With the release of Debian 13 “Trixie”, many users and admins will upgrade existing systems. In principle, this runs – as usual – without problems. However, it was noticeable that the updated systems (bare-metal or VM) were sometimes no longer accessible after a reboot. For example, with VMs, the guest agent only reported 127.0.0.1 or IPv6 and/or existing Docker networks. With bare-metal systems without IPMI or similar, access is then naturally only possible locally.
Simple remedy (cause still unclear)
What ultimately leads to this is still unclear. Newly set up Debian 13 systems do not show this phenomenon, with identical config of /etc/network/interfaces.
An additional entry in /etc/network/interfaces. provides a remedy.
Example with dynamic IP configuration (before):
# The primary network interface
allow-hotplug eno1
iface eno1 inet dhcp
Adding “auto”:
# The primary network interface
allow-hotplug eno1
auto eno1
iface eno1 inet dhcp
Setting the entry before the upgrade
To avoid surprises, you should set the entry for automatically starting the interface at boot time before the upgrade.
