| *** g3ek has joined #openstack-kolla | 00:00 | |
| inc0 | Reepicheep: https://github.com/openstack/kolla-ansible/blob/master/ansible/roles/nova/defaults/main.yml#L13-L23 this is full list of volumes that nova_libvirt mounts | 00:00 |
|---|---|---|
| inc0 | nova_compute is one you want to override for ephemerals if my memory serves | 00:01 |
| *** srwilkers has joined #openstack-kolla | 00:02 | |
| sdake | inc0 can you pleze review the kolla-kubernetes queue | 00:03 |
| sdake | inc0 there are many patches which require rubber staps | 00:03 |
| sdake | inc0 this patch set: https://review.openstack.org/#/c/452851/ | 00:04 |
| sdake | srwilkers ^ | 00:04 |
| sdake | inc0 note anything that is not green don't approve ;) | 00:04 |
| *** Administrator_ has quit IRC | 00:04 | |
| *** britthouser8 has quit IRC | 00:04 | |
| *** britthouser has joined #openstack-kolla | 00:05 | |
| *** Administrator_ has joined #openstack-kolla | 00:05 | |
| sdake | inc0 and this patch set: https://review.openstack.org/#/c/452851/6 | 00:05 |
| inc0 | sdake: ok, I'll check it out, I'm slacking with reviews:( | 00:06 |
| sdake | inc0 kfox1111 srwilkers rather this patchset (there are about 10 patches0; https://review.openstack.org/#/c/454375/ | 00:06 |
| sdake | inc0 the last set is easy- all dead docs removal | 00:06 |
| sdake | the first set is easy too | 00:06 |
| sdake | our docs.oo page is an utter mess | 00:06 |
| inc0 | kfox1111: want me to fix ceph issue or you'll do it? | 00:06 |
| sdake | inc0 also hrw mentioend this morning he was forking kolla because his patches were not being approved in a timely fashion | 00:07 |
| kfox1111 | inc0: I pushed what I think is a fix... | 00:07 |
| srwilkers | sdake, saw that | 00:08 |
| inc0 | I guess we need more reviewers then | 00:08 |
| inc0 | I'll get on that | 00:08 |
| sdake | cool :) | 00:08 |
| sdake | inc0 i think immediately we coul dsolve that pain for hrw, although the gate was busted | 00:08 |
| sdake | inc0 I was on a plane all day so couldn't look at it, I believ msaner fixed the gate | 00:09 |
| sdake | fixed as in submitted a patch | 00:09 |
| sdake | not sure if it has merged | 00:09 |
| openstackgerrit | Merged openstack/kolla-kubernetes master: Remove services/glance https://review.openstack.org/452851 | 00:11 |
| openstackgerrit | Merged openstack/kolla-kubernetes master: Remove services/horizon https://review.openstack.org/452852 | 00:11 |
| sdake | thanks :) | 00:11 |
| sdake | stop when you get to red | 00:11 |
| sdake | i need to fix some of those (i think around mariadb or so) | 00:11 |
| *** rwellum has quit IRC | 00:11 | |
| sdake | inc0: http://eavesdrop.openstack.org/irclogs/%23openstack-kolla/%23openstack-kolla.2017-04-10.log.html#t2017-04-10T08:45:19 | 00:13 |
| *** manheim has joined #openstack-kolla | 00:13 | |
| kfox1111 | this is interesting.... | 00:15 |
| kfox1111 | https://raw.githubusercontent.com/cncf/landscape/master/landscape/CloudNativeLandscape_v0.9.4_darkhighlight.jpg | 00:15 |
| kfox1111 | whats the light blue imply? | 00:15 |
| kfox1111 | seems to imply proposed projects. | 00:16 |
| sdake | kfox1111 not sure | 00:16 |
| sdake | kfox1111 the toc meetings are open - you can find out if you attend :) | 00:17 |
| kfox1111 | the green stuff is all stuff in the tent. | 00:17 |
| kfox1111 | sdake: I've had issues heariing the audio. not sure why. :/ | 00:17 |
| sdake | oh didn't know you tried | 00:17 |
| kfox1111 | the blue stuff has a lot of stuff that was proposed for sure. and some interesting things I hadn't heard were being proposed. | 00:17 |
| *** manheim has quit IRC | 00:18 | |
| *** sayantan_ has joined #openstack-kolla | 00:18 | |
| kfox1111 | interesting. hyper's making a go of it too... | 00:22 |
| *** MasterOfBugs has quit IRC | 00:22 | |
| *** pramodrj07 has quit IRC | 00:23 | |
| *** deadnull has joined #openstack-kolla | 00:24 | |
| *** iceyao has joined #openstack-kolla | 00:26 | |
| *** deadnull has quit IRC | 00:42 | |
| *** duonghq has joined #openstack-kolla | 00:43 | |
| duonghq | morning guys | 00:43 |
| duonghq | eanylin, hi | 00:43 |
| sdake | sup duonghq | 00:46 |
| sdake | kfox1111 kolla-kubernetes is pretty well constructed | 00:47 |
| sdake | kfox1111 the gate is hard to learn but once learned seems fairly well done | 00:47 |
| sdake | kfox1111 the code seems to work well as well | 00:47 |
| sdake | kfox1111 i think our biggest challenge will be ramping new contributors up | 00:47 |
| sdake | kfox1111 i've pretty much ramped up over the last month writing those docs and various other things | 00:48 |
| sdake | kfox1111 i have a really good understnading of the code structure now | 00:48 |
| sdake | looks pretty good :) | 00:48 |
| sdake | sbezverk_ ^^ | 00:48 |
| *** MarginHu has joined #openstack-kolla | 00:49 | |
| *** ssurana has quit IRC | 00:50 | |
| MarginHu | Good morning , Kolla | 00:50 |
| *** daidv_ has joined #openstack-kolla | 00:53 | |
| *** zhurong has joined #openstack-kolla | 00:55 | |
| *** cuongnv has joined #openstack-kolla | 00:57 | |
| duonghq | morning MarginHu | 01:00 |
| daidv_ | morning | 01:03 |
| *** srwilkers has quit IRC | 01:04 | |
| *** daidv has quit IRC | 01:11 | |
| *** daidv_ has quit IRC | 01:11 | |
| *** daidv has joined #openstack-kolla | 01:11 | |
| *** tovin07_ has joined #openstack-kolla | 01:20 | |
| *** guoguo has joined #openstack-kolla | 01:20 | |
| *** MarginHu has quit IRC | 01:23 | |
| *** hrw has quit IRC | 01:26 | |
| masber | good morning all, I am trying to understand kolla-ansible network deployment, could someone please tell me why kolla-ansible is creating a br-ex on my compute node? I am reading this document https://www.rdoproject.org/networking/networking-in-too-much-detail/ and it looks like br-ex is only needed on the neutron node? | 01:29 |
| *** hrw has joined #openstack-kolla | 01:31 | |
| spsurya | morning guys | 01:37 |
| *** yangyapeng has joined #openstack-kolla | 01:41 | |
| *** dixiaoli has joined #openstack-kolla | 01:54 | |
| *** dixiaoli has quit IRC | 01:58 | |
| *** dixiaoli has joined #openstack-kolla | 02:02 | |
| *** unicell has quit IRC | 02:04 | |
| daidv | Jeffrey4l, morning, can I ask you something about default collation for mariadb in Kolla-ansible? | 02:08 |
| *** shashank_t_ has joined #openstack-kolla | 02:11 | |
| *** Pavo has joined #openstack-kolla | 02:13 | |
| eanylin | Hi duonghq, just saw your message | 02:17 |
| duonghq | hi eanylin | 02:18 |
| eanylin | sdake: Should we update the DG for cinder LVM deployment with the changes in https://review.openstack.org/#/c/455032/ ? The current way of deploying is incomplete... | 02:18 |
| eanylin | I made the changes as per what sbezverk suggested... | 02:19 |
| eanylin | hi duonghq | 02:19 |
| duonghq | where do you live? in the US? | 02:20 |
| *** japestinho has joined #openstack-kolla | 02:20 | |
| eanylin | Yeah, in US | 02:20 |
| eanylin | trying to get out of office... still working on some stuff... | 02:20 |
| eanylin | duonghq: You? Where do you live? | 02:21 |
| sdake | eanylin sounds good | 02:22 |
| duonghq | eanylin, I'm from Vietnam | 02:22 |
| eanylin | sdake: Thanks. | 02:22 |
| eanylin | duonghq: Ah, ok, its early there, maybe 10 plus in the morning I think | 02:23 |
| duonghq | eanylin, right | 02:24 |
| *** shashank_t_ has quit IRC | 02:25 | |
| eanylin | yeah | 02:25 |
| *** shashank_t_ has joined #openstack-kolla | 02:25 | |
| *** shashank_t_ has quit IRC | 02:29 | |
| inc0 | good evening | 02:37 |
| duonghq | evening inc0 | 02:37 |
| inc0 | how is everyone? | 02:37 |
| sdake | tired | 02:37 |
| *** iceyao has quit IRC | 02:37 | |
| *** Pavo has quit IRC | 02:38 | |
| eanylin | sdake: Same here. Long day | 02:38 |
| sdake | eanylin long year | 02:38 |
| inc0 | tell me about it | 02:39 |
| eanylin | lol | 02:39 |
| duonghq | it's turning to summer here, not good news | 02:39 |
| eanylin | we got some hail storm just now = / | 02:39 |
| *** eaguilar has quit IRC | 02:39 | |
| inc0 | and I'm surprised that its' not raining in OR | 02:39 |
| inc0 | it was earlier today | 02:40 |
| sdake | inc0 can you ack this patch stream plz | 02:40 |
| sdake | https://review.openstack.org/#/c/454375/ | 02:40 |
| sdake | on the right - the related patches - there are about 10 in the stack | 02:40 |
| *** Pavo has joined #openstack-kolla | 02:40 | |
| *** eaguilar has joined #openstack-kolla | 02:42 | |
| *** Pavo has quit IRC | 02:42 | |
| *** caowei has joined #openstack-kolla | 02:43 | |
| openstackgerrit | Merged openstack/kolla-kubernetes master: Remove outdated DNS guide https://review.openstack.org/454375 | 02:43 |
| daidv | inc0, Hi, can us debate around default collation in Kolla ansible template? like utf8_general_ci orr utf8_unicode_ci? | 02:46 |
| inc0 | hmm, I don't know if I can say much about it, it's up to what ansible handles | 02:47 |
| daidv | I mean default collation which we set in galera template in Kolla ansible | 02:49 |
| daidv | https://review.openstack.org/#/c/455154/3/ansible/roles/mariadb/templates/galera.cnf.j2 | 02:49 |
| inc0 | ahh | 02:49 |
| *** eanylin has quit IRC | 02:49 | |
| inc0 | ok, so we need to be careful | 02:49 |
| inc0 | since if we change it in our templates | 02:50 |
| inc0 | that will be updated after we do upgrade | 02:50 |
| inc0 | and that might cause chaos | 02:50 |
| daidv | Actually, I'm facing with a Panko deployment problem. https://launchpad.net/bugs/1680783 | 02:52 |
| openstack | Launchpad bug 1680783 in kolla-ansible "Illegal mix of collations for operation 'UNION' during Panko project deployment" [Undecided,In progress] - Assigned to Dai Dang Van (daidv) | 02:52 |
| inc0 | yes I saw this bug, looks like bug in Pank | 02:53 |
| inc0 | Panko | 02:53 |
| *** iceyao has joined #openstack-kolla | 02:55 | |
| inc0 | I'm worried about changing defaults in config of database because this can easily break existing deployments during upgrade | 02:55 |
| daidv | IMO, it's OK | 02:55 |
| daidv | because | 02:55 |
| daidv | 1. devstack are using default MYSQL config with utf8_general_ci | 02:56 |
| daidv | If any project can work well via Devstack | 02:56 |
| daidv | it can run via Kolla deployment. | 02:56 |
| inc0 | but if we already created databases | 02:57 |
| inc0 | and upgrade | 02:57 |
| inc0 | therefore reconfigure mariadb | 02:57 |
| inc0 | are you 100% sure it won't break? | 02:57 |
| inc0 | also Panko could be fixed in a way that it will enforce collation during syncdb | 02:57 |
| *** Pavo has joined #openstack-kolla | 02:57 | |
| inc0 | if it can't work otherwise | 02:57 |
| inc0 | changing any configuration of a core service is extremally risky thing | 02:58 |
| inc0 | as we need to ensure that not only deployment work, but also, even more importantly, existing deployment won't be broken during upgrade | 02:59 |
| *** iceyao has quit IRC | 02:59 | |
| inc0 | look, tables can have collation right? so if we have for example nova database, today it's utf8_general_ci | 02:59 |
| inc0 | and nova community creates new table during Pike release | 03:00 |
| *** dixiaoli has quit IRC | 03:00 | |
| inc0 | if we run upgrades, this will reconfigure mariadb, so new table will be created with utf8_unicode_ci and we'll end up having 2 different collations within same database | 03:00 |
| daidv | Yep, but now, In this Jeffrey4l's commit https://review.openstack.org/#/c/365428 | 03:01 |
| *** dixiaoli has joined #openstack-kolla | 03:01 | |
| daidv | we are using unicode collation. | 03:02 |
| *** Pavo has quit IRC | 03:02 | |
| daidv | But I think many developer are working on Devstack to develop something more. | 03:02 |
| inc0 | devstack is not good measure for us | 03:02 |
| *** iceyao has joined #openstack-kolla | 03:03 | |
| inc0 | I'm much more concerned about upgradability | 03:03 |
| daidv | But this mean some new query can work with via Devstack and got merged in other projects. | 03:04 |
| daidv | And can get fail with Kolla deployment | 03:04 |
| daidv | because utf8_general_ci is less accurate | 03:05 |
| inc0 | can we fix it in Panko? | 03:05 |
| daidv | than utf8_unicode_ci | 03:05 |
| daidv | you mean Panko project or Panko ansible? | 03:06 |
| inc0 | I mean frankly, I'm less worried about Panko not being deployable than adding additional risk to existing databases during upgrade | 03:06 |
| inc0 | Panko project | 03:06 |
| inc0 | or I know | 03:06 |
| inc0 | when we create database for Panko | 03:06 |
| inc0 | we can enforce utf8_unicode_ci just for this database right? | 03:07 |
| inc0 | there is no need to do it across the board | 03:07 |
| daidv | Yes. | 03:07 |
| inc0 | which will pose considerably less risk | 03:07 |
| inc0 | https://github.com/openstack/kolla-ansible/blob/master/ansible/roles/panko/tasks/bootstrap.yml#L12-L25 <- this task | 03:08 |
| inc0 | http://docs.ansible.com/ansible/mysql_db_module.html you can enforce collation with mysql_db module | 03:09 |
| daidv | I see it run via mysql_db lib. | 03:09 |
| inc0 | just add collation to module_args and see if it helps | 03:10 |
| daidv | Ok, but are you think in future, we can get the same problem in other project? | 03:10 |
| daidv | meanwhile they are coding and testing via utf8_general_ci collation? | 03:11 |
| inc0 | thats first time I see this problem | 03:11 |
| inc0 | and stable projects usually will be careful about this stuff | 03:11 |
| inc0 | again, I'd rather take that risk than make upgrades more dangerous | 03:13 |
| daidv | I agree with you at this point. But I think when we set default collation is utf8_unicode_ci | 03:16 |
| daidv | this have gotten different with | 03:16 |
