Update network configuration for the Qubes staging environment

Expected behavior

The network configuration described in https://github.com/freedomofpress/securedrop-dev-docs/blob/main/docs/qubes_staging.rst should work out of the box. However, when setting up the staging environment in a 4.1 Qubes (with also the Workstation installed), I had problems connecting the focal VM to the internet.

Actual behavior

When using the IP Addresses provided by Qubes as static addresses and following the guide, the VM does not properly connect to the internet.

Steps to reproduce the problem

1. Follow the guide up until "Initial VM configuration"
2. Log in and test internet connectivity

Additional information

I think that Qubes updated their networking settings. All my domains have their static IP Addresses in the 10.137.0.0/24 range. However the gateway is in the 10.138.0.0/24 subnet. When configuring the network like this, without a static route for the gateway first, the gateway cannot be set properly.

As an example, my personal domain currently has this output from ip route:

default via 10.138.2.138 dev eth0 onlink 
10.138.2.138 dev eth0 scope host onlink 

Thus, to address this issue on Ubuntu, we need to add a static route in netplan after installation.

In /etc/netplan/00-installer-config.yml add after the addresses section:

routes:
- to: 10.138.0.138/32
  scope: link

Where 10.138.0.138 is the gateway IP.

Before opening an actual pull request, I'd like to double-check if it is reproducible for anyone else or if it is due to some Qubes or other misconfiguration on my side.

Start with docs/qubes_staging.rst and follow the guide through “Initial VM configuration” on Qubes 4.1 with the Workstation installed. Check the focal VM’s connectivity and the route output, then review /etc/netplan/00-installer-config.yml and the proposed static route. Done means confirming the cause and updating the staging guide with accurate, reproducible network configuration instructions.