you can use and like and still complain about things that should be better or were better under old.
systemd solved a ton of headaches but also added few more, like inability to express "just shut the fucking system down, you won't have power in 5 minutes" for servers connected to UPS.
> And writing a systemd unit is pretty easy, contrary to the old bash script mess.
We had thousands of lines of "simple" sysv init scripts fixes because apparently even seasoned maintainers or developers of the app can't figure it out. It's huge improvement.
One example: A java app that writes its own pid. The status subcommand relies on the pid existing.
so calling start then status will return that the service haven't started yet. And that is what stuff like for example Pacemaker does so it could just randomly fail under sysv. Under systemd it's all so much simpler
With two “--force” options, that is essentially equivalent to “echo o > /proc/sysrq-trigger”, isn’t it? I would think that one “--force” would have the actually desired effect.
systemd solved a ton of headaches but also added few more, like inability to express "just shut the fucking system down, you won't have power in 5 minutes" for servers connected to UPS.
> And writing a systemd unit is pretty easy, contrary to the old bash script mess.
We had thousands of lines of "simple" sysv init scripts fixes because apparently even seasoned maintainers or developers of the app can't figure it out. It's huge improvement.
One example: A java app that writes its own pid. The status subcommand relies on the pid existing.
so calling start then status will return that the service haven't started yet. And that is what stuff like for example Pacemaker does so it could just randomly fail under sysv. Under systemd it's all so much simpler