| daidv | default config of Mariadb and Devstack as well | 03:17 |
| daidv | https://review.openstack.org/#/c/365428 | 03:17 |
| inc0 | again, I'm not worried about devstack, they made their choices | 03:17 |
| inc0 | also every deployer can override our decision with config overrides | 03:17 |
| inc0 | it's existing deployments I'm worried about | 03:18 |
| *** MarginHu has joined #openstack-kolla | 03:21 | |
| sdake | inc0 few more patches: https://review.openstack.org/#/c/454389/1 | 03:21 |
| daidv | In othercase, when we deploy OPS with Kolla and use uft8_unicode_ci. For example, in Pike cycle, any project have new query | 03:21 |
| daidv | I work with general collation but not with unicode collation | 03:22 |
| daidv | I convern about that. | 03:22 |
| daidv | s/convern/concern/ | 03:22 |
| inc0 | daidv: my only concern are upgrades. If you can tell me that it's 100% safe, then I'll ack it | 03:23 |
| inc0 | sdake: tomorrow plz, now I'm trying to decide whether I need a drink or a painkiller, I'm getting sick:( | 03:23 |
| sdake | painkiller it is! | 03:23 |
| inc0 | or both... | 03:24 |
| inc0 | :P | 03:24 |
| sdake | bad to mix :) | 03:24 |
| inc0 | depends on desired outcome | 03:24 |
| sdake | if death is your desired outcome to mix ahead | 03:24 |
| inc0 | not today | 03:24 |
| duonghq | I'll choose painkiller and some music | 03:26 |
| sdake | rubber stamping dude rubber stamping | 03:26 |
| sdake | its 5 clicks *2 :) | 03:26 |
| sdake | if we can't even get rubber stamp reviews after 5 days, we are in troubel :( | 03:27 |
| daidv | inc0, Ok, I need find out something more and confirm if I follow your way. | 03:27 |
| inc0 | than you daidv | 03:28 |
| inc0 | check with Jeffrey4l since he wrote initial config option | 03:28 |
| duonghq | sdake, what is the origin of "rubber stamp" | 03:28 |
| daidv | Ok, I'm waiting for him. :) | 03:28 |
| sdake | duonghq in the US govt | 03:29 |
| sdake | they make rubber stamps | 03:29 |
| sdake | and they stamp paperwork with them | 03:29 |
| sdake | STAMP | 03:29 |
| sdake | STAMP | 03:29 |
| sdake | STAMP | 03:29 |
| sdake | red tape | 03:29 |
| sdake | it gives it an air of authority :) | 03:29 |
| inc0 | to me it's sound from postal office | 03:30 |
| duonghq | got it | 03:32 |
| daidv | Can anyone explain for me why should we use utf8_unicode_ci instead of utf8_general_ci? for default galera server collation ? | 03:34 |
| daidv | https://review.openstack.org/#/c/365428/2/ansible/roles/mariadb/templates/galera.cnf.j2 | 03:34 |
| inc0 | http://stackoverflow.com/questions/2344118/utf-8-general-bin-unicode reading this utf8_unicode_ci seems to be safer | 03:35 |
| *** eanylin has joined #openstack-kolla | 03:36 | |
| daidv | inc0, but normally, other project still use general collation as a default config via mariadb default config? | 03:37 |
| daidv | https://docs.openstack.org/ocata/install-guide-ubuntu/environment-sql-database.html | 03:38 |
| daidv | And in OpenStack installation docs, I still get utf8_general_ci as a default? | 03:38 |
| daidv | Why didn't we follow that? | 03:38 |
| *** gkadam has joined #openstack-kolla | 03:39 | |
| *** gkadam is now known as gkadam-afk | 03:44 | |
| *** MarginHu has quit IRC | 03:46 | |
| *** MarginHu has joined #openstack-kolla | 03:47 | |
| *** iceyao has quit IRC | 03:49 | |
| openstackgerrit | Steven Dake proposed openstack/kolla-kubernetes master: Placement support in nova service chart https://review.openstack.org/453846 | 03:51 |
| openstackgerrit | Steven Dake proposed openstack/kolla-kubernetes master: Revert "Update to Helm v2.3.0" https://review.openstack.org/455502 | 03:51 |
| sdake | daidv openstack changed to that after kolla had upgrades | 03:52 |
| sdake | daidv iirc there is no way to upgrade a database of one format to another | 03:52 |
| sdake | devstack doesn't do upgrades - so it doesn't have this problem | 03:53 |
| *** manheim has joined #openstack-kolla | 03:55 | |
| *** eaguilar has quit IRC | 03:55 | |
| SamYaple | daidv: it was missed when the configs were initially written. openstack switched to that in the Kilo timeframe | 03:57 |
| daidv | sdake, it just collation of mariadb? so, we can change it during upgrading. | 03:57 |
| sdake | daidv i am not an expert on mariadb- that is jut my recollection | 03:58 |
| SamYaple | daidv: it does affect the tables, but you are correct ti should be fine now | 03:58 |
| SamYaple | well what i mean to say is _openstack_ has fixed the tables so it should be fine now | 03:58 |
| sdake | SamYaple thanks - i had forgotten about that but your right that is basically what happened (kilo) | 03:58 |
| SamYaple | daidv: now that i say that, idk. it might require an ALTER | 03:59 |
| SamYaple | daidv: but it _could_ be upgraded is the point. if you want to look into that | 04:00 |
| daidv | The default collation configs were initially written from Sep 2016? https://review.openstack.org/#/c/365428 | 04:00 |
| *** manheim has quit IRC | 04:00 | |
| sdake | daidv i think we didn't get all of kolla 1.0.0 correct ;) | 04:00 |
| sdake | considering we had to dispatch the tree because of a fatal defect | 04:00 |
| SamYaple | nah 1.0.0 could have been migrated. no one wrote the migration script is all | 04:01 |
| daidv | So, we have no way to implement a script for this change during upgrading? | 04:08 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible master: Detect openstack_release variable in mariadb recoverty playbooks https://review.openstack.org/455505 | 04:11 |
| *** gkadam-afk is now known as gkadam | 04:11 | |
| daidv | Jeffrey4l, Can you explain for me why should we use utf8_unicode_ci instead of utf8_general_ci for default galera server collation meanwhile OpenStack are using general as a default? | 04:12 |
| *** iceyao has joined #openstack-kolla | 04:13 | |
| Jeffrey4l | daidv, sorry , i have no idea for this. when i changing this, just pick one from them. | 04:17 |
| *** dixiaoli has quit IRC | 04:17 | |
| Jeffrey4l | daidv, the original issue is Panko? | 04:18 |
| daidv | Jeffrey4l, yes, Panko query issue. | 04:18 |
| Jeffrey4l | i am facing this issue, too. | 04:19 |
| daidv | But I think it related to default collation | 04:19 |
| daidv | of OpenStack | 04:19 |
| *** dixiaoli has joined #openstack-kolla | 04:19 | |
| Jeffrey4l | but other project is OK. | 04:19 |
| Jeffrey4l | have u tried to use utf8_general_ci collaction? | 04:19 |
| daidv | Yep, but in future, if OpenStack continue use general collation? | 04:20 |
| daidv | Jeffrey4l, I have tried to use utf8_general_ci and it's working well. | 04:20 |
| Jeffrey4l | hrm.. | 04:20 |
| Jeffrey4l | i still doubt that there are some issue in panko or sqlarchemy.. | 04:21 |
| Jeffrey4l | collation is related to how text is sorted and compared. | 04:21 |
| Jeffrey4l | http://stackoverflow.com/questions/766809/whats-the-difference-between-utf8-general-ci-and-utf8-unicode-ci | 04:21 |
| Jeffrey4l | could u explain why utf8-unicode-ci failed? | 04:22 |
| Jeffrey4l | i filed a bug in Panko https://bugs.launchpad.net/panko/+bug/1661847 | 04:23 |
| openstack | Launchpad bug 1661847 in Panko "panko event list got following sql error" [Undecided,New] | 04:23 |
| Jeffrey4l | but there is no comments right now. | 04:23 |
| *** dixiaoli has quit IRC | 04:23 | |
| daidv | I just know unicode get more accurate than general collation. | 04:24 |
| *** jrobinson has quit IRC | 04:24 | |
| daidv | May be, some compare can return true with general but wrong with unicode collation | 04:25 |
| *** shashank_t_ has joined #openstack-kolla | 04:25 | |
| Jeffrey4l | daidv, could u dump the database sql when using utf8_general_ci? | 04:25 |
| daidv | Ok. | 04:26 |
| Jeffrey4l | please upload it to this bug https://bugs.launchpad.net/panko/+bug/1661847 | 04:26 |
| openstack | Launchpad bug 1661847 in Panko "panko event list got following sql error" [Undecided,New] | 04:26 |
| Jeffrey4l | and add comment that" when using utf8_general_ci, it works blabla | 04:26 |
| daidv | Ok, just a minute. | 04:27 |
| Jeffrey4l | thanks ;) | 04:28 |
| *** jrobinson has joined #openstack-kolla | 04:30 | |
| *** lucker has joined #openstack-kolla | 04:33 | |
| *** unicell has joined #openstack-kolla | 04:41 | |
| *** lamt has joined #openstack-kolla | 04:41 | |
| sdake | night | 04:47 |
| *** shashank_t_ has quit IRC | 04:51 | |
| daidv | Jeffrey4l, I have pushed my dump file. | 04:51 |
| daidv | https://bugs.launchpad.net/panko/+bug/1661847 | 04:51 |
| openstack | Launchpad bug 1661847 in Panko "panko event list got following sql error" [Undecided,New] | 04:51 |
| *** shashank_t_ has joined #openstack-kolla | 04:51 | |
| daidv | Please check it for me, thanks. | 04:52 |
| daidv | inc0, I saw your comment in gerrit, but can we confirm with Jeffrey4l that should we continue to use unicode collation or switch to general collation (like OpenStack community now). | 04:53 |
| daidv | ? | 04:54 |
| Jeffrey4l | got. let me check | 04:55 |
| *** shashank_t_ has quit IRC | 04:56 | |
| *** iceyao has quit IRC | 04:57 | |
| *** caowei_ has joined #openstack-kolla | 04:58 | |
| *** caowei has quit IRC | 04:59 | |
| *** caowei_ is now known as caowei | 04:59 | |
| *** klindgren_ has quit IRC | 05:13 | |
| *** iceyao has joined #openstack-kolla | 05:15 | |
| *** iceyao has quit IRC | 05:20 | |
| *** dixiaoli has joined #openstack-kolla | 05:26 | |
| *** klindgren has joined #openstack-kolla | 05:28 | |
| *** lamt has quit IRC | 05:36 | |
| *** manheim has joined #openstack-kolla | 05:44 | |
| *** manheim has quit IRC | 05:48 | |
| openstackgerrit | wangwei proposed openstack/kolla-ansible master: Add remove-garbage function in kolla-ansible https://review.openstack.org/402362 | 05:51 |
| *** lamt has joined #openstack-kolla | 05:53 | |
| *** jaosorior_away is now known as jaosorior | 05:55 | |
| openstackgerrit | Youri Jeong proposed openstack/kolla master: fix typo for releasenotes/notes/non-x86-support-bce168d78db50202.yaml 'targetted' should be 'targeted' https://review.openstack.org/455517 | 05:57 |
| *** iceyao has joined #openstack-kolla | 06:01 | |
| Jeffrey4l | daidv, still around? | 06:03 |
| daidv | I'm here? | 06:04 |
| daidv | How about that? | 06:05 |
| Jeffrey4l | check this table https://mariadb.com/kb/en/mariadb/supported-character-sets-and-collations/ | 06:05 |
| Jeffrey4l | when using utf8， utf8_general_ci collactions is default | 06:06 |
| Jeffrey4l | i guess there are some wrong in sqlalchemy.. | 06:06 |
| daidv | Yep, as I mentioned https://review.openstack.org/#/c/455154/ | 06:06 |
| Jeffrey4l | but anyway, i think your change is ready to go ;) | 06:07 |
| daidv | So, how about inc0 comment in my above commit ? | 06:09 |
| daidv | Do you think we should change for mariadb or panko ansible? | 06:10 |
| *** targon has joined #openstack-kolla | 06:12 | |
| *** unicell has quit IRC | 06:13 | |
| *** unicell has joined #openstack-kolla | 06:14 | |
| Jeffrey4l | changing this won't break anything. the existence table is still using utf8-unicode-ci | 06:14 |
| Jeffrey4l | only new created table is affected. | 06:14 |
| Jeffrey4l | i will leave some comments for thiat. | 06:14 |
| Jeffrey4l | but before that, i will to make another test ;) daidv | 06:15 |
| *** satyar has joined #openstack-kolla | 06:15 | |
| daidv | Jeffrey4l, Ok. | 06:16 |
| *** lamt has quit IRC | 06:19 | |
| *** Serlex has joined #openstack-kolla | 06:22 | |
| *** iceyao has quit IRC | 06:23 | |
| *** wirehead_ has quit IRC | 06:29 | |
| *** skramaja has joined #openstack-kolla | 06:30 | |
| *** wirehead_ has joined #openstack-kolla | 06:30 | |
| *** bogdando has joined #openstack-kolla | 06:34 | |
| *** rwsu has quit IRC | 06:37 | |
| openstackgerrit | Youri Jeong proposed openstack/kolla-ansible master: fix typo https://review.openstack.org/455541 | 06:38 |
| *** lucker has quit IRC | 06:39 | |
| *** sdake has quit IRC | 06:46 | |
| *** kolla-slack has quit IRC | 06:47 | |
| *** pbourke has quit IRC | 06:47 | |
| *** iceyao has joined #openstack-kolla | 06:47 | |
| *** zhuzeyu has joined #openstack-kolla | 06:48 | |
| *** lamt has joined #openstack-kolla | 06:48 | |
| *** pbourke has joined #openstack-kolla | 06:49 | |
| hrw | morning | 06:50 |
| hrw | https://review.openstack.org/#/q/owner:marcin%2540juszkiewicz.com.pl+status:open got to 7 patches ;) Thanks! | 06:50 |
| *** rwsu has joined #openstack-kolla | 06:51 | |
| *** iceyao has quit IRC | 06:52 | |
| *** Serlex has quit IRC | 07:01 | |
| *** lamt has quit IRC | 07:04 | |
| *** Jezekus has joined #openstack-kolla | 07:06 | |
| *** pcaruana has joined #openstack-kolla | 07:06 | |
| *** manheim has joined #openstack-kolla | 07:08 | |
| duonghq | morning hrw | 07:11 |
| *** manheim has quit IRC | 07:12 | |
| *** manheim has joined #openstack-kolla | 07:23 | |
| *** hieulq has quit IRC | 07:27 | |
| *** hieulq has joined #openstack-kolla | 07:27 | |
| *** hieulq has quit IRC | 07:29 | |
| *** hieulq has joined #openstack-kolla | 07:29 | |
| *** hieulq has quit IRC | 07:29 | |
| *** hieulq has joined #openstack-kolla | 07:29 | |
| *** hieulq is now known as Guest33220 | 07:30 | |
| *** hieulq_ has joined #openstack-kolla | 07:30 | |
| *** hieulq_ has quit IRC | 07:30 | |
| *** duonghq has quit IRC | 07:32 | |
| *** mgoddard has joined #openstack-kolla | 07:33 | |
| *** jascott1_ has quit IRC | 07:34 | |
| *** hachi has quit IRC | 07:37 | |
| *** asmaa has quit IRC | 07:37 | |
| *** egonzalez has joined #openstack-kolla | 07:37 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible master: Detect openstack_release variable in mariadb recovery playbooks https://review.openstack.org/455505 | 07:38 |
