On 30.09.2014 17:57, Ruben Safir wrote:
which is the point I was making all along. It doesn't use the damn scripts. It translates them as 'units'
Actually, it doesn't really "translate" the scripts into units. It just looks up the right (existing) unit, and then uses systemd to start the service using the information from that unit, instead of having a shell execute the lines of the script. In fact, both the init.d script and the unit for a service (for example cron, as mentioned by Dirk), are provided by the distribution packagers in order to provide a means to start and stop (etc.) the service (and not by "systemd" or "SysVInit"). For the distribution, this is meant to happen using the init system installed on the computer. So if the distribution is running under systemd, redirecting the rc* scripts to use systemd mechanics is only consistent and to be expected. One more word to units: I've got the impression that you, like some other people on this list, view units like some reimplementation of the service by systemd, which would make systemd a big, monolithic something. However, after looking at some of these units, I see that they are merely configuration files providing the information needed to start,stop, etc. the service. Here you can find the executable that will be run (i. e. /usr/bin/cron), command line parameters, PID file location, etc. This information is used to run a general service control flow, whose logic probably very much resembles what you also find in the init.d scripts (if you compare them, you'll find that most of them are very similar, at least the service control scripts). On top, the unit config file provides dependency information to allow parallel start of independent services. All in all it's mainly "only" a different way to start/stop the same service executables. /Andreas -- Cahn's Axiom: When all else fails, read the instructions. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org