| kfox1111 | arg... | 00:00 |
|---|---|---|
| kfox1111 | so close.... | 00:00 |
| sean-k-mooney | ill leave it run over night i guess. catch everyone tomorow | 00:00 |
| sbezverk_ | kfox1111: by some reason multinode failed again, but all single nodes worked.. | 00:00 |
| kfox1111 | hmm... the pv was still created with the docker ip: | 00:01 |
| kfox1111 | http://logs.openstack.org/22/405422/45/check/gate-kolla-kubernetes-deploy-centos-binary-ceph-multi-nv/e51f217/console.html#_2016-12-06_23_47_02_919505 | 00:01 |
| sbezverk_ | kfox1111: yep | 00:01 |
| *** vilobhmm11 has quit IRC | 00:01 | |
| Pavo | evening | 00:01 |
| *** saneax is now known as saneax-_-|AFK | 00:01 | |
| *** yuanying_ has joined #openstack-kolla | 00:03 | |
| Pavo | so.... having an issue with setting in the inventory file on kolla-ansible | 00:03 |
| kfox1111 | is gerrit having issues for anyone else? | 00:04 |
| Pavo | it says in the file that you can, # When compute nodes and control nodes use different interfaces, you can specify "api_interface" and another interfaces like below: compute01 neutron_external_interface=eth0 api_interface=em1 storage_interface=em1 tunnel_interface=em1 | 00:04 |
| kfox1111 | hmm... finally went through... | 00:05 |
| sbezverk_ | kfox1111: yep it was unresponsive for few minutes | 00:05 |
| kfox1111 | oh. I see the problem. | 00:05 |
| *** yuanying has quit IRC | 00:06 | |
| Pavo | so I tried to put under [control] section controllernode neutron_external_interface=eth1 to use eth1 as the interface for neutron but it didn't work | 00:06 |
| Pavo | it used the neutron_external_interface= setting that is set in globals | 00:06 |
| Pavo | anyone have any suggestions? | 00:06 |
| kfox1111 | sbezverk_: commented inline | 00:06 |
| srwilkers_ | sayantan_, internets being extremely dodgy at work. working on that doc review but might have to postpone finishing it if things dont clear up. :) | 00:07 |
| *** yuanying_ has quit IRC | 00:07 | |
| Pavo | or is that setting only for compute nodes and not controllernode? | 00:08 |
| sayantan_ | srwilkers_ No rush :) | 00:08 |
| *** elynn has joined #openstack-kolla | 00:09 | |
| *** lamt has quit IRC | 00:14 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP Hel-ming Rabbitmq https://review.openstack.org/405422 | 00:14 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Terms https://review.openstack.org/407771 | 00:15 |
| sbezverk_ | kfox1111: if you need to override port then I need to store monitor list in X.X.X.X:6789 format, or you want a separate variable for port? | 00:17 |
| kfox1111 | each mon could potentially have a different port.... | 00:18 |
| kfox1111 | what about the contains function? | 00:18 |
| kfox1111 | something like: {{ if contains ":" . }} {{ . | quote}} {{ else }} {{ printf "%s:6789" . | quote }} | 00:20 |
| kfox1111 | sbezverk_: whats a helm template? | 00:21 |
| *** elynn has quit IRC | 00:21 | |
| sbezverk_ | kfox1111: do we really need to go that deep ?? | 00:21 |
| *** vilobhmm11 has joined #openstack-kolla | 00:21 | |
| openstackgerrit | sayantani proposed openstack/kolla: Minor documentation fixes https://review.openstack.org/406534 | 00:21 |
| kfox1111 | what do you meen deep? its just a single if/else statement. | 00:21 |
| kfox1111 | are you asking if we should support cephs's that have non standard mon ports? | 00:23 |
| sbezverk_ | kfox1111: right I meant, usually people try to follow a specific format X.X.X.X or X.X.X.X:PPPP | 00:23 |
| *** yingjun has joined #openstack-kolla | 00:23 | |
| sbezverk_ | kfox1111: mixing these might be confusing | 00:23 |
| kfox1111 | sbezverk_: yeah. all that if does is tack on the default port if you don't specify one. | 00:23 |
| kfox1111 | so you can use either format. | 00:23 |
| kfox1111 | though, I think if ou don't specify one, ceph itself might just use the default port. | 00:24 |
| kfox1111 | so doing nothing but {{ . | quote }} might work in all cases I guess. | 00:24 |
| *** yingjun_ has joined #openstack-kolla | 00:25 | |
| sbezverk_ | kfox1111: hm never tried but to be on the safe side I will add if | 00:25 |
| kfox1111 | k. thx. | 00:25 |
| *** jtriley has joined #openstack-kolla | 00:28 | |
| *** yingjun has quit IRC | 00:29 | |
| *** Pavo has quit IRC | 00:30 | |
| *** Pavo has joined #openstack-kolla | 00:30 | |
| *** elynn has joined #openstack-kolla | 00:31 | |
| kfox1111 | sbezverk_: how about this: | 00:32 |
| kfox1111 | * Template - Source code that, when processed by Helm along with Values | 00:32 |
