is arriving?
5. What is the difference between a virtual switch uplink and a VM vNIC?