| *** salv-orlando has quit IRC | 00:03 | |
| *** ravig has joined #openstack-kolla | 00:08 | |
| *** ravig has quit IRC | 00:10 | |
| *** ravig has joined #openstack-kolla | 00:10 | |
| *** ravig has quit IRC | 00:14 | |
| *** ravig has joined #openstack-kolla | 00:19 | |
| *** jtriley has quit IRC | 00:21 | |
| *** ravig has quit IRC | 00:23 | |
| *** asalkeld has joined #openstack-kolla | 00:28 | |
| *** godleon has joined #openstack-kolla | 00:35 | |
| godleon | Hi all ~ | 00:47 |
|---|---|---|
| *** weiyu has joined #openstack-kolla | 00:54 | |
| *** yuanying has quit IRC | 00:55 | |
| *** thumpba has joined #openstack-kolla | 00:59 | |
| *** ravig has joined #openstack-kolla | 01:08 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla: Adds Neutron LBaaS Support https://review.openstack.org/318322 | 01:12 |
| *** JoseMello has joined #openstack-kolla | 01:15 | |
| *** tfukushima has joined #openstack-kolla | 01:35 | |
| *** tfukushima has quit IRC | 01:37 | |
| *** sdake has joined #openstack-kolla | 01:39 | |
| *** sdake_ has joined #openstack-kolla | 01:52 | |
| sdake_ | sbezverk any luck? | 01:53 |
| sdake_ | sbezverk i was thinking if lbaas shares /run haproxy may not start upagain because of the pid file | 01:53 |
| *** sdake has quit IRC | 01:53 | |
| *** tfukushima has joined #openstack-kolla | 01:54 | |
| *** yuanying has joined #openstack-kolla | 01:59 | |
| *** rhallisey has joined #openstack-kolla | 02:06 | |
| *** daneyon_ has joined #openstack-kolla | 02:06 | |
| *** rhallisey has quit IRC | 02:07 | |
| *** sdake_ has quit IRC | 02:09 | |
| *** daneyon_ has quit IRC | 02:12 | |
| sbezverk | sdake_ dataplane is working :-) | 02:15 |
| *** ravig has quit IRC | 02:16 | |
| coolsvap | morning | 02:16 |
| *** sbezverk has quit IRC | 02:20 | |
| *** sbezverk has joined #openstack-kolla | 02:20 | |
| *** ravig has joined #openstack-kolla | 02:20 | |
| openstackgerrit | Ryan Hallisey proposed openstack/kolla-kubernetes: WIP - Implement a Replication Controller for Mariadb Bootstrap https://review.openstack.org/320742 | 02:25 |
| *** salv-orlando has joined #openstack-kolla | 02:27 | |
| openstackgerrit | Ryan Hallisey proposed openstack/kolla: WIP - Add kolla-kubernetes bootstrap capability to mariadb https://review.openstack.org/320744 | 02:30 |
| *** salv-orlando has quit IRC | 02:32 | |
| godleon | Is Kolla on Kubernetes working on progress ? | 02:37 |
| *** ravig has quit IRC | 02:42 | |
| *** tfukushima has quit IRC | 02:43 | |
| *** ravig has joined #openstack-kolla | 02:44 | |
| *** unicell has quit IRC | 02:44 | |
| *** vhosakot has joined #openstack-kolla | 02:51 | |
| *** absubram has joined #openstack-kolla | 03:06 | |
| *** absubram_ has joined #openstack-kolla | 03:10 | |
| *** absubram has quit IRC | 03:11 | |
| *** absubram_ is now known as absubram | 03:11 | |
| *** yuanying has quit IRC | 03:14 | |
| *** sdake has joined #openstack-kolla | 03:15 | |
| coolsvap | godleon, yes its in progress | 03:15 |
| coolsvap | sdake, hi | 03:16 |
| godleon | but how about on mesos ? | 03:19 |
| godleon | already been give up ? | 03:19 |
| *** unicell has joined #openstack-kolla | 03:21 | |
| godleon | and.......does Kubernetes support network=host & pid=host already? | 03:23 |
| *** vhosakot has quit IRC | 03:25 | |
| *** vhosakot has joined #openstack-kolla | 03:25 | |
| mark-casey | godleon: Yes I believe mesos was abandoned, and yes Kubernetes added the needed features to make kolla work which brought about the current work | 03:27 |
| *** sdake has quit IRC | 03:27 | |
| *** ssurana has quit IRC | 03:32 | |
| *** ssurana has joined #openstack-kolla | 03:34 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Pin the base distro release version https://review.openstack.org/319992 | 03:36 |
| *** ayoung has quit IRC | 03:42 | |
| *** phuongnh has joined #openstack-kolla | 03:43 | |
| *** sdake has joined #openstack-kolla | 03:44 | |
| sdake | evening folks | 03:44 |
| sdake | hey coolsvap | 03:44 |
| sdake | just got my intel 1.2tb ssd back | 03:45 |
| sdake | restoring/installing my workstation | 03:45 |
| coolsvap | :) | 03:45 |
| sdake | i am doing a fresh install of centos | 03:45 |
| sdake | and reovering windows if that is possible | 03:45 |
| sdake | i have backups but they are not working | 03:45 |
| sdake | so now I ran dd if=/dev/sda of=/dev/nvmen1 | 03:45 |
| sdake | to see if that works correctly | 03:45 |
| sdake | and yay, finally paid my taxes | 03:46 |
| coolsvap | :) | 03:47 |
| coolsvap | sdake, what is your view on https://review.openstack.org/#/c/317642/ ? | 03:48 |
| patchbot | coolsvap: patch 317642 - kolla - Make passwords.yml file generation configurable | 03:48 |
| sdake | coolsvap reviewing moment | 03:48 |
| sdake | coolsvap enjoy | 03:51 |
| godleon | If ansible will play the main deploy role after moving Kolla on Kubernetes? or Kubernetes will take over most of the deploy tasks? | 03:52 |
| sdake | any cores around to merge https://review.openstack.org/#/c/320589/ | 03:55 |
| patchbot | sdake: patch 320589 - kolla - change keyfile to key | 03:55 |
| sdake | its a serious obvious patch | 03:55 |
| sdake | godleon we are leaving ansible implementation as is | 03:55 |
| sdake | godleon kolla-kubernetes is experimental work, may not bear fruit, who knows | 03:55 |
| sdake | its our 3rd attempt at making a k8s based deployment architecture | 03:55 |
| sdake | if it works great, if it doesn't - still have ansible :) | 03:55 |
| sdake | godleon my estimate for kolla-kubernetes is 18 months to production or failure | 03:56 |
| godleon | 3rd ??? | 03:56 |
| *** salv-orlando has joined #openstack-kolla | 03:56 | |
| godleon | sdake: what are the first two ? | 03:57 |
| sdake | our initial attempt was to make openstack run on kubernetes to run on rhel atomic | 03:57 |
| sdake | our second attempt was mesos | 03:57 |
| sdake | the mesos roadmap got foxtrotted by microsoft investment, which triggered a pullout of all of the contributors | 03:57 |
| sdake | so that 2nd attempt failed | 03:57 |
| sdake | the 1st attempt failed because of lacking features in kubernetes | 03:58 |
| sdake | keep in mind it ws 0.97 rc3 or something | 03:58 |
| godleon | sdake: haha..........the main reason is microsoft. | 03:58 |
| sdake | even now kubernetes does not have all of the features we require to deploy openstack | 03:58 |
| openstackgerrit | Merged openstack/kolla: Make passwords.yml file generation configurable https://review.openstack.org/317642 | 03:59 |
| godleon | sdake: oh? Does kubernetes still miss anything? | 03:59 |
| sdake | unfortunately we dont totally understand what those features are | 03:59 |
| sdake | but we are strugglign in a few key areas | 03:59 |
| coolsvap | sdake, can you review(and ack) https://review.openstack.org/#/c/315390 https://review.openstack.org/#/c/317238 | 03:59 |
| sdake | and something needs to be done in kubernetes around there | 03:59 |
| coolsvap | there are 4-5 patches above it which are either acked or +2 waiting on these patches | 04:01 |
| *** salv-orlando has quit IRC | 04:02 | |
| godleon | sdake: Core OS released Stackanetes few weeks ago, have you ever seen that? It's a architecture of openstack on k8s, I am wondering if there any differences between the Stackanetes and what kolla-kubernetes is doing. | 04:02 |
| sdake | coolsvap pronouns | 04:03 |
| sdake | pet peeve ;) | 04:03 |
| coolsvap | sdake, let me do the edits | 04:05 |
| *** vhosakot has quit IRC | 04:06 | |
| *** unicell has quit IRC | 04:06 | |
| sdake | godleon stackanetes is based upon kolla-mesos | 04:06 |
| sdake | they took kolla-mesos, forked it about 6 weeks ago, and started working on it | 04:06 |
| sdake | godleon that is about as much as I know | 04:07 |
| sdake | mostly intel is doing the R&D on the project | 04:07 |
| sdake | they have 6 internal engineers | 04:07 |
| sdake | because of how they have structured their repo, they wont be able to enter the big tent without significant changes | 04:07 |
| sdake | but if they sort that out, what could possibly go wrong ;) | 04:08 |
| *** unicell has joined #openstack-kolla | 04:08 | |
| *** thumpba has quit IRC | 04:08 | |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed openstack/kolla: Simplify image building information in quickstart https://review.openstack.org/315390 | 04:09 |
| godleon | sdake: but core os declared Stakanetes is based on kubernetes...isn`t it? | 04:14 |
| *** sacharya has quit IRC | 04:16 | |
| coolsvap | sdake, https://review.openstack.org/#/c/315390/ | 04:19 |
| patchbot | coolsvap: patch 315390 - kolla - Simplify image building information in quickstart | 04:19 |
| sdake | its a mashup of kolla-mesos and kubernetes | 04:20 |
| *** ravig has quit IRC | 04:28 | |
| godleon | sdake: ok, got it. thanks for your explaination. :) | 04:28 |
| coolsvap | sdake, can you review https://review.openstack.org/#/c/315390 I made the changes for pronouns | 04:29 |
| *** yuanying has joined #openstack-kolla | 04:33 | |
| *** ravig has joined #openstack-kolla | 04:41 | |
| *** asalkeld has quit IRC | 04:49 | |