| kfox1111 | passed during instantiaton, creates the final Kubernetes object. | 00:32 |
| kfox1111 | It is provided as part of a Chart and compiled into a Package. | 00:32 |
| sbezverk_ | kfox1111: looks really good | 00:34 |
| kfox1111 | cool. | 00:35 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Terms https://review.openstack.org/407771 | 00:35 |
| sbezverk_ | kfox1111: where did you find "contains" function in helm? I could not find any references to its existence.. | 00:36 |
| sbezverk_ | kfox1111: I am not syain it does not exist, I just hope that place you used has info for other interesting functions | 00:36 |
| kfox1111 | bmace: I keep going back to: https://github.com/kubernetes/helm/blob/master/docs/charts.md | 00:37 |
| kfox1111 | ther's a section: | 00:37 |
| kfox1111 | "Helm Chart templates are written in the Go template language, with the addition of 50 or so add-on template functions from the Sprig library and a few other specialized functions." | 00:37 |
| kfox1111 | with a link here: https://github.com/Masterminds/sprig | 00:37 |
| kfox1111 | most of the goodies are in there. :) | 00:37 |
| sbezverk_ | kfox1111: it worked finally.. SO I am goint to ack you ps then rebase and final clean up for rabbitmq | 00:38 |
| bmace | i assume the bmace: part of that message was in error kfox1111? :) | 00:38 |
| kfox1111 | bmace: yeah. sorry. bad tab on paste. | 00:39 |
| bmace | kk, np... :) | 00:39 |
| srwilkers_ | sayantan_, going to find stable internet. ran through most of it and posted a few comments. will probably add more once i can finish the work through | 00:39 |
| srwilkers_ | cheers :) | 00:39 |
| kfox1111 | sbezverk_: k. :) | 00:39 |
| kfox1111 | sbezverk_: then I'll go do one final review. everything looked pretty good when I skimmed it last, but havent really looked closely yet. | 00:40 |
| sayantan_ | Thank you srwilkers_ . I'll take a look at the comments. appreciate it :) | 00:40 |
| sbezverk_ | kfox1111: great I would like to close rabbitmq tonight | 00:41 |
| kfox1111 | sbezverk_: +1. | 00:41 |
| openstackgerrit | Merged openstack/kolla-kubernetes: Add ceph values for gate https://review.openstack.org/407672 | 00:44 |
| *** srwilkers_ has quit IRC | 00:44 | |
| *** sdake has joined #openstack-kolla | 00:45 | |
| jascott1 | so element_name it is. wfm | 00:45 |
| *** jascott1 has quit IRC | 00:48 | |
| *** tovin07 has joined #openstack-kolla | 00:52 | |
| openstackgerrit | zhangshuai proposed openstack/kolla-ansible: Add karbor ansible role https://review.openstack.org/402229 | 00:54 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP Hel-ming Rabbitmq https://review.openstack.org/405422 | 00:54 |
| openstackgerrit | Li Yingjun proposed openstack/kolla-kubernetes: Fix different repo issue for documentation https://review.openstack.org/405054 | 00:55 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Hel-ming Rabbitmq https://review.openstack.org/405422 | 00:56 |
| sbezverk_ | kfox1111: jascott1: I changed rabbitmq ps for element_name and som other clean up.. please review when you have time | 00:56 |
| kfox1111 | k. will do that asap. | 00:56 |
| *** jascott1 has joined #openstack-kolla | 00:56 | |
| kfox1111 | Just finising off another revision of terms. | 00:56 |
| duonghq | any core around? | 01:01 |
| duonghq | (kolla docker core) | 01:01 |
| *** zhubingbing has joined #openstack-kolla | 01:01 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Terms https://review.openstack.org/407771 | 01:02 |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: WIP Keystone Helm Packaging https://review.openstack.org/407780 | 01:03 |
| *** yuanying has joined #openstack-kolla | 01:04 | |
| *** dave-mccowan has joined #openstack-kolla | 01:04 | |
| *** tonanhngo has quit IRC | 01:04 | |
| *** tonanhngo has joined #openstack-kolla | 01:05 | |
| *** Jeffrey4l has quit IRC | 01:06 | |
| *** Jeffrey4l has joined #openstack-kolla | 01:06 | |
| *** vilobhmm11 has quit IRC | 01:06 | |
| kfox1111 | sbezverk_: I had previously started reviewion ps 46, so I finished that review. then I'll go through the newest. | 01:07 |
| *** Jeffrey4l has quit IRC | 01:08 | |
| *** Jeffrey4l has joined #openstack-kolla | 01:09 | |
| *** tonanhngo has quit IRC | 01:09 | |
| *** portdirect_ has joined #openstack-kolla | 01:10 | |
| *** lamt has joined #openstack-kolla | 01:11 | |
| *** tonanhngo has joined #openstack-kolla | 01:13 | |
| *** sdake has quit IRC | 01:14 | |