| *** kong has quit IRC | 07:39 | |
| *** hieulq_ has joined #openstack-kolla | 07:41 | |
| *** mgoddard has quit IRC | 07:42 | |
| *** imcsk8 has quit IRC | 07:45 | |
| *** imcsk8 has joined #openstack-kolla | 07:46 | |
| openstackgerrit | Merged openstack/kolla-ansible master: Revert "Remove useless variable glance_backend_file" https://review.openstack.org/454218 | 07:46 |
| *** nathharp has joined #openstack-kolla | 07:50 | |
| *** hachi has joined #openstack-kolla | 07:50 | |
| *** asmaa has joined #openstack-kolla | 07:51 | |
| *** shardy has joined #openstack-kolla | 07:53 | |
| *** magicboiz has joined #openstack-kolla | 07:54 | |
| *** hieulq_ has quit IRC | 07:54 | |
| *** Serlex has joined #openstack-kolla | 07:55 | |
| *** magicboiz has quit IRC | 07:55 | |
| *** hieulq_ has joined #openstack-kolla | 07:56 | |
| *** hieulq_ has quit IRC | 08:01 | |
| *** jrobinson has quit IRC | 08:06 | |
| *** gfidente has joined #openstack-kolla | 08:12 | |
| *** gfidente has quit IRC | 08:12 | |
| *** gfidente has joined #openstack-kolla | 08:12 | |
| spsurya | egonzalez: hi..... | 08:13 |
| spsurya | got the three VM nodes | 08:13 |
| spsurya | of 8GB memory each | 08:13 |
| spsurya | semms like we can test KS rolling upgrade | 08:13 |
| spsurya | for multi node | 08:14 |
| *** magicboiz has joined #openstack-kolla | 08:17 | |
| egonzalez | spsurya, cool | 08:21 |
| spsurya | :) | 08:21 |
| *** athomas has quit IRC | 08:23 | |
| masber | hi | 08:30 |
| *** haplo37_ has quit IRC | 08:30 | |
| masber | is there a reason why kolla-ansible doesn't install port-security as a extension ? | 08:30 |
| egonzalez | masber, it depends if a service who need it enables it, for example tacker and devstack enables port-security | 08:31 |
| *** robcresswell has left #openstack-kolla | 08:36 | |
| *** athomas has joined #openstack-kolla | 08:36 | |
| *** g3ek has quit IRC | 08:39 | |
| *** haplo37 has quit IRC | 08:40 | |
| *** sayantan_ has quit IRC | 08:40 | |
| *** iniazi has quit IRC | 08:42 | |
| *** rmart04 has joined #openstack-kolla | 08:43 | |
| *** derekjhyang has joined #openstack-kolla | 08:43 | |
| *** iniazi has joined #openstack-kolla | 08:44 | |
| egonzalez | is there someone with a multinode deploy UP to check one thing? | 08:45 |
| openstackgerrit | Merged openstack/kolla master: Get correct tag name when job is running in release pipeline https://review.openstack.org/447507 | 08:47 |
| openstackgerrit | Bertrand Lallau proposed openstack/kolla-ansible master: WIP: Enable Ceph input plugin in Telegraf https://review.openstack.org/455602 | 08:49 |
| *** iceyao has joined #openstack-kolla | 08:50 | |
| *** haplo37 has joined #openstack-kolla | 08:54 | |
| *** g3ek has joined #openstack-kolla | 08:54 | |
| *** iceyao has quit IRC | 08:54 | |
| *** haplo37_ has joined #openstack-kolla | 08:54 | |
| *** openstackgerrit has quit IRC | 09:03 | |
| *** openstackgerrit has joined #openstack-kolla | 09:09 | |
| openstackgerrit | shaofeng cheng proposed openstack/kolla-ansible master: Add VMware DataStore support to glance https://review.openstack.org/452176 | 09:09 |
| Jezekus | egonzalez: yup | 09:12 |
| egonzalez | Jezekus, can you check in your compute nodes if there is br-ex created with a "missing interface error/warn", docker exec -ti openvswitch_db ovs-vsctl show | 09:13 |
| openstackgerrit | Bertrand Lallau proposed openstack/kolla-ansible master: Remove kolla_logs_volume unused variable https://review.openstack.org/455616 | 09:16 |
| Jezekus | egonzalez: give me a moment | 09:17 |
| satyar | hi Jeffrey4l | 09:18 |
| satyar | how are you? | 09:18 |
| Jeffrey4l | sup satyar. | 09:18 |
| *** blallau has joined #openstack-kolla | 09:18 | |
| Jeffrey4l | not bad :) | 09:18 |
| satyar | having an issue | 09:18 |
| Jeffrey4l | yep? | 09:18 |
| satyar | i have few VMs up | 09:18 |
| satyar | and then ran kolla upgrade | 09:19 |
| satyar | BTW i have jumbo frame setup | 09:19 |
| satyar | before upgrade VMs MTU was 8950 | 09:19 |
| satyar | after upgrade VMs were not accessable | 09:19 |
| satyar | so i restarted the VMs | 09:19 |
| satyar | then saw VMs MTU is 1500 | 09:19 |
| satyar | any new VMs which i am creating getting 8950 | 09:20 |
| satyar | and accessable | 09:20 |
| Jeffrey4l | satyar, after restart the vms is accessable? | 09:20 |
| satyar | nope | 09:20 |
| Jeffrey4l | which kolla version are u using? | 09:20 |
| satyar | after restart VMs got 1500 MTU so not able to get the ips | 09:20 |
| satyar | i am using the stable/ocata | 09:21 |
| Jeffrey4l | branch? | 09:21 |
| satyar | stable | 09:22 |
| Jeffrey4l | i meant you are using stable/ocata branch? | 09:22 |
| Jeffrey4l | or ocata tag? | 09:22 |
| satyar | yes | 09:22 |
| satyar | stable ocata branch | 09:22 |
| Jeffrey4l | ok. | 09:22 |
| Jeffrey4l | satyar, have u check the ovs ports are OK after upgrade? | 09:23 |
| satyar | how to check that? | 09:23 |
| satyar | ovs-vsctl show looks good | 09:26 |
| Jeffrey4l | yep | 09:26 |
| *** sambetts|afk is now known as sambetts | 09:26 | |
| *** hachi has quit IRC | 09:26 | |
| *** hachi has joined #openstack-kolla | 09:31 | |
| *** guoguo has quit IRC | 09:34 | |
| openstackgerrit | shaofeng cheng proposed openstack/kolla-ansible master: Add VMware DataStore support to glance https://review.openstack.org/452176 | 09:39 |
| *** skramaja_ has joined #openstack-kolla | 09:40 | |
| *** wagiel has joined #openstack-kolla | 09:41 | |
| Jezekus | egonzalez: https://pastebin.com/KUFFm0MV | 09:42 |
| *** skramaja has quit IRC | 09:42 | |
| Jezekus | egonzalez: I dont see such error | 09:42 |
| egonzalez | Jezekus, you have neutron_external_interface present in compute nodes right? that's why the error is not raising | 09:43 |
| Jezekus | egonzalez: Yep, thats right. I have only 3 nodes and control nodes are all of them | 09:44 |
| Jezekus | I'll check the other cluster where the neutron_external interface is not on all nodes (if I remember correctly) | 09:44 |
| Jeffrey4l | daidv, still around? | 09:50 |
| daidv | I'm here | 09:50 |
| daidv | and just see your comment | 09:50 |
| Jeffrey4l | check the comments https://review.openstack.org/#/c/455154/ | 09:50 |
| daidv | Ok. | 09:51 |
| Jeffrey4l | yep. i am telling u ;) | 09:51 |
| Jeffrey4l | thanks for fixing this ;) | 09:51 |
| openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible master: Fix panko mongo idempotency https://review.openstack.org/455624 | 09:52 |
| *** skramaja_ is now known as skramaja | 09:52 | |
| *** Jezekus has quit IRC | 09:53 | |
| *** tovin07_ has quit IRC | 09:55 | |
| daidv | Jeffrey4l, thanks for your review, so detail for me to learn more. | 09:58 |
| Jeffrey4l | aha, i learned this in the past few hours :) | 09:58 |
| daidv | Cool :) | 09:58 |
| egonzalez | someone has been able to deploy panko recently? | 10:02 |
| egonzalez | always fails on bootstrap_service | 10:02 |
| openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible master: WIP: Support osprofile usage https://review.openstack.org/455628 | 10:03 |
| daidv | egonzalez, which branch? | 10:04 |
| egonzalez | daidv, master | 10:04 |
| egonzalez | doing to update images, just in case something was fixed recently | 10:05 |
| *** MarginHu has quit IRC | 10:07 | |
| *** daidv has quit IRC | 10:07 | |
| *** zhurong has quit IRC | 10:10 | |
| *** zhurong has joined #openstack-kolla | 10:11 | |
| *** cuongnv has quit IRC | 10:13 | |
| Jeffrey4l | egonzalez, i am. deployed stable branch recently, which works. | 10:18 |
| egonzalez | Jeffrey4l, for me failed at doing db sync with mongo backend http://paste.openstack.org/show/606128/ | 10:19 |
| egonzalez | the error is the same as this fixed bug, https://bugs.launchpad.net/kolla-ansible/+bug/1668216 | 10:20 |
| openstack | Launchpad bug 1668216 in kolla-ansible ocata "Error in Task: Creating Panko MongoDB database" [High,Fix committed] - Assigned to Eduardo Gonzalez (egonzalez90) | 10:20 |
| Jeffrey4l | ok. i only tried panko + mysql | 10:20 |
| Jeffrey4l | raised ImportError: guess there are some packages are not imported | 10:21 |
| egonzalez | yep, going to try with new images and see what happens, maybe some library failed to install in previous images | 10:22 |
| Jeffrey4l | ok | 10:22 |
| *** kong has joined #openstack-kolla | 10:22 | |
| *** satyar has quit IRC | 10:27 | |
| *** haplo37_ has quit IRC | 10:31 | |
| *** haplo37_ has joined #openstack-kolla | 10:40 | |
| *** iniazi has quit IRC | 10:41 | |
| *** caowei has quit IRC | 10:42 | |
| *** iniazi has joined #openstack-kolla | 10:43 | |
| *** rwallner has joined #openstack-kolla | 10:59 | |
| *** rwallner has quit IRC | 11:03 | |
| *** shardy has quit IRC | 11:07 | |
| *** shardy has joined #openstack-kolla | 11:07 | |
| *** shardy is now known as shardy_lunch | 11:09 | |
| *** tovin07 has joined #openstack-kolla | 11:16 | |
| *** tovin07 has quit IRC | 11:16 | |
| *** dixiaoli has quit IRC | 11:21 | |
| *** dixiaoli has joined #openstack-kolla | 11:21 | |
| sean-k-mooney | Jeffrey4l: egonzalez i was just looking at https://review.openstack.org/#/c/451289/ we dont plan on ever supproting nova-networks in kolla correct | 11:22 |
| egonzalez | sean-k-mooney, I don't think so, also n-network is deprecated | 11:23 |
| sean-k-mooney | egonzalez: yes infact its been deprecated twice now | 11:23 |
| Jeffrey4l | hrm nova-network is mark as deprected and kolla do not supported it since begin. so i do not think we need add this | 11:24 |
| sean-k-mooney | egonzalez: its due for removal in queens when cells v1 is removed | 11:24 |
| *** dixiaoli has quit IRC | 11:25 | |
| sean-k-mooney | ya do you know of another way they can support vsphere? in that patch? | 11:25 |
| sean-k-mooney | apparently networking-vsphere does not support the default vsphere vswitch | 11:26 |
| openstackgerrit | Eduardo Gonzalez proposed openstack/kolla master: Add pymongo in panko images https://review.openstack.org/455667 | 11:30 |
| *** rwallner has joined #openstack-kolla | 11:39 | |
| *** rwallner has quit IRC | 11:40 | |
| *** rwallner has joined #openstack-kolla | 11:41 | |
| Serlex | defualt is nsx? | 11:42 |
| *** Jezekus has joined #openstack-kolla | 11:42 | |
| *** rwallner has quit IRC | 11:43 | |
| *** srwilkers has joined #openstack-kolla | 11:43 | |
| *** jascott1_ has joined #openstack-kolla | 11:44 | |
| egonzalez | iirc, vmware removed support for vsphere in neutron, only support a vcenter cluster and nsx as plugin $$$ | 11:44 |
| sean-k-mooney | egonzalez: that makes sense in a corpearte way... | 11:45 |
| sean-k-mooney | Serlex: i belive so or at least i always assumed that unless you were deploying vios you were useing nsx if you were using vmware based openstack | 11:46 |
| egonzalez | yeah, esxi is free, and people was using it with openstack without paying anything. they decided to only support vcenter which need licenses per core | 11:47 |
| egonzalez | this is from juno at least, maybe something changed till now | 11:48 |
| *** jascott1_ has quit IRC | 11:48 | |
| sean-k-mooney | i dont track vmware much as i generally just work with kvm | 11:48 |
| Serlex | I tried to do some testing with vmware and kolla/openstack...Spent all the time to set up the virtual machines for esxi, vsphere, openstack etc | 11:49 |
| Serlex | then found out nsxi isn't free, not even available for trial...THANKS VMWARE | 11:49 |
| egonzalez | first line in docs: When a cloud operator wants to use OpenStack with vSphere using open source elements, he/she can only do so by relying on nova-network | 11:49 |
| egonzalez | https://wiki.openstack.org/wiki/Neutron/Networking-vSphere | 11:49 |
| sean-k-mooney | ouch $$$ | 11:50 |
| Serlex | there is an alternative, let me find the github page | 11:51 |
| egonzalez | but as in that docs says, can be used OVSvApp, a VM inside esxi | 11:51 |
| Serlex | oh yeah, OVSvApp - https://wiki.openstack.org/wiki/Neutron/Networking-vSphere | 11:51 |
| egonzalez | guess that the path to support vsphere in kolla | 11:51 |
| Serlex | not tried it, it would be good to know how you get on sean-k-mooney :) | 11:51 |
| *** rwallner has joined #openstack-kolla | 11:51 | |
| sean-k-mooney | Serlex: well im not actully trying to enable vsphere in kolla but https://review.openstack.org/#/c/451289 is | 11:52 |
| sean-k-mooney | i just came across it when i saw it as a conflcit for the external ovs patch and was like cool new hypervisor for kolla | 11:55 |
| *** rwallner has quit IRC | 11:58 | |
| *** rwallner has joined #openstack-kolla | 11:59 | |
| Serlex | ah i see, yeah, i rather see end user customize kolla on their end than upstream | 11:59 |
| sean-k-mooney | well i dont think its a bad thing to have support in kolla provided it is posablly to deploy using opensouce componats so we can gate it | 12:00 |
| *** MarginHu has joined #openstack-kolla | 12:00 | |
| *** MarginHu has quit IRC | 12:01 | |
| openstackgerrit | Youri Jeong proposed openstack/kolla-ansible master: fix typo https://review.openstack.org/455541 | 12:08 |
| *** rmart04 has quit IRC | 12:09 | |
| *** shardy_lunch is now known as shardy | 12:10 | |
| *** wagiel has quit IRC | 12:12 | |
| *** Pavo has joined #openstack-kolla | 12:14 | |
| *** iceyao has joined #openstack-kolla | 12:15 | |
| openstackgerrit | Dan Ardelean proposed openstack/kolla-ansible master: Add HyperV role https://review.openstack.org/455684 | 12:16 |
