Wednesday, 2016-01-20

*** markvoelker has quit IRC00:14
*** markvoelker has joined #openstack-ha00:17
*** hoangcx has joined #openstack-ha00:40
*** masahito has joined #openstack-ha01:54
*** markvoelker has quit IRC02:42
*** beekhof has quit IRC03:17
openstackgerritOpenStack Proposal Bot proposed openstack/ha-guide: Updated from openstack-manuals  https://review.openstack.org/26998303:23
openstackgerritMerged openstack/ha-guide: Updated from openstack-manuals  https://review.openstack.org/26998303:55
*** markvoelker has joined #openstack-ha04:14
*** hoangcx has quit IRC06:06
*** hoangcx has joined #openstack-ha06:07
*** Lane_ has joined #openstack-ha06:19
*** wuhao has joined #openstack-ha06:37
*** mobaer has joined #openstack-ha06:46
*** wuhao has quit IRC06:55
*** wuhao has joined #openstack-ha06:55
*** wuhao has quit IRC07:01
*** wuhao has joined #openstack-ha07:02
*** wuhao has quit IRC07:03
*** wuhao has joined #openstack-ha07:03
*** mjura_ has joined #openstack-ha07:08
*** mjura has joined #openstack-ha07:08
*** markvoelker has quit IRC07:11
*** markvoelker has joined #openstack-ha07:12
*** markvoelker_ has joined #openstack-ha07:18
*** mobaer has quit IRC07:19
*** markvoelker has quit IRC07:20
*** mobaer has joined #openstack-ha07:29
*** markvoel_ has joined #openstack-ha07:42
*** markvoelker_ has quit IRC07:43
*** rmart04 has joined #openstack-ha08:10
*** dgurtner has joined #openstack-ha08:20
*** dgurtner has joined #openstack-ha08:20
*** jpena has joined #openstack-ha08:26
*** beekhof has joined #openstack-ha08:33
*** FL1SK has quit IRC09:56
*** mobaer has quit IRC10:14
*** hoangcx has quit IRC10:16
*** mobaer has joined #openstack-ha10:23
bogdandoaspiers, hi10:36
bogdandoaspiers, do you think this link may fit the action item? http://serverfault.com/questions/115838/are-pid-files-reliable-for-determining-whether-a-process-is-running10:37
bogdando(and the same PIDs reusage depends on the system load and intensity of processes spawned around)10:38
aspiershi10:39
aspiersbogdando: remind me which action item?10:42
bogdandobogdando to give some concrete examples where pid files (and perhaps hence systemd) cannot be relied on10:42
bogdandofrom the Monday IRC meeting10:42
aspiersoh10:43
aspiersI don't think that's a strong argument10:43
aspiersare you talking about this scenario:10:43
bogdandoA: a crash-stop failure of a process, leaving the PIDfile AND the process running while stopped function in fact10:45
bogdandoB: a lazy pidfile creation, example: MySQL creates only after replica finished10:45
bogdandoC: a kinda Byzantine failure when a process creates wrong pidfile10:46
aspiersA: crash-stop does not leave the process running?10:46
aspiersB and C are just bugs10:46
bogdandonot always. It just stops the "code flow"10:46
bogdandofor example it may sleep for ever10:46
aspiersa hung process gets caught by the OCF RA monitor action10:47
bogdandoB is not a bug, AFAICT10:47
aspiersif MySQL is running, there should be a pid file, period10:47
bogdandowell, may be it is a bug, but example exists :)10:47
aspierseven stale pid files are not a problem10:48
aspiersas long as you check that the name of the process referred to by the pid file is the one you were expecting10:48
bogdandoonly if you strengthen by the additional procfs checks perhaps10:48
aspiersif it doesn't match, the pid in the stale file got reused by another process => delete stale pid file10:49
bogdandodoes systemd check?10:49
aspiersif it does match, the service is running (maybe again, coincidentally with the same pid) => nothing to do10:49
aspiersI don't know10:49
aspiersbut I guess systemd probably has extra checks relating to cgroups etc.10:49
aspiersagain, this gets covered by a combination of systemctl status and the RA monitor action10:50
bogdandowe could keep in mind the test cases while design the OCF switching to the systemd checks10:50
bogdandoA/B/C is easy to simulate with a simple bash script10:50