| *** daneyon_ has joined #openstack-kolla | 04:49 | |
| *** daneyon_ has quit IRC | 04:53 | |
| *** mark-casey has quit IRC | 05:03 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Pin the base distro release version https://review.openstack.org/319992 | 05:08 |
| *** sdake_ has joined #openstack-kolla | 05:11 | |
| *** dave-mccowan has quit IRC | 05:13 | |
| *** sdake has quit IRC | 05:13 | |
| *** sacharya has joined #openstack-kolla | 05:17 | |
| *** sacharya has quit IRC | 05:22 | |
| *** ravig has quit IRC | 05:24 | |
| *** tfukushima has joined #openstack-kolla | 05:26 | |
| *** salv-orlando has joined #openstack-kolla | 05:26 | |
| *** salv-orlando has quit IRC | 05:35 | |
| coolsvap | sdake_, Jeffrey4l can you review https://review.openstack.org/315390 | 05:40 |
| *** salv-orlando has joined #openstack-kolla | 05:51 | |
| *** salv-orlando has joined #openstack-kolla | 05:51 | |
| *** ssurana has quit IRC | 05:58 | |
| Jeffrey4l | ok | 06:03 |
| *** unicell has quit IRC | 06:15 | |
| *** ssurana has joined #openstack-kolla | 06:18 | |
| *** sacharya has joined #openstack-kolla | 06:18 | |
| *** ssurana has quit IRC | 06:19 | |
| *** sacharya has quit IRC | 06:23 | |
| *** unicell has joined #openstack-kolla | 06:24 | |
| *** shardy has joined #openstack-kolla | 06:41 | |
| *** ssurana has joined #openstack-kolla | 06:43 | |
| *** ssurana has quit IRC | 06:45 | |
| *** cu5 has joined #openstack-kolla | 06:51 | |
| *** unicell has quit IRC | 06:52 | |
| *** unicell has joined #openstack-kolla | 06:52 | |
| *** salv-orlando has quit IRC | 06:55 | |
| coolsvap | Jeffrey4l, any review comments for the change? | 06:56 |
| *** athomas has joined #openstack-kolla | 07:10 | |
| *** mikelk has joined #openstack-kolla | 07:17 | |
| *** JoseMello has quit IRC | 07:49 | |
| *** daneyon has quit IRC | 07:50 | |
| *** sdake_ has quit IRC | 07:56 | |
| *** tfukushima has quit IRC | 08:02 | |
| *** tfukushima has joined #openstack-kolla | 08:02 | |
| *** mgoddard has joined #openstack-kolla | 08:07 | |
| *** phuongnh has quit IRC | 08:09 | |
| *** salv-orlando has joined #openstack-kolla | 08:15 | |
| *** Serlex has joined #openstack-kolla | 08:19 | |
| openstackgerrit | Martin André proposed openstack/kolla: Simplify image building information in quickstart https://review.openstack.org/315390 | 08:19 |
| *** salv-orlando has quit IRC | 08:21 | |
| *** salv-orlando has joined #openstack-kolla | 08:21 | |
| *** dmk0202 has joined #openstack-kolla | 08:28 | |
| *** coolsvap is now known as coolsvap_ | 08:30 | |
| *** brad[] has quit IRC | 08:36 | |
| *** fragatina has quit IRC | 08:36 | |
| *** brad[] has joined #openstack-kolla | 08:37 | |
| *** anbehl has quit IRC | 08:41 | |
| *** sdake has joined #openstack-kolla | 08:43 | |
| *** pbourke_ has quit IRC | 08:47 | |
| *** pbourke_ has joined #openstack-kolla | 08:48 | |
| *** sdake has quit IRC | 08:49 | |
| *** sdake has joined #openstack-kolla | 08:51 | |
| *** daneyon has joined #openstack-kolla | 08:52 | |
| *** ccesario___ has joined #openstack-kolla | 08:53 | |
| *** daneyon has quit IRC | 08:55 | |
| *** mbound has joined #openstack-kolla | 08:55 | |
| *** daneyon has joined #openstack-kolla | 08:56 | |
| *** anbehl has joined #openstack-kolla | 08:58 | |
| sdake | coolsvap_ i added more comments to that last patch set | 09:05 |
| sdake | now I'm off to bed. | 09:05 |
| sdake | enjoy | 09:05 |
| *** sdake has quit IRC | 09:05 | |
| *** ccesario___ has quit IRC | 09:06 | |
| *** yuanying has quit IRC | 09:08 | |
| *** yuanying has joined #openstack-kolla | 09:10 | |
| *** gfidente has joined #openstack-kolla | 09:14 | |
| *** gfidente has joined #openstack-kolla | 09:14 | |
| *** liyi_ has joined #openstack-kolla | 09:30 | |
| liyi_ | Hi all, Just want to share this: https://github.com/kubernetes/kubernetes/pull/20698. It is a missing important feature to get kolla worked with K8s | 09:30 |
| wmiller | Does anyone know if it's intentional that, when running kolla-build, a dict of all the images built/not built is printed (even after they've all been pretty printed already)? It looks to me like a side effect from the setuptools entrypoint running sys.exit(main()). | 09:44 |
| wmiller | I'm interested because I recently hit Launchpad bug 1578404, and am looking for the best way to fix it. | 09:45 |
| openstack | Launchpad bug 1578404 in kolla "kolla-build return wrong exit code" [Undecided,New] https://launchpad.net/bugs/1578404 | 09:45 |
| *** coolsvap_ is now known as coolsvap | 09:50 | |
| *** ccesario___ has joined #openstack-kolla | 09:52 | |
| *** fragatina has joined #openstack-kolla | 09:55 | |
| *** ccesario___ has quit IRC | 09:56 | |
| *** fragatina has quit IRC | 09:59 | |
| *** mummer has joined #openstack-kolla | 10:01 | |
| *** sacharya has joined #openstack-kolla | 10:19 | |
| *** sacharya has quit IRC | 10:24 | |
| *** ccesario___ has joined #openstack-kolla | 10:27 | |
| *** fragatina has joined #openstack-kolla | 10:55 | |
| *** tfukushima has quit IRC | 10:58 | |
| *** tfukushima has joined #openstack-kolla | 10:59 | |
| *** tfukushima has quit IRC | 10:59 | |
| *** fragatina has quit IRC | 11:00 | |
| *** liyi_ has quit IRC | 11:09 | |
| *** openstackgerrit has quit IRC | 11:18 | |
| *** openstackgerrit has joined #openstack-kolla | 11:18 | |
| *** mliima has joined #openstack-kolla | 11:20 | |
| *** weiyu has quit IRC | 11:21 | |
| mliima | morning guys | 11:24 |
| *** ccesario___ has quit IRC | 11:30 | |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed openstack/kolla: Update quickstart guide for formatting https://review.openstack.org/317205 | 11:34 |
| *** fragatina has joined #openstack-kolla | 11:37 | |
| *** sean-k-mooney has quit IRC | 11:42 | |
| *** fragatina has quit IRC | 11:42 | |
| *** tfukushima has joined #openstack-kolla | 11:44 | |
| *** tfukushi_ has joined #openstack-kolla | 11:45 | |
| *** jmccarthy has joined #openstack-kolla | 11:47 | |
| *** tfukushima has quit IRC | 11:48 | |
| coolsvap | mliima, mandre can you approve https://review.openstack.org/#/c/317205 it has 2 +2s just rebased | 11:50 |
| *** vincent_vdk has quit IRC | 11:50 | |
| mliima | coolsvap, done | 11:54 |
| coolsvap | mliima, thanks! | 11:55 |
| *** athomas has quit IRC | 11:58 | |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed openstack/kolla: Update quickstart guide for formatting https://review.openstack.org/317205 | 11:58 |
| *** mliima has quit IRC | 12:00 | |
| *** vincent_vdk has joined #openstack-kolla | 12:02 | |
| *** athomas has joined #openstack-kolla | 12:03 | |
| openstackgerrit | Martin André proposed openstack/kolla: Simplify image building information in quickstart https://review.openstack.org/315390 | 12:07 |
| openstackgerrit | Martin André proposed openstack/kolla: Update the image-building documentation https://review.openstack.org/320933 | 12:07 |
| *** fragatina has joined #openstack-kolla | 12:08 | |
| *** godleon has quit IRC | 12:09 | |
| *** mummer has quit IRC | 12:12 | |
| *** fragatina has quit IRC | 12:12 | |
| *** mliima has joined #openstack-kolla | 12:12 | |
| *** dwalsh has joined #openstack-kolla | 12:16 | |
| openstackgerrit | Merged openstack/kolla: Update quickstart guide for formatting https://review.openstack.org/317205 | 12:16 |
| *** fragatina has joined #openstack-kolla | 12:18 | |
| *** thumpba has joined #openstack-kolla | 12:19 | |
| *** thumpba has quit IRC | 12:19 | |
| *** sacharya has joined #openstack-kolla | 12:21 | |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed openstack/kolla: Update ubuntu dockerfiles for formatting https://review.openstack.org/317238 | 12:24 |
| *** fragatina has quit IRC | 12:24 | |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed openstack/kolla: Add support for magnum ubuntu binary containers https://review.openstack.org/317251 | 12:25 |
| *** sacharya has quit IRC | 12:25 | |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed openstack/kolla: Add support to lint the install commands in Dockerfiles https://review.openstack.org/317364 | 12:25 |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed openstack/kolla: Add ubuntu binary container for zaqar https://review.openstack.org/317276 | 12:25 |
| *** jmccarthy has left #openstack-kolla | 12:26 | |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed openstack/kolla: Add support for murano binary containers for rpm based distros https://review.openstack.org/317270 | 12:26 |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed openstack/kolla: Add sahara containers https://review.openstack.org/317261 | 12:26 |
| *** fragatina has joined #openstack-kolla | 12:29 | |
| *** jmccarthy has joined #openstack-kolla | 12:29 | |
| *** ccesario has quit IRC | 12:33 | |
| *** fragatina has quit IRC | 12:33 | |
| *** godleon has joined #openstack-kolla | 12:36 | |
| openstackgerrit | Bertrand Lallau proposed openstack/kolla: Remove unnecessary executable permissions https://review.openstack.org/320960 | 12:38 |
| *** jmccarthy has left #openstack-kolla | 12:42 | |