| *** zhurong has joined #openstack-kolla | 01:14 | |
| openstackgerrit | Duong Ha-Quang proposed openstack/kolla-ansible: Implement Keystone zero-downtime upgrade https://review.openstack.org/405215 | 01:16 |
| *** cu5 has quit IRC | 01:16 | |
| *** tonanhngo has quit IRC | 01:18 | |
| *** Jeffrey4l has quit IRC | 01:20 | |
| *** Jeffrey4l has joined #openstack-kolla | 01:21 | |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Pull kube-system images as part of CNI setup https://review.openstack.org/407788 | 01:23 |
| *** zhurong has quit IRC | 01:25 | |
| *** zhurong has joined #openstack-kolla | 01:26 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Hel-ming Rabbitmq https://review.openstack.org/405422 | 01:31 |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: WIP Keystone Helm Packaging https://review.openstack.org/407780 | 01:32 |
| *** david-lyle has joined #openstack-kolla | 01:33 | |
| sbezverk_ | kfox1111: I think I addressed most of your comments, the only one I am not so sure is rabbit cluster cookie. I think there should be one in the config file as if it only exist as a cli parameter, it will get lost.. | 01:33 |
| kfox1111 | yeah. but if its set, and a user forgets, it can cause two elements to join together in the same cluster unexpectedly. :/ | 01:34 |
| kfox1111 | so, user launches nova-rabbit, and neutron-rabbit and doesn't override the cookie on one of them, they may end up in a cluster together. | 01:34 |
| *** zhurong has quit IRC | 01:34 | |
| kfox1111 | is it valid to put it in the values as: | 01:35 |
| *** zhurong has joined #openstack-kolla | 01:35 | |
| kfox1111 | rabbit_cookie: | 01:35 |
| kfox1111 | with no value? | 01:35 |
| kfox1111 | it should end up not being assigned a value then, but is still discoverable. | 01:35 |
| sbezverk_ | kfox1111: I do not know, both approaches causes issues.. the only reason I kind of for putting it there, is to be in line with kolla | 01:36 |
| kfox1111 | do they default it to something? | 01:37 |
| kfox1111 | (they don't support multi instance rabbit though, so we're already doing something different) | 01:37 |
| sbezverk_ | yep | 01:37 |
| *** Jeffrey4l has quit IRC | 01:39 | |
| *** Jeffrey4l has joined #openstack-kolla | 01:39 | |
| sbezverk_ | kfox1111: ok I will remove value and add rabbit_cluster_cookie to the command line, but if people complain I will tell them it is your fault ;-) | 01:40 |
| kfox1111 | sbezverk_: heh. I'm sure they will complain about manually orchestrating microservices anway. thats a layer the majority of the folks never want to see anyway. | 01:41 |
| kfox1111 | but thats what operators/service packages are for. to hide that stuff. | 01:41 |
| kfox1111 | k. gota head out. | 01:43 |
| kfox1111 | will try and check back later. | 01:43 |
| *** wwriverrat has quit IRC | 01:44 | |
| *** prameswar has joined #openstack-kolla | 01:44 | |
| openstackgerrit | prameswar proposed openstack/kolla: QSG updated for shared volumes https://review.openstack.org/407061 | 01:47 |
| sbezverk_ | kfox1111: ttyl | 01:55 |
| *** mdnadeem has joined #openstack-kolla | 01:57 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Hel-ming Rabbitmq https://review.openstack.org/405422 | 02:01 |
| *** sdake has joined #openstack-kolla | 02:01 | |
| *** Jeffrey4l has quit IRC | 02:06 | |
| *** Jeffrey4l has joined #openstack-kolla | 02:07 | |
| sdake | yo peeps | 02:08 |
| portdirect_ | yo sdake | 02:08 |
| duonghq | hey sdake | 02:08 |
| sdake | sup portdirect_ | 02:08 |
| sdake | sup duonghq | 02:08 |
| duonghq | I just fly over our code base, seem that we need some kind of bp which purpose is apply upgrade mechanism from OpenStack service in Kolla | 02:09 |
| duonghq | we lack at least neutron and Keystone | 02:10 |
| duonghq | bp or bugs | 02:10 |
| duonghq | I don't sure | 02:10 |
| sdake | duonghq do you mean kolla-kubernetes? | 02:11 |
| duonghq | sdake, no, just kolla-ansible and dockerfile | 02:12 |
| *** sp_ has quit IRC | 02:12 | |
| sdake | kolla-ansible has upgrades | 02:12 |
| sdake | do you mean some upgrades are not implemented? | 02:12 |
| duonghq | sdake, moment, I'm finding the neutron upgrade guide | 02:13 |
| duonghq | sdake, according to this link http://docs.openstack.org/mitaka/networking-guide/migration-database.html | 02:13 |
| duonghq | the neutron db migration contains $ neutron-db-manage upgrade --expand and --contract | 02:14 |
| sdake | if you think action is needed file a blueprint :) | 02:16 |
