| *** prithiv has quit IRC | 00:05 | |
| wirehead_ | rhallisey: if I’m right and it’s just not done, let me know and I can help out. :) | 00:06 |
|---|---|---|
| *** mdnadeem has quit IRC | 00:08 | |
| *** mdnadeem has joined #openstack-kolla | 00:09 | |
| openstackgerrit | Ken Wronkiewicz proposed openstack/kolla-kubernetes: Regression: Kubernetes was rejecting the YAML we were generating https://review.openstack.org/323624 | 00:12 |
| *** inc0 has quit IRC | 00:13 | |
| *** dims has quit IRC | 00:14 | |
| *** sacharya has quit IRC | 00:15 | |
| *** dims has joined #openstack-kolla | 00:19 | |
| *** ravig has quit IRC | 00:29 | |
| *** sdake_ has joined #openstack-kolla | 00:29 | |
| *** sdake has quit IRC | 00:30 | |
| sdake_ | inc0 how about making a spec for your work related to includes and macros and stuff | 00:32 |
| *** pbourke-home has joined #openstack-kolla | 00:35 | |
| *** ravig has joined #openstack-kolla | 00:37 | |
| *** pbourke-home has quit IRC | 00:38 | |
| sdake_ | rhallisey ping | 00:39 |
| sdake_ | jpeeler around? | 00:40 |
| *** wirehead_ has left #openstack-kolla | 00:41 | |
| *** wirehead_ has joined #openstack-kolla | 00:41 | |
| *** ssurana has quit IRC | 00:41 | |
| *** dcwangmit01_ has joined #openstack-kolla | 00:48 | |
| *** dcwangmit01 has quit IRC | 00:50 | |
| *** sdake has joined #openstack-kolla | 00:56 | |
| *** sdake_ has quit IRC | 00:59 | |
| *** sdake has quit IRC | 01:02 | |
| rhallisey | wirehead_, the bootstrap for mariadb should work. What are you seeing? | 01:08 |
| rhallisey | its a little sesitive though' | 01:08 |
| *** weiyu has joined #openstack-kolla | 01:11 | |
| *** sacharya has joined #openstack-kolla | 01:15 | |
| *** rhallisey has quit IRC | 01:19 | |
| *** sacharya has quit IRC | 01:21 | |
| *** zhiwei has joined #openstack-kolla | 01:23 | |
| williamcaban | I just pull the latest version from git, is it me or the "--passwords" option in kolla-ansible is not recognized anymore? | 01:24 |
| williamcaban | Even when is displayed as part of the help, the option is not longer listed in the LONG_OPTS | 01:27 |
| wirehead_ | rhallisey: so, understanding that the mainline is broken and I’ve got a one-line patch to fix things. | 01:35 |
| *** sdake has joined #openstack-kolla | 01:41 | |
| *** sdake has quit IRC | 01:43 | |
| *** sdake has joined #openstack-kolla | 01:47 | |
| *** sdake_ has joined #openstack-kolla | 01:48 | |
| *** thumpba has joined #openstack-kolla | 01:50 | |
| *** sdake has quit IRC | 01:51 | |
| *** tfukushima has joined #openstack-kolla | 01:58 | |
| *** unicell has quit IRC | 02:02 | |
| *** sacharya has joined #openstack-kolla | 02:16 | |
| *** thumpba has quit IRC | 02:16 | |
| *** ravig has quit IRC | 02:22 | |
| *** ravig has joined #openstack-kolla | 02:28 | |
| *** pbourke-home has joined #openstack-kolla | 02:37 | |
| *** pbourke-home has quit IRC | 02:38 | |
| *** tyrola has joined #openstack-kolla | 02:40 | |
| *** tyrola has quit IRC | 02:45 | |
| *** yuanying has quit IRC | 02:50 | |
| openstackgerrit | William Caban proposed openstack/kolla: Closes-Bug: 1587706 https://review.openstack.org/323634 | 02:51 |
| openstack | bug 1587706 in kolla "kolla-ansible missing --password option" [Undecided,New] https://launchpad.net/bugs/1587706 | 02:51 |
| *** ayoung has quit IRC | 03:09 | |
| openstackgerrit | Merged openstack/kolla: Install missing package in spiceproxy container https://review.openstack.org/323054 | 03:16 |
| *** daneyon_ has joined #openstack-kolla | 03:22 | |
| *** daneyon_ has quit IRC | 03:27 | |
| *** ravig has quit IRC | 03:34 | |
| *** ravig has joined #openstack-kolla | 03:37 | |
| *** pbourke-home has joined #openstack-kolla | 03:38 | |
| *** pbourke-home has quit IRC | 03:39 | |
| *** yuanying has joined #openstack-kolla | 03:47 | |
| *** dave-mccowan has quit IRC | 03:48 | |
| *** wirehead_ has quit IRC | 03:50 | |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed openstack/kolla: Install missing package in spiceproxy container https://review.openstack.org/323650 | 03:51 |
| *** wirehead_ has joined #openstack-kolla | 03:52 | |
| *** g3ek has joined #openstack-kolla | 03:57 | |
| *** haplo37 has joined #openstack-kolla | 03:57 | |
| *** ravig has quit IRC | 04:09 | |
| *** david-lyle has quit IRC | 04:21 | |
| *** ssurana has joined #openstack-kolla | 04:48 | |
| *** ssurana has left #openstack-kolla | 04:49 | |
| *** tfukushima has quit IRC | 04:53 | |
| sdake_ | yo | 04:53 |
| sdake_ | Jeffrey4l_ ping | 04:53 |
| Jeffrey4l_ | sdake_, pong | 04:53 |
| *** ssurana has joined #openstack-kolla | 04:54 | |
| sdake_ | coolsvap ping | 04:54 |
| openstackgerrit | Merged openstack/kolla: Add passwords parameter to LONG_OPTS in kolla-ansible script https://review.openstack.org/323585 | 04:56 |
| sdake_ | https://review.openstack.org/#/c/323612 | 04:56 |
| sdake_ | Jeffrey4l_ can you please revview this | 04:56 |
| sdake_ | inc0 wants a spec, and I am blocked on elemental support until the spec is either approved or not approved | 04:56 |
| sdake_ | remember, requires majority | 04:56 |
| Jeffrey4l_ | ok. actually, i am not big fan for this idea. :) | 04:57 |
| coolsvap | sdake_, I will review | 04:57 |
| sdake_ | Jeffrey4l_ what is your proposal ? | 04:58 |
| coolsvap | i am open to experiment with this as we discussed at the start, this looks to solve some major headache's down the line | 04:58 |
| sdake_ | Jeffrey4l_ before we get into your propoosal can you explain why you dont like the approach | 04:59 |
| Jeffrey4l_ | hmm.. let's me think.. i will push something in the ml. | 04:59 |
| sdake_ | inc0 has somethign with include files and jinja2 | 04:59 |
| sdake_ | which I think will require approxximately 1800 top level include files be littered all over the repository | 05:00 |
| sdake_ | https://review.openstack.org/#/c/323589/ | 05:02 |
| patchbot | sdake_: patch 323589 - kolla - Refactor of base Dockerfile to allow custom repo | 05:02 |
| sdake_ | Jeffrey4l_ have a review of that | 05:02 |
| Jeffrey4l_ | ok | 05:02 |
| sdake_ | Jeffrey4l_ i'd appreciate you review on the reviews rather then the ml | 05:02 |
| Jeffrey4l_ | roger that | 05:02 |
| Jeffrey4l_ | i will | 05:02 |
| sdake_ | i'd still like to know what you dont like about the idea | 05:03 |
| sdake_ | in 6 months when dockerfile.j2 are in a total charlie foxtrot state, I won't be re-proposing this idea | 05:03 |
| sdake_ | now or never, so i wnat to get the complaints out on the table | 05:04 |
| sdake_ | so eveeyrone has the correct information to work with | 05:04 |
| sdake_ | nihilifer ping | 05:05 |
| Jeffrey4l_ | "charlie foxtrot state" means? ( sorry for no idea for this.) | 05:05 |
| sdake_ | cluster fuck state | 05:05 |
| Jeffrey4l_ | ok | 05:06 |
| sdake_ | so what are the issues with the approach then? :) | 05:06 |
| Jeffrey4l_ | i need some time for this.. and i will comment what i think in your PS. ok? | 05:09 |
| sdake_ | ok | 05:12 |
| sdake_ | try to do today please | 05:13 |
| openstackgerrit | Merged openstack/kolla: Pin the base distro release version https://review.openstack.org/319992 | 05:13 |
| Jeffrey4l_ | np | 05:13 |
| nihilifer | sdake_: pong | 05:15 |
| sdake_ | nihilifer can you review a spec please | 05:15 |
| nihilifer | sure | 05:15 |
| sdake_ | https://review.openstack.org/#/c/323612/ | 05:16 |
| patchbot | sdake_: patch 323612 - kolla - Introduction of a new DSL: Elemental | 05:16 |
| openstackgerrit | Merged openstack/kolla: Add sahara containers https://review.openstack.org/317261 | 05:16 |
| sdake_ | nihilifer any chance you can do it now :) | 05:22 |
| nihilifer | yes, i'm looking right now | 05:22 |
| sdake_ | thx | 05:23 |
| *** phuongnh has joined #openstack-kolla | 05:23 | |
| openstackgerrit | Steven Dake proposed openstack/kolla: Introduction of a new DSL: Elemental https://review.openstack.org/323612 | 05:29 |
| openstackgerrit | Steven Dake proposed openstack/kolla: Introduction of a new DSL: Elemental https://review.openstack.org/323612 | 05:30 |
| sdake_ | nihilifer ok i'm done making changes for a short bit :) | 05:30 |
| *** reddydodda has joined #openstack-kolla | 05:30 | |
| openstackgerrit | Steven Dake proposed openstack/kolla: Introduction of a new DSL: Elemental https://review.openstack.org/323612 | 05:37 |
| *** pbourke-home has joined #openstack-kolla | 05:39 | |
| *** pbourke-home has quit IRC | 05:40 | |
| sdake_ | coolsvap can you review https://review.openstack.org/#/c/323612 please | 05:41 |
| sdake_ | also please vote one way or another | 05:42 |
| sdake_ | nihilifer ^^ :) | 05:42 |
| sdake_ | every spec revision that is reviewed should always hae a vote associated ith it, so the spec author knows if he is headed in the right or wrong direction | 05:42 |
| *** sdake has joined #openstack-kolla | 05:52 | |
| *** sdake_ has quit IRC | 05:55 | |
| *** coolsvap_ has joined #openstack-kolla | 06:02 | |
| *** unicell has joined #openstack-kolla | 06:03 | |
| *** daneyon_ has joined #openstack-kolla | 06:05 | |
| *** daneyon_ has quit IRC | 06:10 | |
| *** sdake has quit IRC | 06:15 | |
| *** unicell1 has joined #openstack-kolla | 06:20 | |
| *** unicell has quit IRC | 06:20 | |
| *** berendt has joined #openstack-kolla | 06:26 | |
| *** jmccarthy1 has joined #openstack-kolla | 06:38 | |
| *** tyrola has joined #openstack-kolla | 06:41 | |
| *** tyrola has quit IRC | 06:45 | |
| *** tfukushima has joined #openstack-kolla | 06:52 | |
| *** cu5 has joined #openstack-kolla | 06:59 | |
| *** Serlex has joined #openstack-kolla | 07:02 | |
| *** slagle has quit IRC | 07:06 | |
| *** Mr_Broken has joined #openstack-kolla | 07:10 | |
| *** b_bezak has joined #openstack-kolla | 07:10 | |
| *** mikelk has joined #openstack-kolla | 07:21 | |
| *** matrohon has joined #openstack-kolla | 07:29 | |
| Mech422 | pbourke_: Morning :-) | 07:34 |
| Mech422 | pbourke_: if you happen to see sdake_ - could you let him know I think I'm on track with the bonds/vlan stuff... | 07:34 |
| Mech422 | pbourke_: it appears that you can just use a veth pair, and hand 1 end of it to kolla ... | 07:35 |
| Mech422 | and *presto* its happy :) | 07:35 |
| *** pbourke-home has joined #openstack-kolla | 07:39 | |
| *** pbourke-home has quit IRC | 07:40 | |
| *** ssurana has quit IRC | 07:50 | |
| *** fragatin_ has joined #openstack-kolla | 08:03 | |
| *** prithiv has joined #openstack-kolla | 08:05 | |
| *** coolsvap_ has quit IRC | 08:05 | |
| *** fragatina has quit IRC | 08:07 | |
| *** nilk has joined #openstack-kolla | 08:07 | |
| *** fragatin_ has quit IRC | 08:07 | |
| *** fragatina has joined #openstack-kolla | 08:08 | |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed openstack/kolla: Introduction of a new DSL: Elemental https://review.openstack.org/323612 | 08:09 |
| *** cristicalin has joined #openstack-kolla | 08:10 | |
| openstackgerrit | Zhijiang Hu proposed openstack/kolla: Check avahi-daemon is not running when using haproxy in prechecks https://review.openstack.org/323715 | 08:13 |
| *** efoley has joined #openstack-kolla | 08:17 | |
| *** sgordon has quit IRC | 08:19 | |
| *** tyrola has joined #openstack-kolla | 08:19 | |
| *** tyrola has quit IRC | 08:20 | |