| *** absubram has quit IRC | 12:47 | |
| *** ccesario has joined #openstack-kolla | 12:49 | |
| *** absubram has joined #openstack-kolla | 12:50 | |
| openstackgerrit | Mauricio Lima proposed openstack/kolla: Update Documentation https://review.openstack.org/319899 | 12:55 |
| * coolsvap will be back later in evening | 12:59 | |
| *** ayoung has joined #openstack-kolla | 13:01 | |
| *** SiRiuS__ has joined #openstack-kolla | 13:02 | |
| *** mikelk has quit IRC | 13:05 | |
| *** vincent_vdk has quit IRC | 13:05 | |
| *** vincent_vdk has joined #openstack-kolla | 13:06 | |
| *** mikelk has joined #openstack-kolla | 13:06 | |
| *** dave-mccowan has joined #openstack-kolla | 13:07 | |
| openstackgerrit | Mauricio Lima proposed openstack/kolla: Update Documentation https://review.openstack.org/319899 | 13:08 |
| *** sbezverk has quit IRC | 13:10 | |
| *** sbezverk has joined #openstack-kolla | 13:12 | |
| *** anbehl has quit IRC | 13:20 | |
| *** inc0 has joined #openstack-kolla | 13:22 | |
| *** anbehl has joined #openstack-kolla | 13:23 | |
| *** fragatina has joined #openstack-kolla | 13:29 | |
| *** dwalsh has quit IRC | 13:30 | |
| *** fragatina has quit IRC | 13:35 | |
| openstackgerrit | Liyi Meng proposed openstack/kolla-kubernetes: Implment swift bootstrap job https://review.openstack.org/320316 | 13:35 |
| *** inc0 has quit IRC | 13:43 | |
| *** mgoddard_ has joined #openstack-kolla | 13:44 | |
| *** mgoddard has quit IRC | 13:48 | |
| *** salv-orl_ has joined #openstack-kolla | 13:56 | |
| *** SiRiuS__ has quit IRC | 13:59 | |
| *** salv-orlando has quit IRC | 13:59 | |
| *** fragatina has joined #openstack-kolla | 14:01 | |
| *** jtriley has joined #openstack-kolla | 14:02 | |
| *** dwalsh has joined #openstack-kolla | 14:02 | |
| *** inc0 has joined #openstack-kolla | 14:02 | |
| *** fragatina has quit IRC | 14:08 | |
| *** haplo37_ has joined #openstack-kolla | 14:10 | |
| openstackgerrit | Mauricio Lima proposed openstack/kolla: Update Documentation https://review.openstack.org/319899 | 14:12 |
| *** mgoddard_ has quit IRC | 14:13 | |
| *** rhallisey has joined #openstack-kolla | 14:15 | |
| openstackgerrit | Mauricio Lima proposed openstack/kolla: Update Documentation https://review.openstack.org/319899 | 14:15 |
| *** diogogmt has quit IRC | 14:20 | |
| *** sacharya has joined #openstack-kolla | 14:21 | |
| *** liyi has quit IRC | 14:22 | |
| *** mgoddard has joined #openstack-kolla | 14:25 | |
| mgkwill | Can anyone point me to place in source where external ports are defined for kolla containers? Trying to get ironic-inspector working and it has service url fqdn:5050 but it is not exposed when container is deployed. | 14:25 |
| *** Avi82 has joined #openstack-kolla | 14:26 | |
| rhallisey | mgkwill, ansible/group_vars/all.yml | 14:26 |
| *** sacharya has quit IRC | 14:26 | |
| rhallisey | that's from inside the kolla dir | 14:26 |
| mgkwill | rhallisey: thanks, i'll take a look. | 14:26 |
| rhallisey | sure | 14:26 |
| *** inc0_ has joined #openstack-kolla | 14:30 | |
| *** brad[] has quit IRC | 14:30 | |
| *** ravig has joined #openstack-kolla | 14:30 | |
| *** ravig has quit IRC | 14:32 | |
| *** salv-orl_ has quit IRC | 14:33 | |
| *** inc0 has quit IRC | 14:34 | |
| *** ravig has joined #openstack-kolla | 14:35 | |
| *** SiRiuS has joined #openstack-kolla | 14:45 | |
| *** godleon has quit IRC | 14:49 | |
| *** daneyon_ has joined #openstack-kolla | 14:50 | |
| *** jtriley has quit IRC | 14:54 | |
| *** daneyon_ has quit IRC | 14:55 | |
| *** SiRiuS has quit IRC | 14:55 | |
| *** efoley has joined #openstack-kolla | 14:56 | |
| d_code | how can I add a bind mount for nova-compute? I’m trying to do pci passthrough, and nova is failing b/c it cannot access “/dev/vfio/*” | 14:57 |
| d_code | bind mounting /dev/vfio:/dev/vfio might suffice | 14:57 |
| *** ravig has quit IRC | 15:00 | |
| *** cu5 has quit IRC | 15:01 | |
| *** SiRiuS has joined #openstack-kolla | 15:02 | |
| *** fragatina has joined #openstack-kolla | 15:03 | |
| *** vhosakot has joined #openstack-kolla | 15:06 | |
| *** fragatina has quit IRC | 15:08 | |
| dims | rhallisey : sbezverk : how far we away from duplicating the demo at austin? (with kolla-kubernetes) | 15:09 |
| *** sdake has joined #openstack-kolla | 15:09 | |
| rhallisey | haha | 15:09 |
| dims | :) | 15:09 |
| rhallisey | dims, if we wanted to go that route we could run kolla-ansible in pods | 15:09 |
| rhallisey | dims, once we get this bootstrap bit sorted, I think the cluster will start taking form | 15:10 |
| *** mliima has quit IRC | 15:12 | |
| *** mgoddard_ has joined #openstack-kolla | 15:12 | |
| sbezverk | dims: rhallisey: agreed, once mariadb is operational, the rest will be built upon it.. | 15:13 |
| *** sdake_ has joined #openstack-kolla | 15:13 | |
| *** sean-k-mooney has joined #openstack-kolla | 15:13 | |
| *** sdake has quit IRC | 15:13 | |
| *** mliima has joined #openstack-kolla | 15:14 | |
| *** mgoddard has quit IRC | 15:16 | |
| openstackgerrit | Mauricio Lima proposed openstack/kolla: Update Documentation https://review.openstack.org/319899 | 15:16 |
| sean-k-mooney | coolsvap: Quick Q. you were working on adding support for ceilometer,gnocchi and aodh. is there any way to deploy those container currently? | 15:16 |
| d_code | k…figured out how to adjust that in the nova-compute role. I’ll give that a try | 15:18 |
| *** diogogmt has joined #openstack-kolla | 15:19 | |
| *** SiRiuS has quit IRC | 15:20 | |
| dims | rhallisey : the demo was rather simple right? nova starting a vm, nuking horizon to get a newer version of horizon from what i remember | 15:21 |
| rhallisey | dims, ya they had everything deployed. Started a vm and upgraded horizon | 15:22 |
| rhallisey | they also showed how kube will reschedule things | 15:22 |
| dims | sbezverk rhallisey : so when we get there, we can do a milestone of some kind? | 15:23 |
| *** berendt has quit IRC | 15:23 | |
| rhallisey | I'm not how we will track things yet | 15:24 |
| dims | ack thanks | 15:24 |
| rhallisey | we can do that if you like | 15:24 |
| dims | rhallisey : it will help the cause i think (plus i can use it to drum beat ) | 15:25 |
| rhallisey | awesome | 15:25 |
| *** huikang has joined #openstack-kolla | 15:25 | |
| *** SiRiuS has joined #openstack-kolla | 15:27 | |
| *** jtriley has joined #openstack-kolla | 15:28 | |
| *** salv-orlando has joined #openstack-kolla | 15:33 | |
| *** sdake_ has quit IRC | 15:34 | |
| *** efoley has quit IRC | 15:35 | |
| *** ravig has joined #openstack-kolla | 15:35 | |
| *** liyi has joined #openstack-kolla | 15:36 | |
| sbezverk | rhallisey: have you played already with etcd and curl? | 15:37 |
| rhallisey | sbezverk, ya I almost have it working | 15:37 |
| *** salv-orlando has quit IRC | 15:38 | |
| sbezverk | rhallisey: super!!! | 15:38 |
| sbezverk | will wait for your PS then.. | 15:39 |
| *** huikang has quit IRC | 15:41 | |
| *** mikelk has quit IRC | 15:41 | |
| *** huikang has joined #openstack-kolla | 15:41 | |
| *** SiRiuS has quit IRC | 15:43 | |
| *** ssurana has joined #openstack-kolla | 15:45 | |
| *** huikang has quit IRC | 15:46 | |
| *** huikang has joined #openstack-kolla | 15:48 | |
| *** dmk0202 has quit IRC | 15:49 | |
| *** sacharya has joined #openstack-kolla | 15:51 | |
| *** SamYaple has joined #openstack-kolla | 15:51 | |
| *** SamYaple has left #openstack-kolla | 15:51 | |
| *** ravig has quit IRC | 15:54 | |
| *** ravig has joined #openstack-kolla | 15:59 | |
| openstackgerrit | Merged openstack/kolla-kubernetes: Implement neutron pods and service https://review.openstack.org/319293 | 15:59 |
| *** fragatina has joined #openstack-kolla | 16:02 | |
| *** fragatina has quit IRC | 16:06 | |
| *** anbehl has quit IRC | 16:07 | |
| *** jtriley has quit IRC | 16:08 | |
| *** athomas has quit IRC | 16:08 | |
| *** jtriley has joined #openstack-kolla | 16:09 | |
| *** athomas has joined #openstack-kolla | 16:09 | |
| *** athomas has quit IRC | 16:10 | |
| rhallisey | hmm | 16:10 |
| rhallisey | so we need to boostrap mariadb, then run with --wsrep-new-cluster, then start the service | 16:11 |
| *** fbarilla has joined #openstack-kolla | 16:12 | |
| fbarilla | hi, tempest container does not build. Dockerfile has 'ADD tempest-archive /tempest-source' but tempest-archive does not exist.... | 16:17 |
| liyi | hi Ryan, instead of using ectd directory, can be used configmap to exchange data? | 16:17 |
| *** mbound has quit IRC | 16:18 | |
| rhallisey | liyi, hey. We could except that requires ceph | 16:18 |
| rhallisey | because services arrn't guarunteed to be on the same host | 16:18 |
| liyi | yaha I thought configmap is backed with ectd internally | 16:19 |
| liyi | i need to have a look into this. | 16:19 |
| rhallisey | oh sorry you said configmap | 16:20 |
| rhallisey | uh | 16:20 |
| rhallisey | I misunderstood | 16:20 |
| liyi | no problem | 16:20 |
| liyi | just my two cents, I have not tested it out | 16:20 |
| liyi | if my speculation is right, thing might be simplified | 16:21 |
| liyi | I will have a look into this | 16:21 |