| sdake | your part of the drivers group right? | 02:16 |
| duonghq | right, but I don't sure it's worth a bp or just a bugs | 02:17 |
| duonghq | I filed a bug about Keystone case | 02:17 |
| sdake | dunno - bug triage goes poorly in our project unfortunately | 02:17 |
| duonghq | but seem that it's affect other services too | 02:17 |
| sdake | approved blueprints get more attention | 02:17 |
| duonghq | roger | 02:18 |
| duonghq | sdake, but the bp don't have "also affect" like bugs? | 02:19 |
| sdake | is srwilkers about? | 02:19 |
| sdake | duonghq really up to you - if your going to work on it personally you can use bugs or blueprints - whatever your pleasure | 02:20 |
| Pavo | evening sdake | 02:20 |
| sdake | if not, blueprints again get mre attention from the dev team :) | 02:20 |
| sdake | sup pavo | 02:20 |
| sdake | portdirect_ are you on the kolla-kubernetes-drivers team? | 02:20 |
| Pavo | had an issue with kolla-ansible and inventory file today | 02:20 |
| Pavo | it says in the file that you can, # When compute nodes and control nodes use different interfaces, you can specify "api_interface" and another interfaces like below: compute01 neutron_external_interface=eth0 api_interface=em1 storage_interface=em1 tunnel_interface=em1 | 02:21 |
| Pavo | so I tried to put under [control] section controllernode neutron_external_interface=eth1 to use eth1 as the interface for neutron but it didn't work | 02:21 |
| Pavo | it used the neutron_external_interface= setting that is set in globals | 02:21 |
| Pavo | or is that setting only for compute nodes and not controllernode? | 02:21 |
| portdirect_ | sdake: nope | 02:23 |
| openstackgerrit | qiankun.zhang proposed openstack/kolla: Install tacker-horizon into horizon image Closes-Bug: #1647894 https://review.openstack.org/407810 | 02:24 |
| openstack | bug 1647894 in kolla "tacker-horizon is not loaded when enabled tacker" [Undecided,New] https://launchpad.net/bugs/1647894 - Assigned to qiankun.zhang (qiankun.zhang) | 02:24 |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Keystone Helm Packaging https://review.openstack.org/407780 | 02:26 |
| sbezverk_ | kfox1111: when you back please check I hope final version. the gate is all green.. | 02:27 |
| Pavo | any ideas sdake? | 02:30 |
| *** Pavo has quit IRC | 02:30 | |
| *** Pavo has joined #openstack-kolla | 02:30 | |
| duonghq | sdake, can you review the bp: https://blueprints.launchpad.net/kolla/+spec/apply-service-upgrade-procedure | 02:34 |
| sdake | i was afk, reading scrollback | 02:34 |
| portdirect_ | sbezverk_: nice work with rabbit - its looking good dude. I've got a ps that should fix one of the intermitant gate issues you've seen if you have a chance to look at it https://review.openstack.org/#/c/407788/ | 02:34 |
| sdake | portdirect_ happy to invite you - its not a voting thing or any of that | 02:34 |
| sdake | portdirect_ anyone is pretty much welcome in the drivers group that shows interest in a particular deliverable | 02:35 |
| sdake | in terms of reviews or code | 02:35 |
| sdake | portdirect_ drivers group allows you to admin launchpad | 02:35 |
| sdake | portdirect_ triage bugs, move around blueprints | 02:35 |
| sdake | blueprint approval generally reserved for the core reviewer team | 02:35 |
| portdirect_ | sdake: that would be cool - up for getting involved wherever I can be useful :) | 02:35 |
| sdake | cool whats your launchpad id | 02:36 |
| sdake | pavo will be with you in a moment :) | 02:36 |
| sdake | duonghq same :) | 02:36 |
| portdirect_ | can I get a digest/summary of bugs emailed to me? | 02:36 |
| sdake | portdirect_ i'm sure there is a way to do that - but not certain | 02:36 |
| sdake | portdirect_ drivers team is self-admining as well, you can always disable yourself later if you like | 02:37 |
| portdirect_ | sdake: my id is portdirect on launchpad | 02:37 |
| duonghq | portdirect_, you want to subscribe to the bug update? | 02:37 |
| sdake | cool moment | 02:37 |
| sdake | i dont like to guess on that :) | 02:37 |
| portdirect_ | duonghq: that would be great | 02:38 |
| *** hogepodge has quit IRC | 02:38 | |
| *** jascott1 has quit IRC | 02:39 | |
| duonghq | portdirect_, https://bugs.launchpad.net/kolla -> in the right column, "Subscribe to bug mail" link, also for kolla-ansible and kolla-kubernetes if you like | 02:39 |
| sdake | ok apvo - let me look at blueprint real quick - i read your report - and don't hvae any ideas off top of my head but it will take me more time to process | 02:39 |
