| *** akwasnie has quit IRC | 00:01 | |
| *** tzn has joined #kolla | 00:01 | |
| sdake | folks | 00:01 |
|---|---|---|
| sdake | please don't approe any patches until i tag tomororw morning unless they are related to our release announcement features | 00:01 |
| ntpttr | thanks again for the intro all, excited to continue with this, headed out for the day | 00:01 |
| sdake | ntpttr enjoy :) | 00:02 |
| mandre | sdake, patches on stable/liberty still ok? | 00:02 |
| sdake | mandre yes those are fine | 00:02 |
| *** rhallisey has quit IRC | 00:03 | |
| *** tzn has quit IRC | 00:03 | |
| *** achanda has joined #kolla | 00:03 | |
| *** Jeffrey4l has quit IRC | 00:08 | |
| *** dims has joined #kolla | 00:08 | |
| *** dims_ has quit IRC | 00:09 | |
| *** dims has quit IRC | 00:10 | |
| *** AndChat|241401 has quit IRC | 00:11 | |
| *** rajathagasthya has quit IRC | 00:13 | |
| *** alisonh has left #kolla | 00:14 | |
| *** achanda has quit IRC | 00:15 | |
| *** ccesario has joined #kolla | 00:15 | |
| *** nkorabli has quit IRC | 00:18 | |
| *** salv-orl_ has quit IRC | 00:40 | |
| *** inc0 has quit IRC | 00:42 | |
| *** jasonsb has joined #kolla | 00:46 | |
| *** dims has joined #kolla | 00:52 | |
| *** jtriley has joined #kolla | 00:54 | |
| *** diogogmt has joined #kolla | 00:56 | |
| *** diogogmt has quit IRC | 00:56 | |
| *** iceyao has joined #kolla | 00:58 | |
| *** iceyao_ has joined #kolla | 01:02 | |
| *** iceyao has quit IRC | 01:04 | |
| *** iceyao has joined #kolla | 01:08 | |
| *** iceyao_ has quit IRC | 01:12 | |
| *** ssurana has quit IRC | 01:15 | |
| *** sdake has quit IRC | 01:22 | |
| *** achanda has joined #kolla | 01:26 | |
| *** prithiv has quit IRC | 01:30 | |
| *** rstarmer has joined #kolla | 01:35 | |
| *** rstarmer has quit IRC | 01:46 | |
| *** rstarmer has joined #kolla | 01:47 | |
| *** banix has joined #kolla | 01:57 | |
| *** iceyao_ has joined #kolla | 01:57 | |
| *** iceyao__ has joined #kolla | 02:00 | |
| *** iceyao has quit IRC | 02:00 | |
| *** iceyao_ has quit IRC | 02:03 | |
| *** salv-orlando has joined #kolla | 02:03 | |
| *** tobe has joined #kolla | 02:05 | |
| *** unicell has quit IRC | 02:05 | |
| *** allen_gao has quit IRC | 02:08 | |
| *** salv-orlando has quit IRC | 02:12 | |
| *** zaneb has quit IRC | 02:14 | |
| *** dave-mccowan has quit IRC | 02:19 | |
| *** kbyrne has quit IRC | 02:19 | |
| *** Slower has quit IRC | 02:19 | |
| *** cfarquhar has quit IRC | 02:19 | |
| *** dims has quit IRC | 02:20 | |
| *** jtriley has quit IRC | 02:22 | |
| *** dave-mccowan has joined #kolla | 02:25 | |
| *** kbyrne has joined #kolla | 02:25 | |
| *** Slower has joined #kolla | 02:25 | |
| *** cfarquhar has joined #kolla | 02:25 | |
| *** kbyrne has quit IRC | 02:25 | |
| *** jtriley has joined #kolla | 02:26 | |
| *** iceyao has joined #kolla | 02:27 | |
| *** iceyao__ has quit IRC | 02:27 | |
| *** kbyrne has joined #kolla | 02:28 | |
| *** sbezverk has quit IRC | 02:34 | |
| openstackgerrit | Merged openstack/kolla: Remove commented line https://review.openstack.org/293583 | 02:35 |
| *** coolsvap has joined #kolla | 02:37 | |
| *** jtriley has quit IRC | 02:42 | |
| *** ccesario has quit IRC | 02:42 | |
| *** ccesario has joined #kolla | 02:43 | |
| *** jtriley has joined #kolla | 02:43 | |
| *** ntpttr has quit IRC | 02:44 | |
| *** ntpttr has joined #kolla | 02:45 | |
| *** achanda has quit IRC | 02:48 | |
| *** jtriley has quit IRC | 02:49 | |
| *** achanda has joined #kolla | 02:50 | |
| *** achanda has quit IRC | 02:52 | |
| *** achanda has joined #kolla | 02:55 | |
| *** jtriley has joined #kolla | 02:56 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Add generate.py to generate passwords https://review.openstack.org/293728 | 02:56 |
| *** daneyon has joined #kolla | 02:56 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Add generate.py to generate passwords https://review.openstack.org/293728 | 02:57 |
| openstackgerrit | Gerard Braad proposed openstack/kolla: Changed port of docker_registry in globals.yml https://review.openstack.org/293240 | 03:00 |
| coolsvap | SamYaple: you created generate_passwords.yml instead of py | 03:02 |
| *** banix has quit IRC | 03:05 | |
| *** banix has joined #kolla | 03:10 | |
| *** vhosakot has quit IRC | 03:12 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Add generate.py to generate passwords https://review.openstack.org/293728 | 03:12 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Move manila variable https://review.openstack.org/293797 | 03:12 |
| *** salv-orlando has joined #kolla | 03:17 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Add generate_passwords.py to generate passwords https://review.openstack.org/293728 | 03:18 |
| *** iceyao has quit IRC | 03:18 | |
| *** iceyao has joined #kolla | 03:18 | |
| *** salv-orlando has quit IRC | 03:26 | |
| openstackgerrit | Merged openstack/kolla: Changed port of docker_registry in globals.yml https://review.openstack.org/293240 | 03:37 |
| *** jtriley has quit IRC | 03:39 | |
| *** jtriley has joined #kolla | 03:41 | |
| *** tfukushima has joined #kolla | 03:43 | |
| *** ccesario has quit IRC | 03:43 | |
| *** ccesario has joined #kolla | 03:45 | |
| *** jtriley has quit IRC | 03:45 | |
| *** jtriley has joined #kolla | 03:46 | |
| *** rstarmer has joined #kolla | 03:58 | |
| *** dave-mccowan has quit IRC | 04:00 | |
| *** daneyon_ has joined #kolla | 04:00 | |
| *** daneyon has quit IRC | 04:03 | |
| *** Jeffrey4l has joined #kolla | 04:13 | |
| *** jtriley has quit IRC | 04:18 | |
| *** rstarmer has quit IRC | 04:28 | |
| *** banix has quit IRC | 04:31 | |
| *** rstarmer has joined #kolla | 04:35 | |
| *** tfukushima has quit IRC | 04:41 | |
| *** salv-orlando has joined #kolla | 04:48 | |
| *** salv-orlando has quit IRC | 04:51 | |
| *** Jeffrey4l has quit IRC | 05:03 | |
| *** vhosakot has joined #kolla | 05:05 | |
| *** Marga_ has quit IRC | 05:05 | |
| *** unicell has joined #kolla | 05:16 | |
| *** unicell1 has joined #kolla | 05:30 | |
| *** unicell has quit IRC | 05:30 | |
| *** vhosakot has quit IRC | 05:31 | |
| *** vhosakot has joined #kolla | 05:32 | |
| *** mdnadeem has quit IRC | 05:33 | |
| *** achanda has quit IRC | 05:37 | |
| *** AndChat|241401 has joined #kolla | 05:37 | |
| *** ccesario has quit IRC | 05:37 | |
| *** rstarmer has quit IRC | 05:37 | |
| *** rstarmer has joined #kolla | 05:39 | |
| *** tfukushima has joined #kolla | 05:42 | |
| *** sdake has joined #kolla | 05:44 | |
| *** sdake_ has joined #kolla | 05:45 | |
| *** unicell has joined #kolla | 05:49 | |
| *** unicell1 has quit IRC | 05:49 | |
| *** sdake has quit IRC | 05:49 | |
| *** rstarmer has quit IRC | 05:50 | |
| sdake_ | evening folks | 05:51 |
| *** sdake_ is now known as sdake | 05:51 | |
| *** tobe has quit IRC | 06:02 | |
| vhosakot | night! | 06:03 |
| *** vhosakot has quit IRC | 06:04 | |
| *** rstarmer has joined #kolla | 06:07 | |
| *** Marga_ has joined #kolla | 06:12 | |
| *** salv-orlando has joined #kolla | 06:13 | |
| *** Marga_ has quit IRC | 06:17 | |
| *** Marga_ has joined #kolla | 06:19 | |
| *** Marga_ has quit IRC | 06:22 | |
| *** Marga_ has joined #kolla | 06:22 | |
| *** salv-orlando has quit IRC | 06:23 | |
| openstackgerrit | Merged openstack/kolla: MariaDB lights out recovery https://review.openstack.org/293161 | 06:24 |
| coolsvap | evening sdake | 06:25 |
| sdake | hey coolsvap | 06:26 |
| coolsvap | sdake: do you have the link handy for design summit sessions? | 06:27 |
| sdake | https://etherpad.openstack.org/p/kolla-newton-summit | 06:27 |
| coolsvap | got it | 06:29 |
| *** tobe has joined #kolla | 06:29 | |
| coolsvap | thx | 06:29 |
| *** rstarmer has quit IRC | 06:29 | |
| *** rstarmer has joined #kolla | 06:30 | |
| *** tobe has quit IRC | 06:31 | |
| *** Jeffrey4l has joined #kolla | 06:32 | |
| openstackgerrit | MD NADEEM proposed openstack/kolla: Reconfigure for haproxy https://review.openstack.org/288248 | 06:33 |
| *** mdnadeem has joined #kolla | 06:35 | |
| *** akwasnie has joined #kolla | 06:45 | |
| *** tfukushima has quit IRC | 06:55 | |
| *** Marga__ has joined #kolla | 07:00 | |
| *** Marga_ has quit IRC | 07:02 | |
| *** daneyon_ has quit IRC | 07:03 | |
| openstackgerrit | MD NADEEM proposed openstack/kolla: Reconfigure for haproxy https://review.openstack.org/288248 | 07:06 |
| *** tfukushima has joined #kolla | 07:13 | |
| *** mikelk has joined #kolla | 07:15 | |
| *** shakamunyi has quit IRC | 07:18 | |
| openstackgerrit | Merged openstack/kolla-mesos: CLI: implement config commands https://review.openstack.org/289295 | 07:19 |
| openstackgerrit | Merged openstack/kolla-mesos: Convert chronos-cli to kolla-mesos chronos https://review.openstack.org/292738 | 07:20 |
| *** Jeffrey4l has quit IRC | 07:22 | |
| sdake | boy i wish people would triage bugs properly | 07:22 |
| *** akwasnie has quit IRC | 07:22 | |
| *** shakamunyi has joined #kolla | 07:24 | |
| coolsvap | sdake: any issues? | 07:30 |
| sdake | there are 4 fields, all 4 should be set when triaging a bug | 07:31 |
| coolsvap | hehe, yes | 07:31 |
| openstackgerrit | Merged openstack/kolla-mesos: Move Runner classes to service.py https://review.openstack.org/286994 | 07:31 |
| openstackgerrit | Merged openstack/kolla-mesos: Move common config functions into a new file https://review.openstack.org/286995 | 07:31 |
| openstackgerrit | Merged openstack/kolla-mesos: Remove deployment_id_prefix for per-service commands https://review.openstack.org/291054 | 07:31 |
| openstackgerrit | Merged openstack/kolla-mesos: Move write_openrc to a common file https://review.openstack.org/292739 | 07:31 |
| openstackgerrit | Merged openstack/kolla-mesos: Move apply_deployment_vars to a common file https://review.openstack.org/292740 | 07:32 |
| *** salv-orlando has joined #kolla | 07:32 | |
| openstackgerrit | Merged openstack/kolla-mesos: CLI: implement service commands https://review.openstack.org/286996 | 07:32 |
| *** mikelk has quit IRC | 07:33 | |
| sdake | lots of undecided fix commmitted bugs | 07:34 |
| sdake | which have no target assigned | 07:34 |
| sdake | i dont understand how that happens | 07:34 |
| sdake | oh well, can't fix the universe ;) | 07:35 |
| sdake | time for bed | 07:35 |
| sdake | night coolsvap | 07:35 |
| coolsvap | night sdake | 07:35 |
| openstackgerrit | Angus Salkeld proposed openstack/kolla-mesos: Change mysql user from test to nobody https://review.openstack.org/293571 | 07:37 |
| openstackgerrit | Angus Salkeld proposed openstack/kolla-mesos: Add hostname of a node as a lock name. https://review.openstack.org/292878 | 07:37 |
| openstackgerrit | Angus Salkeld proposed openstack/kolla-mesos: run_once commands locking rework https://review.openstack.org/291693 | 07:37 |
| openstackgerrit | Angus Salkeld proposed openstack/kolla-mesos: Create separate ZK lock nodes https://review.openstack.org/292877 | 07:37 |
| *** jhyang_ has joined #kolla | 07:40 | |
| *** salv-orlando has quit IRC | 07:40 | |
| *** rstarmer has quit IRC | 07:45 | |
| *** tfukushima has quit IRC | 07:47 | |
| *** stvnoyes has quit IRC | 08:01 | |
| *** stvnoyes has joined #kolla | 08:02 | |
| *** jmccarthy has quit IRC | 08:04 | |
| *** jmccarthy has joined #kolla | 08:05 | |
| ajafo | no space on jenkins nodes? 2016-03-17 07:39:05.337 | ERROR:kolla.cmd.build:nova-compute-ironic:ApplyLayer exit status 1 stdout: stderr: write /var/lib/rpm/Packages: no space left on device | 08:06 |
| *** akwasnie has joined #kolla | 08:09 | |
| *** mikelk has joined #kolla | 08:12 | |
| *** dwalsh has joined #kolla | 08:13 | |
| *** mbound has joined #kolla | 08:19 | |
| *** salv-orlando has joined #kolla | 08:23 | |
| *** Serlex has joined #kolla | 08:25 | |
| *** akwasnie has quit IRC | 08:25 | |
| *** jhyang_ is now known as jhyang | 08:32 | |
| *** Marga__ has quit IRC | 08:42 | |
| *** Marga_ has joined #kolla | 08:42 | |
| *** sdake has quit IRC | 08:43 | |
| *** barra204 has joined #kolla | 08:43 | |
| *** sdake has joined #kolla | 08:43 | |
| *** shakamunyi has quit IRC | 08:44 | |
| *** cristicalin has joined #kolla | 08:46 | |
| *** akwasnie has joined #kolla | 08:46 | |
| *** sdake has quit IRC | 08:50 | |
| *** mbound has quit IRC | 08:57 | |
| *** dwalsh has quit IRC | 09:00 | |
| *** kproskurin has joined #kolla | 09:01 | |
| openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: Add deployment_id and timestamp to every named volume https://review.openstack.org/291115 | 09:02 |
| openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: Move yaml_utils to type_utils https://review.openstack.org/291114 | 09:02 |
| openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: Cleanup volumes on multiple nodes https://review.openstack.org/291116 | 09:02 |
| *** mbound has joined #kolla | 09:04 | |
| *** rmart04 has joined #kolla | 09:07 | |
| *** salv-orlando has quit IRC | 09:07 | |
| *** athomas has joined #kolla | 09:11 | |
| *** akwasnie has quit IRC | 09:13 | |
