I plan to upgrade a server from Debian stretch to buster. Having read
the release notes I wonder what's the best way to avoid the new scheme
of unpredictable network interface names.
I don't care what PCI bus and what slot my NICs are attached to, I
don't want to learn and don't want to have to remember this hardware
configurationan and I don't want to type these cumbersome and error
prone names. I simply have eth0 for the internal network and eth1 for
my external network to the DSL router. That's easy and I want to keep
it that way.
The Debian wiki on this shows several ways involving kernel cmdline,
udev, and systemd. I've read it, I've also read some of the sparse
and incomplete systemd documentation for almost an hour. Still I
don't know when and what software component (kernel, udev, systemd)
decides the NIC names and whether and in which way these conflict each
other. [1]
Also, after reading the wiki it's still unclear to me, which of the
several ways will survive the next upgrade to bullseye.
The safest way seems to be what's called "Custom schemes" but this
section explicitly states the names eth0, etc. shouldn't be used.
So I'm still confused what to do after the upgrade to buster to keep
my network names.
Steve
[1] Unfortunately, much of the Linux community and many distributions
try to get as far away from its Unix roots as possible, away from
the good KISS principle, and instead repeat all the mistakes
Windows has made in making everything complex, automagic, obscure,
instable, and unpredictable.
