On Tue, 03.02.15 17:26, Martin Pitt (martin.p...@ubuntu.com) wrote: > Hey all, > > I'm currently reviewing our Debian patches for systemd, and came > across this one which sounds important for other distributions, too. > This was reported and fixed two years ago in > https://bugs.debian.org/635777 which has all the details and logs, but > the summary is: > > Distributions have quite a lot of "run scripts in this .d/ directory > if stuff happens"; e. g. the ISC DHCP client has > /etc/dhcp/dhclient-{enter,exit}-hooks.d/, Debian's ifupdown has > /etc/network/if-{up,down}.d/, and of course init.d scripts themselves > also occasionally call "service foo reload" and similar. It can happen > that when requesting a shutdown, a script of the above kind reloads or > restarts another service. In this case, postfix wants to reload itself > after a network interface goes up or down; during runtime that works > fine, but if it happens during shutdown, that "systemctl reload > postfix" will cause the entire shutdown to stall for 90s (the default > timeout).
Hmm, why precisely does this stall for 90s? Isn't this a case where people should just use "--no-block"? Lennart -- Lennart Poettering, Red Hat _______________________________________________ systemd-devel mailing list systemd-devel@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/systemd-devel