| openstackgerrit | MD NADEEM proposed openstack/kolla: Reconfigure for haproxy https://review.openstack.org/288248 | 09:16 |
| *** akwasnie has joined #kolla | 09:30 | |
| *** cristicalin has quit IRC | 09:42 | |
| *** achanda has joined #kolla | 09:44 | |
| *** weiyu has joined #kolla | 09:44 | |
| *** achanda has quit IRC | 09:45 | |
| *** achanda has joined #kolla | 09:46 | |
| *** achanda has quit IRC | 09:46 | |
| *** dims has joined #kolla | 09:57 | |
| *** opennode has joined #kolla | 09:59 | |
| *** pbourke has quit IRC | 10:13 | |
| *** coolsvap has quit IRC | 10:13 | |
| *** pbourke has joined #kolla | 10:13 | |
| *** iceyao has quit IRC | 10:16 | |
| *** iceyao has joined #kolla | 10:16 | |
| openstackgerrit | Merged openstack/kolla-mesos: Change mysql user from test to nobody https://review.openstack.org/293571 | 10:19 |
| openstackgerrit | Merged openstack/kolla-mesos: Create separate ZK lock nodes https://review.openstack.org/292877 | 10:19 |
| *** iceyao has quit IRC | 10:20 | |
| *** iceyao_ has joined #kolla | 10:20 | |
| *** iceyao_ has quit IRC | 10:22 | |
| *** coolsvap has joined #kolla | 10:22 | |
| *** iceyao has joined #kolla | 10:22 | |
| *** tzn has joined #kolla | 10:22 | |
| *** iceyao has quit IRC | 10:23 | |
| *** AndChat|241401 has quit IRC | 10:23 | |
| *** iceyao has joined #kolla | 10:23 | |
| *** ccesario has joined #kolla | 10:24 | |
| *** iceyao has quit IRC | 10:24 | |
| *** ccesario has quit IRC | 10:24 | |
| *** iceyao has joined #kolla | 10:24 | |
| *** ccesario has joined #kolla | 10:24 | |
| *** AndChat|241401 has joined #kolla | 10:25 | |
| openstackgerrit | Andrey Pavlov proposed openstack/kolla-mesos: Fixing run, show, list for interactive mode https://review.openstack.org/292343 | 10:27 |
| *** ccesario has quit IRC | 10:27 | |
| openstackgerrit | Andrey Pavlov proposed openstack/kolla-mesos: Added custom help for interactive mode https://review.openstack.org/292384 | 10:27 |
| *** bradjones has quit IRC | 10:27 | |
| *** AndChat-241401 has joined #kolla | 10:27 | |
| *** opennode has quit IRC | 10:28 | |
| openstackgerrit | Merged openstack/kolla-mesos: run_once commands locking rework https://review.openstack.org/291693 | 10:28 |
| openstackgerrit | Merged openstack/kolla-mesos: Add hostname of a node as a lock name. https://review.openstack.org/292878 | 10:29 |
| *** AndChat|241401 has quit IRC | 10:30 | |
| *** ccesario has joined #kolla | 10:31 | |
| *** AndChat|241401 has joined #kolla | 10:32 | |
| *** ccesario has quit IRC | 10:32 | |
| *** ccesario has joined #kolla | 10:32 | |
| *** AndChat-241401 has quit IRC | 10:34 | |
| *** AndChat-241401 has joined #kolla | 10:34 | |
| *** absubram has quit IRC | 10:34 | |
| *** bradjones has joined #kolla | 10:36 | |
| *** bradjones has quit IRC | 10:36 | |
| *** bradjones has joined #kolla | 10:36 | |
| *** AndChat|241401 has quit IRC | 10:36 | |
| *** ccesario has quit IRC | 10:37 | |
| *** AndChat-241401 has quit IRC | 10:40 | |
| openstackgerrit | Merged openstack/kolla-mesos: Move yaml_utils to type_utils https://review.openstack.org/291114 | 10:42 |
| *** coolsvap has quit IRC | 10:48 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla: [WIP] iscsi container with lvm2 support https://review.openstack.org/291285 | 10:48 |
| openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: [WIP] Fix OpenStack services registration https://review.openstack.org/291251 | 10:50 |
| *** weiyu has quit IRC | 10:50 | |
| *** coolsvap has joined #kolla | 10:51 | |
| *** Jeffrey4l has joined #kolla | 10:53 | |
| *** asalkeld has quit IRC | 11:00 | |
| *** cfarquhar has quit IRC | 11:03 | |
| *** salv-orlando has joined #kolla | 11:07 | |
| *** apuimedo has joined #kolla | 11:12 | |
| *** salv-orlando has quit IRC | 11:18 | |
| *** jtriley has joined #kolla | 11:27 | |
| *** rhallisey has joined #kolla | 11:35 | |
| *** openstackgerrit has quit IRC | 11:48 | |
| *** openstackgerrit has joined #kolla | 11:48 | |
| *** alisonh has joined #kolla | 11:49 | |
| *** akwasnie has quit IRC | 11:50 | |
| openstackgerrit | MD NADEEM proposed openstack/kolla: Manila ubuntu binary container https://review.openstack.org/286379 | 11:50 |
| *** dave-mccowan has joined #kolla | 11:53 | |
| *** iceyao has quit IRC | 11:53 | |
| *** coolsvap has quit IRC | 11:57 | |
| *** mlima has joined #kolla | 11:59 | |
| *** akwasnie has joined #kolla | 12:01 | |
| *** alisonh has quit IRC | 12:01 | |
| *** dwalsh has joined #kolla | 12:02 | |
| *** alisonh has joined #kolla | 12:05 | |
| *** apuimedo has quit IRC | 12:07 | |
| openstackgerrit | Merged openstack/kolla-mesos: Add deployment_id and timestamp to every named volume https://review.openstack.org/291115 | 12:09 |
| *** sbezverk has joined #kolla | 12:10 | |
| *** mlima_ has joined #kolla | 12:13 | |
| *** mlima has quit IRC | 12:16 | |
| *** gfidente has joined #kolla | 12:19 | |
| *** jtriley has quit IRC | 12:20 | |
| openstackgerrit | Allen Gao proposed openstack/kolla-mesos: Fix the deprecated security_group_api and network_api_class https://review.openstack.org/290454 | 12:22 |
| *** dims has quit IRC | 12:25 | |
| *** tzn has quit IRC | 12:27 | |
| *** jtriley has joined #kolla | 12:29 | |
| *** jtriley has quit IRC | 12:34 | |
| *** mlima__ has joined #kolla | 12:45 | |
| openstackgerrit | Andrey Pavlov proposed openstack/kolla-mesos: CLI: log command implementation https://review.openstack.org/291410 | 12:46 |
| *** mlima_ has quit IRC | 12:48 | |
| *** salv-orlando has joined #kolla | 12:48 | |
| *** weiyu_ has joined #kolla | 12:48 | |
| *** mlima_ has joined #kolla | 12:50 | |
| *** alyson_ has quit IRC | 12:52 | |
| *** mlima__ has quit IRC | 12:53 | |
| *** diogogmt has joined #kolla | 12:55 | |
| *** salv-orlando has quit IRC | 12:58 | |
| openstackgerrit | MD NADEEM proposed openstack/kolla: Update README for heka image https://review.openstack.org/294007 | 13:00 |
| *** banix has joined #kolla | 13:02 | |
| *** jtriley has joined #kolla | 13:02 | |
| *** zaneb has joined #kolla | 13:06 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla: [WIP] iscsi container with lvm2 support https://review.openstack.org/291285 | 13:10 |
| *** iceyao has joined #kolla | 13:11 | |
| *** weiyu_ has quit IRC | 13:13 | |
| openstackgerrit | Andrey Pavlov proposed openstack/kolla-mesos: CLI: implement deployment commands https://review.openstack.org/292741 | 13:13 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Reconfigure for haproxy https://review.openstack.org/288248 | 13:13 |
| *** akwasnie has quit IRC | 13:14 | |
| *** salv-orlando has joined #kolla | 13:15 | |
| sbezverk | good morning | 13:15 |
| *** akwasnie has joined #kolla | 13:15 | |
| sbezverk | pbourke, it is you have take care of oracle linux right? | 13:16 |
| *** dwalsh has quit IRC | 13:17 | |
| *** diogogmt has quit IRC | 13:18 | |
| ccesario_ | morning :) | 13:25 |
| openstackgerrit | Artur Zarzycki proposed openstack/kolla-mesos: Connect to mariadb with dnsname. https://review.openstack.org/290176 | 13:26 |
| *** dencaval has joined #kolla | 13:29 | |
| *** dwalsh has joined #kolla | 13:31 | |
| openstackgerrit | MD NADEEM proposed openstack/kolla: Update README for heka image https://review.openstack.org/294007 | 13:33 |
| SamYaple | morning | 13:35 |
| *** vhosakot has joined #kolla | 13:36 | |
| ccesario_ | Hey folks, wht is exactly kolla-mesos ? is it other project!? | 13:36 |
| *** vhosakot has quit IRC | 13:36 | |
| SamYaple | ccesario_: its a project that uses kolla images deployed with mesos | 13:39 |
| *** shardy has joined #kolla | 13:41 | |
| ccesario_ | hmmmm | 13:44 |
| ccesario_ | SamYaple, thanks by info :) | 13:45 |
| dave-mccowan | SamYaple ping | 13:46 |
| dave-mccowan | I have a keystone endpoint question. Looking at open bug where "keystone user-list" is not working when TLS is not. | 13:47 |
| dave-mccowan | When the OS_AUTH_URL is https://172.18.151.92:5000/v2.0, it works. when it is https://172.18.151.92:5000 or https://172.18.151.92:5000/v3 it fails. | 13:48 |
| dave-mccowan | do you know anything about that? | 13:48 |
| SamYaple | dave-mccowan: keystone command will not work at all with v3 keystone | 13:48 |
| SamYaple | ever | 13:48 |
| SamYaple | its deprecated (it should be screaming that at you) | 13:48 |
| SamYaple | use openstack client | 13:49 |
| dave-mccowan | that's what all the deprecated message mean. :-) | 13:49 |
| SamYaple | yup :) | 13:49 |
| dave-mccowan | why doesn't it work with https://172.18.151.92:5000/v2.0 | 13:49 |
| openstackgerrit | Merged openstack/kolla-mesos: CLI: log command implementation https://review.openstack.org/291410 | 13:49 |
| SamYaple | we dont setup a v2.0 endpoint with kolla | 13:49 |
| SamYaple | not for mitaka | 13:49 |
| SamYaple | you as the operator could if you need to | 13:49 |
| *** inc0 has joined #kolla | 13:50 | |
| dave-mccowan | does this happen to work somehow without TLS? I'm trying to make sense of the bug report. | 13:50 |
| SamYaple | it should not, no | 13:51 |
| SamYaple | but i cant say for certain | 13:51 |
| openstackgerrit | Merged openstack/kolla: Libvirt needs openvswitch to connect br-int https://review.openstack.org/288075 | 13:51 |
| SamYaple | what bug report? | 13:51 |
| SamYaple | inc0: you around? | 13:51 |
| dave-mccowan | https://bugs.launchpad.net/bugs/1553577 | 13:51 |
| openstack | Launchpad bug 1553577 in kolla "self-signed certificates don't function with keystone" [Low,Confirmed] - Assigned to Dave McCowan (dave-mccowan) | 13:51 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla: [WIP] iscsi container with lvm2 support https://review.openstack.org/291285 | 13:53 |
| *** tzn has joined #kolla | 13:53 | |
| inc0 | SamYaple, just woken uo | 13:53 |
| inc0 | whats up? | 13:53 |
| *** dims has joined #kolla | 13:53 | |
| openstackgerrit | MD NADEEM proposed openstack/kolla: Added debian as base_distro for heka https://review.openstack.org/294033 | 13:53 |
| SamYaple | dave-mccowan: didnt you document this already? | 13:54 |
| dave-mccowan | SamYaple the use of certificates, twice.. once in the openrc-example, and a second time in advanced-configuration. | 13:55 |
| dave-mccowan | not sure about reported case of keystone CLI not working with TLS. | 13:55 |
| dave-mccowan | i think maybe sdake was hitting two problems at once. v3 and untrusted certs. | 13:57 |
| SamYaple | dave-mccowan: i closed the bug as invalid, sdake can reopen if he things it is not invalid | 13:57 |
| SamYaple | we are already documenting all of this info | 13:57 |
| dave-mccowan | i'll ping him later too. | 13:58 |
| dave-mccowan | from trying it, it does look like we can use v2.0 with a kolla deploy. i don't know why keystone can't find that endpoint on it's own though. | 13:59 |
| SamYaple | they did alot of versionless endpoints in v3 | 14:00 |
| SamYaple | i dont think that stuff existed for v2.0 | 14:00 |
| dave-mccowan | oh, i see. running with --debug, it doesn't even query the version endpoints. it just slaps on /tokens and goes. | 14:00 |
| SamYaple | the openstackclient does not do this | 14:01 |
| SamYaple | just hte keystone one | 14:01 |
| dave-mccowan | yep. mystery solved. bug closed. moving on. :-) | 14:01 |
| SamYaple | deprecation means deprecation | 14:01 |
| *** mbound has quit IRC | 14:02 | |
| *** akwasnie has quit IRC | 14:16 | |
| *** akwasnie has joined #kolla | 14:17 | |
| *** fthiagogv has joined #kolla | 14:17 | |
| *** dims has quit IRC | 14:19 | |
| *** akwasnie has quit IRC | 14:21 | |
| *** akwasnie has joined #kolla | 14:21 | |
| *** dims has joined #kolla | 14:22 | |
| *** dims has quit IRC | 14:25 | |
| *** achanda has joined #kolla | 14:29 | |
| *** spisarski has joined #kolla | 14:30 | |
| *** sdake has joined #kolla | 14:39 | |
| sbezverk | folks, iscsi-lvm patch is ready for "alpha" test as I see consistent positive results in my test bed and I managed to clean up the gate tests. I would appreciate is somebody could volunteer to test it and provide some feedback. | 14:39 |
| *** achanda has quit IRC | 14:40 | |
| sdake | morning | 14:44 |
| sdake | sbezverk did you sort out the sudo thing | 14:44 |
| sbezverk | sdake yep, everything is good know, at least I hope | 14:45 |
| *** ssurana has joined #kolla | 14:52 | |
| openstackgerrit | MD NADEEM proposed openstack/kolla-mesos: tox: delete all pyc before running test https://review.openstack.org/294083 | 14:56 |
| *** ssurana has quit IRC | 14:56 | |
| elmiko | hi all, i'm looking for some debugging advice. i'm getting comfortable with building and deploying the basic parts of the stack through kolla, but now i'm struggling to get swift working. | 14:58 |
| elmiko | i've followed https://github.com/openstack/kolla/blob/master/doc/swift-guide.rst | 14:58 |
| elmiko | but i get errors when running `swift stat`, how can i see the logs produced by swift to debug this? | 14:58 |
| SamYaple | elmiko: those logs will be at /var/lib/docker/volumes/kolla_logs/_data | 14:59 |
| SamYaple | alternatively you can also access them in the swift or heka containers themselves | 14:59 |