| sdake | pavo ^ | 02:40 |
| Pavo | ok no problem | 02:40 |
| portdirect_ | duonghq: thanks man :) | 02:40 |
| duonghq | sdake, thanks | 02:41 |
| sdake | pavo no idea really | 02:43 |
| sdake | pavo i think what would be cool is to ask that quesetion on ask.openstack.org | 02:43 |
| sdake | pavo and i'll make sure it gets answered there | 02:44 |
| sdake | that way we can build up a record of Q&A | 02:44 |
| Pavo | only thing I could think of is that that setting is only for compute nodes | 02:44 |
| Pavo | ok will do | 02:44 |
| sdake | ok need to play IT professional with my parents computers | 02:44 |
| * sdake groan | 02:44 | |
| sdake | s | 02:44 |
| sdake | bbl | 02:44 |
| *** elynn has quit IRC | 02:46 | |
| *** sdake has quit IRC | 02:47 | |
| *** elynn has joined #openstack-kolla | 02:48 | |
| openstackgerrit | Zeyu Zhu proposed openstack/kolla-ansible: Optimize the code https://review.openstack.org/407815 | 02:49 |
| *** hfu has joined #openstack-kolla | 02:50 | |
| *** schwicht has quit IRC | 02:55 | |
| *** sp_ has joined #openstack-kolla | 02:55 | |
| *** yuanying has quit IRC | 02:58 | |
| *** sayantan_ has quit IRC | 02:59 | |
| *** kristian__ has joined #openstack-kolla | 02:59 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: TEST NEW GATE_RUN.SH SCRIPT https://review.openstack.org/407819 | 02:59 |
| openstackgerrit | qiankun.zhang proposed openstack/kolla-ansible: Enable tacker-horizon when enable_tacker is yes https://review.openstack.org/407822 | 03:02 |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Keystone Helm Packaging https://review.openstack.org/407780 | 03:03 |
| *** kristian__ has quit IRC | 03:03 | |
| *** mdnadeem has quit IRC | 03:04 | |
| *** fragatina has quit IRC | 03:06 | |
| *** hfu has quit IRC | 03:07 | |
| *** hfu has joined #openstack-kolla | 03:07 | |
| *** sayantan_ has joined #openstack-kolla | 03:07 | |
| *** srwilkers_ has joined #openstack-kolla | 03:08 | |
| *** v1k0d3n has joined #openstack-kolla | 03:09 | |
| *** tonanhngo has joined #openstack-kolla | 03:10 | |
| *** hfu has quit IRC | 03:19 | |
| *** hfu has joined #openstack-kolla | 03:20 | |
| *** Jeffrey4l has quit IRC | 03:23 | |
| *** Jeffrey4l has joined #openstack-kolla | 03:24 | |
| *** hfu has quit IRC | 03:24 | |
| *** dave-mccowan has quit IRC | 03:25 | |
| *** hfu has joined #openstack-kolla | 03:26 | |
| *** dave-mccowan has joined #openstack-kolla | 03:28 | |
| openstackgerrit | qiankun.zhang proposed openstack/kolla: Install tacker-horizon into horizon image Closes-Bug: #1647894 https://review.openstack.org/407810 | 03:28 |
| openstack | bug 1647894 in kolla "tacker-horizon is not loaded when enabled tacker" [Undecided,In progress] https://launchpad.net/bugs/1647894 - Assigned to qiankun.zhang (qiankun.zhang) | 03:28 |
| openstackgerrit | qiankun.zhang proposed openstack/kolla: Install tacker-horizon into horizon image https://review.openstack.org/407810 | 03:29 |
| *** hfu has quit IRC | 03:37 | |
| *** hfu has joined #openstack-kolla | 03:38 | |
| *** srwilkers_ has quit IRC | 03:38 | |
| *** Jeffrey4l has quit IRC | 03:38 | |
| *** Jeffrey4l has joined #openstack-kolla | 03:38 | |
| openstackgerrit | caoyuan proposed openstack/kolla: Install mistral dashboard into horizon image https://review.openstack.org/406726 | 03:41 |
| *** yuanying has joined #openstack-kolla | 03:43 | |
| *** hfu has quit IRC | 03:43 | |
| *** klindgren_ has quit IRC | 03:43 | |
| *** klindgren_ has joined #openstack-kolla | 03:44 | |
| *** portdirect_ is now known as portdirect_away | 03:45 | |
| *** hfu has joined #openstack-kolla | 03:47 | |
| *** yuanying has quit IRC | 03:48 | |
| *** yuanying has joined #openstack-kolla | 03:49 | |
| *** yuanying has quit IRC | 03:53 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible: Optimized reconfigure action for glance https://review.openstack.org/406978 | 03:56 |
| *** hfu has quit IRC | 03:57 | |
| *** hfu has joined #openstack-kolla | 04:00 | |
| *** hfu has quit IRC | 04:01 | |
| *** sp_ has quit IRC | 04:01 | |
| openstackgerrit | Duong Ha-Quang proposed openstack/kolla: Apply neutron database migration https://review.openstack.org/407840 | 04:03 |
| *** dave-mccowan has quit IRC | 04:06 | |
| *** Jeffrey4l has quit IRC | 04:11 | |
| *** sayantan_ has quit IRC | 04:11 | |
| *** Jeffrey4l has joined #openstack-kolla | 04:11 | |
| *** sp_ has joined #openstack-kolla | 04:13 | |