| Jezekus | egonzalez: unfortunately all nodes in clusters have the neutron_external_interface. But I could create a new cluster without neutron_external_interface on nova nodes during the afternoon | 12:17 |
| *** shashank_t_ has joined #openstack-kolla | 12:17 | |
| egonzalez | Jezekus, no problem, don't waste your time with this | 12:18 |
| *** eaguilar has joined #openstack-kolla | 12:19 | |
| *** iceyao has quit IRC | 12:20 | |
| *** iceyao has joined #openstack-kolla | 12:20 | |
| mnaser | morning folks | 12:20 |
| egonzalez | morning mnaser | 12:21 |
| *** shashank_t_ has quit IRC | 12:22 | |
| egonzalez | is nova master broken? services are not registered in conductor, empty openstack compute service list | 12:22 |
| egonzalez | gates are failing for all source deploy | 12:22 |
| *** jmontleo has joined #openstack-kolla | 12:26 | |
| *** jmontleon has quit IRC | 12:27 | |
| *** nathharp has quit IRC | 12:28 | |
| *** nathharp has joined #openstack-kolla | 12:29 | |
| *** iceyao has quit IRC | 12:31 | |
| *** iceyao has joined #openstack-kolla | 12:32 | |
| *** nathharp has quit IRC | 12:34 | |
| *** nathharp has joined #openstack-kolla | 12:34 | |
| *** nathharp has quit IRC | 12:35 | |
| mnaser | egonzalez ive seen this as a race condition | 12:35 |
| mnaser | got an example of a failed build? | 12:35 |
| mnaser | im pretty sure the reason is the older version of docker which causes issues | 12:36 |
| *** nathharp has joined #openstack-kolla | 12:36 | |
| mnaser | i proposed a fix a while back https://review.openstack.org/#/c/449285/ .. but it hasnt been looked at for a while :\ | 12:36 |
| egonzalez | http://logs.openstack.org/24/455624/1/check/gate-kolla-ansible-dsvm-deploy-centos-source-centos-7-nv/8c74194/ | 12:36 |
| mnaser | yup, that patch was meant to fix it egonzalez | 12:38 |
| mnaser | http://logs.openstack.org/24/455624/1/check/gate-kolla-ansible-dsvm-deploy-centos-source-centos-7-nv/8c74194/logs/system_logs/docker.log | 12:38 |
| mnaser | (pid 9251) has become an orphan, killing it | 12:38 |
| *** nathharp has quit IRC | 12:39 | |
| *** nathharp has joined #openstack-kolla | 12:39 | |
| *** gkadam has quit IRC | 12:40 | |
| openstackgerrit | Merged openstack/kolla-kubernetes master: Remove labels.rst https://review.openstack.org/454376 | 12:40 |
| *** nathharp has quit IRC | 12:40 | |
| egonzalez | mnaser, what version is 1.13.1? I have 17.03.1 locally | 12:41 |
| mnaser | egonzalez 1.13.1 is the one that came before 17.03.1 (before docker changes its numbering system) | 12:41 |
| openstackgerrit | Merged openstack/kolla-kubernetes master: Remove host-setup https://review.openstack.org/454377 | 12:41 |
| egonzalez | upgrading to 17.04 | 12:41 |
| openstackgerrit | Merged openstack/kolla-kubernetes master: Remove kubernetes-setup.rst https://review.openstack.org/454378 | 12:41 |
| mnaser | i wanted to eventually bump to 17.03/17.04 but wanted to do these on a step-by-step basis | 12:41 |
| mnaser | so we dont run into big breakages | 12:41 |
| egonzalez | I had no issue with those versions since an hour ago | 12:42 |
| openstackgerrit | Merged openstack/kolla-kubernetes master: Remove installation.rst https://review.openstack.org/454379 | 12:42 |
| *** iceyao_ has joined #openstack-kolla | 12:42 | |
| openstackgerrit | Merged openstack/kolla-kubernetes master: Remove quickstart.rst https://review.openstack.org/454381 | 12:42 |
| openstackgerrit | Merged openstack/kolla-kubernetes master: Remove keystone guide https://review.openstack.org/454388 | 12:42 |
| *** iceyao has quit IRC | 12:45 | |
| *** nathharp has joined #openstack-kolla | 12:45 | |
| egonzalez | mnaser, with 17.04 is not working, the issue are not launching instances, is nova services not being registered in nova service-list | 12:46 |
| mnaser | egonzalez oh you mean that it's failing locally? | 12:46 |
| egonzalez | yep | 12:46 |
| egonzalez | also in gates | 12:46 |
| openstackgerrit | Dan Ardelean proposed openstack/kolla-ansible master: Add Hyper-V role https://review.openstack.org/455684 | 12:49 |
| *** zhurong has quit IRC | 12:54 | |
| openstackgerrit | Dan Ardelean proposed openstack/kolla-ansible master: Add Hyper-V role https://review.openstack.org/455684 | 12:55 |
| *** lamt has joined #openstack-kolla | 12:57 | |
| openstackgerrit | Dan Ardelean proposed openstack/kolla-ansible master: Add Hyper-V role https://review.openstack.org/455684 | 12:59 |
| *** iceyao_ has quit IRC | 13:00 | |
| *** rwellum has joined #openstack-kolla | 13:02 | |
| *** satyar has joined #openstack-kolla | 13:05 | |
| *** eaguilar has quit IRC | 13:05 | |
| *** eaguilar has joined #openstack-kolla | 13:07 | |
| *** lamt has quit IRC | 13:08 | |
| *** eanylin has quit IRC | 13:16 | |
| *** eaguilar has quit IRC | 13:17 | |
| *** jtriley has joined #openstack-kolla | 13:21 | |
| *** erlon has joined #openstack-kolla | 13:24 | |
| *** iceyao has joined #openstack-kolla | 13:25 | |
| *** iceyao has quit IRC | 13:29 | |
| *** srwilkers has quit IRC | 13:31 | |
| *** eanylin has joined #openstack-kolla | 13:32 | |
| *** al498u has joined #openstack-kolla | 13:35 | |
| openstackgerrit | Dan Ardelean proposed openstack/kolla-ansible master: Add Hyper-V role https://review.openstack.org/455684 | 13:38 |
| *** mbruzek has joined #openstack-kolla | 13:40 | |
| *** mbruzek has joined #openstack-kolla | 13:40 | |
| openstackgerrit | Mark Duggan proposed openstack/kolla-ansible master: Unmount Ceph OSD disks as part of destroy https://review.openstack.org/455714 | 13:41 |
| *** eaguilar has joined #openstack-kolla | 13:42 | |
| *** jascott1_ has joined #openstack-kolla | 13:45 | |
| openstackgerrit | Mark Duggan proposed openstack/kolla-ansible master: Unmount Ceph OSD disks as part of destroy https://review.openstack.org/455714 | 13:45 |
| *** srwilkers has joined #openstack-kolla | 13:50 | |
| *** jascott1_ has quit IRC | 13:50 | |
| *** targon has quit IRC | 13:50 | |
| *** cuongnv has joined #openstack-kolla | 13:54 | |
| *** nathharp_ has joined #openstack-kolla | 13:54 | |
| *** nathharp has quit IRC | 13:56 | |
| *** nathharp_ is now known as nathharp | 13:56 | |
| *** manheim has quit IRC | 13:58 | |
| *** manheim has joined #openstack-kolla | 13:59 | |
| *** mattmceuen has joined #openstack-kolla | 14:02 | |
| *** manheim has quit IRC | 14:04 | |
| *** jaosorior is now known as jaosorior_away | 14:06 | |
| *** esharao has joined #openstack-kolla | 14:08 | |
| *** manheim has joined #openstack-kolla | 14:09 | |
| *** eaguilar has quit IRC | 14:10 | |
| *** lamt has joined #openstack-kolla | 14:11 | |
| *** manheim has quit IRC | 14:12 | |
| *** manheim has joined #openstack-kolla | 14:12 | |
| *** haplo37_ has quit IRC | 14:13 | |
| *** lpetrut has joined #openstack-kolla | 14:14 | |
| *** swinn has joined #openstack-kolla | 14:15 | |
| *** eaguilar has joined #openstack-kolla | 14:15 | |
| *** lrensing has joined #openstack-kolla | 14:16 | |
| *** claudiub has joined #openstack-kolla | 14:16 | |
| *** magicboiz has quit IRC | 14:16 | |
| *** abalutoiu has joined #openstack-kolla | 14:16 | |
| *** eaguilar has quit IRC | 14:17 | |
| *** lucasxu has joined #openstack-kolla | 14:17 | |
| *** lamt has quit IRC | 14:22 | |
| *** haplo37_ has joined #openstack-kolla | 14:22 | |
| *** abalutoiu_ has joined #openstack-kolla | 14:24 | |
| *** jmccarthy has joined #openstack-kolla | 14:25 | |
| -openstackstatus- NOTICE: latest base images have mistakenly put python3 in some places expecting python2 causing widespread failure of docs patches - fixes are underway | 14:26 | |
| *** abalutoiu has quit IRC | 14:27 | |
| *** lamt has joined #openstack-kolla | 14:28 | |
| *** lucasxu has quit IRC | 14:28 | |
| *** lucasxu has joined #openstack-kolla | 14:29 | |
| *** jascott1_ has joined #openstack-kolla | 14:30 | |
| *** jascott1_ has quit IRC | 14:33 | |
| *** dave-mccowan has joined #openstack-kolla | 14:34 | |
| *** lucasxu has quit IRC | 14:38 | |
| *** japestinho has quit IRC | 14:40 | |
| *** lucasxu has joined #openstack-kolla | 14:44 | |
| -openstackstatus- NOTICE: we have rolled back centos-7, fedora-25 and ubuntu-xenial images to the previous days release. Feel free to recheck your jobs now. | 14:47 | |
| *** lamt has quit IRC | 14:50 | |
| *** eaguilar has joined #openstack-kolla | 14:50 | |
| openstackgerrit | Bertrand Lallau proposed openstack/kolla-ansible master: Remove dead code from cleanup-host.sh script https://review.openstack.org/455738 | 14:51 |
| *** rwallner has quit IRC | 14:51 | |
| *** lucasxu has quit IRC | 14:54 | |
| *** lucasxu has joined #openstack-kolla | 14:55 | |
| openstackgerrit | Merged openstack/kolla-ansible master: Use utf8_general_ci collation as a default collation https://review.openstack.org/455154 | 14:55 |
| *** eaguilar has quit IRC | 14:56 | |
| swinn | I’m running a deploy of ocata, built images, ubuntu/source/4.0.0 and hitting an error with nova_compute containers that cells is not configured, nova-api aborts. RPC reports: HostMappingNotFound: Host 'dc01-r01-n02' is not mapped to any cell. What might I be missing in globals.yaml? | 14:58 |
| *** rwallner has joined #openstack-kolla | 14:59 | |
| egonzalez | swinn, did all deployment step succeed? there is one step "simple_cell_setup" to do those mappings | 15:02 |
| swinn | egonzalez: negative, deploy stops waiting for nova-compute to come online and the containers report the error I pasted | 15:03 |
| *** yangyapeng has quit IRC | 15:03 | |
| egonzalez | swinn, failed the 20 tries of waiting for nova-compute? in my env compute appears between 3-6 tries | 15:04 |
| swinn | yep, failed after 20 tries and validating right now that the containers are still not up by re-running the deploy | 15:05 |
| swinn | but based on what I see in the logs, RPC is not happy because no cells exist | 15:05 |
| egonzalez | swinn, can you use nova service-list? is something in the output? | 15:06 |
| egonzalez | is rabbitmq container running and nova services connected? or trying to connect? | 15:07 |
| swinn | I just destroyed the whole thing to rebuild it again. I have no data on these nodes yet so giving that a shot, few mins | 15:08 |
| egonzalez | swinn, are you using ocata installed from pip or source code from github? | 15:09 |
| *** kong has quit IRC | 15:09 | |
| swinn | I’m using ubuntu/source built images, not upstream from hub. Recommended to me yesterday by sdake as I was using ubuntu/binary and hitting all kinds of issues. | 15:09 |
| blallau | @swinn: maybe it's due to cells V2 | 15:10 |
| blallau | you must add the compute node to the cell database | 15:10 |
| blallau | "nova-manage cell_v2 discover_hosts --verbose" | 15:10 |
| egonzalez | yep, but on deployment is failing before doing the simple_cell_setup and create nova_cell0 db | 15:11 |
| swinn | yea I would think the initial deploy would handle that…. | 15:11 |
| blallau | oh sorry, | 15:11 |
| egonzalez | waiting for nova-compute task just wait until a compute node appears registered in nova | 15:11 |
| egonzalez | before doing simple_cell_setup | 15:11 |
| swinn | it’s deploying nova now, will see in a moment if its up or not | 15:12 |
| egonzalez | today i had issues with deployment from master nova | 15:12 |
| swinn | confirmed, one node appears in the cells db and the other does not | 15:16 |
| openstackgerrit | Merged openstack/kolla-ansible master: Remove dead code from cleanup-host.sh script https://review.openstack.org/455738 | 15:17 |
| swinn | err one node reports back, the other does not | 15:17 |
| *** ksumit has joined #openstack-kolla | 15:17 | |
| egonzalez | swinn, with just one node active before simple_cell_setup is enough. after deploy will automatically discover hosts | 15:18 |
| *** cuongnv has quit IRC | 15:20 | |
| *** manheim has quit IRC | 15:22 | |
| *** tovin07 has joined #openstack-kolla | 15:23 | |
| swinn | compute service list is empty | 15:23 |
| *** ksumit is now known as joe4213 | 15:24 | |
| *** joe4213 has left #openstack-kolla | 15:25 | |
| *** swinn has quit IRC | 15:25 | |
| *** skramaja has quit IRC | 15:26 | |
| *** lucasxu has quit IRC | 15:26 | |
| egonzalez | swinn, can you double check if are using ocata code or master? is the same issue i'm facing at master atm | 15:28 |
| openstackgerrit | Dai Dang Van proposed openstack/kolla-ansible stable/ocata: Use utf8_general_ci collation as a default collation https://review.openstack.org/455747 | 15:28 |
| kfox1111 | morning. | 15:28 |
| *** rwallner has quit IRC | 15:35 | |
| *** blallau has quit IRC | 15:37 | |
| *** swinn has joined #openstack-kolla | 15:38 | |
| egonzalez | swinn, can you double check if are using ocata code or master? is the same issue i'm facing at master atm | 15:39 |
| swinn | pretty sure ocata, that would be release 4.0.0 yes? | 15:39 |
| egonzalez | swinn, if do pip freeze | grep -i nova, should be a version like nova==15.0.<something> in current master is 15.0.0.0rc2.dev873 | 15:40 |
| *** lamt has joined #openstack-kolla | 15:41 | |
| *** fooliouno has joined #openstack-kolla | 15:43 | |
| *** daidv has joined #openstack-kolla | 15:44 | |
| openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible stable/ocata: DNM: test ocata branch https://review.openstack.org/455760 | 15:46 |
| *** wagiel has joined #openstack-kolla | 15:47 | |
| *** tonanhngo has joined #openstack-kolla | 15:47 | |
| *** mbruzek has quit IRC | 15:50 | |
| *** mbruzek has joined #openstack-kolla | 15:51 | |
| *** mbruzek has quit IRC | 15:51 | |
| *** mbruzek has joined #openstack-kolla | 15:51 | |
| swinn | egonzalez: nova==15.0.0.0rc2.dev879 | 15:51 |
| egonzalez | swinn, is master then | 15:51 |
| *** Administrator_ has quit IRC | 15:51 | |
