On Tue, 2013-03-12 at 16:35 +0100, Nicolas HAHN wrote: > Hello Rainer, > > We run it only with -d. > I'll try to run it with -dn as you suggest, and I'll come back to the mailing > list :) > > Then, I'll try after that to run it under valgrind as per your second > suggestion. > > Also, I've seen that when I stop rsyslog using the famous well known command > service rsyslog stop (version 7.2.6-2 now), it seems it's not properly > closing the database connections because from time to time, I've this SQL > error in the pg_log directory: > > LOG: unexpected EOF on client connection with an open transaction > > I should probably fill a bug request, no? > not sure - this could probably be related. IF something blocks the output thread, and rsyslog is terminated, and the block remains for longer than the timeout period you have configured, rsyslogd core cancels the thread. That could lead to what you see and would be perfectly fine (of course, except for the question why the output is blocking, but that's probably a different story...).
Rainer > I come back to you at least with the output of rsyslog -dn > > Thanks a lot for your suggestions Rainer. > > Best regards, > Nicolas > - > United Nations International Computing Center > Geneva > > Rainer Gerhards <[email protected]> a écrit : > > > On Tue, 2013-03-12 at 14:57 +0100, Nicolas HAHN wrote: > >> Hello David, > >> > >> As written at the top of my e-mail, I confirm SELINUX is disabled > >> and there is no firewall rules preventing the traffic to come. > >> Furthermore, it works like a charm on the current rsyslog production > >> server we try to replace, and which is using rsyslog 4.8.x on RHEL 5. > >> > >> When syslog stops, nothing shows in the debug logs as like rsyslog > >> "loose" some IMRELP servers, it also looses its debugging ability. > >> That's what I wrote in a previous e-mail: it display tons of things > >> on the screen when ran with the -d flag, then it stops displaying > >> and give you back the hand on the shell. But we can still see > >> rsyslog in the process table running with its -d flag... > >> > > Do you run it with just -d or -dn? Without -n (As in -dn), it > > auto-backgrounds and this means you usually lose the debug information. > > > > Also, I would suggest to run it under valgrind control, as this can > > bring up any misadressing problems. That would be along the lines of > > > > $ valgrind rsyslogd -dn ...other options.... > > > > Rainer > > _______________________________________________ > > rsyslog mailing list > > http://lists.adiscon.net/mailman/listinfo/rsyslog > > http://www.rsyslog.com/professional-services/ > > What's up with rsyslog? Follow https://twitter.com/rgerhards > > NOTE WELL: This is a PUBLIC mailing list, posts are ARCHIVED by a > > myriad of sites beyond our control. PLEASE UNSUBSCRIBE and DO NOT > > POST if you DON'T LIKE THAT. > > > > > ---------------------------------------------------------------- > This message was sent using IMP, the Internet Messaging Program. > > Hello Rainer, > > We run it only with -d. > I'll try to run it with -dn as you suggest, and I'll come back to the > mailing list :) > > Then, I'll try after that to run it under valgrind as per your second > suggestion. > > Also, I've seen that when I stop rsyslog using the famous well known > command service rsyslog stop (version 7.2.6-2 now), it seems it's not > properly closing the database connections because from time to time, > I've this SQL error in the pg_log directory: > > LOG: unexpected EOF on client connection with an open transaction > > I should probably fill a bug request, no? > > I come back to you at least with the output of rsyslog -dn > > Thanks a lot for your suggestions Rainer. > > Best regards, > Nicolas > - > United Nations International Computing Center > Geneva > > > Rainer Gerhards <[email protected]> a écrit : > > > On Tue, 2013-03-12 at 14:57 +0100, Nicolas HAHN wrote: > >> Hello David, > >> > >> As written at the top of my e-mail, I confirm SELINUX is disabled > >> and there is no firewall rules preventing the traffic to come. > >> Furthermore, it works like a charm on the current rsyslog > production > >> server we try to replace, and which is using rsyslog 4.8.x on RHEL > 5. > >> > >> When syslog stops, nothing shows in the debug logs as like rsyslog > >> "loose" some IMRELP servers, it also looses its debugging ability. > >> That's what I wrote in a previous e-mail: it display tons of things > >> on the screen when ran with the -d flag, then it stops displaying > >> and give you back the hand on the shell. But we can still see > >> rsyslog in the process table running with its -d flag... > >> > > Do you run it with just -d or -dn? Without -n (As in -dn), it > > auto-backgrounds and this means you usually lose the debug > information. > > > > Also, I would suggest to run it under valgrind control, as this can > > bring up any misadressing problems. That would be along the lines of > > > > $ valgrind rsyslogd -dn ...other options.... > > > > Rainer > > _______________________________________________ > > rsyslog mailing list > > http://lists.adiscon.net/mailman/listinfo/rsyslog > > http://www.rsyslog.com/professional-services/ > > What's up with rsyslog? Follow https://twitter.com/rgerhards > > NOTE WELL: This is a PUBLIC mailing list, posts are ARCHIVED by a > > myriad of sites beyond our control. PLEASE UNSUBSCRIBE and DO NOT > > POST if you DON'T LIKE THAT. > > > > > ---------------------------------------------------------------- > This message was sent using IMP, the Internet Messaging Program. > > _______________________________________________ rsyslog mailing list http://lists.adiscon.net/mailman/listinfo/rsyslog http://www.rsyslog.com/professional-services/ What's up with rsyslog? Follow https://twitter.com/rgerhards NOTE WELL: This is a PUBLIC mailing list, posts are ARCHIVED by a myriad of sites beyond our control. PLEASE UNSUBSCRIBE and DO NOT POST if you DON'T LIKE THAT.