| *** v1k0d3n has quit IRC | 04:14 | |
| *** mdnadeem has joined #openstack-kolla | 04:18 | |
| *** prameswar has quit IRC | 04:24 | |
| openstackgerrit | caoyuan proposed openstack/kolla: Install mistral dashboard into horizon image https://review.openstack.org/406726 | 04:26 |
| *** zhurong has quit IRC | 04:27 | |
| *** Jeffrey4l has quit IRC | 04:29 | |
| *** Jeffrey4l has joined #openstack-kolla | 04:30 | |
| *** Pavo has quit IRC | 04:30 | |
| *** Jeffrey4l has quit IRC | 04:30 | |
| *** Jeffrey4l has joined #openstack-kolla | 04:31 | |
| *** Pavo has joined #openstack-kolla | 04:34 | |
| *** huhaoran has joined #openstack-kolla | 04:36 | |
| *** yuanying has joined #openstack-kolla | 04:40 | |
| *** huhaoran has quit IRC | 04:41 | |
| *** vilobhmm11 has joined #openstack-kolla | 04:47 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible: Optimized reconfigure action for nova https://review.openstack.org/407416 | 04:47 |
| *** JRobinson__ is now known as JRobinson__afk | 04:48 | |
| openstackgerrit | Duong Ha-Quang proposed openstack/kolla: [WIP] Apply neutron database migration https://review.openstack.org/407840 | 04:53 |
| *** hfu has joined #openstack-kolla | 04:59 | |
| *** sdake has joined #openstack-kolla | 05:01 | |
| *** sdake has quit IRC | 05:01 | |
| *** sdake has joined #openstack-kolla | 05:01 | |
| *** pbourke has quit IRC | 05:01 | |
| sdake | yo | 05:01 |
| *** pbourke has joined #openstack-kolla | 05:01 | |
| *** elynn has quit IRC | 05:03 | |
| *** sdake_ has joined #openstack-kolla | 05:04 | |
| duonghq | sup sdake | 05:04 |
| duonghq | just finished your dinner? | 05:04 |
| *** hfu has quit IRC | 05:05 | |
| *** sdake has quit IRC | 05:07 | |
| *** yuanying has quit IRC | 05:13 | |
| *** yuanying has joined #openstack-kolla | 05:13 | |
| *** vilobhmm11 has quit IRC | 05:14 | |
| *** elemoine__ has joined #openstack-kolla | 05:14 | |
| *** akwasnie_ has joined #openstack-kolla | 05:14 | |
| *** hyakuhei has quit IRC | 05:14 | |
| *** zhenguo_ has joined #openstack-kolla | 05:14 | |
| *** akwasnie has quit IRC | 05:14 | |
| *** sindhu_ has joined #openstack-kolla | 05:14 | |
| *** akwasnie_ is now known as akwasnie | 05:14 | |
| *** zhenguo has quit IRC | 05:15 | |
| *** elemoine_ has quit IRC | 05:15 | |
| *** sindhu has quit IRC | 05:15 | |
| *** harmw has quit IRC | 05:15 | |
| *** macsz_ has joined #openstack-kolla | 05:15 | |
| *** DuncanT_ has joined #openstack-kolla | 05:15 | |
| *** elemoine__ is now known as elemoine_ | 05:15 | |
| *** zhenguo_ is now known as zhenguo | 05:15 | |
| *** sindhu_ is now known as sindhu | 05:15 | |
| *** DuncanT has quit IRC | 05:15 | |
| *** DuncanT_ is now known as DuncanT | 05:16 | |
| *** krtaylor has quit IRC | 05:16 | |
| *** saneax-_-|AFK has quit IRC | 05:16 | |
| *** macsz has quit IRC | 05:16 | |
| *** harmw has joined #openstack-kolla | 05:17 | |
| openstackgerrit | qiankun.zhang proposed openstack/kolla: Install tacker-horizon into horizon image https://review.openstack.org/407810 | 05:17 |
| *** macsz_ is now known as macsz | 05:17 | |
| *** fragatina has joined #openstack-kolla | 05:17 | |
| *** coolsvap has joined #openstack-kolla | 05:19 | |
| *** fragatina has quit IRC | 05:26 | |
| openstackgerrit | caoyuan proposed openstack/kolla: Update "searchlight-1.0.0.tar.gz" to "searchlight-master.tar.gz" https://review.openstack.org/407851 | 05:27 |
| *** fragatina has joined #openstack-kolla | 05:27 | |
| *** zhurong has joined #openstack-kolla | 05:27 | |
| *** saneax-_-|AFK has joined #openstack-kolla | 05:27 | |
| *** krtaylor has joined #openstack-kolla | 05:28 | |
| *** sayantan_ has joined #openstack-kolla | 05:28 | |
| *** zhurong has quit IRC | 05:30 | |
| *** zhurong has joined #openstack-kolla | 05:31 | |
| *** hyakuhei has joined #openstack-kolla | 05:31 | |
| *** Jeffrey4l has quit IRC | 05:35 | |
| *** Jeffrey4l has joined #openstack-kolla | 05:35 | |
| zhubingbing | hi jeffrey41 | 05:36 |
| *** JRobinson__afk is now known as JRobinson__ | 05:37 | |
| sdake_ | sup Jeffrey4l | 05:39 |
| sdake_ | sup zhubingbing | 05:39 |
| Jeffrey4l | hey sdake_ | 05:39 |
| Jeffrey4l | zhubingbing, hi | 05:39 |
| zhubingbing | hey sdake_ | 05:39 |
| sdake_ | re the endless stream of questions we receive on the same topics | 05:39 |
| sdake_ | some of those can be handled with documentation | 05:39 |
| sdake_ | some can be handled with https://ask.openstack.org/en/questions/scope:all/sort:activity-desc/page:1/query:kolla/ | 05:40 |