| *** mkoderer has quit IRC | 08:22 | |
| *** sgordon has joined #openstack-kolla | 08:22 | |
| *** lmiccini has quit IRC | 08:22 | |
| *** mkoderer has joined #openstack-kolla | 08:26 | |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed openstack/kolla: Add support for mistral ubuntu binary containers https://review.openstack.org/317240 | 08:30 |
| *** pbourke-home has joined #openstack-kolla | 08:30 | |
| *** lmiccini has joined #openstack-kolla | 08:31 | |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed openstack/kolla: Updated Gnocchi binary containers https://review.openstack.org/317256 | 08:31 |
| *** jmccarthy has quit IRC | 08:31 | |
| *** jmccarthy has joined #openstack-kolla | 08:31 | |
| openstackgerrit | Merged openstack/kolla: Don't always unpack image build results tuple, as it may be None https://review.openstack.org/323505 | 08:35 |
| *** wmiller has joined #openstack-kolla | 08:36 | |
| *** dmk0202 has joined #openstack-kolla | 08:38 | |
| *** Mr_Broken has quit IRC | 08:47 | |
| *** daneyon_ has joined #openstack-kolla | 08:47 | |
| *** Mr_Broken has joined #openstack-kolla | 08:48 | |
| *** dmk0202 has quit IRC | 08:48 | |
| *** daneyon_ has quit IRC | 08:52 | |
| *** sacharya has quit IRC | 08:57 | |
| *** Lyncos has quit IRC | 09:01 | |
| *** tfukushima has quit IRC | 09:04 | |
| *** Lyncos has joined #openstack-kolla | 09:05 | |
| *** tfukushima has joined #openstack-kolla | 09:05 | |
| *** tfukushima has quit IRC | 09:07 | |
| *** fragatin_ has joined #openstack-kolla | 09:11 | |
| *** Mr_Broken has quit IRC | 09:13 | |
| *** Mr_Broken has joined #openstack-kolla | 09:13 | |
| *** fragatina has quit IRC | 09:15 | |
| *** fragatin_ has quit IRC | 09:16 | |
| *** Mr_Broken has quit IRC | 09:18 | |
| *** Mr_Broken has joined #openstack-kolla | 09:25 | |
| *** fragatina has joined #openstack-kolla | 09:27 | |
| *** fragatina has quit IRC | 09:31 | |
| *** efoley_ has joined #openstack-kolla | 09:37 | |
| *** efoley has quit IRC | 09:38 | |
| *** reddydodda has quit IRC | 09:38 | |
| *** SiRiuS has joined #openstack-kolla | 09:40 | |
| *** cristicalin has quit IRC | 09:46 | |
| *** matrohon has quit IRC | 09:47 | |
| *** Jeffrey4l_ has quit IRC | 09:48 | |
| *** mdnadeem has joined #openstack-kolla | 09:51 | |
| *** reddydodda has joined #openstack-kolla | 09:51 | |
| *** mdnadeem has left #openstack-kolla | 09:55 | |
| *** mdnadeem has joined #openstack-kolla | 09:55 | |
| *** sacharya has joined #openstack-kolla | 09:57 | |
| *** nilk has quit IRC | 10:00 | |
| *** rhallisey has joined #openstack-kolla | 10:02 | |
| *** efoley__ has joined #openstack-kolla | 10:02 | |
| *** sacharya has quit IRC | 10:03 | |
| *** efoley_ has quit IRC | 10:05 | |
| *** tyrola has joined #openstack-kolla | 10:05 | |
| *** nilk has joined #openstack-kolla | 10:07 | |
| *** mdnadeem has quit IRC | 10:07 | |
| *** mdnadeem has joined #openstack-kolla | 10:07 | |
| *** tfukushima has joined #openstack-kolla | 10:07 | |
| *** pbourke_ has quit IRC | 10:09 | |
| *** pbourke_ has joined #openstack-kolla | 10:10 | |
| *** tfukushima has quit IRC | 10:12 | |
| openstackgerrit | Christian Berendt proposed openstack/kolla: Load the metadata definitions during Glance bootstrap https://review.openstack.org/322415 | 10:12 |
| openstackgerrit | Md Nadeem proposed openstack/kolla-kubernetes: WIP: Heat services and pod https://review.openstack.org/316850 | 10:12 |
| pbourke-home | Mech422: hey just saw your messages in the logs | 10:18 |
| pbourke-home | Mech422: that's great news :) maybe you could add your learnings into a section in our advanced configuation docs? | 10:18 |
| *** zhiwei has quit IRC | 10:20 | |
| *** Mr_Broken has quit IRC | 10:21 | |
| *** zhiwei has joined #openstack-kolla | 10:21 | |
| *** reddydodda has quit IRC | 10:24 | |
| *** reddydodda has joined #openstack-kolla | 10:25 | |
| *** SiRiuS has quit IRC | 10:25 | |
| Mech422 | pbourke-home: I'm still getting it to deploy - but it passes the prechecks... | 10:30 |
| Mech422 | pbourke-home: how/where are the advanced configuration docs ? | 10:31 |
| pbourke-home | http://docs.openstack.org/developer/kolla/advanced-configuration.html | 10:31 |
| *** Mr_Broken has joined #openstack-kolla | 10:32 | |
| *** daneyon_ has joined #openstack-kolla | 10:36 | |
| *** reddydodda has quit IRC | 10:38 | |
| *** tfukushima has joined #openstack-kolla | 10:38 | |
| *** daneyon_ has quit IRC | 10:40 | |
| *** SiRiuS has joined #openstack-kolla | 10:42 | |
| *** tfukushima has quit IRC | 10:43 | |
| *** weiyu has quit IRC | 10:47 | |
| *** prithiv has quit IRC | 10:50 | |
| *** prithiv has joined #openstack-kolla | 11:02 | |
| Mech422 | pbourke-home: http://pastebin.com/wfMEwTVC | 11:06 |
| *** mdnadeem has quit IRC | 11:06 | |
| *** mdnadeem has joined #openstack-kolla | 11:07 | |
| pbourke-home | Mech422: nice job! | 11:08 |
| pbourke-home | Mech422: are you familiar with how to submit a change via gerrit? | 11:08 |
| Mech422 | pbourke-home: I suck at writing - but the bash commands are there for a better writer...Umm - no... | 11:08 |
| Mech422 | I've never actually used gerrit | 11:09 |
| pbourke-home | http://docs.openstack.org/developer/kolla/CONTRIBUTING.html | 11:09 |
| *** godleon has quit IRC | 11:09 | |
| pbourke-home | let me know if you need more info | 11:09 |
| Mech422 | pbourke-home: ok - btw, I haven't actually got a cluster up with this yet (still configuring stuff) - just basing this on the fact the pre-checks pass | 11:10 |
| Mech422 | pbourke-home: I actually setup 'gen', 'externl', 'ceph', 'cephrep', and 'db'/'dbrep' VLANS... | 11:12 |
| *** slagle has joined #openstack-kolla | 11:13 | |
| Mech422 | pbourke-home: I thought I could override the 'internal'/'external' networks further down in the config, like you can customize placement of containers | 11:13 |
| pbourke-home | you mean different nics for different hosts? | 11:13 |
| Mech422 | pbourke-home: well - like 'ceph_net = "${{ internal_network }}"' | 11:14 |
| Mech422 | so ceph defaults to using the internal network, but you can provide a different value if you want | 11:15 |
| Mech422 | pbourke-home: ahh - here- from globals.yml: | 11:16 |
| *** reddydodda has joined #openstack-kolla | 11:16 | |
| Mech422 | # These can be adjusted for even more customization. The default is the same as | 11:16 |
| Mech422 | # the 'network_interface'. These interfaces must container an IPv4 address. | 11:16 |
| Mech422 | #kolla_external_vip_interface: "{{ network_interface }}" | 11:16 |
| Mech422 | #api_interface: "{{ network_interface }}" | 11:16 |
| Mech422 | #storage_interface: "{{ network_interface }}" | 11:16 |
| Mech422 | #cluster_interface: "{{ network_interface }}" | 11:16 |
| Mech422 | #tunnel_interface: "{{ network_interface }}" | 11:16 |
| pbourke-home | hmm yeah but you can't go any more fine grained | 11:17 |
| pbourke-home | currently | 11:17 |
| Mech422 | pbourke-home: ahh - still - thats a good start :-) | 11:17 |
| Mech422 | I got confused last night and thought you mean I only got 'internal'/'external' with no option to override | 11:17 |
| Mech422 | this is a nice breakdown :-) | 11:17 |
| *** mliima has joined #openstack-kolla | 11:24 | |
| *** dave-mccowan has joined #openstack-kolla | 11:27 | |
| mliima | morning guys | 11:33 |
| coolsvap | morning mlima | 11:33 |
| *** mdnadeem has quit IRC | 11:36 | |
| *** mdnadeem has joined #openstack-kolla | 11:37 | |
| *** tfukushima has joined #openstack-kolla | 11:39 | |
| *** reddydodda has quit IRC | 11:40 | |
| *** mdnadeem has quit IRC | 11:40 | |
| *** mdnadeem has joined #openstack-kolla | 11:41 | |
| *** mdnadeem has quit IRC | 11:41 | |
| *** mdnadeem has joined #openstack-kolla | 11:42 | |
| *** mdnadeem has quit IRC | 11:42 | |
| *** mdnadeem has joined #openstack-kolla | 11:42 | |
| *** mdnadeem has quit IRC | 11:42 | |
| *** reddydodda has joined #openstack-kolla | 11:43 | |
| *** mdnadeem has joined #openstack-kolla | 11:43 | |
| *** tfukushima has quit IRC | 11:44 | |
| Mech422 | grrr...openvswitch really should be part of the 'networking' task or something in Ubuntu | 11:46 |
| Mech422 | I think I'm getting weird races because Docker is firing off before I've fully configured ovs | 11:46 |
| *** Mr_Broken has quit IRC | 11:50 | |
| *** mdnadeem has joined #openstack-kolla | 11:53 | |
| *** sacharya has joined #openstack-kolla | 11:54 | |
| *** phuongnh has quit IRC | 11:56 | |
| *** dwalsh has joined #openstack-kolla | 11:58 | |
| *** mdnadeem has quit IRC | 11:59 | |
| *** sacharya has quit IRC | 11:59 | |
| *** mdnadeem has joined #openstack-kolla | 11:59 | |
| *** mdnadeem has quit IRC | 12:03 | |
| *** mdnadeem has joined #openstack-kolla | 12:04 | |
| *** mdnadeem has quit IRC | 12:05 | |
| *** mdnadeem has joined #openstack-kolla | 12:06 | |
| *** reddydodda has quit IRC | 12:08 | |
| *** efoley__ has quit IRC | 12:09 | |
| *** efoley__ has joined #openstack-kolla | 12:10 | |
| pbourke-home | sbezverk: ping | 12:12 |
| openstackgerrit | Paul Bourke proposed openstack/kolla: Fix typos in cinder-guide.rst https://review.openstack.org/323827 | 12:13 |
| *** efoley__ has quit IRC | 12:14 | |
| *** openstackgerrit has quit IRC | 12:18 | |
| *** openstackgerrit has joined #openstack-kolla | 12:18 | |
| *** reddydodda has joined #openstack-kolla | 12:19 | |
| *** JoseMello has joined #openstack-kolla | 12:21 | |
| *** tfukushima has joined #openstack-kolla | 12:22 | |
| *** tfukushima has quit IRC | 12:22 | |
| *** tfukushima has joined #openstack-kolla | 12:22 | |
| coolsvap | pbourke-home, mliima can you review patch 315407 | 12:23 |
| patchbot | coolsvap: https://review.openstack.org/#/c/315407/ - kolla - Update Cinder guide | 12:23 |
| coolsvap | pbourke-home, i think i am updating the changes for patch 323827 in the patch set you just submitted i will check and let you know | 12:24 |
| patchbot | coolsvap: https://review.openstack.org/#/c/323827/ - kolla - Fix typos in cinder-guide.rst | 12:24 |
| *** Mr_Broken has joined #openstack-kolla | 12:26 | |
| pbourke-home | coolsvap: "As of Newton-1 milestone, Kolla supports LVM2 as cinder backend." | 12:28 |
| pbourke-home | what did we use prior to this? | 12:28 |
| coolsvap | pbourke-home, there was ceph and LVM did not had support for iscsid and tgtd | 12:31 |
| pbourke-home | how did lvm work without iscsi | 12:32 |
| coolsvap | pbourke-home, i think it was there but not functional | 12:34 |
| coolsvap | scsi-target-utils package was installed in cinder-volume from a long time | 12:35 |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed openstack/kolla: Update Cinder guide https://review.openstack.org/315407 | 12:38 |
| coolsvap | pbourke-home, mliima ^^ | 12:38 |
| pbourke-home | coolsvap: another question | 12:44 |
| pbourke-home | whats the deal with cinder_iscsi_ip_address | 12:44 |
| pbourke-home | does cinder_scheduler not choose this | 12:44 |
| *** ppowell has joined #openstack-kolla | 12:45 | |
| coolsvap | pbourke-home, no | 12:45 |
| coolsvap | its the operator configuration | 12:45 |
| coolsvap | iscsi ip can be the same system | 12:45 |
| pbourke-home | what if you have multiple storage nodes | 12:46 |