| elmiko | SamYaple: so like, attach to the container and just look in /var/log ? | 14:59 |
| SamYaple | you can do ti that way elmiko | 14:59 |
| elmiko | interesting, thanks! | 14:59 |
| SamYaple | i have personally found that to be a pain | 15:00 |
| SamYaple | but you can od it for sure | 15:00 |
| elmiko | yea, i'll try the first way | 15:00 |
| elmiko | i'm just not sure what is happening, but swift is producing some internal error. i imagine i've messed up the ring setup | 15:00 |
| SamYaple | its possible. i will be the first to admit our swift setup is not intuitive at this time. but it is functional | 15:01 |
| SamYaple | if you wouldnt mind documenting the issue you are having elmiko and we can put those up for everyone | 15:01 |
| SamYaple | even if its a simple mistake | 15:01 |
| elmiko | should i make a bug against kolla? | 15:02 |
| SamYaple | elmiko: if its a kolla bug, then yes | 15:02 |
| SamYaple | if its user error, no, but we can still do a docs patch | 15:02 |
| elmiko | ok, sure. i'll finish my due dilligence here then at the least i can make a bug if i'm stumped and we can figure out how to proceed | 15:03 |
| *** achanda has joined #kolla | 15:03 | |
| *** mlima_ has quit IRC | 15:04 | |
| SamYaple | elmiko: if you do get stumped post back here and im sure we can provide more tips | 15:05 |
| *** ssurana has joined #kolla | 15:07 | |
| *** achanda has quit IRC | 15:08 | |
| elmiko | SamYaple: will do, thanks! | 15:10 |
| sdake | elmiko sbezverk or pbourke both have swift expertise | 15:14 |
| sdake | elmiko when you have some logs to point to, might ask help of either of them | 15:15 |
| elmiko | sdake: awesome, thank you | 15:16 |
| ccesario_ | how can I deploy ceilometer ? there is no enable_ceilometer option..... | 15:19 |
| *** mlima has joined #kolla | 15:20 | |
| inc0 | ccesario_, we don't have ceilometer roles | 15:20 |
| ccesario_ | ohhhh, I suspected that :) | 15:23 |
| *** vhosakot has joined #kolla | 15:25 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Allow per node configuration with augments https://review.openstack.org/294098 | 15:31 |
| *** iceyao has quit IRC | 15:32 | |
| openstackgerrit | MD NADEEM proposed openstack/kolla-mesos: Fix order of assertEqual https://review.openstack.org/294100 | 15:33 |
| *** apuimedo has joined #kolla | 15:34 | |
| openstackgerrit | Merged openstack/kolla: Update README for heka image https://review.openstack.org/294007 | 15:34 |
| apuimedo | sdake: Hi. I was trying to find out some example of the cisco configuration for when building neutron with the cisco ml2 plugin but I could not find it | 15:35 |
| sdake | apuimedo no idea how to help with that | 15:36 |
| apuimedo | I'm trying to see if we (kuryr) can use kolla for neutron, keystone, rabbitmq, neutron agents and mysql | 15:36 |
| sdake | apuimedo afaik plugins aren't relaly implemented in kolla - although pbourke has some wayt o get them working | 15:36 |
| sdake | nice | 15:36 |
| apuimedo | I only saw how to get the container to install a pypy package | 15:36 |
| apuimedo | we are building the kubernetes support now | 15:37 |
| sdake | you put your special config files in /etc/kolla/config | 15:37 |
| inc0 | apuimedo, how would you use this? you mean deploy and use for development? | 15:37 |
| SamYaple | apuimedo: we dont copy over the specific cisco_nexus_ml2.conf stuff so youll have to put all those sections in ml2_conf.ini | 15:37 |
| apuimedo | inc0: well, in the future, production. For now I'm trying to build the CI in a way that third party vendors can easily have their own neutron container | 15:37 |
| apuimedo | SamYaple: and for vendors that don't use ml2 there is no mechanism right now | 15:38 |
| apuimedo | (that I could find watching your videos and docs) | 15:38 |
| ccesario_ | inc0, ceilometer roles is it in roadmap!? | 15:38 |
| sdake | ya that is why we need to spend tme on plugins next cycle | 15:38 |
| SamYaple | apuimedo: thats correct | 15:38 |
| sdake | because atm, its not in good shape foryou apuimedo | 15:38 |
| inc0 | ccesario_, yes and no, we want to support big tent, but you're first one asking for ceilometer;) | 15:39 |
| apuimedo | yeah. Sorry that I didn't come earlier in the cycle, we were a bit buried with containers-in-vms and k8s support | 15:39 |
| inc0 | and ceilometer is a bit difficult because it really needs it's own set of nodes | 15:40 |
| inc0 | and we'd need to deploy mangodb as well | 15:40 |
| sdake | apuimedo yes no new features in mitaka - at this point we are ony doing bug fixes | 15:40 |
| SamYaple | inc0: we have mongodb playbooks | 15:40 |
| SamYaple | inc0: we could install ceilometer right now, but we dont have teh playbooks for it | 15:40 |
| apuimedo | I was thinking about building the containers and for mitaka just have a manually derived container that puts the plugin | 15:40 |
| ccesario_ | inc0, the first that asked about it ? :P | 15:41 |
| inc0 | you could do that apuimedo | 15:41 |
| SamYaple | apuimedo: if you are ok with a fork its very easy to do whatr you want | 15:41 |
| sdake | ccesario_ people want it but nobody has implemented it | 15:41 |
| SamYaple | we dont have a non-fork way to do it yet though | 15:41 |
| SamYaple | apuimedo: i would be happy to walk you through the two needed changes | 15:41 |
| apuimedo | SamYaple: sdake: I also wanted to know if there are plans to have builds in the docker and quay registries | 15:41 |
| SamYaple | quay? | 15:42 |
| apuimedo | https://quay.io/ for rkt | 15:42 |
| SamYaple | if you mean dockerhub, then yes | 15:42 |
| sdake | apuimedo the issue there is around secure keys | 15:42 |
| SamYaple | ah. not sure about rkt yet | 15:42 |
| sdake | apuimedo so until that is sorted out, we wont be providing builds in docker hub | 15:42 |
| apuimedo | sdake: which keys? | 15:42 |
| SamYaple | sdake: thats not entirely true | 15:42 |
| sdake | as for rocket, its on our roadmap, however, its not immediately on our roadmap | 15:42 |
| apuimedo | sdake: same here. Since we are building for k8s I try to keep it in mind | 15:43 |
| sdake | apuimedo the security token used to upload to registry - we dont want to just check that into our repo | 15:43 |
| SamYaple | im planning on doing the automated docker builds | 15:43 |
| ccesario_ | sdake, understand.... | 15:43 |
| SamYaple | which is outside infra | 15:43 |
| SamYaple | thats just leg work | 15:43 |
| sdake | it really should be inside infrastructure | 15:43 |
| sdake | in cae, you know, you were hit by a bus or something | 15:44 |
| apuimedo | sdake: couldn't that be held on infra as a jenkins task? | 15:44 |
| SamYaple | sdake: it would be inside our kollaglue repo | 15:44 |
| sdake | apuimedo yes - 24 hours in the day | 15:44 |
| SamYaple | if we had a docker registry in infra this would be a diferetn conversation | 15:44 |
| SamYaple | but for dockerhub, im thinking we just use dockerhub to build much like we us pypi | 15:44 |
| sdake | fungi hasn't said no to that request | 15:44 |
| SamYaple | we use* | 15:44 |
| openstackgerrit | Merged openstack/kolla: Manila ubuntu binary container https://review.openstack.org/286379 | 15:45 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla: iscsi container with lvm2 support https://review.openstack.org/291285 | 15:45 |
| apuimedo | SamYaple: sdake: you should check the team suppor tin quay (it also supports docker) | 15:45 |
| sdake | apuimedo your coming at the wrong time to ask for features - summit i the right time :) | 15:45 |
| SamYaple | apuimedo: im looking now | 15:45 |
| apuimedo | *support | 15:45 |
| sdake | i/is | 15:46 |
| apuimedo | sdake: I'm not asking for features :-) | 15:46 |
| apuimedo | just floating ideas | 15:46 |
| apuimedo | Hopefully we can meet in Austin and plan for Newton ;-) | 15:46 |
| sdake | well rather then focusing on how to build dockerhub images, which we all know needs to be done, we should be focusedon releasing rc2 and stable/liberty | 15:47 |
| apuimedo | sdake: sorry. Didn't mean to disrupt the realease | 15:47 |
| sdake | apuimedo its all good | 15:48 |
| *** blahRus has joined #kolla | 15:49 | |
| apuimedo | SamYaple: Instead of a fork I'll try to build some of the containers from source and then do a "FROM " and add the plugin configuration, see if that flies then (telling ansible to use the containers from a specific namespace where there'll be the modified container) | 15:49 |
| SamYaple | apuimedo: the issue is the kolla config.json files | 15:49 |
| SamYaple | sdake: youre welcome to come up with any solutions here, but the image issue sitll needs to be addressed | 15:50 |
| sdake | agree at the right time, it needs to be addressed | 15:50 |
| sdake | now is not the right time | 15:50 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla: Changing Swift option section in globals.yml https://review.openstack.org/294113 | 15:50 |
| sdake | releasing stable/liberty and rc2 will consume all of our collective capacity | 15:50 |
| SamYaple | if you say so | 15:50 |
| sdake | i do say so | 15:51 |
| apuimedo | SamYaple: do you have any estimation of how much resources are needed to build the images? | 15:51 |
| sdake | apu it takes 15 minutes on my gear to build 120 containers or so | 15:51 |
| sdake | disk sapce is about 12gb i think | 15:51 |
| sdake | but depends on distribution | 15:52 |
| SamYaple | i will still be working on images | 15:52 |
| sdake | instead of stable/liberty? | 15:52 |
| SamYaple | there is no instead of sdake. | 15:52 |
| SamYaple | i do what i say ill do as i always have | 15:53 |
| sdake | well feel free to work in whatever order you like | 15:53 |
| SamYaple | apuimedo: if you mean raw disk space, ubuntu the full stack is ~6GB | 15:53 |
| apuimedo | sorry I brought up a sensitive subject | 15:53 |
| SamYaple | apuimedo: if you mean ram and cpu, that hasnt been quantified | 15:53 |
| SamYaple | apuimedo: :) you didnt its fine | 15:53 |
| apuimedo | SamYaple: I don't mean to bring up more controversy. But have you considered a minimal distro like alpine (like docker uses now for their images) | 15:54 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Add cron to pull playbooks https://review.openstack.org/294116 | 15:54 |
| apuimedo | yeah. I mostly meant disk. My crappy small hard drive is almost full | 15:54 |
| apuimedo | I guess I'll have to use some cloud instance | 15:54 |
| SamYaple | apuimedo: that has also been discussed, im all for some alpine linux! | 15:54 |
| SamYaple | but thats a time issue is all | 15:55 |
| apuimedo | SamYaple: I built the kuryr container on alpine | 15:55 |
| SamYaple | nice | 15:55 |
| apuimedo | (also the midonet ones) | 15:55 |
| SamYaple | did i talk to you about midonet sometime long ago? | 15:55 |
| SamYaple | in kolla? | 15:55 |
| apuimedo | SamYaple: you did | 15:55 |
| SamYaple | ah! | 15:55 |
| sdake | apuimedo here is our roadmap: https://etherpad.openstack.org/p/kolla-mitaka-midcycle-roadmap | 15:55 |
| SamYaple | i knew your name looked familiar | 15:55 |
| apuimedo | :-) | 15:55 |
| apuimedo | all the gates are single node now, then? | 15:56 |
| SamYaple | apuimedo: we do not have a multinode gate right now | 15:58 |
| apuimedo | SamYaple: infra constraint? | 15:58 |
| SamYaple | apuimedo: kind of | 15:59 |
| *** banix has quit IRC | 15:59 | |
| SamYaple | i was planning on waiting for zuulv3 for several reasons but now that that is looking a little ways out ill probably have it done early newton | 15:59 |
| SamYaple | most of the work is done apuimedo, just some data transfer things left | 15:59 |
| apuimedo | good :-) | 16:00 |
| SamYaple | so infra only has two node support atm | 16:00 |
| SamYaple | if we have a stable voting multinode gate and request three, they said yes to that | 16:00 |
| SamYaple | so thats sweet | 16:00 |
| apuimedo | sounds good. We'd also make use of that :-) | 16:01 |
| SamYaple | i mainly want three for proper quorum | 16:01 |
| SamYaple | that way failures can be tested | 16:02 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: After refactor, the upgrade code was missing task https://review.openstack.org/294123 | 16:02 |
| SamYaple | everyone I have done a successful master -> master upgrade on a 1 week old cluster that I am running all my internets through | 16:03 |
| SamYaple | i did not lose internet access (as in the vms didnt lose networking) | 16:03 |
| sdake | nice - good to know that works ;) | 16:03 |
| SamYaple | and all the infra services upgraded as well | 16:03 |
| SamYaple | though i did have to patch a few things | 16:03 |
| apuimedo | SamYaple: we use etcd/zookeeper/consul | 16:03 |
| apuimedo | so yeah | 16:03 |
| SamYaple | apuimedo: we have discussed using those for kolla as well | 16:04 |
| SamYaple | apuimedo: we just havent found a way to do them in a way that they arent a REQUIREMENT | 16:04 |
| *** diogogmt has joined #kolla | 16:04 | |
| SamYaple | but i have a few ideas on the subject | 16:04 |
