Perhaps I'm misunderstanding something in this DDNS discussion . . . it
all sounds so difficult.  Maybe I've got it easy?  Perhaps I am not
doing things professionally?

I have a hobby server, which was OS/2-based from 1997 until last year.
My (several) domain names are defined by dyn.com (formerly dyndns, now
Oracle), for about twenty years.  I pay $15/year (I think).  That gets
me basic DNS<->IP lookup.  I have residential broadband service, whose
IP changes on a whim: it may remain the same for six months, or change
three times in a day.

When I was an OS/2 shop, I ran a java client that monitored changes to
the WAN IP, and it used the dyndns API to update my IP in their DNS,
which propagated out to other DNSs.  This worked well for around fifteen
years.

Then, I changed to a router that offered to do this service on its own,
so I disabled the OS/2 java client and used the router to update my IP
with dyndns' DNS.  Since a year or so ago, that got flaky (probably a
change in dyndns' API, and my old hardware router had no firmware update
available), so I now use a Linux client (supplied as a Debian package, I
believe) on a local Ubuntu server to manage the IP updates to dyndns.

Is what you're discussing some variation of what I'm doing with $15/year
dyn.com and a free Debian IP updater client?  Or something else?

-- Regards,
Al Savage