33598 2017-06-27 15:50:28 +0300 не останавливается через systemd 2017-06-29 19:29:53 +0300 1 1 4 Development Sisyphus pacemaker unstable all Linux CLOSED FIXED RS P3 normal --- 1 ekorneechev cas cas shrek qa-sisyphus oldest_to_newest 164470 0 ekorneechev 2017-06-27 15:50:28 +0300 Альт Сервер СПТ 7.0 # rpm -q pacemaker pacemaker-1.1.15-alt0.M70C.1 # systemctl stop pacemaker # systemctl status pacemaker (вырубается только по таймауту, через 5 мин) pacemaker.service - LSB: Starts and stops Pacemaker Cluster Manager. Loaded: loaded (/etc/rc.d/init.d/pacemaker) Active: failed (Result: timeout) since Tue 2017-06-27 14:21:17 MSK; 886ms ago Process: 1471 ExecStop=/etc/rc.d/init.d/pacemaker stop (code=killed, signal=TERM) Process: 64666 ExecStart=/etc/rc.d/init.d/pacemaker start (code=exited, status=0/SUCCESS) Jun 27 14:07:43 cd5201-mbd03-db01.zags.loc systemd[1]: Starting LSB: Starts and stops Pacemaker Cluster Manager.... Jun 27 14:07:51 cd5201-mbd03-db01.zags.loc pacemaker[64666]: Starting Pacemaker Cluster Manager: [ DONE ] Jun 27 14:07:51 cd5201-mbd03-db01.zags.loc systemd[1]: Started LSB: Starts and stops Pacemaker Cluster Manager.. Jun 27 14:16:17 cd5201-mbd03-db01.zags.loc systemd[1]: Stopping LSB: Starts and stops Pacemaker Cluster Manager.... Jun 27 14:16:17 cd5201-mbd03-db01.zags.loc pacemaker[1471]: Waiting for shutdown of managed resources[ DONE ] Jun 27 14:16:17 cd5201-mbd03-db01.zags.loc pacemaker[1471]: Signaling Pacemaker Cluster Manager to terminate: [ DONE ] Jun 27 14:16:18 cd5201-mbd03-db01.zags.loc pacemaker[1471]: Waiting for cluster services to unload:.[ DONE ] Jun 27 14:21:17 cd5201-mbd03-db01.zags.loc systemd[1]: pacemaker.service stopping timed out. Terminating. Jun 27 14:21:17 cd5201-mbd03-db01.zags.loc systemd[1]: Stopped LSB: Starts and stops Pacemaker Cluster Manager.. Jun 27 14:21:17 cd5201-mbd03-db01.zags.loc systemd[1]: Unit pacemaker.service entered failed state Причем, если запускаем init-скрипт, то все ОК: # /etc/rc.d/init.d/pacemaker stop Waiting for shutdown of managed resources [ DONE ] Signaling Pacemaker Cluster Manager to terminate: [ DONE ] Waiting for cluster services to unload:. [ DONE ] 164471 1 ekorneechev 2017-06-27 16:46:24 +0300 Причем: # service pacemaker start # service corosync status active # service pacemaker status active # /etc/rc.d/init.d/pacemaker status pacemakerd (pid 54979) is running... # /etc/rc.d/init.d/pacemaker stop Waiting for shutdown of managed resources [ DONE ] Signaling Pacemaker Cluster Manager to terminate: [ DONE ] Waiting for cluster services to unload:. [ DONE ] # service pacemaker status active # service corosync status inactive # /etc/rc.d/init.d/pacemaker status pacemakerd is stopped # systemctl status pacemaker pacemaker.service - LSB: Starts and stops Pacemaker Cluster Manager. Loaded: loaded (/etc/rc.d/init.d/pacemaker) Active: active (exited) since Tue 2017-06-27 16:38:16 MSK; 5min ago Process: 54708 ExecStop=/etc/rc.d/init.d/pacemaker stop (code=killed, signal=TERM) Process: 54917 ExecStart=/etc/rc.d/init.d/pacemaker start (code=exited, status=0/SUCCESS) Jun 27 16:38:08 cd5201-mbd01-db02.zags.loc systemd[1]: Starting LSB: Starts and stops Pacemaker Cluster Manager.... Jun 27 16:38:16 cd5201-mbd01-db02.zags.loc pacemaker[54917]: Starting Pacemaker Cluster Manager: [ DONE ] Jun 27 16:38:16 cd5201-mbd01-db02.zags.loc systemd[1]: Started LSB: Starts and stops Pacemaker Cluster Manager.. Если на этом этапе service stop - только по таймауту завершиться все равно... 164497 2 shrek 2017-06-29 10:39:11 +0300 в p7/c7 на серверах нельзя использовать systemd (не зря при установке сервера выбран sysvinit), т.к. демоны с service файлами там скорее исключение. Здесь sysdemd запускает все через init скрипты от sysvinit, в результате получаем то что получаем. Самый простой способ решения - удалить systemd и установить sysvinit 164517 3 repository-robot 2017-06-29 19:29:53 +0300 pacemaker-1.1.16-alt3 -> sisyphus: Thu Jun 29 2017 Denis Medvedev <nbr@altlinux.org> 1.1.16-alt3 - Fix initscript,(ALT #33598).