| apuimedo | SamYaple: docker requires one of the three and k8s requires etcd | 16:04 |
| apuimedo | so we just need them | 16:04 |
| SamYaple | i like etcd personally | 16:04 |
| SamYaple | kolla-mesos is using zookeeper | 16:05 |
| apuimedo | SamYaple: what would you want it for? For the services finding each other? | 16:05 |
| apuimedo | or all the config? | 16:05 |
| apuimedo | (when available) | 16:05 |
| SamYaple | configs | 16:05 |
| apuimedo | I hoped you'd say that | 16:05 |
| SamYaple | we had an amazing patchset for service finding too, but it wasnt configurable | 16:05 |
| SamYaple | https://review.openstack.org/#/c/228719/ | 16:05 |
| SamYaple | Hui worked on that alot | 16:06 |
| apuimedo | SamYaple: Did you check docker 1.10 docker nets included dns service? | 16:06 |
| SamYaple | apuimedo: we dont use docker proxy stuff | 16:06 |
| SamYaple | we use --net=host | 16:06 |
| apuimedo | SamYaple: same here | 16:06 |
| SamYaple | the images themselves could technically work with docker-proxy, but kolla-ansible is all --net=host | 16:07 |
| apuimedo | Consul is indeed the easiest choice at that | 16:07 |
| SamYaple | inc0: do you have the rabbitmq playbooks done? | 16:07 |
| apuimedo | the neutron agents would have to be most likely --net=host as well | 16:07 |
| sdake | rabbitmq playbooks were pushed to rc2 SamYaple | 16:08 |
| sdake | as per last meeting | 16:08 |
| SamYaple | i know sdake, im asking for testing | 16:08 |
| SamYaple | apuimedo: its not technically a requirement to put neutron in net host, but yea so much easier | 16:08 |
| SamYaple | apuimedo: we chose het=host for a few reasons, 1 its faster. 2 its easier to manage. 3 its how operators are used to seeing | 16:09 |
| apuimedo | SamYaple: for the agents it depends on the vendor. Some would need quite a bit of a workaround to work on bridge mode | 16:09 |
| SamYaple | apuimedo: also, you can do 'ip netns exec <neutron network namespace>' from the host! | 16:09 |
| sdake | apuimedo actually the rela reason we chose net=host is becuase kubernetes made using non-host networking impossible to deploy nova or neutron early on in the kolla project | 16:09 |
| apuimedo | SamYaple: I also built the midonet containers using net=host with only CAP_NET_ADMIN | 16:09 |
| sdake | the other reasons SamYaple list are valid points | 16:10 |
| inc0 | SamYaple, not yet, also I think sbezverk also wanted to do them? | 16:10 |
| apuimedo | sdake: you don't use k8s any longer, do you? | 16:10 |
| sdake | but when kolla started kubernetes didn't have net=host | 16:10 |
| sdake | nope no k8s | 16:10 |
| SamYaple | inc0: as in sbezverk wanted to take over? or test them? | 16:10 |
| SamYaple | apuimedo: there is a spec to reintroduce it, but not at this time | 16:10 |
| apuimedo | so now we go full circle, when we use kolla + kuryr to provide Neutron networking to k8s deployments | 16:11 |
| inc0 | take over deployment | 16:11 |
| inc0 | upgrades are still on me | 16:11 |
| sdake | yup personally i'm not convinced underlays are the best choice, but if people want to persue them, more power to em :) | 16:11 |
| sdake | I think bare metal underlays are the simplest way to operate an openstack cloud | 16:12 |
| *** banix has joined #kolla | 16:12 | |
| sdake | and simple is better | 16:12 |
| apuimedo | sdake: you mean running the VMs and containers on the underlay? | 16:12 |
| sdake | apuimedo i mean running under k8s or mesos or swarm | 16:12 |
| SamYaple | inc0: im confused, what about deployment is sbezverk taking over? | 16:12 |
| sdake | these are underlays - just liek baremetal is | 16:12 |
| SamYaple | inc0: i was just asking because i just excerised all the other upgrade code | 16:12 |
| apuimedo | sdake: agreed. I also prefer in kolla is without underlays | 16:12 |
| apuimedo | s/in/if/ | 16:13 |
| sdake | we dont dictate no underlays - but our lead implementation is bare metal based | 16:13 |
| sdake | this is why we have kolla-mesos for example | 16:13 |
| inc0 | SamYaple, we need deployment to use hostnames right? | 16:13 |
| inc0 | IPs* | 16:13 |
| inc0 | it uses hostnames | 16:13 |
| inc0 | my code uses hostnames | 16:13 |
| SamYaple | right so that code needs to be updated to ips, should be really simple | 16:14 |
| SamYaple | like 5 minutes worth of work | 16:14 |
| SamYaple | i tested ips already | 16:14 |
| apuimedo | yes. I saw that | 16:14 |
| inc0 | ips are done? | 16:14 |
| SamYaple | inc0: i tested locally, its just a quick change | 16:15 |
| SamYaple | are you not including that in your patch. im confused? | 16:15 |
| inc0 | SamYaple, upgrade will not be compatible with both ip and hostname | 16:15 |
| inc0 | at the same time | 16:15 |
| SamYaple | inc0: then youre doing it wrong | 16:15 |
| inc0 | its either or | 16:15 |
| SamYaple | it can be made compatible with both man | 16:15 |
| inc0 | no, I don't | 16:15 |
| SamYaple | without ifs or whens or cases | 16:16 |
| inc0 | when: ansible_hostname != (gospel_node.stdout | from_json).hostname | 16:16 |
| SamYaple | right thats doing it wrong | 16:16 |
| SamYaple | that was my point | 16:16 |
| SamYaple | ill show you when i get back from lunch what i mean | 16:16 |
| inc0 | how exactly should I do it then? | 16:16 |
| inc0 | please do | 16:16 |
| SamYaple | yea no problem | 16:16 |
| SamYaple | its really simple | 16:16 |
| elmiko | sbezverk or pbourke, ping | 16:17 |
| *** mbound has joined #kolla | 16:19 | |
| *** akwasnie has quit IRC | 16:25 | |
| *** salv-orlando has quit IRC | 16:25 | |
| *** salv-orlando has joined #kolla | 16:26 | |
| sbezverk | elmiko pong | 16:28 |
| *** jtriley_ has joined #kolla | 16:28 | |
| *** mgoddard has joined #kolla | 16:28 | |
| elmiko | sbezverk: hi, sdake suggested that you might be able to help me with some swift issues | 16:28 |
| sbezverk | elmiko I can try, even though sdake would say "do or do not" :-) | 16:29 |
| sdake | no yoda says that | 16:29 |
| sdake | i just quote him :) | 16:29 |
| elmiko | i've followed the kolla guide for ring setup, but when i run `swift stat` i am getting a 503 from the server. in the swift-object-replicator.log i see this http://paste.openstack.org/show/490939/ | 16:30 |
| elmiko | does that look familiar? | 16:30 |
| elmiko | sdake: lol +1 | 16:30 |
| *** mikelk has quit IRC | 16:30 | |
| sbezverk | elmiko let me check | 16:30 |
| *** jtriley has quit IRC | 16:30 | |
| *** rhallisey has quit IRC | 16:30 | |
| *** rhallisey has joined #kolla | 16:31 | |
| sbezverk | elmiko when you generated rings, have you copied them to /etc/kolla/config/swift and have you run kolla-ansible reconfigure? | 16:32 |
| elmiko | sbezverk: i followed this, https://github.com/openstack/kolla/blob/master/doc/swift-guide.rst | 16:33 |
| elmiko | but didn't copy any files after finishing the ring script | 16:33 |
| elmiko | and i did reconfigure through kolla-asnisble | 16:33 |
| sbezverk | elmiko sorry but it does not answer my question, please respond.. | 16:34 |
| *** jtriley_ has quit IRC | 16:34 | |
| elmiko | i copied nothing into /etc/kolla/config/swift and yes i did run kolla-ansible reconfigure | 16:34 |
| *** v1k0d3n has quit IRC | 16:34 | |
| *** SiRiuS_ has joined #kolla | 16:34 | |
| sbezverk | elmiko, please make sure that generated files are located in /etc/kolla/config/swift | 16:34 |
| elmiko | sbezverk: this is what i have in there, http://paste.openstack.org/show/490941/ | 16:35 |
| sbezverk | elmiko ok, when you run reconfigure have you seen ring files were copied | 16:35 |
| vhosakot | SamYaple: When the enviroment variables OS_PROJECT_DOMAIN_ID and OS_USER_DOMAIN_ID are unset, heat does not work even with the latest heat client (released on March 2nd 2016). I am using a brand new shell with _no_ old auth setting or old env variables in my shell. I have responded to your comments for the review https://review.openstack.org/#/c/293732/1 | 16:35 |
| elmiko | sbezverk: let me look at my backtrace | 16:36 |
| vhosakot | SamYaple: does heat work for you when those env vars are unset ? | 16:36 |
| sbezverk | elmiko, I just want to eliminate basic stuff | 16:36 |
| elmiko | sbezverk: i am not seeing any explicit file copies in the log from kolla-ansible reconfigure | 16:37 |
| sbezverk | elmiko you should see something like that: http://paste.openstack.org/show/490942/ | 16:39 |
| *** unicell has quit IRC | 16:40 | |
| elmiko | sbezverk: i definitely don't see that | 16:40 |
| *** unicell has joined #kolla | 16:41 | |
| sbezverk | elmiko hmm, can you do: docker exec -ti swift_object_server bash | 16:43 |
| elmiko | sbezverk: sure, 1sevc | 16:43 |
| elmiko | sbezverk: ok, got a shell | 16:43 |
| sbezverk | and then check if /etc/swift has ring files and they are identical to yours on your deployment machine | 16:43 |
| elmiko | hmm, i see container.ring.gz and object.ring.gz but no account.ring.gz | 16:45 |
| elmiko | are those the ring files? | 16:46 |
| sbezverk | elmiko that is fine account ring should be in swift_account_server | 16:46 |
| elmiko | ah, ok | 16:46 |
| *** gmmaha has joined #kolla | 16:47 | |
| sdake | ok folks repo is open for reviews for rc2 | 16:47 |
| sbezverk | elmiko please do md5sum on all these rings and compare with ones on your deployment machine | 16:47 |
| sdake | congrats on releasing rc1 on time ;) | 16:48 |
| elmiko | sbezverk: ok | 16:48 |
| gmmaha | Hi, i work with ntpttr on the setup that he has and i had a quick Q.. I am trying to find where the network namespaces get created. I am trying to ssh into the instance i have created by hopping onto the namespace and then into the machine | 16:48 |
| vhosakot | sdake: nice RC1 mail! :)_ | 16:48 |
| sdake | vhosakot the community wrote that yesterday | 16:48 |
| gmmaha | woudl really appreciate anny pointers i could get | 16:48 |
| vhosakot | yep! | 16:48 |
| *** Serlex has quit IRC | 16:48 | |
| vhosakot | gmmaha: AIO or multi ? | 16:48 |
| sdake | just messenger boy | 16:48 |
| gmmaha | vhosakot: multi | 16:48 |
| elmiko | sbezverk: confirmed, they are the same | 16:49 |
| gmmaha | i checked the docker containers for neutron in one of my controller nodes | 16:49 |
| sdake | the namespaces are on the host | 16:49 |
| gmmaha | sdake: ohh cool.. will look for it there.. thanks | 16:50 |
| sdake | iirc you should be able to just connect to the machine | 16:50 |
| sdake | from your host | 16:50 |
| sdake | with neutorn thin containers | 16:50 |
| vhosakot | gmmaha: the namespaces should be crated on the node that runs neutron... since a separate dedicated network nodes does _not_ work due to Ansible bug in 1.9.4, the network node is the controller node in a multinode setup... So, the namespaces would be on the controller node | 16:50 |
| sbezverk | elmiko on target host run mount | grep srv | 16:51 |
| vhosakot | gmmaha: "ip netns list" on controller should show | 16:51 |
| elmiko | sbezverk: i see 3 loop devices mounted on /srv/node/d0 /srv/node/d1 /srv/node/d2 | 16:52 |
| sbezverk | elmiko could you paste the script you use to generate rings? | 16:53 |
| ntpttr | is there a kolla topic that you can subscribe to in the openstack-dev mailing list? I don't see it listed on the page | 16:53 |
| elmiko | sbezverk: sure, 1sec | 16:53 |
| *** rmart04 has quit IRC | 16:53 | |
| gmmaha | sdake: didnt knwo about thin containters.. will cehck them | 16:54 |
| gmmaha | vhosakot: testing it now | 16:54 |
| vhosakot | gmmaha: cool | 16:54 |
| inc0 | gmmaha, hey Ganesh:) | 16:54 |
| gmmaha | Yup.. since the router is not HA its just in one controller vhosakot | 16:54 |
| gmmaha | :) | 16:54 |
| gmmaha | howdy inc0 | 16:54 |
| vhosakot | gmmaha: my middle name is Ganesh too! :) | 16:55 |
| inc0 | gmmaha, you're guys running master right? | 16:55 |
| elmiko | sbezverk: http://paste.openstack.org/show/490948/ | 16:55 |
| gmmaha | vhosakot: really? :) | 16:55 |
| gmmaha | inc0: yeah.. master | 16:55 |
| vhosakot | yep :) | 16:55 |
| *** haplo37 has joined #kolla | 16:55 | |
| gmmaha | vhosakot: nice.. :) | 16:56 |
| inc0 | then you're good, namespaces will be visible on host | 16:56 |
| inc0 | Liberty won't have that, but we're backporting this stuff | 16:56 |
| inc0 | just not merged yet | 16:56 |
| gmmaha | inc0: aah cool... i believe its all master. right ntpttr ? | 16:57 |
| ntpttr | gmmaha: yep all on master, I can see the router on host 4 | 16:57 |
| gmmaha | ntpttr: yeah, host 4 is the one which has it | 16:57 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla: Changing Swift option section in globals.yml https://review.openstack.org/294113 | 16:59 |
| *** jasonsb has quit IRC | 17:02 | |
| sbezverk | elmiko please re-generate rings using this script: http://paste.openstack.org/show/490951/ then rerun reconfigure, make sure new ring files are deployed by comparing md5sum. if after that you still have problem, we will need to enable debug for swift and go through it.. | 17:02 |
| openstackgerrit | Merged openstack/kolla: Increase the docker parition for rht bases in gate https://review.openstack.org/293031 | 17:02 |
