Skip to main content

How to connect a VM to a VPC

This guide explains how to connect a VM to one or more VPC subnets, at creation or on an existing VM, then how to check and configure the interface in the OS. It also covers detaching.

Prerequisites​

  • A Hikube account and a project
  • A VPC with at least one subnet (see Quick start), or the intention to create one from the VM wizard
  • SSH access to the VM for verification

Steps​

1. Connect a new VM​

  1. Open VM Instances > Create an Instance and fill in the first steps.
  2. At the Network step, VPC Networks (Secondary) section, check the VPC you want.
  3. Under Subnets, check one or more subnets.
  4. Finish the wizard: the Summary shows Private Networks with the number of subnets. Click Create instance.

No VPC yet? Click + VPC in the VPC Networks (Secondary) section: the Create VPC dialog lets you create the VPC and its subnets, then checks it automatically. To add a subnet to a checked VPC, click Add subnet, enter a name and a CIDR range, then confirm.

Debian and Ubuntu

If the VM image is Debian or Ubuntu and a VPC is checked, the section shows the Manual setup required on Debian and Ubuntu warning: the VPC address is reserved, but it is not applied automatically in the system. Configure the interface once the VM has started (see Check in the OS). The same warning appears in Edit.

2. Connect an existing VM​

  1. Open the VM's detail page and click Edit.
  2. In Network & Security, under VPC Networks (Secondary), check the VPC, then the subnets.
  3. Click Save.

3. Check in the console​

On the VM's detail page, Network & Security section:

  • VPC Networks lists each VPC with its Connected subnets and their range, for example app (172.16.0.0/24);
  • the Internal zone adds one address per subnet, next to the mgmt IP.

4. Check in the OS​

ip -br addr

Expected result: one additional interface per connected subnet. On Debian and Ubuntu, the OS does not configure it automatically: it first appears without an address, for example:

lo UNKNOWN 127.0.0.1/8 ::1/128
enp1s0 UP 10.x.x.x/xx ...
enp2s0 DOWN

The VPC address is assigned by DHCP. Distributions using NetworkManager (AlmaLinux, Rocky…) pick it up automatically; on Debian and Ubuntu, enable DHCP on the secondary interface. The lease does not provide a gateway: the default route stays on the primary interface.

Example with netplan (Debian, Ubuntu):

/etc/netplan/60-vpc.yaml
network:
version: 2
ethernets:
enp2s0:
dhcp4: true
dhcp4-overrides:
use-routes: false
sudo chmod 600 /etc/netplan/60-vpc.yaml
sudo netplan apply
ip -br addr show enp2s0

Expected result: enp2s0 is UP with the subnet address shown in the Internal zone of the console, for example 172.16.0.11/24.

use-routes: false prevents the VPC interface from replacing the default route: Internet access keeps going through the primary interface.

5. Detach a VM​

  1. VM detail page > Edit.
  2. Under VPC Networks (Secondary), uncheck the subnet, or the whole VPC.
  3. Click Save.

Then remove the corresponding configuration in the OS (for example the netplan file you added).

Verification​

From another VM on the same subnet:

ping -c 3 <secondary-address-of-the-vm>

Further reading​