| rhallisey | so we could do that vs ectd | 16:21 |
| sbezverk | liyi etcd looks as a more reliable option | 16:21 |
| *** anbehl has joined #openstack-kolla | 16:21 | |
| *** jtriley has quit IRC | 16:21 | |
| rhallisey | but I don't think is solves the main issue | 16:21 |
| sbezverk | liyi and what is more important it is persistent | 16:22 |
| liyi | right, it just a way to share state | 16:22 |
| fbarilla | as advertised on the OpenStack Web Site : 'Ceph In Kolla' The out-of-the-box Ceph deployment requires 3 hosts with at least one block device on each host that can be dedicated for sole use by Ceph. However, with tweaks to the Ceph cluster you can deploy a “healthy” cluster with a single host and a single block device. | 16:22 |
| liyi | sezverk, configmao might lost data? | 16:23 |
| rhallisey | liyi, well if it gets deleted | 16:23 |
| fbarilla | is the single system Ceph configuration documented somewhere ? | 16:23 |
| rhallisey | vs killing etcd entirely | 16:23 |
| *** athomas has joined #openstack-kolla | 16:24 | |
| *** jtriley has joined #openstack-kolla | 16:25 | |
| liyi | Ryan, we are supposed to delete a configmap which is still in used, right? If configmap is backed by the internal etcd service, then the difference is not much significant | 16:25 |
| rhallisey | if the configmap recovers it's data thats fine | 16:27 |
| rhallisey | seems like it would, so then that would work | 16:27 |
| liyi | Let me have some investigation into it. will report back to you later. | 16:28 |
| rhallisey | ok cool | 16:28 |
| *** sdake has joined #openstack-kolla | 16:29 | |
| sdake | morning | 16:30 |
| sdake | mliima ping | 16:30 |
| sdake | mliima one thing i forgot to metnion | 16:30 |
| mliima | \o/ | 16:30 |
| mliima | i'm here | 16:30 |
| sdake | youcan't approve your own patches or patches in which your listed on the coauthor line | 16:30 |
| *** ccesario___ has joined #openstack-kolla | 16:30 | |
| sdake | you can +1 only | 16:30 |
| mliima | ok :) | 16:31 |
| sdake | if you cherry-pick to a stable branch and a cherrypick reuslts in no conflicts, you may approve the cherrypick | 16:31 |
| sdake | (but still requires 2 +2s | 16:31 |
| sbezverk | sdake morning, did you see, lbaas dataplane is working :-) | 16:31 |
| sdake | sbezverk rockin dude | 16:31 |
| sdake | sbezverk what was the issue | 16:31 |
| sbezverk | sdake these guys are smart, instead of running centralized daemon, they run haproxy per namespace in foreground :-) to do that haproxy binary must be present in lbaas image | 16:32 |
| sdake | cool | 16:33 |
| sdake | so that sounds like an easy fix | 16:33 |
| sbezverk | sdake oh it is fixed in the latest PS, just need cores to ack it | 16:34 |
| sdake | sbezverk quick questison, will this work with liberty ass is? | 16:34 |
| sbezverk | sdake I do not see why not, lbaasv2 was there also there is no other funcy dependencies | 16:36 |
| sdake | sbezverk link please | 16:36 |
| *** jtriley has quit IRC | 16:36 | |
| sbezverk | sdake https://review.openstack.org/318322 | 16:36 |
| sdake | sbezverk can you cherrypick that intnerally to a liberty branch | 16:38 |
| sdake | and see if it works? | 16:38 |
| sdake | by internally I mean on your local workstation | 16:38 |
| *** daneyon_ has joined #openstack-kolla | 16:38 | |
| sbezverk | sdake I will not be able to test it as it requires a couple of servers behind it | 16:39 |
| sbezverk | to do that I use cinder as I do not use other backends | 16:40 |
| *** unicell has quit IRC | 16:40 | |
| sbezverk | and since iscsi is not in mitake or liberty, it makes my life difficukt | 16:40 |
| *** jtriley has joined #openstack-kolla | 16:41 | |
| sdake | sbezverk shouldn't the lbaas agent run on the networking nodes? | 16:41 |
| sdake | sbezverk if you have an empty disk you can setup ceph single node | 16:42 |
| sdake | it is super easy | 16:42 |
| sdake | sbezverk the /etc/localtime | 16:43 |
| *** daneyon_ has quit IRC | 16:43 | |
| sdake | why is that necessary? | 16:43 |
| sbezverk | sdake yes, it should run on network node | 16:45 |
| *** ccesario___ has quit IRC | 16:46 | |
| sbezverk | sdake it is necessary at least for me, because I do not know the other way of bringing up a web server other than to build centos vm and install apache on it. | 16:46 |
| sdake | sbezverk what i mean is why is /etc/localtime necessary as a bindmount | 16:47 |
| sdake | patch looks really good | 16:47 |
| sdake | missing rpm installs | 16:47 |
| sdake | need that in the code | 16:48 |
| sdake | ubuntu binary is a mess, but if you can figure out how to get that building for lbaas that would be good too | 16:48 |
| sdake | ubuntu binary sort of works in mitaka | 16:48 |
| sdake | as in cmopute kit works iirc | 16:48 |
| sdake | i left review comments sbezverk | 16:48 |
| sdake | mliima can you review sbezverk 'z patch eh linked above plz | 16:49 |
| mliima | this ps https://review.openstack.org/318322 : | 16:49 |
| mliima | ? | 16:49 |
| sdake | yup | 16:49 |
| *** salv-orlando has joined #openstack-kolla | 16:50 | |
| rhallisey | sdake, /etc/localtime has to do with ceph | 16:50 |
| *** Jeffrey4l has quit IRC | 16:51 | |
| *** Serlex has quit IRC | 16:51 | |
| *** huikang has quit IRC | 16:52 | |
| rhallisey | if you change the timezone on your host, galera adn ceph could break since the containers will think it's a different time zone | 16:52 |
| *** athomas has quit IRC | 17:00 | |
| *** huikang has joined #openstack-kolla | 17:01 | |
| *** fragatina has joined #openstack-kolla | 17:03 | |
| *** fragatina has quit IRC | 17:07 | |
| *** mgoddard_ has quit IRC | 17:07 | |
| *** mgoddard_ has joined #openstack-kolla | 17:07 | |
| *** SiRiuS has joined #openstack-kolla | 17:08 | |
| *** unicell has joined #openstack-kolla | 17:10 | |
| *** ccesario___ has joined #openstack-kolla | 17:12 | |
| *** brad[] has joined #openstack-kolla | 17:18 | |
| *** dwalsh has quit IRC | 17:25 | |
| sbezverk | sdake ping | 17:26 |
| *** tfukushi_ has quit IRC | 17:27 | |
| *** berendt has joined #openstack-kolla | 17:28 | |
| *** SiRiuS has quit IRC | 17:29 | |
| sdake | sbezverk shoot | 17:30 |
| *** ravig has quit IRC | 17:32 | |
| *** ravig has joined #openstack-kolla | 17:32 | |
| *** dwalsh has joined #openstack-kolla | 17:32 | |
| *** salv-orlando has quit IRC | 17:32 | |
| *** sdake_ has joined #openstack-kolla | 17:33 | |
| sdake_ | sbezverk sorry got disconnected | 17:34 |
| sdake_ | say I have a 10gig switch which has VLAN support | 17:34 |
| sdake_ | how should I use the VLANs? | 17:34 |
| sdake_ | this is connected to a 5 machine clsuter in my house | 17:34 |
| sbezverk | sdake all these combinations of binary/source and then dependecy that neutron-server needs lbaas makes docker files for neutron-base,neutron-server and lbaas look really spaghetti like | 17:34 |
| *** sdake has quit IRC | 17:35 | |
| sbezverk | sdake you need to configure routing between vlans if you need connectivity between them | 17:36 |
| sbezverk | sdake I need to re-work it again to try to make it more intuitive and simple.. | 17:37 |
| *** ravig has quit IRC | 17:39 | |
| openstackgerrit | Christian Berendt proposed openstack/kolla: Add title to index page of developer documentation https://review.openstack.org/320499 | 17:40 |
| *** ravig has joined #openstack-kolla | 17:40 | |
| sdake_ | sbezverk I don't need routing I am just curious what exactly value a vlan adds | 17:41 |
| sdake_ | is it just hardware based network isolation for particular traffic? | 17:41 |
| sdake_ | and if so, does each interface get its own vlan identifier? | 17:41 |
| *** huikang has quit IRC | 17:41 | |
| sdake_ | sbezverk what is a good linux network speed tester | 17:42 |
| *** mgoddard_ has quit IRC | 17:42 | |
| *** gfidente has quit IRC | 17:42 | |
| sdake_ | i got my 10gig switch just hooked up again, spending this morning getting my office into shape | 17:42 |
| sdake_ | sbezverk something simple that any idiot can cut and paste two commands and have a # come out the other end | 17:43 |
| sbezverk | sdake_ yes it is isolation, you need to define vlan which will be a layer 3 interface, then you assign ports to be a member of a particular VLAN, so hosts in th same vlan can communicate between each other | 17:44 |
| sbezverk | sdake_ I am not aware of any good opensource traffic gen.. | 17:45 |
| *** ravig has quit IRC | 17:45 | |
| sbezverk | sdake_ we use ixia and spirent but they are very expensive commerical products | 17:45 |
| sdake_ | so i could have one port be my management network vlan and another port be my storage network vlan? | 17:46 |
| sdake_ | is it only isolation? | 17:46 |
| sbezverk | sdake_ in general yes | 17:46 |
| sbezverk | sdake_ you can have bunch of other additional/optional features | 17:47 |
| *** ravig has joined #openstack-kolla | 17:48 | |
| sbezverk | sdake_ if you need your management network talking to storage network, then you will need routing | 17:48 |
| sdake_ | what i really want is to run my management and storage networks at large mtu | 17:48 |