| swinn | in globals I have a release set…. | 15:51 |
| egonzalez | building image is not tied to globals.yml | 15:52 |
| *** Administrator_ has joined #openstack-kolla | 15:52 | |
| openstackgerrit | Dai Dang Van proposed openstack/kolla-ansible stable/ocata: Use utf8_general_ci collation as a default collation https://review.openstack.org/455747 | 15:53 |
| egonzalez | swinn, i would do, git chechout origin/stable/ocata ; cd tools ; ./build.py --BUILD_OPTIONS | 15:53 |
| *** mbruzek1 has joined #openstack-kolla | 15:54 | |
| *** mbruzek1 has quit IRC | 15:54 | |
| *** mbruzek1 has joined #openstack-kolla | 15:54 | |
| *** eaguilar has joined #openstack-kolla | 15:54 | |
| swinn | egonzalez: thanks, will give that a shot | 15:55 |
| *** mbruzek has quit IRC | 15:55 | |
| *** Jezekus has quit IRC | 15:55 | |
| *** mbruzek1 has quit IRC | 15:55 | |
| *** mbruzek has joined #openstack-kolla | 15:56 | |
| *** mbruzek has joined #openstack-kolla | 15:56 | |
| *** eaguilar has quit IRC | 15:57 | |
| *** jistr is now known as jistr|mtg | 15:59 | |
| *** rwsu has quit IRC | 16:04 | |
| *** rwsu has joined #openstack-kolla | 16:04 | |
| *** eaguilar has joined #openstack-kolla | 16:04 | |
| *** manheim has joined #openstack-kolla | 16:10 | |
| *** unicell has quit IRC | 16:12 | |
| *** unicell has joined #openstack-kolla | 16:14 | |
| *** tovin07 has quit IRC | 16:14 | |
| *** rwallner has joined #openstack-kolla | 16:14 | |
| *** ggillies has quit IRC | 16:15 | |
| *** manheim has quit IRC | 16:15 | |
| *** ggillies has joined #openstack-kolla | 16:17 | |
| openstackgerrit | Merged openstack/kolla-ansible master: Detect openstack_release variable in mariadb recovery playbooks https://review.openstack.org/455505 | 16:18 |
| *** Serlex has quit IRC | 16:20 | |
| *** swinn has quit IRC | 16:21 | |
| *** wagiel has quit IRC | 16:21 | |
| *** egonzalez has quit IRC | 16:29 | |
| *** pcaruana has quit IRC | 16:29 | |
| *** jistr|mtg is now known as jistr | 16:37 | |
| *** daidv has quit IRC | 16:37 | |
| kfox1111 | so... back to the point where ceph support in kolla 4 seems to be broken... | 16:38 |
| kfox1111 | http://logs.openstack.org/41/454841/7/experimental/gate-kolla-kubernetes-deploy-centos-binary-4-ceph-multi-nv/c4655a6/logs/openstack/kolla-nova-libvirt-lfkjq.txt | 16:39 |
| *** lpetrut has quit IRC | 16:39 | |
| *** swinn has joined #openstack-kolla | 16:41 | |
| kfox1111 | hmm.. same binaries though. how very strange... | 16:45 |
| openstackgerrit | Marcus Williams proposed openstack/kolla-ansible master: Add OpenDaylight role https://review.openstack.org/416367 | 16:46 |
| kfox1111 | welll, between 3.0.3 and 4.0.0... maybe 3.0.2 is different... | 16:46 |
| *** nathharp has quit IRC | 16:47 | |
| kfox1111 | hmm.. versions are a bit different with 3.0.3 -> 3.0.2... lets try and pull 3 up to 3.0.3 too. | 16:49 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes master: WIP: v4 gates. https://review.openstack.org/454841 | 16:49 |
| *** bogdando has quit IRC | 16:56 | |
| *** wagiel has joined #openstack-kolla | 17:05 | |
| *** abalutoiu_ is now known as abalutoiu | 17:08 | |
| openstackgerrit | sean mooney proposed openstack/kolla-ansible master: introduce contrib playbook for ovs-dpdk https://review.openstack.org/408872 | 17:09 |
| *** wagiel has quit IRC | 17:09 | |
| *** eaguilar has quit IRC | 17:11 | |
| *** jascott1_ has joined #openstack-kolla | 17:13 | |
| *** eaguilar has joined #openstack-kolla | 17:13 | |
| jascott1_ | o/ | 17:14 |
| openstackgerrit | sean mooney proposed openstack/kolla-ansible master: This change allow setting kernel commandline args https://review.openstack.org/408876 | 17:14 |
| *** sambetts is now known as sambetts|afk | 17:15 | |
| kfox1111 | jascott1_: hey. | 17:15 |
| jascott1_ | hey hows it going? | 17:16 |
| kfox1111 | jascott1_: sdake thinks 2.3.0 broke the compute kit package. | 17:16 |
| jascott1_ | how so? | 17:16 |
| kfox1111 | haven't seen him to verify why he thinks that. | 17:16 |
| jascott1_ | ah | 17:16 |
| kfox1111 | seeing stuff like this now: | 17:16 |
| openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible master: Temporaly fix deploy gate https://review.openstack.org/455785 | 17:16 |
| kfox1111 | http://logs.openstack.org/80/454380/1/experimental/gate-kolla-kubernetes-deploy-centos-binary-2-helm-compute-kit-nv/388bf97/console.html#_2017-04-08_16_55_18_932113 | 17:17 |
| jascott1_ | "Please upgrade to the MS Helm Pro" | 17:17 |
| kfox1111 | ;) | 17:17 |
| kfox1111 | yeah, we'll see what happens there... | 17:18 |
| jascott1_ | oh yeah saw him talking about that yesterday | 17:18 |
| openstackgerrit | Paul Bourke (pbourke) proposed openstack/kolla-ansible master: [WIP] Add a Kolla 'devstack' mode https://review.openstack.org/454690 | 17:20 |
| *** lamt has quit IRC | 17:23 | |
| *** jascott1_ has quit IRC | 17:27 | |
| *** jascott1_ has joined #openstack-kolla | 17:28 | |
| *** eaguilar has quit IRC | 17:29 | |
| *** jascott1_ has quit IRC | 17:34 | |
| *** eaguilar has joined #openstack-kolla | 17:35 | |
| *** unicell has quit IRC | 17:40 | |
| *** unicell has joined #openstack-kolla | 17:40 | |
| *** bmaltais has joined #openstack-kolla | 17:40 | |
| bmaltais | HI, I am trying to setup AIO kolla kubernetes following https://docs.openstack.org/developer/kolla-kubernetes/deployment-guide.html#introduction | 17:41 |
| *** MasterOfBugs has joined #openstack-kolla | 17:41 | |
| *** pramodrj07 has joined #openstack-kolla | 17:41 | |
| bmaltais | I am running into an issue | 17:41 |
| bmaltais | the following step fails: git review -x 439740 | 17:42 |
| bmaltais | I somehow can't pull the change as I am asked to authenticate | 17:42 |
| bmaltais | But I don't have a userid:password | 17:42 |
| bmaltais | Any tip to get past this? Without that the step "kolla-kubernetes/tools/secret-generator.py create" is failing | 17:43 |
| bmaltais | Anyone? | 17:47 |
| *** lamt has joined #openstack-kolla | 17:48 | |
| *** srwilkers has quit IRC | 17:53 | |
| kfox1111 | bmaltais: https://wiki.openstack.org/wiki/How_To_Contribute | 17:54 |
| openstackgerrit | Marcus Williams proposed openstack/kolla-ansible master: Add OpenDaylight role https://review.openstack.org/416367 | 17:55 |
| bmaltais | I pulled the fix using "git fetch git://git.openstack.org/openstack/kolla-kubernetes refs/changes/40/439740/39 && git cherry-pick FETCH_HEAD" | 17:56 |
| bmaltais | But apparently it does not fix the issue. I keep getting an error when running: "kolla-kubernetes/tools/secret-generator.py create" | 17:57 |
| bmaltais | kolla_kubernetes.exception.KollaDirNotFoundException: Unable to locate None directory in search_paths=[/root/kolla-bringup/kolla-kubernetes/share/kolla, /usr/share/kolla, /usr/lib/python2.7/kolla, /usr/lib/python2.7/site-packages/kolla] | 17:57 |
| *** wagiel has joined #openstack-kolla | 17:59 | |
| kfox1111 | bmaltais: seems like it wasn't applied, as the change changes kolla to kolla-ansible, | 17:59 |
| kfox1111 | but the output says its not looking there. | 18:00 |
| kfox1111 | is there an install step that need to be run? | 18:00 |
| bmaltais | kfox1111: I guess the git fetch did not work | 18:01 |
| bmaltais | I am still not able to do a git review -x 439740 as stated in the instructions | 18:01 |
| bmaltais | Darn | 18:01 |
| esharao | bmaltais have you installed the tool ? | 18:02 |
| esharao | https://docs.openstack.org/infra/git-review/installation.html | 18:02 |
| *** jmccarthy has quit IRC | 18:02 | |
| bmaltais | I did a yum install git-review but did not look at those specific instructions... let me have a look | 18:03 |
| inc0 | kk | 18:04 |
| inc0 | ^ sry wrong window;) | 18:04 |
| inc0 | but hello everyone | 18:04 |
| *** wagiel has quit IRC | 18:05 | |
| *** fooliouno has quit IRC | 18:05 | |
| bmaltais | <kfox1111>: Must be doing something wrong... running "git review -x 439740" result in: | 18:07 |
| *** unicell has quit IRC | 18:07 | |
| bmaltais | Exception: Could not connect to gerrit at ssh://bmaltais@review.openstack.org:29418/openstack/kolla-kubernetes.git | 18:07 |
| *** eaguilar has quit IRC | 18:08 | |
| bmaltais | Perhaps I registered my account in the wrong place... | 18:08 |
| inc0 | hmm...maybe we can just merge this review? | 18:08 |
| inc0 | kfox1111: ^ | 18:08 |
| inc0 | since we need it. | 18:08 |
| bmaltais | kfox111: Trying using the web interface I get: The page you requested was not found, or you do not have permission to view this page | 18:09 |
| inc0 | bmaltais: https://review.openstack.org/#/c/439740/ | 18:09 |
| bmaltais | So is it possible my account does not have read access to pull the commit? | 18:09 |
| inc0 | on right-top corner you have "download" | 18:10 |
| inc0 | try these links | 18:10 |
| bmaltais | inc0: tried the cherry pick link... but it did not work. Should I try the checkout instead? | 18:11 |
| bmaltais | inc0: git fetch ssh://null@review.openstack.org:29418/openstack/kolla-kubernetes refs/changes/40/439740/39 && git checkout FETCH_HEAD | 18:12 |
| bmaltais | result in: Permission denied (publickey). | 18:12 |
| bmaltais | fatal: Could not read from remote repository. | 18:12 |
| *** trozet has joined #openstack-kolla | 18:15 | |
| trozet | hi, can a core reviewer please review https://review.openstack.org/#/c/416369/ ? | 18:15 |
| bmaltais | inc0: git pull git://git.openstack.org/openstack/kolla-kubernetes refs/changes/40/439740/39 | 18:15 |
| bmaltais | That worked | 18:15 |
| bmaltais | But "kolla-kubernetes/tools/secret-generator.py create" still error out | 18:16 |
| *** srwilkers has joined #openstack-kolla | 18:16 | |
| bmaltais | "kolla_kubernetes.exception.KollaDirNotFoundException: Unable to locate None directory in search_paths=[/root/kolla-bringup/kolla-kubernetes/share/kolla, /usr/share/kolla, /usr/lib/python2.7/kolla, /usr/lib/python2.7/site-packages/kolla]" | 18:16 |
| inc0 | trozet: while I'm ok with having ODL containers in Kolla per se (they don't cost us much), it's gonna be interesting discussion whether it wouldn't be better to create kolla-ansible like deployment tool for ODL | 18:17 |
| inc0 | by kolla-ansible like I mean runninc docker containers with ansible | 18:17 |
| trozet | inc0: for my purpose (OOO) i only need the container | 18:18 |
| bmaltais | Perhaps that patch was broken in a recent update? | 18:18 |
| inc0 | OOO? | 18:18 |
| trozet | inc0: tripleO | 18:18 |
| inc0 | ah | 18:18 |
| inc0 | right, ok | 18:18 |
| inc0 | well still good thing to talk about, but for sake of not blocking you I won't cause a fuzz;) | 18:20 |
| inc0 | I'll need to talk with opennvf and odl people about taht tho | 18:20 |
| satyar | Hi Inc0 | 18:20 |
| satyar | how are you? | 18:21 |
| trozet | inc0: thanks. What is opennvf? | 18:21 |
| inc0 | opennfv | 18:21 |
| *** MasterOfBugs has quit IRC | 18:21 | |
| *** pramodrj07 has quit IRC | 18:21 | |
| inc0 | or opnfv | 18:21 |
| trozet | inc0: you mean opnfv? | 18:21 |
| inc0 | my bad | 18:21 |
| inc0 | yeah | 18:21 |
| trozet | inc0: i work on that and ODL | 18:21 |
| trozet | inc0: first time in kolla channel though :) | 18:21 |
| satyar | inc0: i have a strage issue with ocata | 18:21 |
| bmaltais | inc0: OK, found my problem... PBKAC | 18:22 |
| inc0 | cool, then maybe we can start this discussion:) | 18:22 |
| inc0 | trozet: I just talk with few of my collegues about deployment of opnfv | 18:22 |
| satyar | strage -> strange | 18:22 |
| bmaltais | Forgot to re-run sudo pip install -U kolla-ansible/ kolla-kubernetes/ after applying the patch | 18:22 |
| trozet | inc0: which ones? I run the Apex OPNFV project, which is an NFV/SDN installer based on TripleO | 18:22 |
| inc0 | and I suggested that instead of doing this in Kolla, we can do dedicated project with kolla+kolla-ansible methodology, which is proven and I think really good | 18:22 |
| inc0 | I work for Intel so we have some people in this general idea | 18:23 |
| inc0 | in any case, while I'm reluctant to put it all to KOlla (obviously, our mission is about OpenStack) | 18:23 |
| trozet | inc0: oh so you are saying ODL should live it it's own kolla project, because it is not native openstack? | 18:24 |
| inc0 | I think there is tremedous value in our learnings/infrastructure that we can and will share happily | 18:24 |
| inc0 | my question is, should ODL live in openstack namespace at all? | 18:24 |
| *** abalutoiu has quit IRC | 18:25 | |
| inc0 | or rather have sister project inside OPNFV foundation that will work very well with kolla but will be dedicated for deploying OPNFV | 18:25 |
| inc0 | as OPNFV is not exactly OpenStack centric | 18:25 |
| inc0 | satyar: what's the issue? | 18:25 |
| *** lucasxu has joined #openstack-kolla | 18:25 | |
| *** gfidente is now known as gfidente|afk | 18:25 | |
| satyar | inc0: affter upgrading the neutron container not able to ssh to the VM | 18:25 |
| trozet | inc0: well networking-odl lives in openstack, the driver for neutron to use, but ODL itself comes from ODL's repo, and we are moving puppet-opendaylight and ansible-opendaylight into the ODL repos | 18:26 |
| trozet | inc0: i think it makes more sense for kolla-odl or whatever you want to call it to live in ODL repo | 18:26 |
| satyar | and the strange part if the tap device MTU is 8950 but if i check the mtu of the VM its 1500 (i have jumbo frame setup) | 18:26 |
| trozet | inc0; if the kolla team doesnt want to support every possible SDN controller in their own repo | 18:26 |
| satyar | if i create new VM its accessable | 18:26 |
| *** abalutoiu has joined #openstack-kolla | 18:27 | |