aspierssure - I don't think there is any big engineering challenge here10:51
aspiersonly thing worth worrying about is race conditions when checking the pid file10:51
bogdandoyup10:51
bogdandopacemaker's start vs init's respawn?10:51
bogdandowell, I mean RA OCF's start10:53
*** smoriya_ has quit IRC11:00
*** mobaer has quit IRC11:08
*** mobaer has joined #openstack-ha11:23
*** openstackgerrit has quit IRC11:43
*** openstackgerrit has joined #openstack-ha11:44
*** mjura has quit IRC12:15
*** mobaer has quit IRC13:02
*** ddeja has quit IRC13:06
*** ddeja has joined #openstack-ha13:09
*** dguitarbite has quit IRC13:23
*** dguitarbite has joined #openstack-ha13:28
*** dslev has joined #openstack-ha13:34
*** mobaer has joined #openstack-ha13:35
*** FL1SK has joined #openstack-ha13:40
*** dslev has quit IRC13:43
*** rmart04_ has joined #openstack-ha13:48
*** ddeja_ has joined #openstack-ha13:49
*** FL1SK has quit IRC13:50
*** jpena is now known as jpena|lunch13:50
*** bandini_ has joined #openstack-ha13:51
*** raginbaj- has joined #openstack-ha13:55
*** smoriya_afk has joined #openstack-ha13:55
*** xenolog13 has joined #openstack-ha13:55
*** raginbajin has quit IRC13:56
*** _gryf has quit IRC13:56
*** ddeja has quit IRC13:56
*** rmart04 has quit IRC13:56
*** bandini has quit IRC13:56
*** dulek has quit IRC13:56
*** smoriya has quit IRC13:56
*** bogdando has quit IRC13:56
*** mancdaz has quit IRC13:56
*** xenolog- has quit IRC13:56
*** smoriya_afk is now known as smoriya13:56
*** rmart04_ is now known as rmart0413:56
*** dulek has joined #openstack-ha13:56
*** raginbaj- is now known as raginbajin13:56
*** cknight has joined #openstack-ha14:01
*** bogdando has joined #openstack-ha14:02
*** _gryf has joined #openstack-ha14:03
*** ddeja_ is now known as ddeja14:10
*** mobaer has quit IRC14:51
*** cknight has quit IRC14:53
*** cknight has joined #openstack-ha14:54
*** mobaer has joined #openstack-ha14:55
*** jpena|lunch is now known as jpena14:58
*** kgaillot has joined #openstack-ha15:02
*** mancdaz has joined #openstack-ha15:08
*** kgaillot has quit IRC15:13
*** sigmavirus24_awa is now known as sigmavirus2415:13
openstackgerritOpenStack Proposal Bot proposed openstack/ha-guide: Updated from openstack-manuals  https://review.openstack.org/27024515:19
*** markvoel_ has quit IRC15:25
*** fabbione has quit IRC15:36
*** markvoelker_ has joined #openstack-ha15:46
*** yan-gao has quit IRC15:47
*** FL1SK has joined #openstack-ha15:50
*** yan-gao has joined #openstack-ha15:53
bogdandoJFYI, new time for HA guide docs team meeting http://lists.openstack.org/pipermail/openstack-docs/2016-January/008209.html15:58
*** rmart04 has quit IRC16:44
*** mjura_ has quit IRC16:46
*** kgaillot has joined #openstack-ha17:05
openstackgerritMerged openstack/ha-guide: Updated from openstack-manuals  https://review.openstack.org/27024517:06
*** bandini_ is now known as bandini17:12
*** rmart04 has joined #openstack-ha17:21
*** rmart04 has quit IRC17:27
*** rmart04 has joined #openstack-ha17:32
*** zzhou has quit IRC17:39
*** zzhou has joined #openstack-ha17:53
*** mobaer has quit IRC17:55
*** dgurtner has quit IRC18:05
*** rmart04 has quit IRC18:07
*** jpena has quit IRC18:07
*** _gryf has quit IRC18:22
*** bogdando has quit IRC18:22
*** bogdando has joined #openstack-ha18:28
*** _gryf has joined #openstack-ha18:28
*** cknight has left #openstack-ha18:37
*** mobaer has joined #openstack-ha18:40
*** bandini has quit IRC18:56
*** bandini has joined #openstack-ha18:56
*** yan-gao has quit IRC18:58
*** yan-gao has joined #openstack-ha18:58
*** mobaer has quit IRC20:18
*** mobaer has joined #openstack-ha21:07
*** beekhof has quit IRC21:19
*** sigmavirus24 is now known as sigmavirus24_awa22:54
*** kgaillot has quit IRC23:44
*** sigmavirus24_awa is now known as sigmavirus2423:46
*** smoriya_ has joined #openstack-ha23:48
*** sigmavirus24 is now known as sigmavirus24_awa23:59

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!