| sdake_ | those networks are intended to be isolated | 17:48 |
| sdake_ | what I do need is my internet conneccting to my management network | 17:48 |
| sbezverk | sdake_ depending on switch hardware it could support up to 9k mtu | 17:48 |
| sdake_ | but I think what i'll do there is just use the onboard 1gig nic for my api network | 17:49 |
| sdake_ | it does indeed support 9k mtu | 17:49 |
| *** SiRiuS__ has joined #openstack-kolla | 17:49 | |
| sdake_ | after 6 years of writing totem, the #1 thing you could do to improve performance was increasse mtu | 17:50 |
| sdake_ | (at the expense of latency) | 17:50 |
| sbezverk | sdake_ careful with that I had some issue with 9k mtu when I tried to attach 9k enabled interfaces to ovs switch | 17:52 |
| sdake_ | this netgear switch has windows management software | 17:52 |
| *** esharao has joined #openstack-kolla | 17:53 | |
| sdake_ | sbezverk i am just running 9k mtu on my non-neutron interfaces | 17:53 |
| *** salv-orlando has joined #openstack-kolla | 17:53 | |
| sdake_ | sweet i have two 10gig ports now | 17:53 |
| sbezverk | sdake_ got it, have no experience using netgear, I prefer nexus with 32 40g ports ;-) | 17:53 |
| sdake_ | well ya but i had to buy the gear myself | 17:54 |
| sdake_ | i bought this about 6 months ago | 17:54 |
| sdake_ | it was 850$ for 8 ports of 10gig | 17:54 |
| sdake_ | which IMNSHO is a winner for small businesses or home offices | 17:55 |
| sdake_ | [ 3] 0.0-10.0 sec 11.0 GBytes 9.41 Gbits/sec | 17:58 |
| sdake_ | this is without jumbo framess | 17:58 |
| sdake_ | do you think jumbo frames would go faster? | 17:58 |
| openstackgerrit | Ryan Hallisey proposed openstack/kolla-kubernetes: WIP - Implement a Replication Controller for Mariadb Bootstrap https://review.openstack.org/320742 | 17:59 |
| sbezverk | sdake_ you already at 9.41 Gbps, where do you want to go faster? | 18:00 |
| openstackgerrit | Ryan Hallisey proposed openstack/kolla-kubernetes: WIP - Implement a Replication Controller for Mariadb Bootstrap https://review.openstack.org/320742 | 18:00 |
| sdake_ | i dunno I just have thought jumbo frames was sfaster | 18:00 |
| sdake_ | it was when i did years of gige development | 18:00 |
| sdake_ | i mean why bother with jumbo frames if it adds no value? | 18:01 |
| sbezverk | sdake_ I restored default mtu on my network gear, I have not seen obvious perf. advantages, but I surely saw some compatibility issues | 18:02 |
| rhallisey | sbezverk liyi, sitll doesn't work yet :/. I'm not able to get mariadb to recover | 18:03 |
| *** fragatina has joined #openstack-kolla | 18:03 | |
| sbezverk | rhallisey so after bootstrap completion, mariadb is not coming up? | 18:04 |
| rhallisey | get this: mysqld from pid file /var/lib/mysql/mariadb.pid ended | 18:04 |
| sdake_ | this netgear switch has a built in cable tester | 18:05 |
| sdake_ | pretty sweet | 18:05 |
| sbezverk | rhallisey we saw that way back even in kolla, but it was image related problem.. | 18:06 |
| *** fragatina has quit IRC | 18:08 | |
| *** sbezverk has left #openstack-kolla | 18:08 | |
| openstackgerrit | Ryan Hallisey proposed openstack/kolla-kubernetes: WIP - Implement a Replication Controller for Mariadb Bootstrap https://review.openstack.org/320742 | 18:12 |
| *** wirehead_ has quit IRC | 18:15 | |
| d_code | sdake_: got the SR-IOV full pass through to work…after a week of racking my brain (and my keyboard) | 18:16 |
| d_code | I’ll be filing a bug and the docs need…well…we need docs | 18:16 |
| sean-k-mooney | hi | 18:16 |
| *** wirehead_ has joined #openstack-kolla | 18:17 | |
| d_code | hi _o/ | 18:17 |
| sean-k-mooney | has anyone tried to boot a vm via horizon without a cinder volumn lately? | 18:17 |
| *** fragatina has joined #openstack-kolla | 18:17 | |
| sean-k-mooney | i built and deployed kolla from master earlier today and dont think you can | 18:17 |
| sean-k-mooney | not sure if its a horizon bug or a kolla config bug | 18:18 |
| sean-k-mooney | also d_code glad to hear you got it working | 18:19 |
| sean-k-mooney | what was the issue in the end? | 18:19 |
| openstackgerrit | Ryan Hallisey proposed openstack/kolla-kubernetes: Implement a Job for Mariadb Bootstrap https://review.openstack.org/320742 | 18:20 |
| sdake_ | dims ping | 18:20 |
| dims | sdake_ : pong | 18:20 |
| rhallisey | ^^ I think that works | 18:20 |
| rhallisey | testing.. | 18:20 |
| sdake_ | d_code fantastic - wtb docs, I think other people want the same thing, but I have no idea where to begin | 18:20 |
| sdake_ | sean-k-mooney there is a horizon bug where you can't boot at all until you add the _member_ role | 18:21 |
| sean-k-mooney | sdake_: oh ok did not hit that. i just started using the openstack client and it works but im lazy and miss horizon | 18:21 |
| sdake_ | sean-k-mooney all you ahve to do is add a _member_ role and horizon will work | 18:22 |
| sdake_ | v1k0d3n has a patch in teh queue that needs a little R&D attention and then that problem will be fixed | 18:23 |
| *** sdake_ is now known as sdake | 18:23 | |
| sean-k-mooney | sdake_: well for me i can boot but it complain about not being able to allocate the block device | 18:23 |
| sdake | sean-k-mooney oh with cinder | 18:23 |
| sdake | are you using ceph storage? | 18:23 |
| sean-k-mooney | sdake: is that the behavior that is expected without the _member_ role | 18:23 |
| sean-k-mooney | nope | 18:24 |
| sdake | sean-k-mooney not sure about cinder | 18:24 |
| sdake | what type of cinder storage are you using? | 18:24 |
| sean-k-mooney | sdake: i did not deploy cinder | 18:24 |
| *** fragatina has quit IRC | 18:24 | |
| d_code | sdake: yeah, I think there were some updates in Mitaka that changed _slightly_ how it works | 18:24 |
| sdake | d_code if you could add docs for it, atleast get a rough draft up, I'd be happy to take it over for you | 18:24 |
| sean-k-mooney | sdake: the issue is on the source tab in horizon when you select boot form image it still requires a volume for some reason of at least 1G | 18:25 |
| d_code | I can’t really discuss it now…but the “bug” is that either nova-compute or nova-libvirt (or both) requires access to /dev/vfio, so I added /dev:/dev to the mounts for those to containers | 18:25 |
| sdake | sean-k-mooney ya that comes directly from the nova named volume though | 18:25 |
| d_code | I’ll create a bug, tonight or tomorrow adn we can discuss the best approach | 18:25 |
| sean-k-mooney | sdake: nova named volume ? | 18:26 |
| d_code | for the scheduler, I had to add "device_type": "type-PF" to the pci_alias | 18:26 |
| d_code | k…I’m out | 18:26 |
| d_code | bbl | 18:26 |
| sdake | sean-k-mooney so thee is a named volume for nova, which gets imported into the nova vm for storage of disk data | 18:27 |
| sdake | this is not used if you use ceph | 18:27 |
| sdake | but it is used otherwise | 18:27 |
| sdake | the problem with this model is there is no centralized backup | 18:27 |
| sean-k-mooney | sdake: ah you mean a docker volumn not cinder | 18:27 |
| sdake | right docker named volume | 18:27 |
| sean-k-mooney | ya i normally do deploy cinder but i was jsut trying to keep it minimal with out it while i was trying to get multi node to work | 18:28 |
| *** ravig has quit IRC | 18:28 | |
| *** fragatina has joined #openstack-kolla | 18:30 | |
| sean-k-mooney | sdake: http://picpaste.com/horizon-IsnSMR2S.PNG this is the view that i am given when i try to boot without cinder deployed | 18:30 |
| sdake | sean-k-mooney i am not sure, but can you try adding the role? | 18:31 |
| sean-k-mooney | sdake: as you can see horizon still seam to try and allocate a cinder volumn too | 18:31 |
| sdake | its a one liner | 18:31 |
| sdake | openstack role add _member_ | 18:31 |
| sean-k-mooney | sure ill give that a try | 18:31 |
| sdake | sean-k-mooney another optoin is to ask in #openstack-horizon | 18:31 |
| sdake | if that comes sback with config errors on our end, we will fix it | 18:31 |
| sean-k-mooney | yep i did/ will again | 18:32 |
| sdake | if it comes back as "horizon requires cinder" I think that is silly | 18:32 |
| sdake | sean-k-mooney ify ou want to do that now, I can help ;) | 18:32 |
| sean-k-mooney | am sure i just added the role so im quickly going to try and boot and see if its the cause | 18:32 |
| sean-k-mooney | ok its still broken so ill pop back into the horizon channel and ask again | 18:34 |
| *** fragatina has quit IRC | 18:34 | |
| *** jeh has joined #openstack-kolla | 18:34 | |
| *** jeh has quit IRC | 18:34 | |
| *** jeh has joined #openstack-kolla | 18:34 | |
| *** sdake_ has joined #openstack-kolla | 18:34 | |
| *** sdake has quit IRC | 18:37 | |
| *** mark-casey has joined #openstack-kolla | 18:43 | |
| *** jtriley has quit IRC | 18:44 | |
| *** ravig has joined #openstack-kolla | 18:45 | |
| sdake_ | mliima ping re your last eail | 18:46 |
| sdake_ | while you may think its a privilege, its more like a responsibility ;) | 18:47 |
| mliima | i'm here | 18:47 |
| sdake_ | eali/email | 18:47 |
| *** sdake_ is now known as sdake | 18:48 | |
| mliima | "With +2 power, come great responsibility" | 18:48 |