| elmiko | sbezverk: ok, i'll run though that. thanks for the help so far =) | 17:03 |
| vhosakot | gmmaha: FYI, you will see namespaces only for DHCP network (to run dnsmasq)... | 17:03 |
| SamYaple | vhosakot: yes it does. let me see the delta (mind you the source code says heat should be working for you!) | 17:04 |
| gmmaha | vhosakot: ohhh.. | 17:05 |
| vhosakot | SamYaple: oh ok, so, you dont see the error when OS_PROJECT_DOMAIN_ID and OS_USER_DOMAIN_ID are unset ? | 17:05 |
| *** dwalsh has quit IRC | 17:06 | |
| gmmaha | vhosakot: there is a router namesapce as well. ( i do have a router, but otherwise just DHCP namespaces) | 17:06 |
| SamYaple | vhosakot: no but im checking now, might be something im doing | 17:06 |
| elmiko | sbezverk: i'm assuming i need to change those IP addresses to the ones i'm using? | 17:06 |
| vhosakot | gmmaha: well, yes... you see qrouter-* namespace for each router that run neutron-L3-agent doing NAT magic | 17:07 |
| sbezverk | elmiko yes, please | 17:07 |
| elmiko | sbezverk: ack, thanks | 17:07 |
| vhosakot | gmmaha: yes, every router ceates a qrouter-* namesapce and every DHCP network created a qdhcp-* namesapce | 17:08 |
| vhosakot | SamYaple: also, https://review.openstack.org/#/c/293728/ I think changes the workflow/steps for the user when using kolla | 17:09 |
| gmmaha | vhosakot: yup.. | 17:09 |
| vhosakot | cool | 17:10 |
| SamYaple | vhosakot: it does | 17:10 |
| SamYaple | kind of | 17:10 |
| SamYaple | basically the user _should_ be populating his own passords now anyway | 17:10 |
| SamYaple | but rarely anyone does | 17:10 |
| vhosakot | SamYaple: doesn't it force the user to run generate_password.py as a new kolla install will have no passwords for any service | 17:10 |
| SamYaple | this would just mean it would be arequirement | 17:11 |
| vhosakot | I know, for sure, neutron does not like to start without a db apssword | 17:11 |
| SamYaple | vhosakot: like i said, the user should be doing it anyway, it would just force the matter | 17:11 |
| vhosakot | ah ok, cool | 17:11 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: After refactor, the upgrade code was missing task https://review.openstack.org/294123 | 17:13 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Add cron to pull playbooks https://review.openstack.org/294116 | 17:13 |
| *** kproskurin has quit IRC | 17:16 | |
| elmiko | sbezverk: ok, ran the script and i see the updated rings on the host in /etc/kolla/config/swift, but after a `kolla-ansible reconfigure` i am not seeing them in the swift_object_server container | 17:16 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Allow per node configuration with augments https://review.openstack.org/294098 | 17:18 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Refactor glance config tasks https://review.openstack.org/294185 | 17:18 |
| SamYaple | sdake: https://review.openstack.org/#/c/294098 is rebased, please update vote | 17:18 |
| sbezverk | elmiko make sure you place these new rings on DEPLOYMENT server and not on the target. reconfigure will take care of copying these.. | 17:20 |
| elmiko | sbezverk: yea, this is AIO deployment, only one server | 17:20 |
| elmiko | not sure why reconfigure didn't copy them | 17:20 |
| SamYaple | elmiko: im not sure reconfigure is implemented correctly for swift | 17:21 |
| SamYaple | i may not be checking ring files | 17:21 |
| SamYaple | elmiko: do a service restart manual | 17:21 |
| SamYaple | docker restart swift.... | 17:21 |
| elmiko | SamYaple: ok, i'll give that a try, for all the swift containers? | 17:21 |
| SamYaple | elmiko: yes | 17:21 |
| elmiko | thanks | 17:21 |
| SamYaple | elemoine: ping | 17:23 |
| SamYaple | does heka support config hot config reload? | 17:23 |
| elemoine_ | no | 17:24 |
| elemoine_ | SamYaple: https://github.com/mozilla-services/heka/issues/38 | 17:24 |
| SamYaple | elemoine_: ok so the question is, now that we have reconfigure should we by default be copying in swift stuff now? | 17:24 |
| SamYaple | remember i said we should remove the bool when we didnt have reconfigure | 17:25 |
| SamYaple | but now.... | 17:25 |
| SamYaple | i only bring it up because we arent copying elasticsearch | 17:25 |
| SamYaple | so we need to change one of them | 17:25 |
| SamYaple | for consistency | 17:25 |
| elmiko | SamYaple, sbezverk, i restarted those containers, but no copy of the ring files. is there another step i am missing? | 17:27 |
| elemoine_ | SamYaple: +1 for not copying in Swift when enable_swift is false | 17:28 |
| SamYaple | do the ring files exist in /etc/kolla/swift*/<rings> ? | 17:28 |
| *** Marga_ has quit IRC | 17:28 | |
| SamYaple | elemoine_: ok let me see how that looks | 17:28 |
| elmiko | SamYaple: looks like the old ones are there | 17:29 |
| elmiko | the new ones are in /etc/kolla/config/swift/ | 17:29 |
| SamYaple | elmiko: hmmm let me lookup the swift code | 17:30 |
| SamYaple | i didnt write that code so itll be a second | 17:30 |
| elmiko | SamYaple: thanks, no rush | 17:30 |
| SamYaple | elmiko: the new ones should be in /etc/kolla/swift/*rings* | 17:31 |
| SamYaple | do you see them there? | 17:32 |
| elmiko | SamYaple: no, they got put in /etc/kolla/config/swift/*rings* | 17:32 |
| SamYaple | so after you rebuild, but them in /etc/kolla/config/swift/*rings* | 17:32 |
| SamYaple | then run reconfig | 17:32 |
| SamYaple | then new ones should be in /etc/kolla/swift/*rings* | 17:32 |
| SamYaple | (this makes more sense when your deploy host is not your target host btw) | 17:32 |
| elmiko | ok, i'll run reconfig again | 17:32 |
| elmiko | oh really? i was just following the AIO stuff | 17:33 |
| SamYaple | elmiko: no no its fine | 17:33 |
| SamYaple | i was just saying the confusion with /etc/kolla/ and /etc/kolla/config | 17:33 |
| SamYaple | it works perfectly fine with AIO | 17:33 |
| elmiko | ah, ok | 17:33 |
| SamYaple | but its sometimes confusing, it wouldnt be confusing with a seperate host | 17:33 |
| sbezverk | SamYaple quick question, could you point me to the code where kolla_logs volume has rights assigned so other containers can write their logs.. | 17:34 |
| elmiko | SamYaple: so, re-ran `kolla-ansible reconfigure` but it doesn't seem to want to copy the files over :/ | 17:36 |
| SamYaple | sbezverk: its in the base container, the group kolla has rights, and all users are part of group kolla | 17:37 |
| SamYaple | elmiko: ok just move them into place manually for now and fill a bug | 17:38 |
| SamYaple | pbourke will be able to help im sure | 17:38 |
| *** jtriley has joined #kolla | 17:38 | |
| *** dwalsh has joined #kolla | 17:38 | |
| elmiko | SamYaple: k, thanks | 17:38 |
| *** daneyon has joined #kolla | 17:40 | |
| sdake | SamYaple let me do an actual review now :) | 17:40 |
| *** rajathagasthya has joined #kolla | 17:42 | |
| sdake | SamYaple nice work on that per node augment | 17:43 |
| *** tzn has quit IRC | 17:46 | |
| sbezverk | SamYaple I am converting two host mount volumes into named volumes which will be shared between three containers.. I need to make sure respective users has corrects rights.. I still cannot locate how it was done for heka.. coudl you please point me to the actual file where it is implamented. | 17:47 |
| elmiko | sbezverk, SamYaple, ok, i've rebuilt the rings, reloaded the containers, confirmed that the new rings are in the containers, still getting 503 from swift. | 17:47 |
| *** tzn has joined #kolla | 17:47 | |
| *** fbarilla has joined #kolla | 17:48 | |
| ntpttr | hey, we're creating networks and then creating VMs to associate to those networks, but every time the VM gets allocated two IPs from the network rather than 1, and only one of them works | 17:48 |
| ntpttr | is this a known issue with fresh kolla installations, or maybe an incorrect config somewhere? | 17:49 |
| openstackgerrit | Andrey Pavlov proposed openstack/kolla-mesos: CLI: implement deployment commands https://review.openstack.org/292741 | 17:50 |
| *** dims has joined #kolla | 17:50 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Rework heka log copying https://review.openstack.org/294206 | 17:50 |
| SamYaple | sbezverk: https://github.com/openstack/kolla/blob/master/docker/heka/extend_start.sh#L8 looks like it got moved | 17:52 |
| sbezverk | elmiko enable debug/verbose in swift and get logs | 17:52 |
| sbezverk | SamYaple thank you | 17:52 |
| SamYaple | ntpttr: ive never heard of that | 17:52 |
| elmiko | sbezverk: ack | 17:52 |
| *** tzn has quit IRC | 17:52 | |
| SamYaple | ntpttr: ill check if its happening to me | 17:52 |
| *** jtriley has quit IRC | 17:52 | |
| SamYaple | elemoine_: https://review.openstack.org/294206 | 17:52 |
| ntpttr | SamYaple: it happened using the networks set up by the post-install scripts provided by kolla and on networks we created ourselves | 17:53 |
| SamYaple | ntpttr: as in these extra ports show up in neutron port-list? | 17:53 |
| SamYaple | or its just skipping a port? | 17:54 |
| ntpttr | SamYaple: yep they show up | 17:54 |
| ntpttr | SamYaple: but only one of the two show up when you're in the VM and run 'ip a' | 17:54 |
| SamYaple | ntpttr: nope not on my deploy | 17:54 |
| SamYaple | its skipping a port | 17:54 |
| SamYaple | but its not showing up in port list | 17:55 |
| SamYaple | skipping an ip* | 17:55 |
| SamYaple | so it assigns .9 then .11 then .13 | 17:55 |
| SamYaple | the skipping port behaviour is new | 17:55 |
| SamYaple | to me at least | 17:55 |
| ntpttr | SamYaple: oh interesting, could it be like a failover/ha thing? one is down and one is active | 17:56 |
| ntpttr | we're testing it out | 17:56 |
| SamYaple | thanks sdake its helpful | 17:56 |
| SamYaple | ntpttr: i... am doubtful | 17:56 |
| *** tzn has joined #kolla | 17:56 | |
| SamYaple | do you have l3ha turned on? | 17:56 |
| ntpttr | SamYaple: where is that in the config again? | 17:56 |
| SamYaple | or are you legacy routering it | 17:56 |
| *** jtriley has joined #kolla | 17:56 | |
| ntpttr | SamYaple: in the globals config? | 17:56 |
| SamYaple | no | 17:56 |
| SamYaple | ntpttr: thats in the neutron configs. if you didnt put it in the augment files you didnt turn it on by default | 17:57 |
| *** daneyon has quit IRC | 17:57 | |
| SamYaple | if you ran init-runonce you didnt create an l3ha router either | 17:57 |
| ntpttr | no then I guess it's not on | 17:57 |
| SamYaple | just checking | 17:57 |
| openstackgerrit | Andrey Pavlov proposed openstack/kolla-mesos: CLI: implement deployment list command https://review.openstack.org/293444 | 17:57 |
| *** tzn has quit IRC | 17:57 | |
| openstackgerrit | Andrey Pavlov proposed openstack/kolla-mesos: CLI: implement commands list command https://review.openstack.org/293081 | 17:57 |
| SamYaple | fyi guys, this needs to merge before mitaka is released (perferably asap) https://review.openstack.org/#/c/293106/ | 17:58 |
| SamYaple | and please review the volume names because we cannot rename those ever after mitaka | 17:58 |
| SamYaple | without downtime | 17:58 |
| *** tzn has joined #kolla | 17:59 | |
| SamYaple | ntpttr: 2016-03-14 15:37:42.946 35 DEBUG neutron.db.ipam_non_pluggable_backend [req-e21442a9-79b1-4002-bb16-40d732d062fa 3134a264c46d4236aef2f8557376ac2d 704a862ee32c48a4baf24d38f25d5867 - - -] Allocated IP - 172.16.100.7 from 172.16.100.8 to 172.16.199.255 _try_generate_ip /var/lib/kolla/venv/local/lib/python2.7/site-packages/neutron/db/ipam_non_pluggable_backend.py:80 | 18:01 |
| SamYaple | it is allocating two ips | 18:01 |
| ntpttr | SamYaple: one other thing, we have ovs enabled | 18:01 |
| SamYaple | ntpttr: so do it | 18:01 |
| *** ccesario has joined #kolla | 18:01 | |
| SamYaple | so do i* | 18:01 |
| *** haplo37 has quit IRC | 18:02 | |
| SamYaple | ntpttr: its allocating two ips, but its deleting the unused one for me | 18:02 |
| ntpttr | hmm never seen that behavior before, and I wonder why yours is being deleted while ours isn't | 18:02 |
| SamYaple | ntpttr: youre telling me this is all new | 18:03 |
| SamYaple | what is this ipman stuff | 18:03 |
| SamYaple | ipam* | 18:03 |
| ntpttr | ip a, we were just listing all the IPs in the vm. It's just like ifconfig | 18:03 |
| ntpttr | only one of the two is showing up within the VM, the active one | 18:03 |
| SamYaple | no i mean from the debug message | 18:03 |
| ntpttr | oh sorry | 18:03 |
| ntpttr | no clue hah | 18:03 |
| SamYaple | let me check the actual instances i have | 18:03 |
| rhallisey | SamYaple, re naming, is this only for liberty? | 18:04 |
| rhallisey | libvirt* | 18:04 |
| rhallisey | not liberty | 18:04 |
| SamYaple | rhallisey: this is for master | 18:04 |
| SamYaple | yes | 18:04 |
| *** ccesario has quit IRC | 18:05 | |
| SamYaple | i only see libvirt needs a change | 18:05 |
| *** ccesario has joined #kolla | 18:05 | |
| SamYaple | so its not "nova_libvirt" but rather "libvirtd" | 18:05 |
| rhallisey | reason I'm asking is the change, I forgot who, wanted to parameterize the volumes names | 18:05 |
