Hello again,
I've worked around ruby's daemonize implementation by rewriting the
app to be of type simple. However, now I'm seeing the long delay on
stop after the application it executes has run once. The application is
a real time app utilizing RTAI. The real time application is started by
a scripted web daemon (the one I recently modified because of problems
between it and systemd.) It forks and then calls exec to start the real
time application, but until told to do so.
The scripted web daemon is started by systemd. Systemd keeps track of
every forked child process, right? Here is a status report showing what
its status is shortly after normal startup:
--------------------------------------------------------------------------------------
[rtuser@localhost ~]$ sudo systemctl status webapp.service
webapp.service - scripted daemon
Loaded: loaded (/etc/systemd/system/webapp.service)
Active: active (running) since Tue, 22 Jan 2013 01:42:30
-0700; 7min ago
Process: 553
ExecStartPre=/home/rtuser/rtapp/scripts/preStart.sh (code=exited,
status=0/SUCCESS)
Main PID: 906 (ruby)
CGroup: name=systemd:/system/webapp.service
â 906 /usr/bin/ruby /home/rtuser/rtapp/bin/webapp.rb
--------------------------------------------------------------------------------------
At this point, everything is fine. It will stop almost
instantaneously. And here it is after the real time app has been
started and then subsequently successfully stopped after all calls to
free up and shutdown real time resources have completed and all
processes (except the webapp) have exited.
--------------------------------------------------------------------------------------
[rtuser@localhost ~]$ sudo systemctl status webapp.service
webapp.service - scripted daemon
Loaded: loaded (/etc/systemd/system/webapp.service)
Active: active (running) since Tue, 22 Jan 2013 01:42:30
-0700; 17min ago
Process: 553
ExecStartPre=/home/rtuser/rtapp/scripts/preStart.sh (code=exited,
status=0/SUCCESS)
Main PID: 906 (ruby)
CGroup: name=systemd:/system/webapp.service
â 906 /usr/bin/ruby /home/rtuser/rtapp/bin/webapp.rb
â 1040 [RTAI_KTHRD_M:0]
â 1041 [F:HARD:0:1]
â 1042 [RTAI_KTHRD_M:1]
â 1043 [F:HARD:1:1]
â 1044 [F:HARD:0:2]
â 1045 [F:HARD:1:2]
â 1046 [F:HARD:0:3]
â 1047 [F:HARD:1:3]
â 1048 [F:HARD:0:4]
â 1049 [F:HARD:1:4]
â 1050 [F:HARD:0:5]
â 1051 [F:HARD:1:5]
â 1052 [F:HARD:0:6]
â 1053 [F:HARD:1:6]
--------------------------------------------------------------------------------------
Subsequent starts and stops of the application of course do not change
those process ID and kernel threads because, I can only assume, they are
reused. A kill -9 as root will not even kill them. The problem I have
is that system shutdown times take quite a few minutes unless systemd
has done everything it can to kill these, correct? This mostly results
in a very long delay on a timeout waiting for dependent processes to die
when they never will on shutdown or restart of the computer or just the
webapp service.
I've already posted to the rtai mailing list about this. I've tried
using RemainAfterExit=yes (and no) in the webapp.service file as
previously suggested with no avail. The service file currently looks like:
[Unit]
Description=scripted daemon
After=network.target
[Service]
Type=simple
WorkingDirectory=/home/rtuser/rtapp/bin
PIDFile=/home/rtuser/rtapp/data/logs/webapp.pid
ExecStartPre=/home/rtuser/rtapp/scripts/preStart.sh
ExecStart=/home/rtuser/rtapp/bin/webapp.rb
StandardInput=null
StandardOutput=syslog
StandardError=syslog
User=sim
Group=sim
ControlGroup=cpu:/
[Install]
WantedBy=multi-user.target
I think those threads not getting shutdown then RTAI may be designed as
such and if so, saying RTAI needs to change it's design wouldn't be very
helpful. I'm trying to get this figured out so system shutdowns and
service restarts will not take forever compared to startup. Any help
would be immensely appreciated!
Thanks,
JB
_______________________________________________
systemd-devel mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/systemd-devel