| mliima | :) | 18:48 |
| sdake | ;-) | 18:49 |
| *** ravig has quit IRC | 18:52 | |
| *** ravig has joined #openstack-kolla | 18:52 | |
| *** daneyon_ has joined #openstack-kolla | 18:53 | |
| *** daneyon_ has quit IRC | 18:57 | |
| *** rmart04 has joined #openstack-kolla | 18:58 | |
| *** ravig has quit IRC | 18:59 | |
| *** ravig has joined #openstack-kolla | 19:00 | |
| *** mgoddard has joined #openstack-kolla | 19:00 | |
| sdake | funny story | 19:01 |
| sdake | wife on the phone | 19:01 |
| sdake | friend from emc calls | 19:01 |
| sdake | "so how is emc, did they get aquired i cant even keep track anymore. " :) | 19:01 |
| *** rmart04 has quit IRC | 19:02 | |
| *** fragatina has joined #openstack-kolla | 19:07 | |
| *** fragatina has quit IRC | 19:07 | |
| *** ravig has quit IRC | 19:07 | |
| *** alyson_ has quit IRC | 19:07 | |
| *** fragatina has joined #openstack-kolla | 19:08 | |
| *** jtriley has joined #openstack-kolla | 19:09 | |
| *** ravig has joined #openstack-kolla | 19:20 | |
| *** godleon has joined #openstack-kolla | 19:25 | |
| *** daneyon has quit IRC | 19:26 | |
| v1k0d3n | hey sdake anything i can do on that _member_ issue, or is it still in the backlog for now. | 19:28 |
| v1k0d3n | ? | 19:28 |
| sdake | v1k0d3n i have it in my backlog | 19:29 |
| sdake | i w ill get to it either thursday or friday | 19:29 |
| sdake | fwiw master works for me without that patch | 19:29 |
| sdake | (centos/source) | 19:30 |
| sean-k-mooney | sdake: its working for me also on ubuntu source. | 19:30 |
| sdake | nothing like seeing 20 reviews i have outstanding with the words 'Merge Conflict" next to them | 19:31 |
| sean-k-mooney | the other horizon issue i hit is actully a bug | 19:31 |
| *** daneyon has joined #openstack-kolla | 19:33 | |
| *** sbezverk has joined #openstack-kolla | 19:46 | |
| *** daneyon_ has joined #openstack-kolla | 19:47 | |
| rhallisey | sbezverk, db is up, creating the users with kolla_toolbox is tricky :) | 19:50 |
| *** ravig has quit IRC | 19:52 | |
| *** daneyon_ has quit IRC | 19:52 | |
| sbezverk | rhallisey nice!! what was the issue with bootstrap.. | 19:53 |
| rhallisey | command: ["usr/bin/ansible", "localhost"] | 19:54 |
| rhallisey | ERROR! Missing target hosts :( | 19:54 |
| rhallisey | sbezverk, the issue was wrong dir | 19:55 |
| rhallisey | dirs being mounted | 19:55 |
| rhallisey | it was never recovering because it was looking in the wrong place | 19:55 |
| *** ravig has joined #openstack-kolla | 19:55 | |
| vhosakot | sbezverk: can I pull PS 8 of the lbaas review and test it ? | 19:56 |
| sbezverk | rhallisey ok, can you try instead of localhost to use ip address assigned to mariadb pod | 19:56 |
| vhosakot | sbezverk: I need lbaas PS working and will rebase my magnum changes on top of your PS | 19:56 |
| sbezverk | vhosakot PS8 is working I tested it last night, even dataplane was working | 19:57 |
| vhosakot | sbezverk: great, did you test on Ubuntu source on 14.04 ? | 19:57 |
| sbezverk | vhosakot it is not the final version as I will try to simplify it | 19:57 |
| vhosakot | cool.. it's good as long it is working :) | 19:58 |
| sbezverk | vhosakot Honor testing on Ubuntu is your :-) | 19:58 |
| vhosakot | sbezverk: cool, I'll use the simplified version later.. cool | 19:58 |
| v1k0d3n | sdake: ...interesting... | 19:58 |
| vhosakot | sbezverk: ccool :) which distro did you tets on ? | 19:58 |
| sbezverk | centos/source | 19:58 |
| vhosakot | sbezverk: cool | 19:58 |
| sdake | vhosakot it does not work for ubuntu | 19:59 |
| vhosakot | sdake: ah... | 19:59 |
| sdake | vhosakot it is missing ubuntu-isms for haproxy i suspect | 19:59 |
| sdake | in the lbaas container | 19:59 |
| vhosakot | sdake: ah ok | 19:59 |
| vhosakot | sdake: I will pull the PS and test it.. cool | 20:00 |
| sdake | the lbaas container is only "build from source" atm - sbezverk and I will take a look after i get off my 1:1 | 20:00 |
| vhosakot | sdake: cool | 20:00 |
| vhosakot | sdake: should I move to centos source for magnum ?> | 20:00 |
| vhosakot | sdake: since lbaas is working on centos source | 20:00 |
| sdake | vhosakot if its not terribly inconvient | 20:00 |
| vhosakot | sdake: yes, I have a centos setup... will move today | 20:00 |
| sdake | i am not sure when the ubuntu implementation will land | 20:00 |
| vhosakot | sdake: got it.. | 20:00 |
| vhosakot | sbezverk: sdake: thanks! | 20:01 |
| sdake | vhosakot if you want to implement the from source part of ubuntu it is probably a one liner | 20:01 |
| sdake | well 3 liner because you need an if conditional | 20:01 |
| vhosakot | for lbaas ? | 20:01 |
| sdake | right | 20:01 |
| vhosakot | ah, then that sounds easier than moveing to cent :) | 20:01 |
| sdake | have to apt-get install haproxy | 20:01 |
| sbezverk | sdake: I tried to do everything symetrically, if I added something to centos same went to ubuntu | 20:02 |
| sbezverk | sdake: it is already there, check neutron-base | 20:02 |
| sdake | roger | 20:02 |
| sbezverk | I just did not test ubuntu | 20:02 |
| sdake | so ubuntu is good to go then | 20:02 |
| sdake | good it would help if vhosakot could test that briefly | 20:02 |
| vhosakot | sdake: cool, I'll test it... saw the PS | 20:03 |
| sdake | sbezverk if you could provide vhosakot with a couple commands you run to exercise things that may be helpful ;) | 20:03 |
| vhosakot | sdake: sbezverk: cool, I see haproxy is installed for ubuntu.. nice | 20:04 |
| vhosakot | sbezverk: yes, thanks for symmetrically doing.. cool! | 20:04 |
| sdake | sbezverk lets chat now, I think patrick is busy | 20:04 |
| *** ayoung has quit IRC | 20:04 | |
| rhallisey | sbezverk, nah didn't work | 20:05 |
| sdake | sbezverk ok lets do webex now to debug your issue | 20:05 |
| sdake | cisco.webex.com/join/stdake | 20:06 |
| vhosakot | sbezverk: yes, send some lbaas commands I need to test/run please | 20:06 |
| sdake | sbezverk time running low - have 50 minutes :) | 20:08 |
| *** ravig has quit IRC | 20:12 | |
| *** ravig has joined #openstack-kolla | 20:16 | |
| fbarilla | what's the best way to attach Kolla to an existing Ceph Cluster ? | 20:18 |
| sbezverk | sdake, man you said in 1 hour | 20:19 |
| *** inc0_ has quit IRC | 20:19 | |
| sdake | sbezverk yes i know a meeting got moved to 2pm | 20:19 |
| sdake | that was supposed to happen at 1 | 20:19 |
| sdake | so its either now or 3pm ;) | 20:19 |
| *** ravig has quit IRC | 20:22 | |
| *** mliima has quit IRC | 20:24 | |
| *** ravig has joined #openstack-kolla | 20:26 | |
| *** jtriley has quit IRC | 20:27 | |
| sean-k-mooney | out of interest would anyone object to me adding support to build from source from a gerrit review referecne | 20:35 |
| *** shardy has quit IRC | 20:35 | |
| sean-k-mooney | e.g. type gerrit source https://git.openstack.org/openstack/<project> reference refs/changes/x/y/z | 20:36 |
| sean-k-mooney | actully i could leave type as git and just check to see if the reference containerd refs/changes as devstack does | 20:38 |
| sean-k-mooney | it would make testing proposed changes simpler. | 20:38 |
| *** mummer has joined #openstack-kolla | 20:41 | |
| sdake | sean-k-mooney i don't understand what your after but dont sak permission ask forgiveness :) | 20:41 |
| *** ravig has quit IRC | 20:41 | |
| *** ravig has joined #openstack-kolla | 20:43 | |
| sean-k-mooney | sdake: :) | 20:46 |
| *** sdake_ has joined #openstack-kolla | 20:46 | |
| sean-k-mooney | sdake: i just want to add a new option to the kolla-build so that you can use a gerrit review url when building images | 20:46 |
| sean-k-mooney | sdake: so it quicker to test other peoples chages | 20:47 |
| sean-k-mooney | sdake: currently i just checkout the change to a local branch and point the build file to that but wanted to skip that step | 20:47 |
| *** sdake has quit IRC | 20:48 | |
| *** ravig has quit IRC | 20:49 | |
| *** fbarilla has quit IRC | 20:53 | |
| *** ccesario___ has quit IRC | 20:53 | |
| *** ravig has joined #openstack-kolla | 20:55 | |
| *** ravig has quit IRC | 20:56 | |
| *** ravig has joined #openstack-kolla | 20:56 | |
| harlowja | woah sdake_ kolla-build ha alot of output :-P | 20:58 |
| harlowja | *has | 20:58 |
| harlowja | nothing in that output really is broken up into sections right, i think i can sort of see various projects getting built | 20:58 |
| harlowja | but nothing quite like sections 'aka, starting build of X' ... | 20:58 |
| harlowja | seems like its threaded so might not be easy to have more of a clean(er) log anyway | 21:00 |
| *** haplo37_ has quit IRC | 21:01 | |
| harlowja | maybe possible to have individual streams per component that go into different output files? | 21:01 |
| harlowja | thoughts? | 21:02 |
| harlowja | then i can more easily trace wtf is going on for a single component :-P | 21:02 |
| harlowja | other idea, have stdout be 'simple' output and write 'detailed' stuff to those files | 21:04 |
| *** mliima has joined #openstack-kolla | 21:06 | |
