Dear Armin,
thanks for reporting this. I'm member of the upstream maintainer team
and not a Debian person. Please be aware that the current maintainer
team is quit new to BIT and we are not yet familiar with all parts of
the code and the behavior of BIT. But I'll try to assist you. I need to
investigate the issue.
Let me try to rephrase. You have a snapshot profile schedule "Every
day". This results in a crontab line like "00 12 * * *"
In Debian 11 (Bullseye, oldstable) the job was started at the first boot
on each day and at a specified time (e.g. 12 o'clock). Correct?
In Debian 12 (Bookworm, stable) the job do not start at boot but only at
the specified time (e.g. 12 o'clock).
This makes me wonder. I would assume that the Debian 12 behavior is
"correct". I wonder why Debian 11 to run that job at boot when there is
just a crontab line. On GNU Linux it is often that cron and anacron are
somehow connected in several ways. Anacron is often involved while
booting. Maybe something here changed between 11 and 12. I'll
investigate that.
On Debian 11 you run BIT 1.2.1-3 from the official Debian repository?
Do you still have a working Debian 11 machine? Or did you updated all
your machines?
If you have Debian 11 and 12 available please provide the following
information from both machines. If not Debian 12 only is also OK.
Run "backintime --diagnostics" (only with BIT 1.3.4) in a terminal and
give the output.
Show roots crontab with "sudo crontab -l" (as user) or "crontab -l" (as
root).
Show the users crontab entries related to BIT with "crontab -l | grep
backintime".
How did you verify (via "journalctl" ?) if BIT was run or not? Maybe it
run but decided nothing changed and there is no need for a new snapshot.
Maybe you can provide the relevant syslog entries via
journalctl --since "two weeks" | grep -i -e backintime -e boot
Kind
Christian