| coolsvap | as long as you are having single block storage provider its managed by cinder plugin | 12:46 |
| pbourke-home | that's not HA | 12:47 |
| coolsvap | because most cinder plugin use the api provided by block storage device | 12:47 |
| coolsvap | they do not directly interact with the device | 12:47 |
| mliima | coolsvap, done :) | 12:47 |
| sbezverk | pbourke-home in this case you need to configure multiple lvm section in cinder.conf | 12:48 |
| sbezverk | pbourke-home also very important is NOT use virtual ip address | 12:48 |
| sbezverk | in cinder.conf for lvm backend | 12:48 |
| pbourke-home | ah I see | 12:49 |
| sbezverk | pbourke-home in normal opensatck there is no HA for this backend either so we are not makeing anything worse than already is | 12:50 |
| pbourke-home | sbezverk: sure, Im just trying to figure out how some of this hangs together | 12:51 |
| *** dave-mccowan has quit IRC | 12:52 | |
| pbourke-home | friend is telling me that cinder_scheduler can see existing cinder_volumes without needing addtional sections | 12:52 |
| pbourke-home | sbezverk: ^ | 12:53 |
| pbourke-home | does make sense? | 12:53 |
| *** Jeffrey4l has joined #openstack-kolla | 12:53 | |
| coolsvap | pbourke-home, i think it cannot afaik | 12:53 |
| * coolsvap not sure of any recent updates | 12:54 | |
| pbourke-home | :/ | 12:54 |
| coolsvap | pbourke-home, i might be bit lagging on cinder updates | 12:54 |
| sbezverk | pbourke-home to create a volume you do not need much, so it might work | 12:54 |
| sbezverk | pbourke-home the trick we accomplished was to allow nova-compute to attach a volume to the instance, here you absolutely need iscsi/tgtd | 12:55 |
| pbourke-home | yes | 12:55 |
| *** stvnoyes has quit IRC | 13:03 | |
| *** stvnoyes has joined #openstack-kolla | 13:04 | |
| *** dave-mccowan has joined #openstack-kolla | 13:07 | |
| *** sdake has joined #openstack-kolla | 13:09 | |
| coolsvap | pbourke-home, i did not find any documentation similar to what you mentioned i ll dig around and let you know if i find something | 13:09 |
| pbourke-home | coolsvap: thanks I'm also digging around | 13:10 |
| *** b_bezak has quit IRC | 13:11 | |
| *** ayoung has joined #openstack-kolla | 13:12 | |
| *** ravig has joined #openstack-kolla | 13:17 | |
| *** inc0 has joined #openstack-kolla | 13:17 | |
| inc0 | good morning | 13:18 |
| *** daneyon_ has joined #openstack-kolla | 13:18 | |
| coolsvap | inc0, good morning | 13:21 |
| berendt | inc0 morning. where are you from? it is nearly end of business here :) | 13:21 |
| inc0 | berendt, Texas now;) | 13:21 |
| *** daneyon_ has quit IRC | 13:23 | |
| *** SiRiuS has quit IRC | 13:29 | |
| *** jtriley has joined #openstack-kolla | 13:32 | |
| *** cu5 has quit IRC | 13:32 | |
| *** sdake has quit IRC | 13:34 | |
| *** rmart04 has joined #openstack-kolla | 13:37 | |
| *** tfukushima has quit IRC | 13:37 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: DO NOT MERGE: TEST MASTER BRANCH https://review.openstack.org/294906 | 13:43 |
| *** mgoddard_ has joined #openstack-kolla | 13:44 | |
| *** mgoddard has quit IRC | 13:47 | |
| nilk | what is the point of doing a "Prepare volumes list" and setting a fact instead of directly assigning? | 13:48 |
| inc0 | nilk, which role? I don't recall this task | 13:51 |
| nilk | nova_compute for example | 13:52 |
| nilk | with the reject("equalto", "") filter | 13:52 |
| nilk | prevent errors if a variable is empty= | 13:52 |
| *** sdake has joined #openstack-kolla | 13:55 | |
| inc0 | hold on, let me take a look | 13:55 |
| *** sacharya has joined #openstack-kolla | 13:56 | |
| sdake | morning | 13:56 |
| inc0 | ahh so prepare volume list is a hacky way to make conditional volumes in ansible | 13:56 |
| coolsvap | morning sdake | 13:56 |
| inc0 | the reject(equalto, "") won't silence errors - it will remove empty string from list | 13:57 |
| nilk | I see | 13:57 |
| inc0 | unfortunately we couldnt easily put "if" in volume list | 13:58 |
| inc0 | hey sdake | 13:58 |
| sdake | hey inc0 | 13:58 |
| inc0 | nilk, https://github.com/openstack/kolla/blob/master/ansible/roles/nova/tasks/start_compute.yml#L31 that's the reason it exists | 13:58 |
| nilk | n | 13:59 |
| nilk | ah I see | 13:59 |
| sdake | inc0 fwiw you can't -2 a spec | 13:59 |
| nilk | the Docker module would fail on the empty string | 14:00 |
| sdake | you can -1, or +2, or even +1 although that has no impact | 14:00 |
| inc0 | I can change to -1 and comment that it's equivalent to -2 in my mind | 14:00 |
| *** sacharya has quit IRC | 14:00 | |
| sdake | I can explain why if you like, but majority votes are veto free | 14:00 |
| sdake | let me expalin why majority votes are veto free | 14:00 |
| sdake | i think tha twill help | 14:00 |
| prithiv | is anyone still facing issue with ansible 2.1.0 like this http://paste.openstack.org/show/506958/ | 14:01 |
| sdake | a majority vote is the last line of tools we have in the community to resolve conflict | 14:01 |
| sdake | if someone can veto that, then they can keep the conflict | 14:01 |
| nilk | another question | 14:01 |
| nilk | if I modify a single Docker image | 14:01 |
| sdake | if there is not a majority vote achieved on any spec for this problem space, then the work just wont be done | 14:02 |
| sdake | it is that simple | 14:02 |
| nilk | is there a built-in way to redeploy it or do I manually remove the images? | 14:02 |
| sdake | inc0 also can you put up a spec please for your proposal | 14:02 |
| *** reddydodda has quit IRC | 14:03 | |
| inc0 | it's an alternative to yours | 14:04 |
| inc0 | but I need to work more on ideas and get some input | 14:04 |
| inc0 | for now I'll do lots of PoC, same way as you did | 14:04 |
| sdake | nihilifer had an idea as well | 14:05 |
| sdake | in one of his reviews of m work | 14:05 |
| sdake | he added in jinja2 code | 14:05 |
| sdake | and I thought "thats interesting, he thought we were using jinja2 to parse the data" | 14:05 |
| *** reddydodda has joined #openstack-kolla | 14:05 | |
| sdake | so there is somewhere in the middle where jinja2 holds the metadata and has a generic parser | 14:05 |
| sdake | because really mmy python could be written in jinja | 14:05 |
| inc0 | well, not really | 14:06 |
| inc0 | I mean if we make jinja2 we keep jinja syntax | 14:06 |
| inc0 | not yaml | 14:06 |
| inc0 | and it is kinda DSL, only one we already use and is fully functional and documented | 14:06 |
| sdake | oh right | 14:06 |
| sdake | i just woke up | 14:06 |
| sdake | anyway mashup - might be an idea ;) | 14:06 |
| inc0 | well, I'm confident we can solve all the problems you mentioned with jinja | 14:07 |
| inc0 | and it won't be ugly jinja | 14:07 |
| sdake | pbourke-home around? | 14:07 |
| sdake | jpeeler mandre ping | 14:07 |
| nilk | as an operator, I really like that I can easily see and modify the dockerfiles | 14:08 |
| mandre | hey sdake | 14:08 |
| jpeeler | hi | 14:08 |
| nilk | actually, I'm doing that right now | 14:08 |
| sdake | mandre long time no chat | 14:08 |
| sdake | hey eff | 14:08 |
| sdake | can you guys review this specification: | 14:08 |
| inc0 | nilk, yeah, I expect lots of people to do it | 14:08 |
| mandre | yeah sdake, been busy :( | 14:08 |
| inc0 | nilk, what do you change there | 14:08 |
| pbourke-home | hey sdake | 14:08 |
| inc0 | ? | 14:08 |
| sdake | https://review.openstack.org/#/c/323612 | 14:08 |
| sdake | i'd appreciate a vote one way or another - if it needs further attention i'd like to get it done asap | 14:09 |
| nilk | modify Cinder to use a commercial storage system | 14:09 |
| inc0 | yeah, drivers are one of priorities for this cycle | 14:09 |
| coolsvap | inc0, from what we have seen in the first draft of mandre's patch it created a lot of different paths for different set of plugins | 14:09 |
| inc0 | one way or another, but in N you'll get to do that without modyfying code | 14:09 |
| mandre | hmm, at first glance, I'm not super hot on the idea of the dsl | 14:10 |
| *** SiRiuS has joined #openstack-kolla | 14:10 | |
| sdake | mandre i wasn't either - you probbaly saw my post to that mirantis cat | 14:11 |
| sdake | mandre its the only viable mechanism i see to do plugins | 14:11 |
| *** thumpba has joined #openstack-kolla | 14:11 | |
| sdake | pbourke-home did some magic with neutron plguins | 14:12 |
| inc0 | not true sdake, we can make it work in number of ways | 14:12 |
| sdake | but tbh its kind of bolted on | 14:12 |
| mandre | sdake: not yet, I've been living under a rock for quite some time | 14:12 |
| coolsvap | extensible and operator friendly solution is what i am after, jinja did appeal to me during our summit discussion, but the dsl looks more promising to me, yeah its more effort but its more promising | 14:12 |
| mandre | actually, I disagree on your assertion about the dsl being the only viable mechanism | 14:12 |
| inc0 | coolsvap, why it's more promising? | 14:13 |
| mandre | I'm quite pleased with the PoC I posted | 14:13 |
| sdake | rhallisey inc0 i have a dr appointment and have to skip our meeting this mroning | 14:13 |
| sdake | can one of you run it please? | 14:13 |
| inc0 | uhh so we won't discuss it on the meeting? | 14:13 |
| inc0 | I can run it, but we need to have this discussion | 14:13 |
| sdake | mandre have a link? | 14:13 |
| sdake | mandre inc0 is also working on a poc | 14:13 |
| inc0 | well, not exactly plugins | 14:13 |
| coolsvap | inc0, unified extension | 14:14 |
| inc0 | but extensible jinja2 at large | 14:14 |
| inc0 | coolsvap, unified as one file? | 14:14 |
| mandre | sdake: I think you looked at it | 14:14 |
| sdake | mandre i have not | 14:14 |
| mandre | sdake: https://review.openstack.org/#/c/315027/ | 14:14 |
| patchbot | mandre: patch 315027 - kolla - [WIP] Add support for third party plugins | 14:14 |
| coolsvap | single format | 14:14 |
| coolsvap | which I would like to compare both from the jinja2 and yaml | 14:15 |
| nilk | I've just been inserting jinja2 conditionals in the Dockerfile for my plugin | 14:15 |
| inc0 | coolsvap, yes, we need this comparison, so let's trade ideas | 14:15 |
| nilk | works pretty well | 14:15 |
| inc0 | nilk, yup, and it's much more powerful than that | 14:16 |
| nilk | but it might get ugly if there are ten more of them | 14:16 |
| inc0 | then you stop using conditionals and move to more advanced structures | 14:16 |
| coolsvap | inc0, yes, the way sdake described i could imagine how it can be extended, TBH I am not able to do the same with jinja2 that easily | 14:16 |
| inc0 | coolsvap, which part you're not able to do? | 14:17 |
| sdake | mandre this question has been asked and answered via vote | 14:17 |
| sdake | To answer your question, I suppose the containers for 3rd party services such as midonet-api could live in the kolla repo with the condition that the code is open source and we can build the images freely, though we haven't talked about it specifically in the core team yet. | 14:17 |
| inc0 | mandre, ad your idea, instead of macro I was thinking about block | 14:17 |
| sdake | mandre your idea looks like it works - what is the special magic text required? | 14:18 |