| SamYaple | but please have a look at the volumes we create and tell me if you see another change | 18:05 |
| SamYaple | rhallisey: that sounds like a good idea, but proably too big a change for now | 18:05 |
| rhallisey | ya totally, I just wanted to understand if there will be a problem with that in the future | 18:06 |
| SamYaple | parameterized volume names could work fine, if you need to rename volume then you have down time. thats it though | 18:06 |
| SamYaple | nope shouldnt affect that | 18:06 |
| rhallisey | gotcha | 18:06 |
| SamYaple | ntpttr: so _in_ hte instance you see two ips | 18:06 |
| rhallisey | more specifically though, the reason for the downtime is because of the libvirt volume name change? | 18:06 |
| SamYaple | yes | 18:07 |
| *** ccesario has quit IRC | 18:07 | |
| *** ccesario has joined #kolla | 18:07 | |
| SamYaple | change volume names isnt really supported, but it would require a full container stop of all containers using htat volume and redeploy after data move | 18:07 |
| ntpttr | SamYaple: no, just one | 18:07 |
| rhallisey | k will keep that noted | 18:07 |
| *** tzn has quit IRC | 18:08 | |
| SamYaple | ntpttr: can you give me `nova list --all-` `neutron port-list` `nova service-list` `neutron agents-list` outputs? | 18:08 |
| gmmaha | SamYaple: we went haead and rebooted the host on which the VM was running and started the instance again. Now the VM is using the second IP that was allocated and the first IP& port is showing down | 18:09 |
| gmmaha | this all feels like allocation for HA without the actual HA | 18:09 |
| SamYaple | gmmaha: super wierd | 18:09 |
| SamYaple | i asking in #openstack-neutron | 18:09 |
| ntpttr | SamYaple: http://paste.openstack.org/show/490966/ | 18:09 |
| SamYaple | that log entry doesnt lie, its neutron doing it | 18:10 |
| gmmaha | SamYaple: :) cool.. let me join and listen in | 18:10 |
| SamYaple | kolla may be able to stop it with a config change, but i dont know why its doing it in the first place | 18:10 |
| vhosakot | gmmaha: I've seen it happen if the DHCP of the port lease expires _during_ reboot | 18:10 |
| sbezverk | mdnadeem sdake you provided contradictive comments on swift doc change, could you please make up your mind :-) | 18:10 |
| SamYaple | ntpttr: thats wierd. you did nova list --all-? | 18:11 |
| gmmaha | vhosakot: aaah.. but thats only if you have multiple IPs assigned to the VM right? | 18:11 |
| ntpttr | SamYaple: the other weird thing it, on the networks created by the kolla script it was assigning 3 IPs, now it's 2 | 18:11 |
| ntpttr | SamYaple: yeah I did | 18:11 |
| SamYaple | oh nevermind | 18:11 |
| SamYaple | those are for gateway and dhcp serve | 18:12 |
| SamYaple | r | 18:12 |
| vhosakot | gmmaha: not really, DHCP lease expires even with one interface (even if it or its VM is down)... the lease expiry is know and triggerred by teh DHCP server (dnsmaq) | 18:12 |
| *** ccesario has quit IRC | 18:12 | |
| *** AndChat|241401 has joined #kolla | 18:12 | |
| SamYaple | ntpttr: `neutron net-list` `neutron subnet-list` | 18:12 |
| vhosakot | gmmaha: it is the respoinsiblity of the DHCP client to renew the lease well before expiry,.. otherwise, VM gets a new IP | 18:12 |
| gmmaha | vhosakot: aaah.. didnt know that | 18:13 |
| *** AndChat|241401 has quit IRC | 18:13 | |
| ntpttr | SamYaple: http://paste.openstack.org/show/490968/ | 18:13 |
| SamYaple | vhosakot: thats not true for neutron though | 18:13 |
| SamYaple | thats true for dhcp in general | 18:13 |
| SamYaple | vhosakot: dhcp is based on port-list which is static | 18:14 |
| vhosakot | gmmaha: yep, for neutron and any cleint, the expiry time can be configuyred | 18:14 |
| vhosakot | SamYaple: yep agreed | 18:14 |
| vhosakot | SamYaple: port-list is static, IP of those ports may not be if not renewed in time | 18:14 |
| gmmaha | SamYaple: vhosakot: so in this scenario, the IP should be renewed for the port given that its static right?? | 18:14 |
| SamYaple | ntpttr: welp got the conversation going in neutron channel, well see what they say..... | 18:15 |
| gmmaha | i agree on how DHCP works.. but very puzzled neutron does the same | 18:15 |
| SamYaple | vhosakot: youre wrong | 18:15 |
| SamYaple | port has an ip period | 18:15 |
| SamYaple | the dhcp agents give out ips based on that port list | 18:15 |
| vhosakot | SamYaple: does it override lease time ? | 18:15 |
| SamYaple | vhosakot: the leases list is statically built | 18:15 |
| SamYaple | it may give out a 24h lease, but thats just dnsmasq configuration | 18:16 |
| vhosakot | SamYaple: right, but, dopesn;t the leae ever expire ? | 18:16 |
| SamYaple | its never going to change | 18:16 |
| SamYaple | dnsmasq by default does set a lease time HOWEVER not renewing wont change your ip | 18:16 |
| sdake | sbezverk could you be moer specific | 18:16 |
| SamYaple | its yours for the life of the port | 18:16 |
| vhosakot | the default is 1 day 86,400 seconds | 18:16 |
| vhosakot | SamYaple: I gree that point | 18:17 |
| vhosakot | agree* | 18:17 |
| *** haplo37 has joined #kolla | 18:17 | |
| openstackgerrit | Merged openstack/kolla: After refactor, the upgrade code was missing task https://review.openstack.org/294123 | 18:19 |
| SamYaple | thats so wierd ntpttr but i think i read the debug message wrong | 18:19 |
| SamYaple | https://github.com/openstack/neutron/blob/master/neutron/db/ipam_non_pluggable_backend.py#L49 | 18:19 |
| SamYaple | thats the code.... | 18:19 |
| sbezverk | sdake in you original comment you suggested to add space to this line #swift blah blah, I have done it, second review from mdnadeem suggests to remove space from line # swift blah blah, what should I do? | 18:20 |
| SamYaple | ntpttr: can you give me `nova show <instance>` ? | 18:20 |
| ntpttr | SamYaple: sure one moment | 18:20 |
| sdake | sbezverk follow the rest of the document which has a space between the # and the content | 18:20 |
| sdake | if you hae a link to hte review i can clarify | 18:20 |
| SamYaple | rhallisey: please rereview the libvirt patch | 18:21 |
| rhallisey | SamYaple, I just did | 18:21 |
| ntpttr | SamYaple: http://paste.openstack.org/show/490971/ | 18:21 |
| rhallisey | SamYaple, your container name is still nova_libvirt | 18:21 |
| rhallisey | idk if that was dropped or intended | 18:21 |
| SamYaple | rhallisey: i know please rereview the patch | 18:21 |
| sbezverk | sdake https://review.openstack.org/#/c/294113/2/etc/kolla/globals.yml | 18:21 |
| SamYaple | ntpttr: is there a floating ip attached to that? | 18:22 |
| ntpttr | SamYaple: yeah it looks like in your log it's only actually assigning *.7, from the range .8 to .255 | 18:22 |
| ntpttr | SamYaple: yes there is, in the 30.0.0.0 range | 18:22 |
| ntpttr | SamYaple: that one is unrelated | 18:22 |
| SamYaple | can you create another instance without a floating ip attached please? | 18:22 |
| SamYaple | also do you have debug on in your logs by chance? | 18:23 |
| ntpttr | SamYaple: here's the new instance, no floating. http://paste.openstack.org/show/490972/. I don't think we have debug enabled | 18:24 |
| sdake | sbezverk I guess I was wrong | 18:24 |
| elmiko | sbezverk: having a little trouble, is there a shortcut to turn on debug? | 18:24 |
| SamYaple | ntpttr: so wierd! | 18:25 |
| SamYaple | ntpttr: the code says that only happens if you have two networks created | 18:25 |
| SamYaple | two subnets* on your network | 18:25 |
| gmmaha | SamYaple: dnsmasq logs are showing that its reading 2 addresses http://paste.openstack.org/show/490973/ | 18:25 |
| SamYaple | hows your rabbitmq look? | 18:26 |
| SamYaple | split brained? | 18:26 |
| SamYaple | what about the database? | 18:26 |
| sbezverk | sdake np, fixing it now.. | 18:26 |
| SamYaple | ntpttr: `rabbitmqctl eval 'rabbit_clusterer:status().` `mysql -uroot -ppassword -e 'show status like "%wsrep%";'` | 18:27 |
| SamYaple | ntpttr: assuming you setup the aliases from before ^ | 18:27 |
| SamYaple | if not youll need to docker exec those commands | 18:27 |
| vhosakot | SamYaple: haproxy reconfig also does not have section based ini file.. so, just wanted to let you know there are 3 services with non-ini - rabbitmq, mariadb and haproxy | 18:28 |
| vhosakot | SamYaple: are there more ? | 18:28 |
| SamYaple | vhosakot: mariadb is ini based | 18:28 |
| SamYaple | we do merge configs on it already | 18:28 |
| SamYaple | vhosakot: horizon is one | 18:28 |
| vhosakot | SamYaple: horizon horizon.. right.. thanks! :) | 18:28 |
| ntpttr | SamYaple: the rabbit cluster is appearing down - Error: unable to connect to node rabbit@jfz1r02h03: nodedown | 18:29 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla: Changing Swift option section in globals.yml https://review.openstack.org/294113 | 18:30 |
| vhosakot | SamYaple: the veri first line in galera.cnf.j2 is _not_ under a section.. won't Python ConfigParser/merge_config complain ? https://github.com/openstack/kolla/blob/master/ansible/roles/mariadb/templates/galera.cnf.j2#L1 | 18:30 |
| vhosakot | very* | 18:30 |
| SamYaple | ntpttr: `rabbitmqctl cluster_status` on all the rabbitmq nodes please | 18:30 |
| vhosakot | SamYaple: the rest of galera.cnf.j2 is section-based and ini-based | 18:31 |
| SamYaple | vhosakot: it gets templated first :) | 18:31 |
| SamYaple | its valid ini by the time it merges | 18:31 |
| SamYaple | all configs go through template engine first | 18:31 |
| ntpttr | SamYaple: same everywhere | 18:31 |
| vhosakot | SamYaple: so, merge_config templates as well.. cool! thanks ! :) | 18:31 |
| SamYaple | ntpttr: so rabbitmq is down. this my explain your issues! | 18:31 |
| SamYaple | how its working at all is the real question | 18:32 |
| vhosakot | SamYaple: so, rabbit, haproxy and horizon...3 services are non-ini | 18:32 |
| ntpttr | SamYaple: no kidding! how are vms coming up, we're able to ssh into them, everything appears to be working | 18:32 |
| ntpttr | SamYaple: in the dhcp_agent logs I remember seeing that it connected to rabbit | 18:32 |
| SamYaple | ntpttr: `ansible -i <your multinode inventory> -m shell -a "docker ps -a | grep rabbitmq" all` | 18:33 |
| SamYaple | please post that output | 18:33 |
| *** jtriley has quit IRC | 18:34 | |
| ntpttr | 172.16.2.3 | FAILED | rc=1 >> | 18:34 |
| ntpttr | 172.16.2.5 | FAILED | rc=1 >> | 18:34 |
| ntpttr | 172.16.2.4 | FAILED | rc=1 >> | 18:34 |
| ntpttr | 172.16.2.11 | FAILED | rc=1 >> | 18:34 |
| ntpttr | 172.16.2.10 | FAILED | rc=1 >> | 18:34 |
| ntpttr | 172.16.2.6 | FAILED | rc=1 >> | 18:34 |
| ntpttr | 172.16.2.12 | FAILED | rc=1 >> | 18:34 |
| ntpttr | 172.16.2.7 | FAILED | rc=1 >> | 18:34 |
| ntpttr | 172.16.2.8 | FAILED | rc=1 >> | 18:34 |
| ntpttr | 172.16.2.9 | FAILED | rc=1 >> | 18:35 |
| *** unicell1 has joined #kolla | 18:35 | |
| ntpttr | not looking good | 18:35 |
| SamYaple | ntpttr: thats all your hosts? | 18:35 |
| ntpttr | yeah | 18:35 |
| SamYaple | do you have rabbitmq on any of your nodes....? | 18:35 |
| sdake | ntpttr would you midn using a paste service please | 18:35 |
| sdake | ls | 18:35 |
| *** manjeets has joined #kolla | 18:35 | |
| SamYaple | rhallisey: the VOLUME name is changing. it has to change now we cant change it later without downtime. all of this is explained in the commit | 18:35 |
| ntpttr | SamYaple: should be .3 .4 and .5 that are controllers running rabbit | 18:35 |
| SamYaple | ntpttr: can you show my docker ps -a form those nodes? | 18:35 |
| ntpttr | SamYaple: ah wait small type in the command | 18:36 |
| ntpttr | 172.16.2.3 | success | rc=0 >> | 18:36 |
| ntpttr | 4bfcbf57711f 172.16.2.2:4000/kollaglue/ubuntu-source-rabbitmq:2.0.0 "kolla_start" 19 hours ago Up 19 hours rabbitmq | 18:36 |
| ntpttr | 172.16.2.4 | success | rc=0 >> | 18:36 |
| ntpttr | b55e434dbb96 172.16.2.2:4000/kollaglue/ubuntu-source-rabbitmq:2.0.0 "kolla_start" 19 hours ago Up 19 hours rabbitmq | 18:36 |
| ntpttr | 172.16.2.5 | success | rc=0 >> | 18:36 |
| ntpttr | 8fe2eb204fa8 172.16.2.2:4000/kollaglue/ubuntu-source-rabbitmq:2.0.0 "kolla_start" 19 hours ago Up 19 hours rabbitmq | 18:36 |
| rhallisey | SamYaple, then why are you remove libvirtd in cleanuip containers | 18:36 |
| ntpttr | there's the three nodes that run it | 18:36 |
| rhallisey | that is looking at container names | 18:36 |
| SamYaple | rhallisey: look again please | 18:36 |
| ntpttr | sdake: sure sorry | 18:36 |
| sdake | ntpttr no reason to apologize | 18:37 |
| sdake | just hard to keep up with irc with all the scroll | 18:37 |
| rhallisey | nevermind I realize my mistake.. | 18:37 |
| SamYaple | ntpttr: ok we need to ifigure out your rabbitmq problem (this is why i wanted to do that the other day :P) | 18:37 |
| SamYaple | rhallisey: no problem :) | 18:37 |
| rhallisey | I'm outdated on that script.. | 18:37 |
| vhosakot | I cherry-picked and tested the haproxy reconfig and it works fine with no issues... so, gave +1 for it... https://review.openstack.org/#/c/288248/14 | 18:37 |
| *** unicell has quit IRC | 18:37 | |
| SamYaple | ntpttr: lets dig into the rabbitqm logs please | 18:37 |