| satyar | now i am not getting any way to find whats wrong with it why the VM which is created earlier having 1500 MTU after upgrade | 18:27 |
| inc0 | trozet: exactly my point, while I'd be happy to have overlap of communities, I'd rather have ODL to be separate core team and stuff like that | 18:27 |
| inc0 | anyway, I +2'd this docker change, since it's not decided either way and won't make us any less stable | 18:27 |
| trozet | inc0: makes sense, i'm just responsible for getting it containerized in tripleO. So I need a container, and I saw Marcus was already doing that with that patch | 18:28 |
| inc0 | and maybe we can start talking between communities | 18:28 |
| trozet | inc0: but we should start a discussion | 18:28 |
| trozet | inc0: yeah | 18:28 |
| inc0 | who should I talk to about that from ODL/OPNFV side? | 18:28 |
| trozet | inc0: let me see if i can get him to come in here :) | 18:29 |
| inc0 | thank you good sir | 18:29 |
| satyar | Sorry to interrupt inc0 | 18:30 |
| inc0 | np satyar, just juggling multiple things around | 18:30 |
| satyar | yeah was reading that :) | 18:31 |
| *** abalutoiu_ has joined #openstack-kolla | 18:31 | |
| satyar | this gonna be a nice thing... | 18:31 |
| inc0 | satyar: so that's something we definetly need to look at, would you mind publishing a bug for it? | 18:31 |
| inc0 | mtu issue? | 18:31 |
| satyar | no the ODL nfv | 18:31 |
| inc0 | not sure if it's bug in kolla, bug or neutron or not a bug at all | 18:31 |
| satyar | definately a bug | 18:32 |
| satyar | but not sure where :P | 18:32 |
| inc0 | maybe some strange conf | 18:32 |
| inc0 | in any case, please submit it to our tracker | 18:32 |
| satyar | the conf is exactly what kolla provides | 18:32 |
| inc0 | not of everything:) we don't configure your jf;) | 18:33 |
| satyar | only addition is the MTU in neutron and ml2 | 18:33 |
| inc0 | yeah I remember that there was sth about it | 18:33 |
| satyar | yeah that i added to my internal kolla build :P | 18:33 |
| inc0 | if it's in build time, then it won't carry over to deploy | 18:33 |
| satyar | its in the ansible | 18:34 |
| inc0 | also all our containers use net=host so we don't mess with MTU anywhere | 18:34 |
| satyar | when i deploy its good | 18:34 |
| satyar | but after upgrade old VMs not not happy as their MTU is 1500 | 18:34 |
| satyar | but tap devide MTU is 8950 | 18:34 |
| *** abalutoiu has quit IRC | 18:34 | |
| trozet | inc0: he isn't responding right now. I would think this repo should live in ODL. Daniel Farrell dfarrell07 is the guy you want to talk to. We could add a new repo for it under the integration & test ODL repo (where puppet/ansible opendaylight projects will live) | 18:34 |
| satyar | same | 18:34 |
| satyar | my deployment also having net=host | 18:35 |
| inc0 | trozet: which irc channel? | 18:35 |
| trozet | inc0: you can use #opendaylight, i can also give you his Red Hat email | 18:35 |
| satyar | just to have VMs to get 8950 MTU having changes in 3 places nova.conf, neutron.conf and ml2.conf.ini | 18:35 |
| inc0 | I'll figure it out:) thanks trozet for making connection | 18:36 |
| trozet | inc0: thanks for the +2! | 18:36 |
| inc0 | todat I have fairly busy day but I'll try to reach out tomorrow | 18:36 |
| inc0 | or just answer here if Daniel happends to join us:) | 18:36 |
| *** bmaltais has quit IRC | 18:37 | |
| inc0 | satyar: make sure that your changes to confs landed in container as it should | 18:38 |
| inc0 | just docker exec into container and confirm configs are right | 18:39 |
| SamYaple | satyar: the mtu thing is neutron not being smart. or trying ot be _too_ smart | 18:40 |
| SamYaple | satyar: you can make the instances have whatever you want for mtu, but it requires some work | 18:41 |
| SamYaple | personally, i set my instances to 1500 mtu for compatibility | 18:41 |
| satyar | yes | 18:42 |
| satyar | inc0: yes its landed in the container | 18:42 |
| inc0 | yeah that sounds like neutron issue to me too | 18:43 |
| kfox1111 | inc0: no we can't just merge that review. its breaks all the gates. | 18:43 |
| satyar | SamYaple: initial deployment works fine only issue with upgrade (old VMs getting set to 1500 where new VMs getting 8950) | 18:43 |
| kfox1111 | the docs can be updated to actually just use the symlink like the gates do until that merges. | 18:44 |
| inc0 | yeah...maybe | 18:44 |
| kfox1111 | or we can finish getting the config to kolla-kubernetes. | 18:44 |
| kfox1111 | whatever's faster. | 18:44 |
| inc0 | well, I thing this is about kollakube resource create | 18:44 |
| SamYaple | satyar: yea i know | 18:44 |
| SamYaple | satyar: thats a new neutron "feature" | 18:45 |
| satyar | SamYaple: while with initial deployment VMs were getting 8950 after upgrade restarted the VMs and saw them having 1500 | 18:45 |
| kfox1111 | inc0: its about finding the kolla-ansible bits. | 18:45 |
| inc0 | https://review.openstack.org/#/c/447733/ | 18:45 |
| kfox1111 | which should be all gone once the config is finished being forked. | 18:45 |
| satyar | :O feature | 18:45 |
| satyar | i thought its a bug | 18:45 |
| SamYaple | satyar: yea i have an old ML thread about this (from a 9 months ago or so) | 18:45 |
| satyar | please post me that will help | 18:46 |
| kfox1111 | inc0: I prefer this one https://review.openstack.org/#/c/449262/ | 18:46 |
| SamYaple | satyar: ill go over your configs if you want, can you post your ml2_ini.conf and neutron.conf (removing passwords) from your containers | 18:46 |
| satyar | sure | 18:46 |
| kfox1111 | no extra ansible required then. | 18:46 |
| SamYaple | satyar: no there was no resolution from that thread. just acknoleging the issue and saying "its not an issue" | 18:46 |
| SamYaple | so now its just working around :/ | 18:47 |
| SamYaple | satyar: what would you _like_ to see? 8950 mtu, or 1500 mtu? | 18:47 |
| satyar | 8950 | 18:47 |
| SamYaple | version of openstack you are upgrading to? | 18:47 |
| SamYaple | newton or ocata? | 18:47 |
| *** shashank_t_ has joined #openstack-kolla | 18:48 | |
| *** shashank_t_ has quit IRC | 18:49 | |
| satyar | yes | 18:49 |
| satyar | Ocata to ocata | 18:49 |
| satyar | http://paste.openstack.org/show/606196/ my neutron.conf | 18:49 |
| *** lamt has quit IRC | 18:50 | |
| satyar | SamYaple: my ml2.conf.ini http://paste.openstack.org/show/606197/ | 18:51 |
| SamYaple | satyar: what version of neutron ocata to what version? are you saying within the *same* *version* this is changing? | 18:51 |
| satyar | it was 3 weeks back to 1 week back | 18:52 |
| satyar | almost same version | 18:52 |
| satyar | yes SamYaple | 18:52 |
| satyar | BTW i saw your blogs and loved those :) | 18:53 |
| SamYaple | uhh. hmm. if this was the tagged stable... that shouldnt be happening | 18:53 |
| SamYaple | satyar: yea ive got to update it with all the newton/ocata changes | 18:53 |
| inc0 | ha! Kolla-ansible is 6th most reviewed project in OpenStack | 18:53 |
| SamYaple | its got misinformation right now :( | 18:53 |
| inc0 | all glory to Kolla! | 18:53 |
| *** wagiel has joined #openstack-kolla | 18:54 | |
| satyar | with 1500 MTU i dont see any issues | 18:54 |
| satyar | only seeing the issue when jumbo frame is setup | 18:54 |
| SamYaple | satyar: can you start a container with each version of kolla neutron and pip list to get me the neutron version in each? | 18:54 |
| satyar | sure | 18:54 |
| satyar | i have 2 setup | 18:54 |
| SamYaple | cool. yea i just want to make sure you are runnning a stable version, it shouldnt be changing unless you did a reboot of the host and the MTUs changed on startup or something | 18:55 |
| *** lamt has joined #openstack-kolla | 18:55 | |
| satyar | http://paste.openstack.org/show/606198/ the one which is upgraded | 18:55 |
| SamYaple | :/ its 0.0.0 | 18:56 |
| *** abalutoiu has joined #openstack-kolla | 18:56 | |
| SamYaple | thats a kolla build bug there | 18:56 |
| inc0 | how fun | 18:56 |
| satyar | http://paste.openstack.org/show/606199/ the one from which i am upgrading | 18:56 |
| inc0 | hmm, I wonder why tho, I mean we don't pip install neutron | 18:56 |
| inc0 | we downtload tarball | 18:57 |
| inc0 | and pip install -e . | 18:57 |
| SamYaple | inc0: yea but it requires a history | 18:57 |
| SamYaple | so if neutron stopped including the .git dir.... | 18:57 |
| SamYaple | or isnt made into an sdist | 18:57 |
| SamYaple | then it wont work | 18:57 |
| SamYaple | not sure which it is | 18:57 |
| inc0 | hold on, neutron checks it's own version for mtu, and tarballs-installed is no-bueno? | 18:57 |
| inc0 | pbr is being clever? | 18:58 |
| *** wagiel has quit IRC | 18:58 | |
| inc0 | I don't think tarballs.o.o has all the .git stuff | 18:58 |
| inc0 | to me it's not kolla issue...it's pbr | 18:58 |
| inc0 | or tarballs.o.o | 18:59 |
| inc0 | I'll validate this one | 18:59 |
| *** abalutoiu_ has quit IRC | 19:00 | |
| SamYaple | inc0: well its something a kolla build could _fix_ | 19:00 |
| SamYaple | thats probably a better way to put it | 19:00 |
| satyar | i used source based installation | 19:00 |
| SamYaple | but i dont disagree with it being something that can be fixed in tarballs.o.o | 19:00 |
| inc0 | agree | 19:00 |
| inc0 | and yes, tarballs doesn't hold .git | 19:01 |
| SamYaple | right, but it should be an sdist | 19:01 |
| inc0 | which means pbr will break | 19:01 |
| SamYaple | no if its an sdist pbr still works | 19:01 |
| SamYaple | if it was a full clone when the sdist was made, it should work | 19:01 |
| inc0 | let me check if I can run pbr in freshly untared neutron | 19:02 |
| *** eaguilar has joined #openstack-kolla | 19:02 | |
| inc0 | ok it works | 19:03 |
| inc0 | so it might be kolla bug | 19:03 |
| inc0 | right, satyar did you publish a bug? | 19:05 |
| inc0 | I'll bump it to high | 19:05 |
| inc0 | and link this convo to desc | 19:06 |
| satyar | for neutron version? | 19:06 |
| *** lamt has quit IRC | 19:06 | |
| inc0 | well mtu | 19:06 |
| inc0 | it might be due to version info for neutron | 19:06 |
| satyar | ohh ok | 19:06 |
| satyar | raising it | 19:06 |
| *** Manheim has joined #openstack-kolla | 19:07 | |
| *** lucasxu has quit IRC | 19:08 | |
| *** abalutoiu_ has joined #openstack-kolla | 19:08 | |
| SamYaple | inc0: there was recently that pbourke change with git+sdist, i didnt dig into it, perhaps that changed something? | 19:09 |
| inc0 | I must've missed that, possibly | 19:09 |
| SamYaple | satyar: its not because neutron is _missing_ version, but if i dont know what version is installed, i dont know i can help | 19:09 |
| inc0 | yes, there was sth | 19:09 |
| inc0 | well, that's a bug either way | 19:10 |
| satyar | yeah raising 2 issues | 19:10 |
| satyar | 1 for MTU other for version | 19:10 |
| satyar | is it fine? | 19:10 |
| inc0 | yeah | 19:11 |
| *** lucasxu has joined #openstack-kolla | 19:11 | |
| *** abalutoiu has quit IRC | 19:12 | |
| *** abalutoiu_ has quit IRC | 19:12 | |
| satyar | inc0: SamYaple: https://bugs.launchpad.net/kolla/+bug/1681919 | 19:19 |
| openstack | Launchpad bug 1681919 in kolla "After upgrade VMs getting 1500 MTU although jumbo frame issetup" [Undecided,New] | 19:19 |
| *** dave-mccowan has quit IRC | 19:21 | |
| satyar | inc0: SamYaple: for version issue https://bugs.launchpad.net/kolla/+bug/1681920 | 19:22 |
| openstack | Launchpad bug 1681920 in kolla "Neutron version is missing from pip list" [Undecided,New] | 19:22 |
| *** jascott1_ has joined #openstack-kolla | 19:27 | |
| *** bmace has quit IRC | 19:30 | |
| *** bmace has joined #openstack-kolla | 19:30 | |
| *** lamt has joined #openstack-kolla | 19:31 | |
| *** jascott1_ has quit IRC | 19:31 | |
| *** jascott1_ has joined #openstack-kolla | 19:33 | |
| *** pramodrj07 has joined #openstack-kolla | 19:36 | |
| *** MasterOfBugs has joined #openstack-kolla | 19:36 | |
| *** jascott1_ has quit IRC | 19:36 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes master: WIP: v4 gates. https://review.openstack.org/454841 | 19:39 |
| *** sayantan_ has joined #openstack-kolla | 19:39 | |
| *** jascott1_ has joined #openstack-kolla | 19:40 | |
| *** jascott1_ has quit IRC | 19:43 | |
| mgkwill | inc0 + trozet FYI one of key use cases I'm trying to enable is Intel reference architecture which has adopted Kolla for deploy. They have ODL on roadmap for next release and ease of use is key. While I agree that ODL might not belong long term in current location, testing/gate, repo etc would have to be duplicated elsewhere and lots worked out to use outside | 19:45 |
| mgkwill | of Kolla community. I liked original idea of overrides, where changes don't affect stability of core project but allow expansion using outside tools, with eventual graduation to stable from experimental. | 19:45 |
| *** satyar has quit IRC | 19:45 | |
| *** jascott1_ has joined #openstack-kolla | 19:46 | |
| *** wagiel has joined #openstack-kolla | 19:48 | |
| *** manheim_ has joined #openstack-kolla | 19:49 | |
| mgkwill | inc0 trozet: dfarrell07 is indeed good resource on this to get opinion (I added him to all patches) But I bet the person doing the work would be me. | 19:49 |
| *** esharao has quit IRC | 19:50 | |
| *** jascott1_ has quit IRC | 19:50 | |
| *** shardy has quit IRC | 19:50 | |
| mgkwill | inc0: I'd love to meet up with you in JF for lunch this week to discuss if it would be helpful. | 19:51 |
| trozet | mgkwill: well get on it ;) | 19:52 |
| trozet | mgkwill: while your at it do the OOO pathc for me too :) | 19:52 |
| *** jascott1_ has joined #openstack-kolla | 19:52 | |
| trozet | patch* | 19:52 |
