From this week's Distrowatch. Could be a problem in some circumstances.
The systemd project tends to stir up debate with each new feature it implements. One of the most recent changes to systemd (available in version 230), forces user processes to terminate when the user logs out. On some systems, such behaviour makes sense and effectively cleans up misbehaving processes when the user leaves. However, many administrators and developers rely on processes continuing to run to perform tests, backups or other tasks after they log off. Guus Sliepen summed up the concerns of the latter group quite nicely in a Debian bug report: It is now indeed the case that any background processes that were still running are killed automatically when the user logs out of a session, whether it was a desktop session, a VT session, or when you SSHed into a machine. Now you can no longer expect a long running background processes to continue after logging out. I believe this breaks the expectations of many users. For example, you can no longer start a screen or tmux session, log out, and expect to come back to it. For this reason, I think it is a bad decision on the part of the systemd maintainers to enable this feature by default, and it should rather be disabled by default in Debian." A similar discussion is taking place on a Fedora mailing list. The new default systemd behaviour can be disabled (either by distributions or system administrators) by editing the logind configuration file.
On Mon, May 30, 2016 at 10:22:13AM +0000, Ian Young wrote:
It is now indeed the case that any background processes that were still running are killed automatically when the user logs out of a session, whether it was a desktop session, a VT session, or when you SSHed into a machine. Now you can no longer expect a long running background processes to continue after logging out. I believe this breaks the expectations of many users.
Yes, it breaks the well established practice that the controlling terminal sends all its child processes a HUP signal on logout, and that the default signal handler will kill such processes, and if the process wants to remain running it therefore establishes a signal handler to properly receive and handle the HUP signal so that it is not terminated. If processes are not being terminated on logout when they should be, then it is the non-terminating processes fault as the process has taken action to set up a non-default signal hangler to receive the HUP signal to achieve this effect. It is most certainly not the responsibility of systemd to kill processes on logout! With these critical and non-sensical changes to system behaviour that is well without the established remit of the init process the systemd developers are fast losing what little respect I might have had for them. Cheers Michael.
On Mon, 30 May 2016 10:22:13 +0000, Ian Young wrote:
From this week's Distrowatch. Could be a problem in some circumstances.
One of the most recent changes to systemd (available in version 230), forces user processes to terminate when the user logs out. On some systems, such behaviour makes sense and effectively cleans up misbehaving processes when the user leaves. However, many administrators and developers rely on processes continuing to run to perform tests, backups or other tasks after they log off.
I’m still not sure exactly what the problem is. I normally use setsid(1) <http://linux.die.net/man/1/setsid> if I want to run such background processes, and I think this will still work. I asked on this bug report <https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=825394> whether this is the case, but it appears my question has got lost in the usual anti-systemd noise...
From this week's Distrowatch. Could be a problem in some circumstances.
One of the most recent changes to systemd (available in version 230), forces user processes to terminate when the user logs out. On some systems, such behaviour makes sense and effectively cleans up misbehaving processes when the user leaves. However, many administrators and developers rely on processes continuing to run to perform tests, backups or other tasks after they log off.
I’m still not sure exactly what the problem is. I normally use setsid(1) <http://linux.die.net/man/1/setsid> if I want to run such background processes, and I think this will still work. I asked on this bug report <https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=825394> whether this is the case, but it appears my question has got lost in the usual anti-systemd noise...
Since systemd is generating quite some interest, I've turned the meetup suggestion for systemd into the topic for our meeting in June: http://www.meetup.com/WaikatoLinuxUsersGroup/events/231330392/ Cheers, Peter -- Peter Reutemann Dept. of Computer Science University of Waikato, NZ +64 (7) 858-5174 http://www.cms.waikato.ac.nz/~fracpete/ http://www.data-mining.co.nz/
participants (4)
-
Ian Young -
Lawrence D'Oliveiro -
Michael Cree -
Peter Reutemann