| sdake | I don't like allt he dropping we have all over the build.py for things unrelated to building :( | 14:18 |
| *** berendt has quit IRC | 14:18 | |
| inc0 | one way to solve complexity of build.py is to use less flags and more config | 14:19 |
| inc0 | configs can be huge, that's acceptable if it's huge and well documented | 14:19 |
| sdake | ok,, i'll stop harping on build.py | 14:19 |
| sdake | I think we all kknow and agree there is a problem there that needs a refactor when time permits | 14:20 |
| rhallisey | inc0, can you run the meeting. I'll be in 2x meetings | 14:20 |
| inc0 | sure | 14:21 |
| sdake | inc0 spend 20 minutes on midcycle brainstorming - please put in etherpad | 14:21 |
| mandre | sdake: basically you invoke a macro to insert snippets of dockerfile | 14:21 |
| sdake | mandre i see how the iplementation works | 14:22 |
| sdake | can you tell me what the ucustom snippts are that you put | 14:22 |
| sdake | i am curious about user experience | 14:22 |
| inc0 | https://review.openstack.org/#/c/315027/4/docker/plugins/plugins_helper.j2 it's all ther | 14:22 |
| patchbot | inc0: patch 315027 - kolla - [WIP] Add support for third party plugins | 14:22 |
| inc0 | e | 14:22 |
| sdake | inc0 mandre said something must be added to build-kolla.conf | 14:22 |
| sdake | but then doesns't say what :) | 14:23 |
| mandre | this file is purely for developers, no user need to look at it | 14:23 |
| sdake | kolla_build.conf? | 14:23 |
| sdake | operators need to enable and disable plugins | 14:23 |
| sdake | mandre what i'm asking is how do i turne it on | 14:23 |
| inc0 | mandre, https://review.openstack.org/#/c/323589/3/docker/base/custom-dockerfile.j2 line 9 | 14:23 |
| patchbot | inc0: patch 323589 - kolla - Refactor of base Dockerfile to allow custom repo | 14:23 |
| mandre | so sdake, you have to enable plugins in kolla_build.conf | 14:23 |
| inc0 | can do the same | 14:23 |
| inc0 | you put empty blocks in key places of a dockerfile and people can inject whatever they want there | 14:24 |
| inc0 | I'm thinking, if we, instad of "enabling neutron plugins" enable code injection in crucial places | 14:25 |
| *** thumpba has quit IRC | 14:25 | |
| mandre | inc0: yeah, I'm not familiar with blocks | 14:25 |
| sdake | inc0 you recognize that is a dsl on dsl spread out all over the filesystem instead of one file? | 14:25 |
| mandre | looks like it will work too | 14:25 |
| inc0 | that will allow people to go all batshit crazy about dockerfiles, and even post their working somewhere and it keeps compatibilisty | 14:25 |
| inc0 | sdake, we can make it work however we want, it's about designing good structure which needs more work and thought | 14:26 |
| sdake | i had thought about the putting inclludes in | 14:26 |
| sdake | I identified 8 sections that require inclusion | 14:26 |
| inc0 | they could provide 3rd party containers based on containers we have easily | 14:26 |
| mandre | inc0: do we really need people to inject code in more than 1 place per dockerfile? | 14:26 |
| coolsvap | inc0, i agree on it's about designing good structure which needs more work and thought | 14:26 |
| sdake | but I thought 8 files * 115 containers = 900 include files | 14:26 |
| sdake | is a bit much | 14:26 |
| coolsvap | and i think its a more refactor than for dsl | 14:27 |
| sdake | mandre if you want to do it properly with jinja2, the answer unfortunately is yes | 14:27 |
| inc0 | mandre, well, custom apt sources.list would another place | 14:27 |
| *** diogogmt has joined #openstack-kolla | 14:27 | |
| inc0 | sdake, people have to put more stuff in dockerfile... | 14:27 |
| inc0 | proxy settings is another | 14:27 |
| sdake | can handle proxy settings with the merge feqature | 14:28 |
| inc0 | how so? | 14:28 |
| mandre | inc0, sdake: right, we've switched the discussion to extensibility rather than purely plugins | 14:28 |
| sdake | put in elemental-merge.yml | 14:28 |
| sdake | let the merging take place | 14:29 |
| sdake | mandre extensibility was where the oirginal discussion was | 14:29 |
| sdake | I just made the statemeent plugins cant be done any other way | 14:29 |
| sdake | and the answer is perhaps they can | 14:29 |
| inc0 | sdake, how do you ensure that it is called BEFORE all other? | 14:29 |
| sdake | mandre what triggered my interest in this work is a need to support custom repos per container | 14:30 |
| inc0 | you want to do external_commands >> repos >> packages >>external_commands? | 14:30 |
| sdake | inc0 that is precisely what the parser does now | 14:30 |
| inc0 | so you have multiple external commands? | 14:30 |
| sdake | two groups yes | 14:30 |
| sdake | all_common_commands | 14:31 |
| sdake | all_commands | 14:31 |
| sdake | or bin or src | 14:31 |
| sdake | its right there in the spec | 14:31 |
| mandre | sdake: I see where the dsl shines, but I wonder if it's flexible enough | 14:31 |
| mandre | we often had the case where we needed to tweak out dockerfiles to remove conflicting packages, etc. | 14:31 |
| nilk | can I prevent kolla-build from re-building all of the base images? | 14:31 |
| nilk | I only want to re-build a single container | 14:32 |
| sdake | nilk no but it doesn't matter, the build is cached | 14:32 |
| pbourke-home | nilk: use the regex feature | 14:32 |
| sdake | to build a single container specify it at the end of build | 14:32 |
| mandre | nilk: pass that image name to the command line | 14:32 |
| nilk | I did but it's still rebuilding all dependencies | 14:32 |
| sdake | mandre I have heard this same concern a couple times | 14:32 |
| sdake | mandre that we will end up in a situation in which we are unable to proceed because the dsl is not sufficient | 14:33 |
| pbourke-home | nilk: post the full command you're using? | 14:33 |
| openstackgerrit | Merged openstack/kolla: Update Cinder guide https://review.openstack.org/315407 | 14:34 |
| inc0 | nilk, it will rebuild deps | 14:34 |
| sdake | mandre to that i'd say - its possible, but hihgly unlikely we will ever be etotally cornered with a dsl | 14:34 |
| inc0 | but it will be quick as you use cache | 14:34 |
| sdake | mandre at worst case we can add new top level keys to uncorner ourselves | 14:34 |
| pbourke-home | sdake: one concern is it makes it harder to modify for operators | 14:34 |
| pbourke-home | sdake: if what they want to do isn't in the dsl, they sol | 14:35 |
| mandre | sdake: by using dockerfile templates, we're sure it's not going to happen | 14:35 |
| sdake | makes which harder to modify? | 14:35 |
| nilk | pbourke-home: kolla-build cinder-volume | 14:35 |
| pbourke-home | nilk: is it building stuff other than base and cinder-base? | 14:35 |
| pbourke-home | (and openstack-base) | 14:35 |
| sdake | mandre is that the only tradeoff you considering atm? | 14:35 |
| nilk | no but that's exactly the issue | 14:35 |
| mandre | also, I find the dsl hard to read and understand what's really happening underneath | 14:35 |
| nilk | it rebuilds openstack-base each time and that takes a while | 14:35 |
| pbourke-home | nilk: it has to rebuild dependencies, that's how docker works. but it should be using the cache as sdake mentioned | 14:36 |
| *** thumpba has joined #openstack-kolla | 14:36 | |
| mandre | you'd better have a small jinja2 dockerfile template that you render thanks to your dsl parser | 14:36 |
| sdake | mandre huh? | 14:37 |
| sdake | mandre actually i think nihilfer suggested that yesterday but my initial poc was in python ;) | 14:37 |
| mandre | sdake: let me paste an example | 14:37 |
| sdake | mandre i implementd base already | 14:38 |
| sdake | and openstack-base | 14:38 |
| sdake | is that what you mean? | 14:38 |
| sdake | and mariadb | 14:38 |
| sdake | and aodh | 14:38 |
| mandre | sdake: no, I'd like a stripped down dockerfile that shows the structure of the dockerfile | 14:39 |
| mandre | right now there is no way to know which group of instructions is executed first without looking at the python source code | 14:40 |
| sdake | #. is an ordered list | 14:40 |
| sdake | the order is right there in the spec | 14:40 |
| sbezverk | nilk make sure you do not use --no-cache option, in this case it will not rebuild the base | 14:42 |
| *** mgoddard_ has quit IRC | 14:42 | |
| nilk | no we're not | 14:42 |
| sdake | mliima around? | 14:42 |
| nilk | but we're using stable/mitaka | 14:42 |
| *** mgoddard has joined #openstack-kolla | 14:42 | |
| mandre | sdake: I don't expect the spec to remain up to date with the code | 14:42 |
| mandre | it's a spec, not a doc | 14:42 |
| sdake | while I'm busy harassing people, also please use review ryan's patch on kolla-kubernetes | 14:43 |
| sdake | mandre i did write documentation as well | 14:43 |
| sdake | mandre https://review.openstack.org/#/c/322345/ | 14:43 |
| patchbot | sdake: patch 322345 - kolla - Documentation for elemental DSL | 14:43 |
| mandre | ok, I'll have to read it I guess :) | 14:43 |
| sdake | the docs need to be true uped with the spec | 14:44 |
| sdake | but there is extra info in there | 14:44 |
| sdake | i keep specs lean and clean | 14:44 |
| *** vhosakot has joined #openstack-kolla | 14:44 | |
| sdake | mandre obviously I pla nto keep the docs u pto date | 14:45 |
| sdake | but nott the spec - the spec is an artifcat | 14:45 |
| sdake | ct | 14:45 |
| *** ppowell has quit IRC | 14:48 | |
| *** diogogmt has quit IRC | 14:48 | |
| *** coolsvap is now known as coolsvap|afk | 14:51 | |
| *** ppowell has joined #openstack-kolla | 14:51 | |
| openstackgerrit | Marc-Andre Gatien proposed openstack/kolla: adding Note for ubuntu 16.04 https://review.openstack.org/323519 | 14:52 |
| mag009_ | i've removed my repo any idea how I can continue my commit when you delete your repo ? | 14:55 |
| inc0 | mag009_, you mean you want to make changes to it? | 14:55 |
| mag009_ | yes | 14:55 |
| inc0 | you can clone fresh master | 14:55 |
| mag009_ | I looked at openstack doc but couldn't found it | 14:56 |
| mag009_ | yes sure but how do I continue my commit | 14:56 |
| inc0 | then do git-review -s | 14:56 |
| mag009_ | I don't want to redo a new change I want to continue my old change | 14:56 |
| inc0 | and then on the right side of https://review.openstack.org/323519 this page there is "Download" button | 14:56 |
| inc0 | there is checkout line, copy and execute it | 14:56 |
| mag009_ | lol | 14:56 |
| mag009_ | oh god | 14:56 |
| inc0 | then commit --amend will work on existing commit | 14:57 |
| mag009_ | man I was looking everywhere | 14:57 |
| inc0 | :D | 14:57 |
| mag009_ | lol feel so dumb.. | 14:57 |
| mag009_ | thx! | 14:57 |
| inc0 | so they changed gerrit layout couple months ago | 14:57 |
| Lyncos | I feel dumb for you | 14:57 |
| inc0 | you wouldn't believe how long it took for me to find this one | 14:57 |
| dave-mccowan | also "git review -d ######" will checkout a branch with the change. combining git and gerrit commands in a repo scares me. :-) | 14:59 |
| mag009_ | thanks thats a good one | 15:00 |
| *** diogogmt has joined #openstack-kolla | 15:00 | |
| *** diogogmt has quit IRC | 15:01 | |
| *** Mr_Broken has quit IRC | 15:06 | |