| sp_ | hi sdake_ Jeffrey4l | 05:40 |
| sdake_ | Jeffrey4l mind taking that up at the next team meeting tomorrow? | 05:40 |
| sdake_ | sup sp_ | 05:40 |
| Jeffrey4l | np. | 05:41 |
| Jeffrey4l | sdake_, optimize glance is almost done could u take a look if u have some time :) https://review.openstack.org/406978 | 05:43 |
| sdake_ | Jeffrey4l sure i'll have a look - don't expect greatness from me - been up since 5am as usual and its 11pm | 05:43 |
| sdake_ | Jeffrey4l can you set the blueprint to the appropriate state for better reconfig? | 05:46 |
| Jeffrey4l | you are always an early bird ;) | 05:46 |
| sdake_ | and night owl | 05:46 |
| sdake_ | bad combo :) | 05:46 |
| Jeffrey4l | done | 05:47 |
| *** absubram has joined #openstack-kolla | 05:49 | |
| sdake_ | done what Jeffrey4l ? | 05:50 |
| Jeffrey4l | made the better reconfig bp to appropriate state | 05:50 |
| sdake_ | oh right | 05:52 |
| sdake_ | Jeffrey4l re your review, it lgtm | 05:52 |
| sdake_ | the best I can really offer is a +1 at this point | 05:52 |
| Jeffrey4l | roger. thanks. | 05:53 |
| *** hfu has joined #openstack-kolla | 05:53 | |
| sdake_ | Jeffrey4l there are some new features being used like notify - which i'm unfamiliar with | 05:53 |
| sdake_ | i could guess - but obviously don't want to break reconfigure | 05:54 |
| Jeffrey4l | notify works like: when this task is changed, run another task defined in handlers/main.yml file. | 05:54 |
| Jeffrey4l | it won't break reconfigure. but after this is done, we do not need reconfigure anymore. | 05:54 |
| Jeffrey4l | because reconfigure works like deploy. | 05:54 |
| sdake_ | so reconfigure can be deprecated? | 05:55 |
| Jeffrey4l | yep. | 05:55 |
| sdake_ | alot of people use reconfigure expecting globals.yml modifications to be applied | 05:55 |
| Jeffrey4l | i'd like to mark it deprecated when this is done. and remove it in the future. | 05:55 |
| sdake_ | i dont even know wher eto begin on that prbolem | 05:55 |
| sdake_ | Jeffrey4l cool - there is a deprecation process for that | 05:56 |
| Jeffrey4l | after made some change, run deploy will work as what u think by using reconfigure. | 05:56 |
| sdake_ | when you get there i can guide ou through it | 05:56 |
| Jeffrey4l | thanks. | 05:56 |
| Jeffrey4l | after this changed is done, run deploy willl work as what u think by using reconfigure. | 05:56 |
| *** absubram has quit IRC | 05:57 | |
| Jeffrey4l | and now implementation is faster. and configure file will copy to the hosts it needed ( this is bug in current implementaion ) | 05:57 |
| *** hfu has quit IRC | 05:57 | |
| Jeffrey4l | so it will speed up a little during deploy ( not need copy so much configuration file ) | 05:57 |
| sdake_ | right | 05:58 |
| *** hfu has joined #openstack-kolla | 05:58 | |
| sdake_ | sounds pretty good to me ;) | 05:58 |
| Jeffrey4l | this is what i always expect and is optimal to me ;) | 05:58 |
| sdake_ | then deploy will take into account changes in globals.yml of existing deployment? | 05:59 |
| Jeffrey4l | yep. | 05:59 |
| sdake_ | perfectly? | 05:59 |
| Jeffrey4l | when something is changed, it will restart container or recreate container. | 05:59 |
| Jeffrey4l | yes. | 05:59 |
| Jeffrey4l | perfect. | 05:59 |
| sdake_ | sweet | 05:59 |
| *** JRobinson__ is now known as JRobinson__afk | 06:00 | |
| *** kristian__ has joined #openstack-kolla | 06:00 | |
| Jeffrey4l | the number lines of the code will be decrease too. check the nova's change. https://review.openstack.org/407416 +308 -647 | 06:00 |
| *** kristian__ has quit IRC | 06:00 | |
| *** hfu has quit IRC | 06:01 | |
| sdake_ | nice! | 06:01 |
| *** kristian__ has joined #openstack-kolla | 06:01 | |
| sdake_ | well that sounds super optimal | 06:01 |
| sdake_ | and solves a whole slew of problems with how reconfigure works today | 06:01 |
| sdake_ | which is to say, not very well at all :) | 06:01 |
| Jeffrey4l | yep. | 06:02 |
| sdake_ | Jeffrey4l we typically dont use the review stte in launchpad | 06:03 |
| sdake_ | Jeffrey4l we use discussion and approved | 06:03 |
| sdake_ | discussion = thinking about it/backlog | 06:03 |
| Jeffrey4l | sdake_, got. i will change it. | 06:03 |
| sdake_ | approved = lets get moving :) | 06:03 |
| Jeffrey4l | will mark it as approved. | 06:04 |
| sdake_ | cool thanks | 06:04 |
| sdake_ | my karma on launchpad is too high ;) | 06:04 |
| sdake_ | need better balance among core team on karma launchpad | 06:04 |