| openstackgerrit | Ryan Hallisey proposed openstack/kolla-kubernetes: Changes Glance bootstrap to use kolla-toolbox https://review.openstack.org/320022 | 21:08 |
| *** inc0 has joined #openstack-kolla | 21:09 | |
| *** rhallisey has quit IRC | 21:10 | |
| harlowja | also another question, so after i run `kolla-build --type source` on centos7 | 21:12 |
| harlowja | then i can see stuff via `docker images` | 21:12 |
| harlowja | which is cool | 21:12 |
| harlowja | can i just run one of those containers locally (is that supposed to work?) | 21:12 |
| *** mummer has quit IRC | 21:13 | |
| harlowja | say i try to run the one built named 'kollaglue/centos-source-nova-api' | 21:13 |
| harlowja | http://paste.openstack.org/show/505449/ | 21:13 |
| harlowja | maybe that's just not supposed to work | 21:14 |
| *** inc0 has quit IRC | 21:16 | |
| openstackgerrit | Christian Berendt proposed openstack/kolla: Workaround ansible bug related to delegate_to in Glance tasks https://review.openstack.org/321241 | 21:16 |
| *** dcwangmit01 has quit IRC | 21:19 | |
| *** dcwangmit01 has joined #openstack-kolla | 21:20 | |
| sean-k-mooney | harlowja: you can set the threading to 1 to be able to read the output cleanly but then you have to wait longer | 21:21 |
| harlowja | right | 21:21 |
| harlowja | seems like most logs in kolla do put the image name as the prefix | 21:21 |
| harlowja | maybe just create a collection of loggers instead of using 1 | 21:21 |
| harlowja | and then can easily redirect each image 'name' to its own file (for example) | 21:22 |
| sean-k-mooney | you could. you can also just build the image you want with its dependeces | 21:22 |
| sean-k-mooney | the build script take a regex as the first optionaly positional argument | 21:23 |
| *** mummer has joined #openstack-kolla | 21:23 | |
| *** rhallisey has joined #openstack-kolla | 21:23 | |
| sean-k-mooney | so tools/build.py horizon or kolla-build horizon | 21:23 |
| sean-k-mooney | if you just want to build horzion | 21:23 |
| *** mgoddard has quit IRC | 21:24 | |
| sean-k-mooney | as for running the image locally other then docker run there is no easy way to run just one image however you can run a singel tag | 21:24 |
| sean-k-mooney | by using "kolla-ansible deploy --tages nova" in your cas | 21:25 |
| *** ccesario___ has joined #openstack-kolla | 21:25 | |
| sean-k-mooney | harlowja: there is no reason you cant do docker run kollaglue/centos-source-nova-api but the container expect some addtional argument to run correctly | 21:26 |
| *** ravig has quit IRC | 21:29 | |
| harlowja | ah | 21:29 |
| *** ravig has joined #openstack-kolla | 21:29 | |
| harlowja | my cas is just a VM that i built the containers on, lol | 21:30 |
| harlowja | whats the 'the container expect some addtional argument to run correctly' | 21:30 |
| harlowja | guess that argument is the config.json file? | 21:31 |
| sean-k-mooney | harlowja: for example docker run -v /lib/modules:/lib/modules -v your/config:/etc/nova/nova.conf --net=host kollaglue/centos-source-nova-api nova-api --config-file=/etc/nova/nova.conf | 21:32 |
| harlowja | k | 21:32 |
| harlowja | gotcha | 21:32 |
| *** ccesario___ has quit IRC | 21:32 | |
| sean-k-mooney | not sure if that would work but something along those lines | 21:32 |
| harlowja | kk | 21:32 |
| *** ccesario___ has joined #openstack-kolla | 21:37 | |
| vhosakot | sbezverk: ping | 21:38 |
| *** sdake_ has quit IRC | 21:45 | |
| *** ravig has quit IRC | 21:48 | |
| openstackgerrit | Joshua Harlow proposed openstack/kolla: Use a image specific logger in pushing thread https://review.openstack.org/321251 | 21:48 |
| harlowja | let me know what u guys think about ^ | 21:48 |
| harlowja | and i can just go through the rest of the stuff and do that | 21:49 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP Adds Dynamic DNS functionality to k8s cluster https://review.openstack.org/321252 | 21:49 |
| *** daneyon has quit IRC | 21:49 | |
| *** daneyon has joined #openstack-kolla | 21:50 | |
| *** sdake has joined #openstack-kolla | 21:51 | |
| sbezverk | sdake ping | 21:53 |
| sdake | sbezverk shoot | 21:53 |
| *** Avi82 has quit IRC | 21:54 | |
| sdake | harlowja yes the logs are extensive | 21:54 |
| harlowja | sdake ya :) | 21:54 |
| harlowja | i can tell, ha | 21:54 |
| sdake | harlowja we could probably tone it down a little bit | 21:54 |
| sdake | one thing i'd like is all output to be synced in one go | 21:54 |
| harlowja | or have more modes -v -v | 21:54 |
| openstackgerrit | Ryan Hallisey proposed openstack/kolla-kubernetes: Changes Glance bootstrap to use kolla-toolbox https://review.openstack.org/320022 | 21:55 |
| sbezverk | it appears plugin goes only into neutron-server container.. I ma trying to build lbaas using neutron-server as a base to see if I get installed plugin that way.. | 21:55 |
| sdake | sbezverk you should - isn't that what we talked about on the phone? | 21:55 |
| sdake | I thought it already used neutron-base | 21:55 |
| sdake | oh | 21:55 |
| sdake | only goes into neutron-server not neutron-base | 21:55 |
| harlowja | i'm also guessing the threads created aren't going to handle ctrl-c correctly, has anyone tried that? | 21:56 |
| sdake | maybe you can insert it the same way | 21:56 |
| sdake | harlowja ctrl-c kills the build | 21:56 |
| *** mark-casey has quit IRC | 21:56 | |
| sdake | by the correct way, what do you mean - a join operation? | 21:56 |
| sdake | if so, no | 21:56 |
| sean-k-mooney | you should be able to use the neutron plugin support in the kolla build to inject it too no? | 21:56 |
| *** mark-casey has joined #openstack-kolla | 21:56 | |
| sdake | sean-k-mooney i was thinking just that | 21:56 |
| sbezverk | sdake no, on the phone we saw that it was based on neutron-base | 21:57 |
| harlowja | ya, something akin to have the thread die when ctrl-c, wait till its dead (join); then exit | 21:57 |
| sbezverk | not the neutron-server | 21:57 |
| sdake | sbezverk riht i remember | 21:57 |
| sdake | sbezverk i'd just revert back to what you had | 21:57 |
| sdake | that seems better then overloading it in kolla-build.conf | 21:57 |
| sbezverk | sdake if it works, would you be ok to use it this way? | 21:57 |
| sdake | what you had also seems better then basing it on neutron-server | 21:58 |
| sbezverk | sdake ok, then I will restore old version | 21:58 |
| sdake | there were other thigns that needed attention iirc | 21:58 |
| sdake | the commit message | 21:59 |
| sdake | and ubuntu haproxy install | 21:59 |
| *** esharao has quit IRC | 21:59 | |
| sbezverk | sdake I showed you that haproxy is installed for ubuntu | 21:59 |
| *** asalkeld has joined #openstack-kolla | 21:59 | |
| sdake | right | 21:59 |
| sdake | so just the commit message | 21:59 |
| sdake | rhallisey ping | 22:00 |
| sean-k-mooney | ok got to run. o/ | 22:00 |
| sdake | sean-k-mooney take care dude | 22:00 |
| rhallisey | sdake, what's up? | 22:00 |
| harlowja | also qq, is there any scratch dir that all the j2 files get written into (that is saved post-run) | 22:01 |
| harlowja | or is it all basically in memory | 22:01 |
| harlowja | guess there is a setup_working_dir that does this | 22:02 |
| sdake | harlowja there is a scratch dir | 22:02 |
| harlowja | ya | 22:02 |
| sdake | you can also make it write out a copy off the built docker files | 22:02 |
| sdake | built/jinja2-ified | 22:02 |
| harlowja | kk, just was thinking that something akin to the following would be useful | 22:02 |
| harlowja | http://logs.openstack.org/75/308075/3/check/gate-anvil-rpms-dsvm-centos-7/83073f8/ | 22:03 |
| harlowja | which has all the log files of all the things exectuted | 22:03 |
| harlowja | something similar could be done for kolla | 22:03 |
| harlowja | (instead of just all over stdout) | 22:04 |
| sdake | per file build logging seems good, but when you look at the big picture of all the things we have on our plate, its not even in the top50% :) | 22:05 |
| harlowja | that's ok | 22:05 |
| harlowja | i do it :-P | 22:05 |
| sdake | wfm | 22:05 |
| harlowja | lol | 22:05 |
| sdake | one of the things that is becoming a problem is the config options to build.py | 22:05 |
| harlowja | ?? | 22:05 |
| sdake | so if you could minimize the amount of stuff you add as config optiosn that would be good | 22:05 |
| harlowja | k | 22:06 |
| sdake | build.py is meant to be used as a library | 22:06 |
| sdake | atleast that is where i'd like to take it | 22:06 |
| sdake | to integrate more thouruoghly with a ci/cd pipeline | 22:07 |
| sdake | maybe the current file usage of kolla-build..py and just running the cli would do the job | 22:07 |
| sdake | i honestly dont know for certain | 22:07 |
| sdake | kolla-build.conf that is | 22:07 |
| harlowja | ya, the whole queue thing is really a DAG right? | 22:08 |
| harlowja | or processed like a DAG? | 22:08 |
| sdake | DAG? | 22:08 |
| harlowja | directed graph | 22:08 |
| harlowja | seems like there is some image parent finding and such | 22:08 |
| sdake | yup that is correct parent deps are always rebuilt | 22:08 |
| *** ravig has joined #openstack-kolla | 22:09 | |