| *** wagiel has quit IRC | 19:52 | |
| *** manheim_ has quit IRC | 19:53 | |
| *** jascott1_ has quit IRC | 19:56 | |
| mgkwill | trozet: lol. I already have a bunch of OPNFV guys (Chris Price, Fatih etc) and some internal guys yelling at me to do ansible stuff for OPNFV master/master test. :) | 19:57 |
| *** lamt has quit IRC | 19:59 | |
| *** jascott1_ has joined #openstack-kolla | 20:00 | |
| trozet | mgkwill: for what? dovetail? | 20:00 |
| *** lamt has joined #openstack-kolla | 20:00 | |
| inc0 | mgkwill: oh you're in JF? | 20:02 |
| mgkwill | trozet: not sure about project. What was discussed with me was using ansible to deploy master branch of odl and master of openstack in test to catch bugs so they don't have to wait six months for them to reach stable branch and then take six more months to fix. Or something like that. Something Fatih is working on. | 20:02 |
| inc0 | mgkwill: where do you sit? | 20:02 |
| *** jmccarthy has joined #openstack-kolla | 20:02 | |
| *** sayantan_ has quit IRC | 20:03 | |
| *** jmccarthy has quit IRC | 20:03 | |
| mgkwill | inc0: Yes :) We marveled at that in Atlanta PTG. Usually on JF3-4 but renovations in progress, all work from home until next week, when we return to finished reno. But would make special trip to office. | 20:04 |
| inc0 | ahh, I didn't connect the dots:) | 20:04 |
| *** jascott1_ has quit IRC | 20:04 | |
| inc0 | mgkwill: you me and Rob should definetly sit together and talk | 20:05 |
| *** sayantan_ has joined #openstack-kolla | 20:05 | |
| mgkwill | inc0: Sounds good. Which Rob? | 20:05 |
| inc0 | Armstrong, from OTC | 20:06 |
| mgkwill | Ah yes. Was just talking to him about this work yesterday. | 20:06 |
| inc0 | he was in room with me at that very moment;) | 20:06 |
| *** jascott1_ has joined #openstack-kolla | 20:06 | |
| inc0 | discusing this very thing | 20:06 |
| mgkwill | inc0: Ha. Yeah I figured there was a reason he was asking me the questions he was. | 20:07 |
| mgkwill | inc0: OK i'll try to set something up. I'll look at both your calendars | 20:07 |
| inc0 | cool, thank you good sir | 20:07 |
| inc0 | I was thinking that we could replicate kolla-ansible methodology | 20:07 |
| inc0 | as it's proven and I really like it | 20:08 |
| *** Pavo has quit IRC | 20:09 | |
| mgkwill | inc0: yeah it's pretty slick. After I started to figure it out - i was pretty impressed. Only thing i'm worried about is overhead involved in starting something new, even with established base of methodology. But it might just be worth doing. | 20:09 |
| *** lamt has quit IRC | 20:09 | |
| inc0 | mgkwill: from my experience it's quicker to start something new than to push every change through reviews of stable project | 20:10 |
| inc0 | less red tape | 20:10 |
| *** jascott1_ has quit IRC | 20:11 | |
| inc0 | core teams of stable project need to learn the stuff you want to push | 20:12 |
| inc0 | and spend time reviewing each change | 20:12 |
| inc0 | odl being prime example - look how long it takes to merge just containers | 20:12 |
| mgkwill | inc0: Interesting. Didn't think of it that way. But now that I do, I see what you mean. | 20:13 |
| *** jascott1_ has joined #openstack-kolla | 20:14 | |
| mgkwill | inc0: sent meeting invite that should work for all of us on Thursday. | 20:14 |
| inc0 | if you start fresh you also don't have "openstack bagage" so people who wants opnfv for non-openstack things will also be interested | 20:14 |
| inc0 | that being said, nothing stops us from making kolla synchronize with this "new project" easily to deploy openstack+odl with just a few commands | 20:15 |
| *** jascott1_ has quit IRC | 20:19 | |
| *** sayantan_ has quit IRC | 20:20 | |
| mgkwill | inc0: At the moment in OPNFV there are 6 'installers' (trozet's APEX is one example) that do most of the deploy now and are the bulk of integration/testing work. If we pulled in kolla + kolla ansible + kolla odl (lacking a better name if we started a new thing) - I bet it would mostly be used for master source testing at first. But I take your point and I | 20:21 |
| mgkwill | bet Intel Reference Arch would use for other things. To your second point - very interested in making this happen. Was original point of my patch sets. | 20:21 |
| inc0 | I think we could get something pretty up and running fairly soon | 20:22 |
| inc0 | we have all the knowledge required | 20:22 |
| *** sdake has joined #openstack-kolla | 20:24 | |
| *** sdake has quit IRC | 20:24 | |
| *** sdake has joined #openstack-kolla | 20:24 | |
| *** lucasxu has quit IRC | 20:27 | |
| sdake | sup peeps | 20:28 |
| sdake | kfox1111 you around | 20:28 |
| sdake | inc0 you around | 20:28 |
| inc0 | for 20 more min | 20:28 |
| *** eaguilar has quit IRC | 20:28 | |
| *** lucasxu has joined #openstack-kolla | 20:30 | |
| *** kolla-slack has joined #openstack-kolla | 20:32 | |
| *** rwallner has quit IRC | 20:33 | |
| *** lucasxu has quit IRC | 20:34 | |
| *** lucasxu has joined #openstack-kolla | 20:35 | |
| sdake | kfox1111 need a merge on https://review.openstack.org/#/c/455502/ | 20:38 |
| sbezverk_ | sdake: it cannot be merged | 20:39 |
| sbezverk_ | it is failing on ironic centos | 20:39 |
| sbezverk_ | which was always working | 20:40 |
| *** lamt has joined #openstack-kolla | 20:40 | |
| *** wagiel has joined #openstack-kolla | 20:42 | |
| *** eaguilar has joined #openstack-kolla | 20:42 | |
| *** sayantan_ has joined #openstack-kolla | 20:43 | |
| sdake | sbezverk_ thiw is a straight revert on master | 20:43 |
| sdake | I think ironic is brokne for a different reason | 20:43 |
| sdake | sbezverk_ with helm 2.3.0 computekit is busted | 20:43 |
| *** wagiel_ has joined #openstack-kolla | 20:43 | |
| *** wagiel_ has quit IRC | 20:44 | |
| openstackgerrit | Bertrand Lallau proposed openstack/kolla-ansible master: WIP: Enable Ceph input plugin in Telegraf https://review.openstack.org/455602 | 20:44 |
| *** wagiel_ has joined #openstack-kolla | 20:44 | |
| sbezverk_ | sdake: computekit is not even part of main gate, so sucrificing a good gate job which is I see working in kfox1111 runs is not good option.. | 20:45 |
| sdake | sbezverk_ which good gate job are you speaking of? | 20:45 |
| *** Pavo has joined #openstack-kolla | 20:45 | |
| sbezverk_ | ironic centos works in kfox1111 runs but is failing in your ps.. I do not see any good reason for merging your ps | 20:47 |
| *** wagiel has quit IRC | 20:47 | |
| *** dmellado has quit IRC | 20:48 | |
| sdake | sbezverk_ computekit doesn't work | 20:48 |
| *** wagiel_ has quit IRC | 20:48 | |
| sdake | sbezverk_ ironic centos gate was promoted by me yesterday | 20:48 |
| sdake | computekit worked prior to kevin's fix | 20:49 |
| *** jascott1_ has joined #openstack-kolla | 20:49 | |
| sdake | rathe rprior to kevin's change to helm 2.3.0 | 20:49 |
| *** eaguilar has quit IRC | 20:49 | |
| sbezverk_ | sdake: merging something which starting breaking things is not right approach. When kfox1111's 2.3.0 ps got merged the gate was ok, otherwise we would not merge it. Maybe you are looking at the wrong place? | 20:50 |
| *** athomas has quit IRC | 20:51 | |
| sdake | sbezverk_ kfox1111's patch never ran a check experimental. | 20:51 |
| sdake | where ironci-4 and computekit were preiviosuly | 20:51 |
| sdake | hence neither of those gate jobs wer erun nor tested | 20:51 |
| sdake | sbezverk_ i agree merging 2.3.0 was the wrong approach as it was not checked against the working experimental gates | 20:52 |
| *** jascott1_ has quit IRC | 20:54 | |
| sdake | i just checked - it was tested against ironic | 20:54 |
| sdake | I think seomthing else broke in ironic | 20:54 |
| sdake | it is clear kfox1111 's patch broke computekit | 20:55 |
| sbezverk_ | sdake: why break ironic for "restoration" os compute kit? | 20:55 |
| sbezverk_ | would it not be better to get the root cause?? | 20:56 |
| sdake | sbezverk_ indeed it would | 20:57 |
| sdake | sbezverk_ wait_for_vm I think doesn't have a long enough timer for ironic | 20:58 |
| sdake | so the ironic gate sometimes fails prematurely | 20:58 |
| *** eaguilar has joined #openstack-kolla | 20:58 | |
| sdake | computekit on the other hand always fails | 20:58 |
| *** eaguilar has quit IRC | 20:58 | |
| sdake | i've got to catch an uber back to the hotel - bbi20mins | 20:58 |
| *** ssurana has joined #openstack-kolla | 20:59 | |
| *** sbezverk_ is now known as sbezverk | 21:04 | |
| *** jascott1_ has joined #openstack-kolla | 21:04 | |
| *** mbruzek has quit IRC | 21:05 | |
| *** jtriley has quit IRC | 21:06 | |
| *** Pavo has quit IRC | 21:07 | |
| *** lucasxu has quit IRC | 21:12 | |
| *** lucasxu has joined #openstack-kolla | 21:16 | |
| *** gfidente|afk has quit IRC | 21:16 | |
| *** eaguilar has joined #openstack-kolla | 21:17 | |
| *** lamt has quit IRC | 21:21 | |
| *** g3ek has quit IRC | 21:25 | |
| *** lamt has joined #openstack-kolla | 21:26 | |
| *** vcn[m] has quit IRC | 21:30 | |
| *** kong has joined #openstack-kolla | 21:30 | |
| *** eaguilar has quit IRC | 21:34 | |
| *** g3ek has joined #openstack-kolla | 21:34 | |
| sdake | hey folks | 21:37 |
| *** wagiel has joined #openstack-kolla | 21:38 | |
| *** jascott1_ has quit IRC | 21:39 | |
| *** lamt has quit IRC | 21:40 | |
| bmace | hey sdake | 21:42 |
| *** wagiel has quit IRC | 21:42 | |
| *** lamt has joined #openstack-kolla | 21:44 | |
| sdake | sup bmace | 21:44 |
| bmace | sdake: you know, the usual.. code code code :) you? halcyon ever un-break? | 21:51 |
| kolla-slack | <sdake> nope | 21:51 |
| bmace | sdake: :( | 21:52 |
| *** unicell has joined #openstack-kolla | 21:52 | |
| *** lamt has quit IRC | 21:53 | |
| *** Manheim has quit IRC | 21:54 | |
| sdake | bmace ya - not maintained | 21:55 |
| sdake | getting ejected from our codebase I think | 21:56 |
| sdake | things that are not maintained go away | 21:56 |
| sdake | simpel as that | 21:56 |
| sdake | i was told to contribute patches to it if I wanted it maintained | 21:56 |
| sdake | its not my upstream to maintain | 21:56 |
| sdake | nor can I contribute to it if I wanted to do so | 21:56 |
| bmace | sdake: yeah, seems like we would need to fork + maintain it at this point | 21:57 |
| *** lamt has joined #openstack-kolla | 21:57 | |
| sdake | bmace CLA doesn't permit such a thing | 21:58 |
| sdake | bmace nobody really likes vagrant anyway | 21:58 |
| sdake | bmace the dev guide will have to do for the moment | 21:58 |
| sdake | rather the deploymnet guide | 21:58 |
| *** lucasxu has quit IRC | 21:58 | |
| bmace | sdake: well, if there is alternative to bring up a dev env quick that is fine | 21:58 |
| sdake | bmace yup there is - working on it | 21:58 |
| bmace | sdake: we ended up doing something similar with vagrant for spin of up a kolla environment for our cli work and devstack and it is really nice, and a no reliance on all the halcyon stuff, much simpler. | 21:59 |
| bmace | of up = up of | 22:00 |
| *** manheim has joined #openstack-kolla | 22:01 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes master: Do not merge: Chasing helm 2.3.0 compute-kit failure https://review.openstack.org/455839 | 22:08 |
| openstackgerrit | Steven Dake proposed openstack/kolla-kubernetes master: Revert "Update to Helm v2.3.0" https://review.openstack.org/455840 | 22:13 |
| openstackgerrit | Steven Dake proposed openstack/kolla-kubernetes master: Revert "Update to Helm v2.3.0" https://review.openstack.org/455841 | 22:20 |
| openstackgerrit | Steven Dake proposed openstack/kolla-kubernetes master: Test without all the extra crap in the revert. https://review.openstack.org/455842 | 22:20 |
| sdake | sbezverk i did notice the helm template version was wrong in helm_versions.sh | 22:20 |
| *** dgonzalez has quit IRC | 22:24 | |
| *** shashank_t_ has joined #openstack-kolla | 22:26 | |
| *** dgonzalez has joined #openstack-kolla | 22:28 | |
| *** wagiel has joined #openstack-kolla | 22:32 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes master: Do not merge: Chasing helm 2.3.0 compute-kit failure https://review.openstack.org/455839 | 22:34 |
| *** wagiel has quit IRC | 22:37 | |
| *** eanylin has quit IRC | 22:37 | |
| swinn | seems I have a mostly working kolla environment, dashboard panels are the only thing remaining which are bugged | 22:37 |
| swinn | and I’m not sure if my compute containers should be qemu based or not | 22:38 |
| swinn | but that’s what they report | 22:38 |
| *** dixiaoli has joined #openstack-kolla | 22:43 | |
| *** al498u has quit IRC | 22:43 | |
| *** lrensing has quit IRC | 22:45 | |
| *** manheim has quit IRC | 22:48 | |
| *** bachp has quit IRC | 22:48 | |
| swinn | seems I had to rebuild the horizon container and update it, multiple layers that hadn’t previously completed | 22:54 |
| kfox1111 | back. | 22:59 |
| *** srwilkers has quit IRC | 23:06 | |
| sdake | kfox1111 sup - helm 2.3.0 is busted | 23:12 |
| kfox1111 | ok. | 23:13 |
| kfox1111 | have you filed a bug report with helm? | 23:13 |
| sdake | i am not sure what is busted about it | 23:13 |
| sdake | henc ehave no data to file a bug report with | 23:13 |
| sdake | othe rthen it is BDWNF | 23:13 |
| kfox1111 | should still file one with as much as you know. can always fill in more later. | 23:14 |
| *** erlon has quit IRC | 23:15 | |
| sdake | helm 2.3.0 broken | 23:15 |
| sdake | Details :TBD | 23:15 |
| sdake | not helpful | 23:16 |
| openstackgerrit | Marcus Williams proposed openstack/kolla-ansible master: Split Openvswitch into own role https://review.openstack.org/416793 | 23:16 |
| sdake | since TBD will never happen | 23:16 |
