https://bugzilla.novell.com/show_bug.cgi?id=846564 https://bugzilla.novell.com/show_bug.cgi?id=846564#c1 Lars Marowsky-Bree changed: What |Removed |Added ---------------------------------------------------------------------------- Priority|P5 - None |P3 - Medium Status|NEW |NEEDINFO InfoProvider| |lzhong@suse.com --- Comment #1 from Lars Marowsky-Bree 2013-10-18 07:54:30 UTC --- ha-vm-04:~ # sbd -d /dev/vdb1 message ha-vm-01 test 2013-10-18T07:51:34.231495+00:00 ha-vm-01 sbd: [1686]: info: Received command test from ha-vm-04 on disk /dev/vdb1 Seems to be working fine for me, and I definitely also see nodes getting fenced. What's your /etc/sysconfig/sbd? What does "systemctl status sbd" say for you? Should look something like this: ha-vm-04:~ # systemctl status sbd sbd.service - Shared-storage based fencing daemon Loaded: loaded (/usr/lib/systemd/system/sbd.service; enabled) Active: active (running) since Thu 2013-10-17 19:03:03 UTC; 12h ago Process: 1660 ExecStart=/usr/share/sbd/sbd.sh start (code=exited, status=0/SUCCESS) Main PID: 1666 (sbd) CGroup: /system.slice/sbd.service ├─1666 sbd: inquisitor ├─1667 sbd: watcher: /dev/vdb1 - slot: 3 - uuid: 78397543-f754-4c37-876b-c172fb8e7668 └─1668 sbd: watcher: Pacemaker Oct 17 19:03:02 ha-vm-04 sbd[1668]: [1668]: info: Monitoring Pacemaker health Oct 17 19:03:02 ha-vm-04 sbd[1667]: [1667]: info: Device /dev/vdb1 uuid: 78397543-f754-4c37-876b-c172fb8e7668 Oct 17 19:03:03 ha-vm-04 sbd[1666]: [1666]: notice: Using watchdog device: /dev/watchdog Oct 17 19:03:03 ha-vm-04 sbd[1666]: [1666]: info: Set watchdog timeout to 5 seconds. .. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=846564 https://bugzilla.novell.com/show_bug.cgi?id=846564#c3 --- Comment #3 from Lars Marowsky-Bree 2013-10-18 08:36:21 UTC --- Correct, sbd cannot be manually started/stopped (to prevent it from being stopped while pacemaker is still up). And neither should you have a need to start either sbd or corosync manually; systemd will start/stop them as needed when you start/stop pacemaker. "systemctl start pacemaker.service" should bring them all online. And "stop" should stop pacemaker->sbd->corosync as well. Is that not happening? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=846564 https://bugzilla.novell.com/show_bug.cgi?id=846564#c5 --- Comment #5 from Lars Marowsky-Bree 2013-10-18 09:08:37 UTC --- Factory code behaves a bit different. It defaults to enabling the watchdog for safety. And it expands a few options from SBD_OPTS to individual ones. (If you install sbd on a fresh system, you should get /etc/sysconfig/sbd with a template that has these options populated. Alas, they're not getting merged in on an upgrade yet.) If you want to explicitly disable it - which I'd strongly discourage, better load an appropriate watchdog driver -, set SBD_WATCHDOG="false". -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.