| *** Mr_Broken has joined #openstack-kolla | 15:06 | |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla: Refactor of base Dockerfile to allow custom repo https://review.openstack.org/323589 | 15:08 |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla: DO NOT MERGE: demo of how drivers could work https://review.openstack.org/323974 | 15:08 |
| *** rhallisey has quit IRC | 15:08 | |
| *** JoseMello has quit IRC | 15:09 | |
| sdake | inc0 can you write a spec for your work plesae | 15:09 |
| sdake | inc0 then it will appear we have dueling specs vying for core reviewer votes :) | 15:09 |
| inc0 | told you, first I want to make some code to get to a point when I think it makes sense | 15:10 |
| sdake | and I guess if neither passes wee will do nothing | 15:10 |
| sdake | inc0 sounds good | 15:10 |
| inc0 | one poc is worth 100 specs imho | 15:10 |
| sdake | inc0 you didn't htink so and forced me to write a spec :) | 15:10 |
| inc0 | but when I'll figure it out, I'll write a spec with end result | 15:10 |
| *** Mr_Broken has quit IRC | 15:11 | |
| inc0 | well, you were pretty confident already about what you had | 15:11 |
| inc0 | as poc goes | 15:11 |
| inc0 | so I want to do the same;) | 15:11 |
| sdake | yup its incomplete but it work | 15:11 |
| sdake | s | 15:11 |
| sdake | understood | 15:11 |
| inc0 | also if you have any feedback and such, it's a discussion | 15:12 |
| sdake | yup | 15:12 |
| sdake | i wouldl like to see one common mechanism used for everything | 15:12 |
| inc0 | take a look at neutron plugin patch | 15:12 |
| sdake | it shortens the learning curve | 15:12 |
| sdake | inc0 one thing I can say about either your or mandre's plugin approach is its better then what exists | 15:13 |
| sdake | another thing I can say is the jinja2 stuff will confuse operators | 15:14 |
| inc0 | let me remind you that I wanted to use full fledged jinja2 from day 1 | 15:14 |
| inc0 | I don't think so | 15:14 |
| inc0 | I think DSL will | 15:14 |
| inc0 | as it looks simple | 15:14 |
| inc0 | but you dont really control what it produces | 15:14 |
| inc0 | more than anything ops like transparency and being able to easily figure out what's happening | 15:15 |
| sdake | the definition of a dsl is it takes an input and produces an output which matches what does the job | 15:15 |
| inc0 | Lyncos, isn't that right? | 15:15 |
| sdake | transparency - I get that | 15:15 |
| Lyncos | Sorry I didn't follow ... let me catch-up | 15:15 |
| inc0 | DSL will obsuce stuff | 15:15 |
| inc0 | I said that operators like transparency and jinja2 isn't too complicated for them | 15:16 |
| Lyncos | Right | 15:16 |
| inc0 | you guys wrote your own jinja2 Dockerfile, was that painful? | 15:16 |
| sdake | the answer is no | 15:17 |
| sdake | because it only used conditionals | 15:17 |
| Lyncos | Very hard | 15:17 |
| Lyncos | ;-) | 15:17 |
| sdake | i guess the answer is yes :) | 15:17 |
| mag009_ | yes well I prefer jinja2 over any puppet | 15:17 |
| Lyncos | lol | 15:17 |
| inc0 | haha | 15:17 |
| inc0 | Dockerfile is a little more than a bash script | 15:17 |
| sdake | mag009_ we aren't talking puppet here, we are talking this dsl: https://review.openstack.org/#/c/323612/ | 15:17 |
| patchbot | sdake: patch 323612 - kolla - Introduction of a new DSL: Elemental | 15:17 |
| Lyncos | It's a love affair between us and Puppet ;-) | 15:18 |
| mag009_ | yes I know ;) just a hater for puppet | 15:18 |
| sdake | oh well if you wrote your own puppet | 15:18 |
| sdake | you guys must be seriously hard core :) | 15:18 |
| sdake | I can't do that myself :) | 15:18 |
| inc0 | been there...not a pretty place | 15:18 |
| inc0 | puppet is magic sometimes | 15:18 |
| inc0 | ansible is so much simpler | 15:18 |
| mag009_ | but my point is puppet has dsl and I hate that from it | 15:19 |
| wirehead_ | The problem with the DSL is that it requires a certain structure for the generated Dockerfiles but it will be much harder to write inscruitable Elemental. The problem with jinja2 is that if you aren't carefull, you can cross over into being inscruitable, but you won't worry about things not representable in Elemental down the road. | 15:19 |
| Lyncos | Just an out of topic question... is Kolla tested on RedHat enterprise ? | 15:19 |
| mag009_ | its a blocker most of the time because you can do simple loop in your manifest etc... | 15:19 |
| mag009_ | sort of thing.. | 15:19 |
| *** zhiwei has quit IRC | 15:20 | |
| inc0 | Lyncos, ask rhallisey for that | 15:20 |
| mag009_ | its definitely simpler in jinja if you want my opinion | 15:20 |
| inc0 | I'm sure he did all sorts of redhatish crazy | 15:20 |
| nilk | dumb question | 15:20 |
| pbourke-home | I think both approaches should be able to build a working base image and service image before arguing any further | 15:20 |
| nilk | if I want to interface with an external storage system | 15:20 |
| nilk | where does the cinder-volume service go? controller or compute node? | 15:21 |
| sdake | pbourke-home i'd agree to that if people don't wait to review until both specs are complete | 15:21 |
| sdake | because who knows when inc0 is done | 15:21 |
| inc0 | nilk, I'd go for controller | 15:21 |
| sdake | I needed this done yesterday | 15:21 |
| sdake | and it takes me about a week to implement | 15:21 |
| sdake | now I have to jerk around with a 2-3 weeks specs process | 15:21 |
| pbourke-home | no | 15:21 |
| pbourke-home | just finish the poc | 15:21 |
| nilk | it also attempts to use iscsiadm to do things like copying volumes | 15:21 |
| pbourke-home | for base + keystone | 15:21 |
| nilk | how does that work with the iscsi container? | 15:21 |
| pbourke-home | if it works it will be much easier to convince people | 15:21 |
| inc0 | sdake, well, and you'll have to jerk around for a full release at least to bring it into working order imho, it's a lot of work | 15:22 |
| *** SiRiuS has quit IRC | 15:22 | |
| mliima | sdake, i'm here | 15:22 |
| sdake | mliima please review https://review.openstack.org/#/c/323612/ | 15:22 |
| patchbot | sdake: patch 323612 - kolla - Introduction of a new DSL: Elemental | 15:22 |
| nilk | I backported everything from the iscsi blueprint to stable/mitaka but I don't understand how cinder-volume is supposed to access the iSCSI container | 15:22 |
| nihilifer | sdake: inc0: about the jinja2 stuff, sorry, i was barely awake when writing this comment | 15:22 |
| pbourke-home | right now it only half works so its much more speculation how well it work | 15:22 |
| inc0 | nilk, alias iscsiadm=docker exec -it iscsi iscsiadm? | 15:22 |
| pbourke-home | for all the problems presented in the spec | 15:22 |
| nihilifer | every solution that will handle pkg versioning is fine for me | 15:22 |
| nilk | inc0: from inside another container? | 15:23 |
| inc0 | nihilifer, pkg versioning? | 15:23 |
| inc0 | uhh, that'll be harder | 15:23 |
| inc0 | sbezverk, can you help nilk | 15:23 |
| inc0 | ? | 15:23 |
| sdake | pbourke-home fair enough - still please review the spec for errors, so when I finish the prototype in 1-2 days I am not jerking around waiting 2-3 weeks for specs approval | 15:23 |
| nihilifer | inc0: it means, handling the pins we have currently in apt_preferences or version.lock | 15:23 |
| inc0 | nihilifer, is customizing apt_preferences bad? | 15:24 |
| nihilifer | for now we have only ceph pinned, but maybe we'll pin other infra services as well | 15:24 |
| inc0 | if you could easily provide your own apt_preferences | 15:24 |
| nihilifer | well, if dsl wants to provide pkg list via yaml, then pkg versions should be also provided via yaml | 15:24 |
| inc0 | but apart from dsl | 15:24 |
| inc0 | if you can provide your own apt_prefs | 15:25 |
| inc0 | would that work? | 15:25 |
| inc0 | nihilifer, https://review.openstack.org/#/c/323589/4 | 15:25 |
| patchbot | inc0: patch 323589 - kolla - Refactor of base Dockerfile to allow custom repo | 15:25 |
| inc0 | however I think I have better idea for file organization | 15:25 |
| inc0 | so just look at custom-dockerfile and imagine it's placed somewhere nice | 15:26 |
| sdake | wirehead_ hit it right on the hed, relating to the tradeoffs of one approach to another | 15:26 |
| nihilifer | inc0: is this patch an alternative to sdake's idea? | 15:26 |
| inc0 | nihilifer, this and few other I'm working on | 15:26 |
| nihilifer | k | 15:26 |
| inc0 | basically demo what we could do with jinja alone | 15:26 |
| nihilifer | have to take a look then | 15:26 |
| sdake | inc0 one fundamental problem I see with the approach is it uses a piecemeal bolt on approach for each type of customization | 15:27 |
| *** SiRiuS has joined #openstack-kolla | 15:27 | |
| inc0 | what's a piecemeal bolt? | 15:28 |
| sdake | we already have that and it works | 15:28 |
| sdake | but its HARD to USE | 15:28 |
| sdake | we have stuff to handle rpms | 15:28 |
| inc0 | but it's flexible | 15:28 |
| sdake | we have stuff to handle include headers | 15:28 |
| sdake | we have stuff to handle include footers | 15:28 |
| inc0 | we can make it better | 15:28 |
| sdake | we have stuff to handle from source installs of plugins | 15:28 |
| inc0 | by refactoring | 15:28 |
| sdake | and we have a fith thing iirc | 15:28 |
| pbourke-home | how do we stop the dsl from becoming bolt on | 15:28 |
| pbourke-home | as we cant anticpate everything up front | 15:29 |
| sdake | pbourke-home worst csae we make a new dsl | 15:29 |
| sdake | and we write a python converter | 15:29 |
| sdake | which takes all of 1 hour at most | 15:29 |
| pbourke-home | i think that's a little optimistic :) | 15:29 |
| sdake | depends on how complex the new dsl change is | 15:29 |
| sdake | but you get the drift, at most under a week for a ocmplete restructure of the dsl | 15:30 |
| inc0 | devils in the details | 15:30 |
| sdake | by the very nature of the dsl, it doesn't need to bolt on | 15:30 |
| inc0 | we craeted initial build script in few hrs too | 15:30 |
| inc0 | look where are we now | 15:30 |
| Lyncos | I would prefer keeping J2 | 15:31 |
| sdake | i have doubts, serious doubts, we can unify the structure of the customizations in jinja2 | 15:31 |
| sdake | without having include sections for each type of include we want | 15:31 |
| sdake | that is the only viable path forward with jinja2 | 15:31 |
| sdake | which ends up with 900 customizatoin files | 15:31 |
| williamcaban | From the maintenance point of view (long term), does it really make sense to maintain another piece of code for the DSL (adding a new definition with any future need) than using jinja? | 15:31 |
| nihilifer | irc meeting now? | 15:31 |
| sdake | ya i've got to jet | 15:31 |
| sdake | inc0 enjoy meeting - 20 minutes on midcycle brainstorming plelase | 15:32 |
| *** rhallisey has joined #openstack-kolla | 15:32 | |
| pbourke-home | in another hours time | 15:32 |
| sdake | oh right 9:30 | 15:32 |
| vhosakot | yeah... 58 more mins to meeting :) | 15:32 |
| *** rhallisey has quit IRC | 15:32 | |
| sdake | williamcaban we are ejecting jinja2 with this proposal | 15:32 |
| sdake | so we aren't maintaining another, we are maintaing a different | 15:32 |
| *** rhallisey has joined #openstack-kolla | 15:32 | |
| wirehead_ | I hate time zone math. Always afraid I'll screw it up. | 15:32 |