| rhallisey | cleanup-containers removes volumes.. | 18:37 |
| rhallisey | I didn't know that.. | 18:37 |
| SamYaple | rhallisey: who would have guessed with a name like "cleanup-containers" right? | 18:37 |
| rhallisey | XD | 18:39 |
| *** jtriley has joined #kolla | 18:43 | |
| *** rajathagasthya has quit IRC | 18:49 | |
| *** Marga_ has joined #kolla | 18:51 | |
| inc0 | SamYaple, sooo....upgrades | 18:53 |
| inc0 | you promised me sth | 18:53 |
| SamYaple | inc0: yea i havent looked at it since i got back from lunch | 18:55 |
| SamYaple | triaging other things | 18:55 |
| sdake | builds in flagstaff = slow compared to gige | 18:55 |
| * sdake sadface | 18:55 | |
| SamYaple | inc0: let me do it now, the triaging is directly related to that issue! ntpttr | 18:55 |
| sdake | ladies and gentlemen | 18:57 |
| sdake | the threat analysis work will get kicked off once we wrap with mitaka and stable/liberty | 18:58 |
| *** mgoddard has quit IRC | 18:58 | |
| sdake | until then feel free to just read through those meeting notes and the gerrit review | 18:58 |
| sdake | and please do read the thread analysis html page | 18:59 |
| sdake | thread/threat | 18:59 |
| vhosakot | kernel thread analysis is probably needed threat analysis | 19:02 |
| openstackgerrit | Steven Dake proposed openstack/kolla: Do not create user home directory in toolbox https://review.openstack.org/290280 | 19:02 |
| openstackgerrit | Steven Dake proposed openstack/kolla: Reorder operations of mkdir and useradd in toolbox https://review.openstack.org/290280 | 19:03 |
| *** athomas has quit IRC | 19:10 | |
| *** salv-orl_ has joined #kolla | 19:16 | |
| *** salv-orlando has quit IRC | 19:19 | |
| mlima | hello guys, i have a aio deployment, but i cant ping any instance | 19:21 |
| mlima | host -> instance | 19:21 |
| mlima | but, inside of router, i can ping any instance | 19:22 |
| inc0 | mlima, sounds like your host doesn't have routing to instances:) | 19:32 |
| *** Jeffrey4l_ has joined #kolla | 19:33 | |
| *** Jeffrey4l has quit IRC | 19:34 | |
| *** haplo37 has quit IRC | 19:34 | |
| *** shardy has quit IRC | 19:36 | |
| mlima | inc0, i can use there network interfaces? eth0,eth1 and eth2? | 19:39 |
| mlima | eth0 and eth2 have ip, but eth1 not have an ip | 19:40 |
| inc0 | so neutron works in a way that it creates namespace and routes there | 19:41 |
| inc0 | what you can do is to create tap interface inside namespace and outside namespace and add it as gateway to vm subnet | 19:42 |
| SamYaple | mlima: or, and preferably, dont do that! | 19:43 |
| SamYaple | ip netns exec <namespace> ping <vm> | 19:43 |
| SamYaple | this is where ip netns shines | 19:43 |
| SamYaple | you can even do: | 19:43 |
| SamYaple | ip netns exec <namespace> bash | 19:43 |
| inc0 | yeah,that's proper way | 19:44 |
| inc0 | but if you need to run stuff outside ns that has access to vms | 19:44 |
| inc0 | you need a bit of tinkering | 19:44 |
| SamYaple | im not sure what things those would be, but there are many ways to change your namespace | 19:45 |
| mlima | ok | 19:45 |
| mlima | i understood | 19:45 |
| mlima | Thank guys | 19:45 |
| mlima | one more doubt, can i use there network interfaces? eth0,eth1 and eth2? | 19:46 |
| mlima | in globals.yml i have two network interfaces | 19:47 |
| *** haplo37 has joined #kolla | 19:47 | |
| mlima | but, i was thinking about the possibillity to have another network interface | 19:48 |
| mlima | SamYaple, inc0 ^ | 19:48 |
| SamYaple | mlima: you can, but it has to be done with augment files | 19:51 |
| SamYaple | and you have to setup the bridge yourself (check out playbooks for the ovs bridge command) | 19:51 |
| SamYaple | its not trivial in kolla atm, but it is possible | 19:51 |
| mlima | I know SamYaple, i tried to do it here and not working well | 19:53 |
| mlima | haha | 19:53 |
| *** Jeffrey4l_ has quit IRC | 19:58 | |
| SamYaple | inc0: i have digggggggeeedd into the cluster code | 20:06 |
| SamYaple | yay! | 20:06 |
| inc0 | what do you mean? | 20:07 |
| SamYaple | well originally i was going to be dirtier about the whole thing, but instead i went the otherway. i turned the erlang into a python dict | 20:07 |
| SamYaple | well list of dicts technically | 20:07 |
| inc0 | I thought about it | 20:08 |
| inc0 | but well, it still needs regex | 20:08 |
| inc0 | or sth like that | 20:08 |
| inc0 | so let me put it this way, getting hostname or ip is not hard | 20:08 |
| SamYaple | yes and regex is what i did, but now its robust and doesnt care | 20:08 |
| SamYaple | it could be hostname or ip | 20:08 |
| SamYaple | and its accurate | 20:08 |
| inc0 | not in when | 20:08 |
| SamYaple | (cause i diggegeegeed into the code) | 20:09 |
| inc0 | that's what I wanted to know | 20:09 |
| SamYaple | in the when conditional you need to not be using ansible_hostname | 20:09 |
| SamYaple | thats not even always accurate | 20:09 |
| SamYaple | you need to be using something like {{ rabbitmq_hostname|default(inventory_hostname) }} | 20:10 |
| SamYaple | so the deployer can set rabbitmq_hostname to wherever they want, default will use whatevers in the inventory | 20:10 |
| SamYaple | which can be ip address | 20:10 |
| SamYaple | actually instead of inventory_hostname itlll be api_internal ip | 20:10 |
| SamYaple | but same thing applies | 20:10 |
| SamYaple | api_interface* | 20:10 |
| inc0 | ok, but what I mean is we need to set up rabbitmq_hostname to be somethign | 20:11 |
| SamYaple | no we dont | 20:11 |
| SamYaple | the deployer can if he wants | 20:11 |
| inc0 | changing my code to extract api and hostname is super easy | 20:12 |
| SamYaple | no its super dirty in your code | 20:12 |
| inc0 | ehh, feel free to upload your then | 20:12 |
| inc0 | regex aren't dirty | 20:12 |
| SamYaple | :P you were doing lookback regex for middle of string stuff | 20:12 |
| SamYaple | and im using regex to parse to erlang | 20:12 |
| inc0 | ok...it's still a valid regex | 20:13 |
| SamYaple | i relaly didnt want to go down the rabbit hole (heh) or erlang parsing, but thats what happened | 20:13 |
| inc0 | but by all means, upload your code | 20:13 |
| inc0 | I just grab a line with gospel in it and find hostname:) | 20:13 |
| SamYaple | so my big problem with what you had was it was actually looking for letters/numbers | 20:14 |
| SamYaple | what happens if the hostname is gospel? | 20:14 |
| SamYaple | if its ip then it doesnt work | 20:14 |
| SamYaple | lots of chance for breakage | 20:14 |
| inc0 | well, it will find first gospel | 20:14 |
| inc0 | second gospel will be used as normal string | 20:14 |
| inc0 | which meets my regex | 20:14 |
| SamYaple | first string would match too though | 20:15 |
| inc0 | can you paste the case you're concerned about plz? | 20:15 |
| inc0 | let's run test and fix if it's broken | 20:15 |
| inc0 | or just upload your code | 20:16 |
| inc0 | I'm just afraid of parsing erlang -> python because I'm fairly sure it's a complicated piece of code | 20:16 |
| inc0 | and simple is better than complex | 20:16 |
| SamYaple | sort of, its clusterer, and its not true erlang | 20:17 |
| inc0 | I know | 20:17 |
| SamYaple | this covers what clusterer _could_ output | 20:17 |
| SamYaple | but more importantly it catches on to the fact that it _doesnt_ understand the code | 20:18 |
| inc0 | mine does too | 20:18 |
| SamYaple | i know | 20:18 |
| SamYaple | your logic is sound | 20:18 |
| SamYaple | its the regex i took issue with | 20:18 |
| inc0 | regex can be changed to capture hostname or ip | 20:18 |
| inc0 | I just wanted to wait with it till we have ip-based deploy | 20:19 |
| inc0 | because even if it catches IP, when statement later on will be disfunctional | 20:19 |
| inc0 | so whole code will be disfunctional | 20:19 |
| inc0 | so if you have idea how to change this when statement so it catches both ip and hostname, I'm all ears | 20:20 |
| SamYaple | youre thinking of it wrong. i just told you how above | 20:22 |
| inc0 | rabbit_hostname is set? | 20:23 |
| inc0 | on the nodes? | 20:23 |
| inc0 | then it's easy | 20:23 |
| inc0 | and let me get to implementation | 20:23 |
| inc0 | of a regex because it's literally dead simple | 20:24 |
| SamYaple | im about to push up the code, just testing it in deploy now | 20:24 |
| SamYaple | youll see what i mean | 20:24 |
| inc0 | but please before you -1 my code because "it's not robust" find a case in which it's not robust | 20:26 |
| inc0 | because it will be simplier than what you have, I can bet on it | 20:26 |
| SamYaple | i just told you where | 20:26 |
| SamYaple | if gospel is your hostname you are screwed | 20:26 |
| inc0 | rabbit gospel rabbit@gospel gospelrabbit <- I tested with this | 20:26 |
| inc0 | worked perfectly | 20:26 |
| inc0 | did you actually try to use gospel as your hostname | 20:27 |
| inc0 | or just think that it don't work | 20:27 |
| openstackgerrit | Artur Zarzycki proposed openstack/kolla-mesos: Connect to mariadb with dnsname. https://review.openstack.org/290176 | 20:29 |
| openstackgerrit | Steven Dake proposed openstack/kolla: Reorder operations of mkdir and useradd in toolbox https://review.openstack.org/290280 | 20:30 |
| SamYaple | inc0: im guessing you didnt actually test what youre saying | 20:31 |
| SamYaple | because, you know, it doesnt work in the real world which was my point | 20:31 |
| inc0 | SamYaple, could you please give me example of output which breaks? | 20:31 |
| inc0 | so I can fix it? | 20:32 |
| inc0 | that's all I'm asking | 20:32 |
| SamYaple | dude it doesnt work, run it please | 20:32 |
| SamYaple | on real output | 20:32 |
| SamYaple | any output | 20:33 |
| SamYaple | if you have gospel in the host name you are broken | 20:33 |
| openstackgerrit | Artur Zarzycki proposed openstack/kolla-mesos: Connect to mariadb with dnsname. https://review.openstack.org/290176 | 20:33 |
| *** mbound has quit IRC | 20:34 | |
| SamYaple | but its cool, i show you my fix as soon as its done testing. it is more complicated. its also more secure because im not string searching | 20:34 |
| vhosakot | SamYaple: sdake: is kolla Liberty stable and tested with Linux Bridge.. I think yes as I was able to deploy Linux Bridge successfully.. somebody is asking me... | 20:35 |
| sdake | as long as you dont care about data loss from data containers yes | 20:35 |
| sdake | april 1 is our published deadline for fixing this problem | 20:36 |
| vhosakot | sdake: is your answer to me ? :) | 20:36 |
| inc0 | ok, I see it now, should be simple tweak | 20:36 |
| sdake | http://docs.openstack.org/developer/kolla/liberty-deployment-warning.html | 20:36 |
| sdake | vhosakot yes | 20:36 |
| vhosakot | cool, thanks for the link sdake! | 20:37 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Playbook for rabbitmq upgrade https://review.openstack.org/287969 | 20:46 |
| SamYaple | ok inc0 that was tested with ips too, but that doesnt _include_ the ips | 20:46 |
| SiRiuS_ | mlima: Are you using virtual machines for AIO deployment? And are you using virtualbox ? | 20:47 |
| inc0 | SamYaple, that I know | 20:47 |
| SamYaple | oops inc0 wrong code | 20:47 |
| *** mlima has quit IRC | 20:47 | |
| SamYaple | moment, bad rebase | 20:47 |
| inc0 | and I'm about to push code which works on both cases | 20:47 |
| SamYaple | that code works in both cases and it properly parses erland | 20:47 |
| inc0 | and gospel in hostname | 20:47 |
| SamYaple | why are you going to push new regex? | 20:48 |
| inc0 | and is super complicated | 20:48 |
| SamYaple | that one actually pythonizes it | 20:48 |
| inc0 | because it's simplier and it works? | 20:48 |
| inc0 | no it does not | 20:48 |
| SamYaple | your code was already shown to be broken! | 20:48 |
| inc0 | which I'm about to fix | 20:48 |
| SamYaple | new code might fix that one issue, but what about unforseen issues? | 20:48 |
| inc0 | by adding few characters | 20:48 |
| SamYaple | this actually parses the erlang properly | 20:48 |
| inc0 | ehh, I give up | 20:48 |
| inc0 | do whatever you like | 20:48 |
| SamYaple | normally you would be arguing from my side though | 20:49 |
| SamYaple | this is the type of thing you want since its more robust | 20:49 |
| SamYaple | i could actually trim off some of those regexs, they are just there to match the code based on clusterer source | 20:49 |
| inc0 | so as for unforseeble problems | 20:50 |
| inc0 | try it with return with less than 3 lines | 20:50 |
| inc0 | your code | 20:50 |
| SamYaple | will be perfectly fine | 20:50 |
| SamYaple | if it returns in less than three lines something went wrong anyway (which i already would have caught btw) | 20:50 |
| inc0 | try it | 20:52 |
| sdake | which erlang is being parsed preciesely? | 20:52 |
| sdake | i dont think we need to invent an erlang parser ;) | 20:52 |
| SamYaple | inc0: if its less than three lines its not valid output which means it would have already rasied an error inc0 | 20:52 |
| SamYaple | sdake: you would be wrong unfortunately | 20:52 |