| kfox1111 | sdake: its got to if we are to make progress. | 23:17 |
| kfox1111 | just going back to 2.2 forever is not a very good plan. | 23:17 |
| sdake | kfox1111 there are 1300 people in the helm channel | 23:18 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes master: WIP: v4 gates. https://review.openstack.org/454841 | 23:18 |
| sdake | surely one of them will find the probelm and fix it | 23:18 |
| kfox1111 | sdake: your assuming they are as crazy as we are in our usage. ;) | 23:18 |
| kfox1111 | assuming bad. | 23:19 |
| sdake | how would they duplicate hte issue? | 23:19 |
| kfox1111 | we can reproduce it in our gate by just enabling it. so thats a start. | 23:19 |
| sdake | yup - in the gate | 23:19 |
| sdake | helm devs are not going to debug helm inside our gate :) | 23:20 |
| kfox1111 | or on their own machine following your docs. | 23:20 |
| sdake | yup tha twould work - if the docs documented cmopute kit which they do not | 23:20 |
| sdake | i plan to ge tthere | 23:20 |
| sdake | atm - higher priority things then using latest bleeding edge of helm | 23:20 |
| sdake | like 4.0.0 gates for eg | 23:21 |
| kfox1111 | whatever. if you don't report bugs, they don't get fixed. just saying. | 23:21 |
| sdake | i dont report bugs when i have no data to report | 23:21 |
| sdake | you can report a bug as easily as I could with no data | 23:21 |
| kfox1111 | just knowing there's brokenness, not just that things are all perfect is still valueable. | 23:21 |
| sdake | well i did revert all your other changes for helm | 23:21 |
| sdake | for 2.3.0 | 23:21 |
| sdake | to see if those triggered teh breakage | 23:21 |
| sbezverk | kfox1111: I am on jascott1 and myself are on something for 2.3.0 | 23:22 |
| sdake | if they do, I'll report that | 23:22 |
| kfox1111 | sbezverk: ok. thanks. | 23:22 |
| sbezverk | there is no urgent need for us to jump back to 2.2.3 | 23:22 |
| sbezverk | let's get to the root cause | 23:22 |
| sbezverk | with helm | 23:22 |
| kolla-slack | <jascott1> which looks like 1mb limit on configmap that Helm creates for the release in Tiller | 23:23 |
| sdake | I see this failure 9 times out of 10: http://logs.openstack.org/42/455842/1/check/gate-kolla-kubernetes-deploy-centos-binary-3-ceph-multi-nv/0f8dd0b/console.html#_2017-04-11_22_48_42_831732 | 23:23 |
| sdake | the urgent need is it is blocking development sbezverk | 23:23 |
| sbezverk | really who? | 23:24 |
| sdake | sbezverk the gates dont work for compute kit - hence I can't do compute kit dev | 23:24 |
| sdake | its blocking *me* | 23:24 |
| sbezverk | I mean who is blocked? | 23:24 |
| sbezverk | what exactly is blocked ? | 23:24 |
| sbezverk | what are you developing which is blocked? | 23:24 |
| *** jtriley has joined #openstack-kolla | 23:25 | |
| kfox1111 | jascott1: interesting. that might explain it. | 23:25 |
| kfox1111 | jascott1: should be easy enough to test I think? can you make a test tiller with the 1m hard coded to 2m instead? | 23:25 |
| *** claudiub has quit IRC | 23:25 | |
| sdake | sbezverk https://review.openstack.org/#/c/453846/ | 23:26 |
| kfox1111 | or is it harder then that? | 23:26 |
| kolla-slack | <jascott1> sure! (i think) | 23:26 |
| kfox1111 | sdake: that review is for service charts. you can still work on that while we figure out the compute kit stuff. service charts != compute kit. | 23:26 |
| *** wagiel has joined #openstack-kolla | 23:26 | |
| sbezverk | sdake: centos ironic is broken, I already told you | 23:26 |
| sdake | kfox1111 compute kit uses service charts | 23:26 |
| sbezverk | in that PS | 23:26 |
| sdake | sbezverk jesus fucking christ sbezverk | 23:27 |
| sbezverk | I do not see why would we swap one problem for another | 23:27 |
| sdake | helm is broke n as well | 23:27 |
| sdake | i am not swapping | 23:27 |
| sdake | I am reducing the problem space | 23:27 |
| sdake | the patch is incomplete in some way | 23:27 |
| *** lamt has quit IRC | 23:27 | |
| sdake | here is the deal guys | 23:28 |
| sdake | my actions 1. deploy compute kit 2. conditionalize compute kit 3. victory | 23:28 |
| sbezverk | when kfox1111 runs his PS I see no more red than in yours. what is advantage then? | 23:28 |
| sdake | i can't do 1 or 2 if the thing doesn't launch in the gate | 23:28 |
| openstackgerrit | Marcus Williams proposed openstack/kolla-ansible master: Add OpenDaylight role https://review.openstack.org/416367 | 23:29 |
| kfox1111 | sdake: 1. conditionalize service. 2. we fix helm, 3 victory. | 23:29 |
| sdake | kfox1111 service is conditionalized | 23:29 |
| sdake | 2. what I mean here is conditionalize each service | 23:29 |
| kfox1111 | whats there to conditionalize in the compute kit? | 23:29 |
| kfox1111 | maybe I'm missing something? | 23:29 |
| sdake | i actually want to make a new chart helm install openstack | 23:29 |
| sdake | where openstack has all the crap disabled | 23:29 |
| kfox1111 | you want to make it possible to not deploy neutron say, when installing computekit? | 23:29 |
| sdake | and you enable | 23:29 |
| sdake | yup | 23:29 |
| sdake | but not computekit | 23:29 |
| sdake | for a new chart | 23:30 |
| kfox1111 | ah. ok. | 23:30 |
| sdake | i cant' do that as currently computekit is broken | 23:30 |
| kfox1111 | if its afecting you in that way, we can revert it I guess. | 23:30 |
| kfox1111 | unless 2.3 fixed something. | 23:31 |
| kfox1111 | sbezverk: did 2.3 fix anything? | 23:31 |
| *** rwellum has quit IRC | 23:31 | |
| sbezverk | did it not enabled conditionals? | 23:31 |
| *** wagiel has quit IRC | 23:31 | |
| sdake | 2.2.2 has conditionals | 23:31 |
| sdake | conditionals were enabled in 2.2.0 | 23:32 |
| sdake | 2.3.0 has nothing valueable for us that I am aware of | 23:32 |
| kfox1111 | sbezverk: 2.3 only added parent includes. which we haven't merged any yet. | 23:32 |
| kfox1111 | sbezverk: and faster builds, which is in the same ps. | 23:32 |
| sdake | i took out the faster builds | 23:32 |
| sdake | to see if that is the problem | 23:32 |
| kfox1111 | ok. so then reverting it allows dev to continue. no known drawbacks for now. | 23:33 |
| kfox1111 | I'm ok reverting then. | 23:33 |
| sdake | also helm_templates is the wrong version in your patch | 23:33 |
| sdake | kfox1111 i have a patch in teh queue to revert already | 23:33 |
| sbezverk | kfox1111: reverting breaks centos ironic job which was very stable | 23:34 |
| sbezverk | kfox1111: we might be swapping one problem for another | 23:34 |
| sdake | https://review.openstack.org/#/c/455502/ | 23:34 |
| sbezverk | since we do not investigate but just jungling with versions | 23:34 |
| sdake | sbezverk the helm 2.3.0 patch went in this morning | 23:34 |
| sdake | sbezverk ironic is broken for other reasons | 23:35 |
| sbezverk | sdake: really ?!?!?!?!?! | 23:35 |
| sbezverk | commit c2d7ec9bc3835a2763f7ca0ea04d57c0b03f6bb9 | 23:35 |
| sbezverk | Author: Kevin Fox <kevin@efox.cc> | 23:35 |
| sbezverk | Date: Thu Apr 6 15:33:51 2017 -0700 | 23:35 |
| sbezverk | Update to Helm v2.3.0 | 23:35 |
| sdake | sbezverk really | 23:36 |
| sdake | sbezverk that is the date it was committed | 23:36 |
| *** dixiaoli has quit IRC | 23:36 | |
| sdake | not the date it merged | 23:36 |
| sdake | commit 9224e8ceec284af8f446476820f4117e106ba235 | 23:36 |
| sdake | Merge: 87e0f9d c2d7ec9 | 23:36 |
| sdake | Author: Jenkins <jenkins@review.openstack.org> | 23:36 |
| sdake | Date: Mon Apr 10 14:02:33 2017 +0000 | 23:36 |
| sdake | Merge "Update to Helm v2.3.0" | 23:36 |
| *** tonanhngo has quit IRC | 23:36 | |
| kfox1111 | sbezverk: oh, it does? what part of 2.3 is required for ironic? | 23:37 |
| sdake | april 10th 2pm | 23:37 |
| sdake | sorry I guess it merged yesterday | 23:37 |
| sdake | the ironic gate was working premerge of that patch as well | 23:37 |
| sdake | which indicates ironic is broken for some other reason | 23:37 |
| kfox1111 | I wouldn't think anything in 2.3 would fix ironic. | 23:37 |
| sbezverk | kfox1111: it is not required, but on sdake's PS ironic is always failing | 23:37 |
| sbezverk | it makes me worrying | 23:37 |
| sdake | sbezverk i ran one gate run | 23:37 |
| sdake | sbezverk i have a recheck going now | 23:37 |
| kfox1111 | sbezverk: I'm seeing failures on multinode as well all day today. | 23:37 |
| sbezverk | kfox1111: when you run your test ps, ironic is ok | 23:38 |
| kfox1111 | been starting to track it on the etherpad. | 23:38 |
| kfox1111 | was wondering if infra changed anything. | 23:38 |
| sdake | i made this job non-voting (from experimental): gate-kolla-kubernetes-deploy-centos-source-4-ironic-nvFAILURE in 31m 50s (non-voting) | 23:38 |
| sbezverk | sdake: let's run test gate now to confirm. | 23:38 |
| sdake | as I thought it was stable | 23:38 |
| sdake | it appears it is not | 23:38 |
| sdake | sbezverk just recheck on my gatetest patch | 23:39 |
| sdake | sbezverk it passes there | 23:39 |
| sbezverk | so then run your PS several times let's see if the failure is consistent | 23:39 |
| sdake | https://review.openstack.org/#/c/422380/ | 23:39 |
| sdake | sbezverk sounds good | 23:39 |
| sdake | sbezverk I am only running the revert | 23:39 |
| sdake | on master | 23:39 |
| sdake | not my patch | 23:39 |
| sdake | the revert should have no bearing | 23:39 |
| sbezverk | kfox1111: when I see the failure on ironic it makes me worrying because it is realted to cell0 | 23:40 |
| sdake | sbezverk https://review.openstack.org/#/c/455842/ | 23:40 |
| sdake | sbezverk there is 2.3.0 of helm WITH passing ironic WITHOUT the optimizations in the compilation | 23:41 |
| sbezverk | kfox1111: I posted logs which indicating that to sdake to analyse | 23:41 |
| sdake | sbezverk posted where? | 23:41 |
| sbezverk | sdake: to the review | 23:41 |
| sbezverk | your previous revert ps | 23:41 |
| sdake | sbezverk riddle me this | 23:42 |
| sdake | https://review.openstack.org/#/c/455839/2 | 23:42 |
| sdake | gate-kolla-kubernetes-deploy-centos-source-4-ironic-nvFAILURE in 32m 19s (non-voting) | 23:42 |
| sdake | why would source 4 ironic fail in your patchset? | 23:43 |
| *** tonanhngo has joined #openstack-kolla | 23:43 | |
| sbezverk | sdake: because I have debugging version of helm | 23:44 |
| sdake | not sure whta that has to do with anything | 23:44 |
| kfox1111 | and another multinode failure.... rax again... | 23:45 |
| kfox1111 | starting to see a trend here... | 23:45 |
| kolla-slack | <jascott1> @sbezverk where is that stack trace? | 23:46 |
| sdake | kfox1111 ya - when I worked on saturday rax-iad failed 9 times out of 10 on multinode | 23:46 |
| sdake | kfox1111 i would have said something but you said at one point "lets stop blaming rax" | 23:46 |
| *** eaguilar has joined #openstack-kolla | 23:46 | |
| sbezverk | jascott1: it did not get generated | 23:46 |
| sdake | so - I stopping blaming them :) | 23:46 |
| sbezverk | jascott1: at last I could not find it | 23:47 |
| kfox1111 | sdake: I'm not blaming rax... just saying there is a pattern that may point to the problem. | 23:47 |
| kfox1111 | rax vms are different then most in having two nics. so the problem may be related to that. | 23:47 |
| sdake | kfox1111 yup rax is definately different - fungi complains about it constantly :) | 23:47 |
| *** tonanhngo has quit IRC | 23:48 | |
| sdake | sbezverk to get a core dump you ahve to enable them in the system | 23:48 |
| kfox1111 | yeah. not claiming rax is blameless ither. just wanting to know for sure if its a rax issue or not before appliying blame. :) | 23:48 |
| sdake | kfox1111 seems like it from my 20 or so patch revs | 23:49 |
| kfox1111 | 9/10 its been our code assuming things about infra that isn't true at rax. | 23:49 |
| sbezverk | sdake: we are not talking about core dump | 23:49 |
| sdake | kfox1111 that may be | 23:49 |
| sbezverk | sdake: go can print the stack with panic call.. | 23:49 |
| sdake | sbezverk yup | 23:50 |
| sbezverk | sdake: yup what? | 23:50 |
| sdake | i was agreeing with you | 23:50 |
| sdake | go can print a stacktrace with a panic call | 23:50 |
| spsurya | morning all | 23:51 |
| sbezverk | sdake: how then core dump comment fit into this ;) ?? | 23:51 |
| sdake | sbezverk i thought you guys wer etyring to generate a core dump to debug | 23:51 |
| sdake | check out this gate run: | 23:51 |
| sdake | https://review.openstack.org/#/c/455840/ | 23:51 |
| sdake | ironic = passing + revert applied | 23:51 |
| *** swinn has quit IRC | 23:52 | |
| sdake | another revert - completely green gate: https://review.openstack.org/#/c/455841/ | 23:53 |
| sdake | helm version has no bearing on whether ironic gate works or doesn't work | 23:53 |
| *** japestinho has joined #openstack-kolla | 23:54 | |
| *** jtriley has quit IRC | 23:54 | |
| sdake | kfox1111 all the optimizations you did in the helm 2.3.0 patch have no bearing on computekit: https://review.openstack.org/#/c/455842/1 | 23:54 |
| sdake | 2.3.0 is straight up broken | 23:54 |
| sdake | with compute kit | 23:54 |
| *** srwilkers has joined #openstack-kolla | 23:55 | |
| *** mattmceuen has quit IRC | 23:56 | |
| *** lamt has joined #openstack-kolla | 23:56 | |
| sbezverk | last PS looks good | 23:57 |
| sbezverk | sdake: how do you know computekit is ok? | 23:57 |
| sdake | sbezverk you mean this one? https://review.openstack.org/#/c/455842/1 | 23:58 |
| sbezverk | I do not see experimental | 23:58 |
| sbezverk | sdake: https://review.openstack.org/#/c/455841/1 | 23:58 |
| sdake | just ran check experimetnal on it | 23:58 |
| *** daidv has joined #openstack-kolla | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