| vhosakot | sdake: sbezverk: I'm still resoving magnum issues... will keep you guys posted.. lbaas is rocking in Ubuntu now | 15:32 |
| *** ravig has quit IRC | 15:33 | |
| vhosakot | saw this issue twice --> https://answers.launchpad.net/ubuntu/+question/292723 | 15:33 |
| nihilifer | wirehead_: yes. and i hate daylight savings even more... | 15:33 |
| *** ravig has joined #openstack-kolla | 15:33 | |
| *** sacharya has joined #openstack-kolla | 15:34 | |
| wirehead_ | Oh, I hate DST. Although one of my better moments in my programming career was when my coworkers messaged me after they changed the date DST changed in the US thanking me for making that easy to update without releasing a new version. | 15:35 |
| wirehead_ | That made no sense. Messaged me thanking me for making it easy to update DST without releasing a new version. Mind you this ran on Windows at banks so a new version was uncool. | 15:35 |
| *** mbound has joined #openstack-kolla | 15:36 | |
| *** diogogmt has joined #openstack-kolla | 15:37 | |
| *** rhallisey has quit IRC | 15:44 | |
| *** rhallisey has joined #openstack-kolla | 15:45 | |
| *** rhallisey has quit IRC | 15:45 | |
| *** ppowell has quit IRC | 15:45 | |
| *** rhallisey has joined #openstack-kolla | 15:45 | |
| *** coolsvap|afk is now known as coolsvap | 15:46 | |
| *** berendt has joined #openstack-kolla | 15:46 | |
| *** mikelk has quit IRC | 15:54 | |
| *** sdake has quit IRC | 15:54 | |
| nilk | anyone here who can help me with iSCSI? | 15:55 |
| nilk | I don't get how other containers are supposed to use the iSCSI one | 15:55 |
| sbezverk | nilk sure, what do you need? | 15:56 |
| nilk | we're trying to interface with a HP 3par storage | 15:57 |
| nilk | as far as we understood, cinder_volume is only managing the storage using its API in this scenario, so it's running on the controller nodes | 15:57 |
| nilk | now, the 3par plugin attempts to run iscsiadm, which fails | 15:57 |
| sbezverk | nilk are you on master or stable/mitaka? | 15:58 |
| *** ravig has quit IRC | 15:58 | |
| nilk | stable/mitaka | 15:59 |
| nilk | backported everything from master which is related to iscsi | 15:59 |
| nilk | except the target, of course | 15:59 |
| sbezverk | nilk target? are you refering to tgtd? | 15:59 |
| nilk | yes | 16:00 |
| *** daneyon_ has joined #openstack-kolla | 16:01 | |
| sbezverk | nilk ok that is the mistake as both tgtd and iscsid are required at least in the case when volume group is on kolla's managed storage node | 16:03 |
| sbezverk | nilk I have never tried to use 3rd party iscsi storage, but I would suggest to port ALL iscsi related from master and then in cinder.conf try to use ip address of your external iscsi server | 16:04 |
| sbezverk | again, it has never been tested ;-) | 16:04 |
| nilk | yes we tried that, it even talks to the storage already | 16:05 |
| nilk | we also have a custom cinder.conf | 16:05 |
| *** daneyon_ has quit IRC | 16:05 | |
| nilk | now cinder wants to use iscsiadm and can't | 16:05 |
| *** mdnadeem has quit IRC | 16:05 | |
| nilk | what's the idea with the iscsid container? | 16:06 |
| nilk | I assume nova talks to it, right? | 16:06 |
| *** mdnadeem has joined #openstack-kolla | 16:06 | |
| *** mdnadeem has quit IRC | 16:07 | |
| nilk | tgtd isn't necessary with a 3rd party storage I believe | 16:07 |
| *** mdnadeem has joined #openstack-kolla | 16:07 | |
| sbezverk | nilk correct, because we do not provide iscsiadm binaries to cinder_volume container, it uses ONLY tgtdadm helper, there is no lio based helper support for now.. | 16:09 |
| *** mdnadeem has quit IRC | 16:09 | |
| nilk | lio? | 16:09 |
| *** mdnadeem has joined #openstack-kolla | 16:09 | |
| inc0 | nilk, you can simply install iscsiadm in cinder-volume | 16:10 |
| sbezverk | iscsi_helper = lioadm | 16:10 |
| sbezverk | inc0 have you tried it? I seriously doubt it will work.. | 16:11 |
| *** fragatina has joined #openstack-kolla | 16:12 | |
| *** prithiv has quit IRC | 16:12 | |
| inc0 | no, just a thought | 16:12 |
| inc0 | I'm not good at iscsi stuff.. | 16:12 |
| inc0 | I deploy ceph;) | 16:12 |
| *** sacharya has quit IRC | 16:13 | |
| sbezverk | inc0 :-) when you have unlimited number of servers, why not ;-) | 16:14 |
| inc0 | nah, ceph is not that bad;) | 16:15 |
| inc0 | I used 3 vms for dev | 16:15 |
| inc0 | now I use 10 | 16:15 |
| inc0 | because I can mwhahaha | 16:15 |
| *** fragatina has quit IRC | 16:17 | |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla: DO NOT MERGE: demo of how drivers could work https://review.openstack.org/323974 | 16:17 |
| *** sdake has joined #openstack-kolla | 16:19 | |
| nilk | sbezverk: I set iscsi_helper to tgtadm, but it still fails | 16:20 |
| nilk | ImageCopyFailure: Failed to copy image to volume: Could not login to any iSCSI portal. | 16:20 |
| sbezverk | nilk usually it means nova compute does not use correct ip address of iscsi server. | 16:22 |
| *** sdake has quit IRC | 16:23 | |
| nilk | I did end up installing iscsi-initiator-utils in the cinder-volume container | 16:23 |
| nilk | it wouldn't work otherwise | 16:23 |
| *** prithiv has joined #openstack-kolla | 16:23 | |
| sbezverk | nilk Since I have never tried it, I would need to setup something with external iscsi and test it.. | 16:24 |
| sbezverk | nilk when you create cinder volume, does it get created properly? | 16:25 |
| nilk | yes | 16:25 |
| nilk | but it's not actually doing anything | 16:25 |
| nilk | as soon as I try to clone an image, it attempts to connect from cinder_storage | 16:26 |
| nilk | there's no iscsid running on that host | 16:26 |
| sbezverk | nilk but do you see lv volume gets created on remote iscsi server? | 16:27 |
| *** prithiv has quit IRC | 16:28 | |
| nilk | no | 16:28 |
| sbezverk | nilk ok it makes sense, because cinder_volume executes lvcreate on the local host where it is running | 16:29 |
| nilk | we set volume_driver=cinder.volume.drivers.san.hp.hp_3par_iscsi.HP3PARISCSIDriver | 16:29 |
| sbezverk | nilk I see.. at this point I would first try to figure out why HP iscsi driver does not create volume.. | 16:31 |
| inc0 | meeting my dear kolla'egues | 16:31 |
| vhosakot | meeting time | 16:32 |
| *** sdake has joined #openstack-kolla | 16:32 | |
| *** ssurana has joined #openstack-kolla | 16:43 | |
| *** ssurana has quit IRC | 16:48 | |
| *** rmart04 has quit IRC | 16:52 | |
| *** unicell has joined #openstack-kolla | 16:54 | |
| *** unicell1 has quit IRC | 16:55 | |
| *** ssurana has joined #openstack-kolla | 16:56 | |
| *** thumpba has quit IRC | 16:56 | |
| *** unicell has quit IRC | 16:57 | |
| *** ravig has joined #openstack-kolla | 16:58 | |
| *** thumpba has joined #openstack-kolla | 17:01 | |
| *** mbound has quit IRC | 17:03 | |
| *** ssurana has quit IRC | 17:03 | |
| *** ssurana has joined #openstack-kolla | 17:04 | |
| *** thumpba has quit IRC | 17:14 | |
| *** ppowell has joined #openstack-kolla | 17:15 | |
| *** sacharya has joined #openstack-kolla | 17:15 | |
| *** ppowell has quit IRC | 17:20 | |
| *** unicell has joined #openstack-kolla | 17:30 | |
| rhallisey | wirehead_, we'll get to gating soon enough. Hopefully we can settle on our bootstrapping/deployment method/tooling then gating shoudl follow | 17:30 |
| rhallisey | plus kolla-kube doesn't deloy enough services yet | 17:30 |
| rhallisey | wirehead_, I do agree though | 17:31 |
| inc0 | pbourke-home, mandre rhallisey quick question, is number of files even an issue? | 17:31 |
| *** unicell1 has joined #openstack-kolla | 17:31 | |
| rhallisey | not that i know of | 17:31 |
| inc0 | I mean we can have one giant file, but that doens't really help with readibility | 17:31 |
| rhallisey | I guess it's whatever brings readability | 17:32 |
| inc0 | yeah, you can't really measure that | 17:32 |
| inc0 | but it's refactoring | 17:32 |
| inc0 | mandre, btw this is another approach to drivers https://review.openstack.org/#/c/323974/1 | 17:33 |
| patchbot | inc0: patch 323974 - kolla - DO NOT MERGE: demo of how drivers could work | 17:33 |
| inc0 | on example of neutron | 17:33 |
| *** ppowell has joined #openstack-kolla | 17:34 | |
| *** thumpba has joined #openstack-kolla | 17:34 | |
| *** unicell has quit IRC | 17:34 | |
| wirehead_ | At the very least, there should be one file per service, perhaps more. | 17:36 |
| inc0 | well we have direcory per container | 17:37 |
| inc0 | I don't think this is bad | 17:37 |
| inc0 | but I might be wrong | 17:37 |
| openstackgerrit | Merged openstack/kolla: adding Note for ubuntu 16.04 https://review.openstack.org/323519 | 17:38 |
| *** pbourke-home has quit IRC | 17:39 | |
| nilk | sbezverk: why isn't /var/lib/iscsi/ shared? | 17:40 |
| nilk | unnecessary? | 17:41 |
| *** daneyon_ has joined #openstack-kolla | 17:49 | |
| Mech422 | pbourke_: so no joy with the VETH stuff yet - it appears when Kolla fires up the openvswitch containers, it kills the existing openvswitch config on the host... | 17:52 |
| Mech422 | pbourke_: I actually lose connectivity to the boxes as my mgmt net. is configured via ovs | 17:53 |
| *** ccesario has quit IRC | 17:53 | |
| *** daneyon_ has quit IRC | 17:54 | |
| *** tyrola has quit IRC | 17:54 | |
| *** jmccarthy1 has left #openstack-kolla | 17:54 | |
| *** thumpba has quit IRC | 17:55 | |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla: DO NOT MERGE: demo of how drivers could work https://review.openstack.org/323974 | 18:01 |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla: Refactor of base Dockerfile to allow custom repo https://review.openstack.org/323589 | 18:01 |
| *** dwalsh has quit IRC | 18:02 | |
| inc0 | mandre, still up? | 18:03 |
| *** ccesario has joined #openstack-kolla | 18:06 | |
| *** Serlex has quit IRC | 18:06 | |
| *** tyrola has joined #openstack-kolla | 18:07 | |
| * Mech422 wonders if OVS can actually run in a Container and still play nice with host OVS | 18:09 | |
| Mech422 | there's only 1 kernel module after all...they duel for it ? :-P | 18:09 |
| *** neophy has joined #openstack-kolla | 18:11 | |
| *** ccesario has quit IRC | 18:11 | |
| inc0 | Mech422, well both would need to share ovs db | 18:13 |
| inc0 | otherwise it will get messy as one ovs will use override flows of another...and I don't even know when that happends | 18:14 |
| Mech422 | inc0: if you have an 'existing' OVS setup on the host | 18:15 |
| *** thumpba has joined #openstack-kolla | 18:15 | |
| inc0 | you could share existing ovsdb with ovs in container | 18:16 |
| Mech422 | inc0: when you fire up the openvswitch-db and openvswitch-switch containers, it wipes out the existing config.. | 18:16 |
| Mech422 | inc0: I wonder if that would be enough | 18:16 |
| inc0 | should be imho | 18:16 |
| Mech422 | inc0: the 'host' side should be doing any changes - its all configured at boot | 18:16 |
| inc0 | ovs client is just glorified editor of ovsdb | 18:16 |
| Mech422 | s/should/shouldn't/ | 18:16 |
| *** zhiwei has joined #openstack-kolla | 18:17 | |
| Mech422 | inc0: yeah - but it appears the kernel module is propagating changes back | 18:17 |
| Mech422 | inc0: I see the 'containerized' config when I run ovs-vsctl show from the 'host' | 18:17 |