| inc0 | yes | 20:52 |
| inc0 | and error is json error | 20:52 |
| inc0 | and it is not caught | 20:53 |
| SamYaple | sdake: but this is just subset which i just when through erlang for | 20:53 |
| inc0 | it will log as "no json could be parsed" | 20:53 |
| inc0 | which is not helpful | 20:53 |
| *** unicell has joined #kolla | 20:53 | |
| SamYaple | inc0: oh trust me it is cause | 20:53 |
| SamYaple | look at the line that says "if "Rabbit is running in cluster configuration" not in raw_status:" | 20:53 |
| SamYaple | you cant have valid 3 line output and get that line | 20:53 |
| *** fbarilla has quit IRC | 20:53 | |
| SamYaple | i checked the source | 20:53 |
| SamYaple | its not alot of code, you can check it too | 20:54 |
| inc0 | so it it's the only line for whatever reason | 20:54 |
| inc0 | or well...I'm not sure if clusterer uses any translations | 20:54 |
| *** unicell1 has quit IRC | 20:54 | |
| inc0 | but if it uses, your code is broken in any non-english setup | 20:54 |
| SamYaple | what are you tring to argue for here? do you want a check for len before spliting? i wont fight you there | 20:54 |
| SamYaple | thats safe | 20:55 |
| SamYaple | im trying to get more safety here man | 20:55 |
| inc0 | no, you're saying that your complicated stuff is better | 20:55 |
| sdake | hey guys guess what - rdo is going to start gating on kolla | 20:55 |
| sdake | and kolla gates on rdo | 20:55 |
| sdake | circular gating ftw! | 20:55 |
| SamYaple | im saying what you had already had a bug inc0 | 20:55 |
| inc0 | and I'm showing you it is not, you changed one potential breakage to another | 20:55 |
| inc0 | yes, which I fixed | 20:55 |
| inc0 | locally | 20:55 |
| SamYaple | i disagree | 20:55 |
| inc0 | you haven't seen the code | 20:56 |
| SamYaple | what potential breakage is there? | 20:56 |
| SamYaple | dude i trust you, but this is actually parsing erlang proper | 20:56 |
| inc0 | so let say I have polish language in clustere | 20:56 |
| SamYaple | that wouldnt be valid but go ahead | 20:56 |
| inc0 | so there is no magical sentence "Rabbit is running in cluster configuration" | 20:56 |
| SamYaple | we dont have that | 20:56 |
| inc0 | so it raises attribute error | 20:57 |
| SamYaple | we pull down a static binary and build a container though | 20:57 |
| SamYaple | that wont happen | 20:57 |
| inc0 | unless they add localization to it | 20:57 |
| inc0 | sometime in the future | 20:57 |
| SamYaple | which _if_ they do, then well deal with it at the end of a cycle | 20:57 |
| *** ccesario has joined #kolla | 20:57 | |
| SamYaple | it wont creep up on us | 20:57 |
| inc0 | you're saying that my code will break on unforseeble stuff | 20:58 |
| inc0 | and your not | 20:58 |
| inc0 | and I just proven you it might | 20:58 |
| sdake | ya double standard ftw ;) | 20:58 |
| SamYaple | im saying your code may return incorrect stuff | 20:58 |
| inc0 | and I agree...before I fixed regex | 20:58 |
| SamYaple | its only a double standard if i said that sdake | 20:58 |
| sdake | when you guys get done brawling can you post your reviews | 20:58 |
| inc0 | that's what reviews are for | 20:58 |
| inc0 | I'm done brawling, I just don't want to brawl | 20:58 |
| sdake | and let the core review team decide which is preferable | 20:58 |
| sdake | inc0 sweet type git review then and be done with it :) | 20:59 |
| inc0 | but then it will override Sams work | 20:59 |
| inc0 | which I don't want | 20:59 |
| inc0 | but hold on | 20:59 |
| inc0 | I'll do it after second thought and comment so people will also look at older code | 20:59 |
| inc0 | just let me address other reviewers comments | 21:00 |
| SamYaple | can you post your regex here real quick inc0? | 21:00 |
| SamYaple | so i can test it | 21:01 |
| inc0 | http://paste.openstack.org/show/490992/ | 21:02 |
| SamYaple | see inc0 another bug | 21:03 |
| SamYaple | it doesnt work with hostnames | 21:03 |
| sdake | looks nice and tidy - test cases will help | 21:04 |
| SamYaple | it doesnt work with ips*** works fine with hostnames | 21:04 |
| inc0 | SamYaple which ips doesn't work? | 21:05 |
| SamYaple | youre pastebin | 21:05 |
| sdake | 0-9 and . are in the regex | 21:05 |
| SamYaple | go ahead and try to run it, youll see the issue | 21:05 |
| inc0 | it will catch ip like 333.333.333.333...that's true | 21:05 |
| inc0 | byt well | 21:05 |
| SamYaple | no no | 21:05 |
| SamYaple | # python test.py | 21:06 |
| SamYaple | 192.168.0.1 | 21:06 |
| SamYaple | 192 | 21:06 |
| SamYaple | it catches multiple | 21:06 |
| SamYaple | if we parse the erlang it either parses or it doesnt | 21:06 |
| inc0 | http://paste.openstack.org/show/490995/ | 21:06 |
| vhosakot | SamYaple: isn't the "tools" directory installed when kolla is installed ? (pip install kolla) | 21:06 |
| SamYaple | inc0: im testing raw output form clusterer | 21:06 |
| vhosakot | SamYaple: I don't se tools in /usr/local/share/kolla | 21:07 |
| vhosakot | see* | 21:07 |
| sdake | vhosakot no | 21:07 |
| vhosakot | sdake: is that expected behavior ? | 21:07 |
| *** AndChat|241401 has joined #kolla | 21:08 | |
| inc0 | SamYaple, this is output I got and play with it | 21:08 |
| sdake | vhosakot yes only cmd directory is installed | 21:08 |
| *** ccesario has quit IRC | 21:08 | |
| openstackgerrit | Artur Zarzycki proposed openstack/kolla-mesos: Connect to mariadb with dnsname. https://review.openstack.org/290176 | 21:08 |
| inc0 | copied straight from clusterer | 21:08 |
| vhosakot | sdake: cool... | 21:08 |
| inc0 | what's output that breaks? | 21:08 |
| sdake | and ansible and docke directory | 21:08 |
| vhosakot | yep, I see them.. cool... just wanted to confirm that missing tools is ok | 21:09 |
| sdake | vhosakot we do install kolla-ansible | 21:10 |
| sdake | vhosakot the problem is there are things in tools related to the gate | 21:10 |
| sdake | for what reason i dont nkow | 21:10 |
| sdake | they used to bei n the test directory | 21:10 |
| sdake | but were moved | 21:10 |
| SamYaple | inc0: maybe you need to strip \r\n because the string you had isnt valid cluster output (because of those chars) | 21:10 |
| *** banix has quit IRC | 21:10 | |
| sdake | so we cant just wholesale install the tools directory | 21:10 |
| vhosakot | sdake: yep, I see kolla-build and kolla-ansible prperly installed in /usr/local/bin/ | 21:11 |
| SamYaple | yea inc0 stripping \r\n worked. thats wierd | 21:11 |
| vhosakot | sdake: right, I don't think the user needs gate scripts in tools like deploy_aio.sh and setup_gate.sh... | 21:11 |
| inc0 | well, I copypasted it | 21:11 |
| inc0 | whitespaces are bitch | 21:11 |
| SamYaple | inc0: so i mean i know you can make one that works in a chop up the string fashion, but are you saying you really think thats better than parsing the string proper? | 21:12 |
| vhosakot | inc0: I have "set tabstop=8 expandtab shiftwidth=4 softtabstop=4 " in my .vimrc so it highlights when I type a whitespece | 21:12 |
| inc0 | SamYaple, your code is magical and complicated in terms of parsing erlang to python | 21:13 |
| inc0 | so if something breaks it will be harder to pinpoint | 21:13 |
| SamYaple | its not magical, it just wraps the strings | 21:13 |
| inc0 | mine is one regex | 21:13 |
| sdake | i'd like to see sam's code | 21:13 |
| SamYaple | its really simple code though | 21:13 |
| SamYaple | https://review.openstack.org/#/c/287969/7..8/docker/rabbitmq/get_gospel_node.py | 21:13 |
| sdake | we should avoid magic tricks in the code base if possible | 21:14 |
| SamYaple | its not magic. it just wraps the strings in quotes | 21:14 |
| inc0 | https://review.openstack.org/#/c/287969/7 sdake | 21:14 |
| SamYaple | those are super super standard regexes | 21:14 |
| SamYaple | the readbehind regex you have is not standard | 21:14 |
| sdake | both seem straightforward code | 21:14 |
| sdake | i dont see why you guys are burning so many cycles over arguing about this minor point | 21:15 |
| sdake | we need to stop being so nitpicky | 21:15 |
| SamYaple | sdake: its what we do | 21:15 |
| inc0 | SamYaple, how about you rebase your code correctly (there are stuff in Dockerfile) add yourself as CoAuthor and we ack your code> | 21:16 |
| inc0 | ? | 21:16 |
| *** AndChat|241401 has quit IRC | 21:16 | |
| inc0 | or sorry | 21:16 |
| inc0 | it's not ever rebase issue | 21:16 |
| inc0 | just add yourself as coauthor | 21:16 |
| SamYaple | are you ok with rasieing atrributeerror so i can reuse that json exit? | 21:17 |
| inc0 | well, one thing tho | 21:17 |
| inc0 | if add attributeerror or jsonerror | 21:17 |
| inc0 | and change when statement | 21:18 |
| *** jtriley has quit IRC | 21:18 | |
| SamYaple | well the when statement isnt going to work ever without ips (reliably) | 21:18 |
| inc0 | so whole point of extracting gospel node is this when statement | 21:19 |
| inc0 | one way or another | 21:19 |
| SamYaple | the ansible_hostname isnt always the hostname for rabbit was my point | 21:19 |
| SamYaple | can i rebase this patch ontop of my rabbitmq ips patch? | 21:19 |
| inc0 | sure | 21:20 |
| inc0 | that's what I wanted to do anyway | 21:20 |
| openstackgerrit | Hui Kang proposed openstack/kolla: Set keepalived conf state to MASTER https://review.openstack.org/294291 | 21:23 |
| SamYaple | can someone on centos deploy AIO? | 21:28 |
| *** sdake has quit IRC | 21:28 | |
| SamYaple | Hui just submitted https://review.openstack.org/294291 where he says MASTER is needed, which can't be true. our gate is working | 21:29 |
| *** diogogmt has quit IRC | 21:30 | |
| *** sdake has joined #kolla | 21:30 | |
| *** inc0 has quit IRC | 21:33 | |
| *** ccesario has joined #kolla | 21:33 | |
| vhosakot | yep, I reviewed it.. nopreempt in keepalived needs both VIP instances to be in BACKUP state, and the master role is determined by the priority value of the VIP instances.. | 21:34 |
| vhosakot | SamYaple: so, in multinode, is the first node under the [network] group is chosen as the "high priority node" as it will have lower index ? | 21:37 |
| SamYaple | vhosakot: yes | 21:38 |
| vhosakot | SamYaple: cool, thanks | 21:39 |
| SamYaple | we try not to force an indexed node wherever possible, but we do for bootstraping | 21:39 |
| vhosakot | SamYaple: cool | 21:42 |
| *** sdake has quit IRC | 21:45 | |
| *** hongbin has joined #kolla | 21:51 | |
| *** hongbin has left #kolla | 21:51 | |
| SamYaple | well shoot. | 21:57 |
| *** asalkeld has joined #kolla | 21:57 | |
| SamYaple | ips arent going to work with all versions of rabbitmq | 21:57 |
| *** dims has quit IRC | 22:01 | |
| *** dims_ has joined #kolla | 22:01 | |
| *** haplo37 has quit IRC | 22:12 | |
| *** vishwanathj has joined #kolla | 22:16 | |
| *** vishwanathj has left #kolla | 22:16 | |
| SamYaple | on multinode rabbitmq breaks with ip addresses GAH | 22:22 |
| *** dims has joined #kolla | 22:23 | |
| vhosakot | due to rabbit version mis-match ?... isn;t the IP seen in the playbook ? | 22:23 |
| SamYaple | its erlang | 22:23 |
| *** inc0 has joined #kolla | 22:24 | |
| vhosakot | ah ok.. is the output of "rabbitmqctl eval 'rabbit_clusterer:status()" in erlang format ? | 22:25 |
| *** dims_ has quit IRC | 22:26 | |
| *** mbound has joined #kolla | 22:26 | |
| *** vhosakot has quit IRC | 22:26 | |
| *** alisonh has quit IRC | 22:31 | |
| *** spisarski has quit IRC | 22:31 | |
| sbezverk | SamYaple could you please review this trivial fix, it is missing 1 more core to be merged. https://review.openstack.org/#/c/294113/ | 22:31 |
| *** jasonsb has joined #kolla | 22:36 | |
| openstackgerrit | Merged openstack/kolla: Changing Swift option section in globals.yml https://review.openstack.org/294113 | 22:37 |
| *** gfidente has quit IRC | 22:39 | |
| *** diogogmt has joined #kolla | 22:51 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Playbook for rabbitmq upgrade https://review.openstack.org/287969 | 23:00 |
| *** alisonh has joined #kolla | 23:00 | |
| *** mbound has quit IRC | 23:02 | |
| *** inc0 has quit IRC | 23:04 | |
| *** cfarquhar has joined #kolla | 23:05 | |
| *** cfarquhar has quit IRC | 23:05 | |
| *** cfarquhar has joined #kolla | 23:05 | |
| *** dims has quit IRC | 23:08 | |
| *** blahRus has quit IRC | 23:20 | |
| *** jasonsb has quit IRC | 23:22 | |
| *** sdake has joined #kolla | 23:24 | |
| *** dims has joined #kolla | 23:28 | |
| *** banix has joined #kolla | 23:29 | |
| *** dims has quit IRC | 23:32 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla: iscsi container with lvm2 support https://review.openstack.org/291285 | 23:34 |
| *** apuimedo has quit IRC | 23:39 | |
| *** manjeets has left #kolla | 23:44 | |
| *** weiyu_ has joined #kolla | 23:47 | |
| *** dwalsh has quit IRC | 23:51 | |
| sdake | hey folks | 23:59 |
| sdake | magnum is probably just going to be doa this release | 23:59 |
| sdake | since it has a hard dependency on barbican | 23:59 |
| sdake | which i dont think we want to introduce at this late stage | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!