We are evolving an existing embedded product line to incorporate the use of systemd as a part replacement for a proprietary start-up system. While it is possible to handle service failures through use of the OnFailure directive, we have an additional requirement for which we can find no clean (direct) solution in systemd: applications in the system should not stop for any reason - any termination must be handled as a failure.
We have worked through the relevant manpages and have eliminated ExecStartPost (semantic is not appropriate) and Restart (we don't want to restart the service, we want a failure notification). SuccessExitStatus cannot be used to turn success codes/signals into failures. We can use a script in ExecStart to turn normal terminations into failures or we could consider some special small wrapper executable. These are the kind of hacks that systemd is supposed to render unneccessary, so we come to this list for help. Is there something we have missed? If not, is there any objection in principle to an extension to systemd to cover this requirement, perhaps a NonStop directive? Regards Richard Spence Under contract to Nokia in Ulm, Germany.
_______________________________________________ systemd-devel mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/systemd-devel