| Mech422 | and all the 'host' bridges are gone | 18:18 |
| inc0 | well, you have some serius mess out there man;) | 18:18 |
| Mech422 | but Kolla doesn't delete them explicitly as far as I can find. I think its just when kolla fires up with an empty ovs db, the kernel module reports that state to the host as well | 18:19 |
| nilk | the network is not containerized | 18:19 |
| Mech422 | inc0: I don't think its really a 'mess' - as you said, if they shared the DB then the state should be the same between them | 18:19 |
| nilk | it's running in the same namespace | 18:19 |
| Mech422 | nilk: oh...hmm.. | 18:20 |
| inc0 | well either way ovs is in kernel, flows are really below namespaces too | 18:20 |
| Mech422 | nilk: so just passing the existing ovs db to the container should be fine...it can already see/change everythign on the host anyway | 18:20 |
| inc0 | yeah, if you need your host ovs you can try passing it | 18:21 |
| inc0 | but I generally suggest wiping host ovs and use kolla's | 18:21 |
| *** zhiwei has quit IRC | 18:21 | |
| *** ppowell has quit IRC | 18:21 | |
| Mech422 | inc0: I could do that for my vlan stuff easy enough - just use native vlans | 18:22 |
| Mech422 | inc0: not sure how this is all gonna play with midonet though - thats my ultimate goal | 18:22 |
| inc0 | so issue is that ansible doesn't see eth0.100 (or whatever your vlan is)? | 18:23 |
| Mech422 | inc0: no - actually that all seems to be fine.... | 18:24 |
| *** ccesario has joined #openstack-kolla | 18:24 | |
| Mech422 | inc0: its just OVS gets 'confused' once Kolla starts driving | 18:24 |
| inc0 | but what would be issue with replacing from host with kolla's one? | 18:24 |
| Mech422 | inc0: you lost me ? Once kolla takes over, I lose the management connection to the box... | 18:25 |
| inc0 | ahh you're tagging vlans on ovs | 18:26 |
| inc0 | ? | 18:26 |
| Mech422 | actually, if networking is not name spaced - it prolly makes sense to _always_ share the ovs db | 18:26 |
| Mech422 | inc0: yeah | 18:26 |
| ccesario | hello folks.... | 18:26 |
| inc0 | even if it's namespaced | 18:26 |
| inc0 | ovs is not namespaced | 18:26 |
| inc0 | you can't have ovs per namespace | 18:26 |
| ccesario | does someone could confirm if deploy/image build it works with Ubuntu 16.04 ?! | 18:27 |
| inc0 | but try sharing ovsdb, give me a moment | 18:27 |
| inc0 | as host or inside docker ccesario ? | 18:27 |
| Mech422 | inco0: yeah - since they 'both' control the same kernel module, and 'both' see the state - might as well just use 1 state db? | 18:27 |
| inc0 | inside docker I don't think it works | 18:27 |
| inc0 | exactly | 18:27 |
| inc0 | and you really want to have only one | 18:27 |
| ccesario | inc0, as host | 18:28 |
| *** tyrola has quit IRC | 18:28 | |
| inc0 | ccesario, should work.. | 18:28 |
| inc0 | Mech422, https://github.com/openstack/kolla/blob/master/ansible/roles/neutron/tasks/start.yml | 18:28 |
| Mech422 | inc0: is there a 'generic' way to add shares to the containers - or should I start looking for what to hack ? | 18:28 |
| inc0 | so first 3 tasks starts ovsdb, you want to remove them | 18:29 |
| ccesario | inc0, stable/mitaka and trunk ... right!? | 18:29 |
| inc0 | ccesario, yeah | 18:29 |
| ccesario | inc0, thanks by info... ;) | 18:29 |
| inc0 | hmmm | 18:31 |
| inc0 | and that might be it Mech422 | 18:31 |
| *** mark-casey has joined #openstack-kolla | 18:31 | |
| Mech422 | inc0: ok - I'll config. 'br-ex' on the 'host' side and rebuild the images | 18:31 |
| inc0 | simply not starting container | 18:31 |
| inc0 | but not sure, this is new to me | 18:31 |
| Mech422 | inc0: yeah - that looks like an elegant way to fix it | 18:31 |
| inc0 | elegant it is not, it's kinda hacky, but it might be only way all things considered | 18:32 |
| Mech422 | inc0: since the next task starts vswitchd in a seperate container | 18:32 |
| inc0 | yeah, I just don't know how they communicate with each other | 18:32 |
| Mech422 | inc0: eh ... seems elegant - 1 less container to run ? :-) No other changes, etc etc | 18:32 |
| inc0 | if it works as I hope it works | 18:33 |
| inc0 | ;) | 18:33 |
| *** mbound has joined #openstack-kolla | 18:35 | |
| *** mbound has quit IRC | 18:40 | |
| *** ccesario has quit IRC | 18:41 | |
| williamcaban | Is there a way to disable the deployment of haproxy + keepalived when doing an external all-in-one deployment? (i.e. when not using the all-in-one with local targets) | 18:42 |
| *** Jeffrey4l has quit IRC | 18:42 | |
| *** Jeffrey4l has joined #openstack-kolla | 18:43 | |
| inc0 | globals.yml: enable_haproxy: no | 18:43 |
| williamcaban | inc0: cool. does that disable keepalived as well? | 18:44 |
| inc0 | yeah | 18:44 |
| williamcaban | thx | 18:44 |
| inc0 | remeber that you need to change kolla_internal_address to existing one as well | 18:44 |
| inc0 | as no more floating ip exists | 18:44 |
| williamcaban | the internal or the vip or both? | 18:45 |
| inc0 | internal will be vip | 18:45 |
| inc0 | external as well in that matter, but since we don't use external that doesn;t really matter | 18:45 |
| williamcaban | got it | 18:45 |
| inc0 | https://github.com/openstack/kolla/blob/master/etc/kolla/globals.yml#L26 | 18:46 |
| inc0 | this one | 18:46 |
| inc0 | rest will follow | 18:46 |
| *** JoseMello has joined #openstack-kolla | 18:49 | |
| williamcaban | BTW, since few days ago the "kolla-ansible prechecks" always fails with the verification of the "Docker version" (I'm using 1.11.1 in CentOS7) even though the deploy's work | 18:49 |
| williamcaban | not sure if because of the regex_replace or the version_compare | 18:51 |
| williamcaban | here https://github.com/openstack/kolla/blob/master/ansible/roles/prechecks/tasks/service_checks.yml#L13 | 18:52 |
| *** ccesario has joined #openstack-kolla | 18:53 | |
| *** daneyon_ has joined #openstack-kolla | 18:56 | |
| *** thumpba has quit IRC | 18:59 | |
| *** thumpba has joined #openstack-kolla | 19:00 | |
| *** daneyon_ has quit IRC | 19:01 | |
| *** papacz has quit IRC | 19:02 | |
| *** ssurana has quit IRC | 19:04 | |
| *** mdnadeem has quit IRC | 19:06 | |
| *** mdnadeem has joined #openstack-kolla | 19:07 | |
| *** mdnadeem has joined #openstack-kolla | 19:07 | |
| *** mdnadeem has quit IRC | 19:07 | |
| *** mdnadeem has joined #openstack-kolla | 19:08 | |
| *** mdnadeem has joined #openstack-kolla | 19:08 | |
| *** mdnadeem has quit IRC | 19:08 | |
| *** mdnadeem has joined #openstack-kolla | 19:09 | |
| *** mdnadeem has joined #openstack-kolla | 19:09 | |
| *** mdnadeem has quit IRC | 19:09 | |
| *** mdnadeem has joined #openstack-kolla | 19:10 | |
| *** mdnadeem has joined #openstack-kolla | 19:10 | |
| *** mbound has joined #openstack-kolla | 19:12 | |
| *** Jeffrey4l_ has joined #openstack-kolla | 19:31 | |
| *** godleon has joined #openstack-kolla | 19:33 | |
| *** Jeffrey4l has quit IRC | 19:34 | |
| inc0 | guys, I need a brainstorm | 19:39 |
| inc0 | who wants a brainstorm? | 19:39 |
| mag009_ | about? | 19:39 |
| inc0 | most convenient way to customize dockerfiles | 19:39 |
| inc0 | so I have an idea | 19:40 |
| inc0 | create /etc/kolla/dockerfile-customizations | 19:40 |
| mag009_ | isn't related to what you guys been talking about DSL | 19:40 |
| inc0 | that will look somehow like that | 19:40 |
| inc0 | that's exactly this topic | 19:40 |
| inc0 | https://review.openstack.org/#/c/323589/3/docker/base/custom-dockerfile.j2 | 19:40 |
| patchbot | inc0: patch 323589 - kolla - Refactor of base Dockerfile to allow custom repo | 19:40 |
| inc0 | so if we make it in /etc | 19:41 |
| inc0 | and you'll have a lots and lots of blocks like that | 19:41 |
| mag009_ | thats if i want to add some stuff to the base image | 19:41 |
| inc0 | so single file with {% nova_compute_additional_packages %} and {% neutron_ovs_agent_post_install %} | 19:41 |
| inc0 | well I can have single file with ability to fill every hook we put in every dockerfile | 19:42 |
| inc0 | would that be convenient? | 19:42 |
| mag009_ | look neat to me | 19:43 |
| mliima | why make it in /etc? | 19:43 |
| inc0 | why not? it's custom | 19:43 |
| mag009_ | for my calico stuff I've added a wget .deb | 19:43 |
| inc0 | we can use --customization-file=/etc/myprettyfile | 19:44 |
| mliima | oh | 19:44 |
| mliima | :) | 19:44 |
| mag009_ | are you guys going forward with your DSL for the dockerfiles? | 19:44 |
| inc0 | I hope not | 19:44 |
| inc0 | I'm strongly against it, but we'll vote on it | 19:45 |
| mag009_ | i dont think its worth the effort to be honest.. | 19:45 |
| inc0 | yeah, I'm afraid that it will be more effort than people realize | 19:45 |
| mliima | we'll choose between them mag009_ | 19:45 |
| openstackgerrit | Ryan Hallisey proposed openstack/kolla-kubernetes: Break deployment procedure into two steps https://review.openstack.org/320668 | 19:45 |
| *** dave-mccowan has quit IRC | 19:46 | |
| inc0 | disadvantage of this solution would be that myprettyfile will not be pretty at all | 19:47 |
| inc0 | it will be huge mix of code from multiple containers | 19:48 |
| inc0 | but we can work on that too | 19:48 |
| *** ravig has quit IRC | 19:48 | |
| rhallisey | wirehead_, ^^ try out that patch | 19:50 |
| rhallisey | that will split the cli | 19:50 |
| *** mdnadeem has quit IRC | 19:52 | |
| *** mdnadeem has joined #openstack-kolla | 19:53 | |
| *** mdnadeem has quit IRC | 19:53 | |
| *** mdnadeem has joined #openstack-kolla | 19:53 | |
| openstackgerrit | Ken Wronkiewicz proposed openstack/kolla-kubernetes: Replacing the bare pod with a replication controller https://review.openstack.org/324077 | 19:53 |
| wirehead_ | rhallisey ^^ That’s the Replication Controller change I was talking about. | 19:55 |
| rhallisey | wirehead_, awesome :) | 19:55 |
| rhallisey | wirehead_, I left a comment | 19:57 |
| openstackgerrit | Ken Wronkiewicz proposed openstack/kolla-kubernetes: Replacing the bare pod with a replication controller https://review.openstack.org/324077 | 19:57 |
| wirehead_ | Okay, didn’t realize that there was a BP | 19:58 |
| *** mdnadeem has quit IRC | 19:58 | |
| *** mdnadeem has joined #openstack-kolla | 19:59 | |
| rhallisey | wirehead_, ya -> https://blueprints.launchpad.net/kolla-kubernetes | 19:59 |
| wirehead_ | My prediction is that we’ll be happier if we do that BP sooner rather than later. :) | 20:00 |
| rhallisey | for sure | 20:00 |
| wirehead_ | It’ll catch all of the hardcoded pod names before they have a chance to roost. | 20:00 |
| *** SiRiuS__ has joined #openstack-kolla | 20:01 | |
| *** mdnadeem has quit IRC | 20:02 | |
| *** dave-mccowan has joined #openstack-kolla | 20:04 | |
| wirehead_ | rhallisey: like, after the replication-controllers BP is done, you are going to do ‘bootstrap’ once and then ‘run’ once. | 20:05 |
| wirehead_ | rhallisey: there should never be a case where you ‘run’ twice. | 20:05 |
| *** david-lyle has joined #openstack-kolla | 20:05 | |
| wirehead_ | rhallisey: which is, obviously, totally opposite kolla-ansible. | 20:05 |
| rhallisey | wirehead_, you will bootstrap again for an upgrade | 20:05 |
| wirehead_ | Yeah. | 20:06 |
| rhallisey | but you won't need to create eh users, ect.. | 20:06 |