| harlowja | k | 22:09 |
| sdake | in a ci/cd system, i ike the build the world model | 22:09 |
| sdake | because its fast - 15 minutes on my gear | 22:09 |
| sdake | i also like push the world model which takes about 10 minutes | 22:09 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla: Adds Neutron LBaaS Support https://review.openstack.org/318322 | 22:10 |
| *** mliima has quit IRC | 22:10 | |
| sbezverk | sdake here is updated version, I suggest to merge it asap before somebody will start suggesting modifications ;-) | 22:11 |
| harlowja | cool, ya, i could see how taskflow could work in this, it seems like there is a <build-workflow> stage, then <process-workflow> stage in this build.py | 22:11 |
| harlowja | with the work to do somewhat intermixed in build.py | 22:11 |
| harlowja | anyway, i'll see if i can get this log output better | 22:12 |
| harlowja | so that i can figure it out more easily | 22:12 |
| *** ravig has quit IRC | 22:20 | |
| *** sdake has quit IRC | 22:25 | |
| *** dwalsh has quit IRC | 22:25 | |
| *** diogogmt has quit IRC | 22:27 | |
| *** ayoung has joined #openstack-kolla | 22:30 | |
| *** ccesario___ has quit IRC | 22:36 | |
| *** weiyu has joined #openstack-kolla | 22:38 | |
| *** mark-casey has quit IRC | 22:42 | |
| vhosakot | meeting in 15 mins | 22:46 |
| *** salv-orl_ has joined #openstack-kolla | 22:50 | |
| *** weiyu has quit IRC | 22:51 | |
| *** salv-orlando has quit IRC | 22:53 | |
| *** stvnoyes has quit IRC | 22:55 | |
| vhosakot | sbezverk: ping | 22:55 |
| *** stvnoyes has joined #openstack-kolla | 22:56 | |
| vhosakot | sbezverk: I cherry-picked your PS #8 and see that the neutron-server is flapping on Ubuntu 14.04 | 22:56 |
| vhosakot | d079c735a14b 192.168.122.201:4000/kollaglue/ubuntu-source-neutron-server:2.0.0 "kolla_start" 2 minutes ago Restarting (1) 57 seconds ago neutron_server | 22:56 |
| vhosakot | sbezverk: I do see these lines in /etc/neutron/neutron.conf | 22:56 |
| vhosakot | service_plugins = router,neutron_lbaas.services.loadbalancer.plugin.LoadBalancerPluginv2 | 22:56 |
| vhosakot | service_provider = LOADBALANCERV2:Haproxy:neutron_lbaas.drivers.haproxy.plugin_driver.HaproxyOnHostPluginDriver:default | 22:57 |
| sbezverk | vhosakot 1 - you need this file kolla-build.conf | 22:57 |
| sbezverk | 2 - you need to rebuild neutron | 22:57 |
| vhosakot | sbezverk: I did | 22:57 |
| sbezverk | 3 - I suggest take PS 9 | 22:57 |
| sbezverk | how about 1? | 22:57 |
| sbezverk | ;-) | 22:58 |
| vhosakot | sbezverk: should we change kolla-build.conf for lbaas ? | 22:58 |
| vhosakot | sbezverk: I'm using the default build params | 22:58 |
| vhosakot | sbezverk: works fine for me without lbaas | 22:58 |
| sbezverk | yes, here is the file you are missing | 22:59 |
| vhosakot | sbezverk: I also see the neutron_lbaas_agent container running fne | 22:59 |
| sbezverk | http://paste.openstack.org/show/505459/ | 22:59 |
| *** asalkeld has quit IRC | 23:00 | |
| vhosakot | sbezverk: ah, can't the lbaas tar be added to https://github.com/openstack/kolla/blob/master/kolla/common/config.py#L155 ? | 23:00 |
| vhosakot | meeting time | 23:01 |
| sbezverk | nope as we need it at two places, 1 - neutron-server container 2 - lbaas container | 23:01 |
| sbezverk | and if you check PS the change in config for lbaaas is already there | 23:02 |
| vhosakot | sbezverk: ah ok | 23:02 |
| vhosakot | sbezverk: doesn't the neutron server source come from https://github.com/openstack/kolla/blob/master/kolla/common/config.py#L216-L219 ? | 23:02 |
| vhosakot | sbezverk: I see you added lbaas in config.py in PS 9.. cool | 23:03 |
| vhosakot | sbezverk: so, I dont need kolla-build.conf with PS 9, right ? | 23:04 |
| sbezverk | vhosakot you need .conf file | 23:04 |
| vhosakot | sbezverk: even with PS 9 ? | 23:04 |
| sbezverk | as we add lbaas to neutron-server as plugin | 23:04 |
| sbezverk | since we cannot install from two source | 23:05 |
| vhosakot | service_plugins = router,neutron_lbaas.services.loadbalancer.plugin.LoadBalancerPluginv2 ? | 23:05 |
| vhosakot | sbezverk: do I need http://paste.openstack.org/show/505459/ when using PS 9? | 23:05 |
| sbezverk | yes | 23:06 |
| vhosakot | sbezverk: isn't https://review.openstack.org/#/c/318322/9/kolla/common/config.py enough to install lbaas from source ? | 23:06 |
| patchbot | vhosakot: patch 318322 - kolla - Adds Neutron LBaaS Support | 23:06 |
| *** akscram has quit IRC | 23:07 | |
| *** Avi82 has joined #openstack-kolla | 23:07 | |
| sbezverk | vhosakot I am confused, I have already told you exactly what you need to make it working?!? Let me know where I was not clear | 23:08 |
| vhosakot | sbezverk: I need to know why http://paste.openstack.org/show/505459/ is needed when https://review.openstack.org/#/c/318322/9/kolla/common/config.py is there ? :) | 23:08 |
| patchbot | vhosakot: patch 318322 - kolla - Adds Neutron LBaaS Support | 23:08 |
| *** godleon has quit IRC | 23:09 | |
| sbezverk | because config.py defines installation from source, .conf is needed because I also install lbaas as a plugon | 23:09 |
| sbezverk | plugin | 23:09 |
| *** Avi82 has quit IRC | 23:10 | |
| *** Avi82 has joined #openstack-kolla | 23:10 | |
| vhosakot | sbezverk: ah ok... can you show me in the review where/how http://paste.openstack.org/show/505459/ is used as a plugin | 23:11 |
| vhosakot | sbezverk: so, this means, laaas has a depedency on having the right kolla-build.conf, is this right ? | 23:11 |
| *** harlowja has quit IRC | 23:12 | |
| sbezverk | vhosakot that part is implemented somewhere in build.py, I have not implemented and unfortunately have no understanding how it works, it just does. | 23:12 |
| vhosakot | sbezverk: ah ok cool.. | 23:12 |
| sbezverk | you are correct lbaas is dependent on kolla-build.conf file | 23:13 |
| vhosakot | sbezverk: ah ok cool... | 23:13 |
| vhosakot | sbezverk: thanks for all the clarification and sorry :) | 23:13 |
| *** d_code has quit IRC | 23:13 | |
| sbezverk | I suggested to sdake to add it in PSm but he told me not to and to give it to you as a file | 23:13 |
| vhosakot | sbezverk: cool, that is fine... | 23:13 |
| vhosakot | sbezverk: when lbaas is enabled and built without kolla-build.conf, kolla-ansible must flag an error.. that can be an enhancement | 23:14 |
| vhosakot | sbezverk: I now need to pull PS 9 and use http://paste.openstack.org/show/505459/, right ? | 23:14 |
| rhallisey | no sdake for the meeting? | 23:14 |
| vhosakot | rhallisey: yeah.. looks like it | 23:14 |
| rhallisey | I can run it see who's around | 23:15 |
| *** d_code has joined #openstack-kolla | 23:15 | |
| vhosakot | rhallisey: cool, I'm around | 23:15 |
| sbezverk | vhosakot yes and rebuild whole neutron | 23:16 |
| vhosakot | sbezverk: yep, I'll do it.. thanks for the clarification.. | 23:16 |
| vhosakot | sbezverk: so, can you tell me how other packages are built form source tar without build.conf ? | 23:17 |
| sbezverk | vhosakot building from source works exactly the same way for everything using config.py definitions | 23:19 |
| sbezverk | but since lbaas on neutron-server is installed as pluging, then we use kolla-build.conf | 23:19 |
| sbezverk | it is exaplained in kolla build-image doc | 23:20 |
| vhosakot | sbezverk: ah got it... | 23:20 |
| *** Jeffrey4l has joined #openstack-kolla | 23:20 | |
| vhosakot | sbezverk: https://github.com/openstack/kolla/blob/master/doc/image-building.rst#plugin-functionality ? | 23:20 |
| sbezverk | yep | 23:21 |
| vhosakot | does it make sense for kolla-build and kolla-ansible to flag an error if lbaas is enabled and built/deployed with kolla-build.conf ? otherwise, the issue (what I'm seeing - neutorn-server flapping) is seen | 23:22 |
| vhosakot | sbezverk: ^^ | 23:22 |
| vhosakot | without* I meant sbezverk | 23:22 |
| sbezverk | vhosakot we need to decide in which form this file will be delivered/generated, based on this we will decide.. | 23:25 |
| vhosakot | sbezverk: right, because, in a brand new kolla repo, the user will not have kolla-build.conf, plus "tox -e genconfig" generates an empty kolla-build.conf.. | 23:26 |
| vhosakot | sbezverk: cool, thanks for you help! will keep you posted :) | 23:27 |
| *** salv-orl_ has quit IRC | 23:27 | |
| *** vhosakot has quit IRC | 23:28 | |
| *** sdake has joined #openstack-kolla | 23:28 | |
| sbezverk | vhosakot that is exactly what sdake was going to figure out, there is no any good doc | 23:28 |
| sdake | not sur ewhat your talking about | 23:29 |
| sbezverk | if it were up to me, I would just add this file into PS | 23:29 |
| sdake | if your talking about kolla-buiold.conf then ya | 23:29 |
| sdake | that file needs to end up in the repo in some way | 23:29 |
| sbezverk | kolla-build.conf | 23:29 |
| sdake | not the repo but the pip file | 23:29 |
| sdake | i wonder if its possible to hardcode config options in build.py | 23:30 |
| sdake | btw - highdpi now works on centos 7.2 | 23:30 |
| * sdake rejoices | 23:30 | |
| sdake | my 4k/2k monitor is now usable :) | 23:31 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