| wirehead_ | That’s probably ‘upgrade’ instead of ‘bootstrap' | 20:06 |
| rhallisey | ideally we want to have a feedback look where as kubernetes can run a pod that will do the db_sync for bootstrapping and again for upgrades | 20:08 |
| *** SiRiuS__ has quit IRC | 20:10 | |
| Mech422 | pbourke_: I can't say that commenting out openvswitch-db 'works' - but the deploy seems to be getting farther and I didn't lose access to the boxes | 20:13 |
| Mech422 | pbourke_: I'm gonna hit the hay and let it deploy while I sleep | 20:13 |
| wirehead_ | Yah. I think there will be some operator resistance to fully hands-off upgrades, but for the smaller clouds, it’ll go over like chicken fried bacon in Texas. | 20:13 |
| Mech422 | pbourke_: oh - I lie - the deploy just finished... | 20:14 |
| mag009_ | when I'm using the master repo I use it with ansible 2.0.0 it fail at : Create log volume | 20:22 |
| mag009_ | master repo is compatible with ansible 2.0.0 no ? | 20:22 |
| mag009_ | guess not :) just tried with 1.9.4 | 20:23 |
| mag009_ | and it worked. | 20:23 |
| inc0 | hmm | 20:31 |
| inc0 | master should be ansible2 | 20:32 |
| *** ssurana has joined #openstack-kolla | 20:32 | |
| inc0 | sorcery! | 20:32 |
| mag009_ | i think its my bad | 20:33 |
| Mech422 | inc0: so looks like just commenting out those 3 tasks worked - haproxy is reporting everything up | 20:33 |
| mag009_ | I pulled it like last week or so | 20:33 |
| inc0 | Mech422, glad to hear that | 20:33 |
| Mech422 | inc0: I'll play with it more tonight - might just be able to put a conditional on those tasks not to run if 'br-ex' already exists | 20:33 |
| Mech422 | inc0: then it should work with no-existing config, or if you pre-configured OVS for it | 20:34 |
| rhallisey | wirehead_, rc looks good :) | 20:34 |
| rhallisey | works like a charm | 20:34 |
| inc0 | rather than that add a conf variable external_ovs | 20:34 |
| inc0 | that will default to false | 20:35 |
| inc0 | and if true don't run ovsdb | 20:35 |
| *** fragatin_ has joined #openstack-kolla | 20:35 | |
| *** fragatin_ has quit IRC | 20:35 | |
| inc0 | not sure if we need to run ovs-vswitchd as well | 20:35 |
| inc0 | probably not | 20:35 |
| Mech422 | inc0: that'd work - I was trying to avoid another setting. I am running ovs-switchd currently | 20:35 |
| inc0 | well, I think it's ok to add this devaulting to false | 20:36 |
| *** fragatina has joined #openstack-kolla | 20:36 | |
| inc0 | since it will be highly unlikely scenerio | 20:36 |
| inc0 | or rather to keep our standard, enable_openvswitch=true by default | 20:36 |
| Mech422 | inc0: it will also help with midonet :-) | 20:38 |
| Mech422 | inc0: Ok - now it really is time for bed...g'nite | 20:38 |
| inc0 | bye | 20:39 |
| *** mliima has quit IRC | 20:46 | |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla: DO NOT MERGE: demo of how drivers could work https://review.openstack.org/323974 | 20:47 |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla: Refactor of base Dockerfile to allow custom repo https://review.openstack.org/323589 | 20:47 |
| *** sdake has quit IRC | 20:49 | |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla: DO NOT MERGE: demo of how drivers could work https://review.openstack.org/323974 | 20:50 |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla: Refactor of base Dockerfile to allow custom repo https://review.openstack.org/323589 | 20:50 |
| *** david-lyle has quit IRC | 20:51 | |
| *** neophy has quit IRC | 20:52 | |
| *** neophy has joined #openstack-kolla | 20:53 | |
| openstackgerrit | Ken Wronkiewicz proposed openstack/kolla-kubernetes: [doc] change Ansible version to exactly 2.0.x in quickstart. https://review.openstack.org/322361 | 20:54 |
| *** thumpba has quit IRC | 20:54 | |
| *** ayoung has quit IRC | 20:55 | |
| inc0 | rhallisey ping | 20:58 |
| mag009_ | http://pastebin.com/Z0LkcmxJ | 20:58 |
| rhallisey | inc0, yo | 20:58 |
| inc0 | https://review.openstack.org/#/c/323589/ check this out | 20:58 |
| patchbot | inc0: patch 323589 - kolla - Refactor of base Dockerfile to allow custom repo | 20:58 |
| mag009_ | i'm getting this when using ansible 2.0.2 | 20:58 |
| rhallisey | inc0, do the packages have to be indented like that? | 20:59 |
| inc0 | nah, just trying to break this up | 21:00 |
| inc0 | so we won't have one giant dockerfile | 21:00 |
| inc0 | but refactoring isn't something I'm 100% sure about, we might think more how to break this up | 21:00 |
| rhallisey | I mean in redhat-binary-packages file | 21:00 |
| inc0 | but I'm curious what you think about template-overrides mechanism | 21:00 |
| inc0 | well, right now it's literally cut-paste from what we have now | 21:01 |
| inc0 | I can make it prettier when we decide what is prettier;) | 21:01 |
| *** david-lyle has joined #openstack-kolla | 21:01 | |
| inc0 | https://review.openstack.org/#/c/323589/7/docker/base/debian-dockerfile.j2 line 34 | 21:01 |
| patchbot | inc0: patch 323589 - kolla - Refactor of base Dockerfile to allow custom repo | 21:01 |
| rhallisey | inc0, I feel like whoever gets the gate passing will win :) | 21:01 |
| inc0 | we'll create bunch of places like that | 21:01 |
| rhallisey | I do like it much better | 21:02 |
| inc0 | https://review.openstack.org/#/c/323589/7/etc/kolla/custom-dockerfile-example.j2 then you override it like in line 11 | 21:02 |
| patchbot | inc0: patch 323589 - kolla - Refactor of base Dockerfile to allow custom repo | 21:02 |
| inc0 | and you pass path to this file in build.py | 21:02 |
| *** neophy has quit IRC | 21:05 | |
| rhallisey | inc0, oh nice. So you can import install_yum macro in the service containers | 21:09 |
| inc0 | that's gonna be doable easily | 21:09 |
| inc0 | but rather than that we'll have common include file with all that macros and such | 21:09 |
| inc0 | just need to figure it out | 21:09 |
| inc0 | and would like to wait till build.py refactor | 21:10 |
| inc0 | so we can do total refactor at once | 21:10 |
| mag009_ | inc0: I have the lastest version of the master repo and it still fail with ansible 2.0.X | 21:10 |
| inc0 | mag009_, hmm...I'll try it out tomorrow | 21:10 |
| inc0 | I'm a bit burned out today | 21:10 |
| inc0 | this DSL thing consumed most of my energy | 21:11 |
| mag009_ | yes its fine Im doing another cleanup and starting from scratch just to see | 21:11 |
| mag009_ | lol | 21:11 |
| rhallisey | inc0, left you some comments | 21:13 |
| inc0 | thanks | 21:13 |
| inc0 | yeah I need to clean this one up | 21:13 |
| inc0 | but how do you like general idea behind it? | 21:14 |
| *** jtriley has quit IRC | 21:20 | |
| *** mbound has quit IRC | 21:22 | |
| *** ravig has joined #openstack-kolla | 21:28 | |
| openstackgerrit | Ken Wronkiewicz proposed openstack/kolla-kubernetes: Replication controllers for Keystone, Memcached, RabbitMQ. https://review.openstack.org/324106 | 21:28 |
| wirehead_ | thanks, dims! Mainline being broken was bothering my fastidious side. :) | 21:30 |
| dims | wirehead_ : i gotta add a functional test!! | 21:31 |
| wirehead_ | dims: yep, trying to avoid straying too far from mainline. Saves merge time. :) | 21:33 |
| *** ravig has quit IRC | 21:34 | |
| openstackgerrit | Ryan Hallisey proposed openstack/kolla-kubernetes: Break deployment procedure into two steps https://review.openstack.org/320668 | 21:40 |
| rhallisey | crap did that already verify | 21:40 |
| rhallisey | if it did, I'll add the last bit in a new patch | 21:40 |
| *** v1k0d3n has quit IRC | 21:45 | |
| *** sdake has joined #openstack-kolla | 21:50 | |
| openstackgerrit | Ken Wronkiewicz proposed openstack/kolla-kubernetes: Adding documentation for labels https://review.openstack.org/324110 | 21:51 |
| sdake | vhosakot ping | 21:58 |
| openstackgerrit | Merged openstack/kolla-kubernetes: Regression: Kubernetes was rejecting the YAML we were generating https://review.openstack.org/323624 | 21:58 |
| openstackgerrit | Merged openstack/kolla-kubernetes: Replacing the bare pod with a replication controller https://review.openstack.org/324077 | 21:58 |
| openstackgerrit | Merged openstack/kolla-kubernetes: Adding the password generation step to the docs. https://review.openstack.org/322362 | 21:59 |
| *** sdake_ has joined #openstack-kolla | 22:01 | |
| *** david-lyle has quit IRC | 22:01 | |
| *** mdnadeem has joined #openstack-kolla | 22:03 | |
| *** sdake has quit IRC | 22:03 | |
| openstackgerrit | Ryan Hallisey proposed openstack/kolla-kubernetes: Break deployment procedure into two steps https://review.openstack.org/320668 | 22:06 |
| openstackgerrit | Ryan Hallisey proposed openstack/kolla-kubernetes: Add bootstrap method to the quickstart doc https://review.openstack.org/324113 | 22:06 |
| openstackgerrit | Ryan Hallisey proposed openstack/kolla-kubernetes: Break deployment procedure into two steps https://review.openstack.org/320668 | 22:07 |
| vhosakot | sdake_: ping | 22:07 |
| *** david-lyle has joined #openstack-kolla | 22:16 | |
| *** sdake_ has quit IRC | 22:21 | |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla: DO NOT MERGE: demo of how drivers could work https://review.openstack.org/323974 | 22:21 |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla: Refactor of base Dockerfile to allow custom repo https://review.openstack.org/323589 | 22:21 |
| *** ravig has joined #openstack-kolla | 22:22 | |
| *** inc0 has quit IRC | 22:23 | |
| *** mbound has joined #openstack-kolla | 22:23 | |
| openstackgerrit | Ken Wronkiewicz proposed openstack/kolla-kubernetes: Added documentation for building Kolla containers. https://review.openstack.org/324116 | 22:24 |
| *** ravig has quit IRC | 22:24 | |
| *** mbound has quit IRC | 22:28 | |
| openstackgerrit | Ken Wronkiewicz proposed openstack/kolla-kubernetes: Adding documentation for labels https://review.openstack.org/324110 | 22:29 |
| *** vhosakot has quit IRC | 22:30 | |
| openstackgerrit | Ken Wronkiewicz proposed openstack/kolla-kubernetes: Added documentation for building Kolla containers. https://review.openstack.org/324116 | 22:31 |
| *** sdake has joined #openstack-kolla | 22:31 | |
| *** daneyon_ has joined #openstack-kolla | 22:32 | |
| *** zhiwei has joined #openstack-kolla | 22:34 | |
| *** daneyon_ has quit IRC | 22:37 | |
| *** zhiwei has quit IRC | 22:38 | |
| *** intr1nsic has quit IRC | 22:43 | |
| *** intr1nsic has joined #openstack-kolla | 22:45 | |
| *** sacharya has quit IRC | 22:45 | |
| *** mark-casey has quit IRC | 22:56 | |
| openstackgerrit | Ken Wronkiewicz proposed openstack/kolla-kubernetes: Added documentation for building Kolla containers. https://review.openstack.org/324116 | 22:57 |
| *** godleon has quit IRC | 22:59 | |
| *** berendt has quit IRC | 22:59 | |
| *** JoseMello has quit IRC | 23:09 | |
| *** diogogmt has quit IRC | 23:33 | |
| *** v1k0d3n has joined #openstack-kolla | 23:40 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: DO NOT MERGE: TEST MASTER BRANCH https://review.openstack.org/294906 | 23:51 |
| *** sdake_ has joined #openstack-kolla | 23:54 | |
| *** sdake has quit IRC | 23:58 | |
| *** david-lyle has quit IRC | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!