| *** kristian__ has quit IRC | 06:05 | |
| sdake_ | https://launchpad.net/kolla-ansible/+topcontributors | 06:06 |
| sdake_ | possibly another topic for tomorrow's meeting | 06:06 |
| sdake_ | unless you think agenda is full | 06:06 |
| sdake_ | Jeffrey4l say, can you modify the agenda for those two things - ask.openstack.org usage and balancing core team karma in launchpad? | 06:16 |
| Jeffrey4l | what balancing karma means. let core team do more work in launchpad? | 06:17 |
| sdake_ | right | 06:17 |
| Jeffrey4l | roger. np | 06:17 |
| sdake_ | the cor eteam can already do more work in launchpad | 06:17 |
| sdake_ | it just isn't happening :) | 06:17 |
| Jeffrey4l | yep | 06:18 |
| sdake_ | it may be a training issue | 06:18 |
| sdake_ | worth discussing | 06:18 |
| sdake_ | if its a training issue that is easily solvable | 06:18 |
| sdake_ | if its a "dont want to do the job" issue that is harder to solve ;) | 06:18 |
| Jeffrey4l | someone may have no idea how to handle such thing properly. | 06:19 |
| sdake_ | if thats the case lets figure that out | 06:19 |
| sdake_ | only cores shoudl really handle blueprints imo | 06:19 |
| Jeffrey4l | i.e. need some training, how to use/ how to manage those. | 06:19 |
| sdake_ | but the bugs can be triaged by anyone | 06:19 |
| sdake_ | well lets figure out if its a training problem or a dont want to do the job problem | 06:20 |
| sdake_ | i'm honestly not sure which it is | 06:20 |
| Jeffrey4l | does kolla-drivers guys can do what he he to do. kolla-drivers != core team right now. | 06:20 |
| Jeffrey4l | ok. will talk in the meeing. | 06:20 |
| sdake_ | right - i like that kolla-drivers is an extended team | 06:20 |
| Jeffrey4l | meeting* | 06:20 |
| Jeffrey4l | it is now. | 06:20 |
| sdake_ | just that kolla-drivers really shouldn't adjust blueprint status | 06:20 |
| sdake_ | only core team should | 06:21 |
| sdake_ | most projects have core team == drivers team | 06:21 |
| sdake_ | we have no separate bug triage team | 06:21 |
| sdake_ | maybe we need one | 06:21 |
| sdake_ | not sure ;) | 06:21 |
| sdake_ | we have a bajilion bugs in our backlog | 06:21 |
| *** saneax-_-|AFK is now known as saneax | 06:24 | |
| Jeffrey4l | yes. | 06:24 |
| *** hfu has joined #openstack-kolla | 06:26 | |
| *** fragatina has quit IRC | 06:26 | |
| *** hfu has quit IRC | 06:27 | |
| *** jtriley has quit IRC | 06:29 | |
| *** Pavo has quit IRC | 06:30 | |
| *** huhaoran has joined #openstack-kolla | 06:30 | |
| *** hfu has joined #openstack-kolla | 06:33 | |
| *** Pavo has joined #openstack-kolla | 06:35 | |
| *** sdake_ has quit IRC | 06:36 | |
| *** sdake has joined #openstack-kolla | 06:36 | |
| openstackgerrit | Zeyu Zhu proposed openstack/kolla-ansible: Optimize the code https://review.openstack.org/407815 | 06:39 |
| *** jgriffith is now known as jgriffith_away | 06:39 | |
| *** hfu has quit IRC | 06:40 | |
| duonghq | zhubingbing, congratulation | 06:41 |
| zhubingbing | thanks | 06:42 |
| zhubingbing | thank all | 06:42 |
| *** huhaoran has quit IRC | 06:46 | |
| openstackgerrit | pangliye proposed openstack/kolla: Fix typo error fix typo error "libary" -> "library" https://review.openstack.org/407876 | 06:46 |
| *** sdake_ has joined #openstack-kolla | 06:47 | |
| *** kristian__ has joined #openstack-kolla | 06:48 | |
| *** huhaoran has joined #openstack-kolla | 06:49 | |
| *** hfu has joined #openstack-kolla | 06:49 | |
| *** sdake has quit IRC | 06:50 | |
| *** kristian__ has quit IRC | 06:51 | |
| *** kristian__ has joined #openstack-kolla | 06:51 | |
| *** jtriley has joined #openstack-kolla | 06:54 | |
| openstackgerrit | pangliye proposed openstack/kolla: Fix type error fix type error "managment" -> "management" https://review.openstack.org/407884 | 06:55 |
| *** kristian__ has quit IRC | 06:55 | |
| *** msimonin has joined #openstack-kolla | 06:55 | |
| *** msimonin has quit IRC | 06:56 | |
| *** jtriley has quit IRC | 07:01 | |
| *** zhurong has quit IRC | 07:01 | |
| *** zhurong has joined #openstack-kolla | 07:02 | |
| *** yuanying has quit IRC | 07:05 | |
| *** kristian__ has joined #openstack-kolla | 07:05 | |
| *** sdake_ has quit IRC | 07:08 | |
| openstackgerrit | qiankun.zhang proposed openstack/kolla: Install tacker-horizon into horizon image https://review.openstack.org/407810 | 07:09 |
| *** hfu has quit IRC | 07:09 | |
| *** hfu has joined #openstack-kolla | 07:10 | |
| *** jascott1 has joined #openstack-kolla | 07:13 | |
| *** yuanying has joined #openstack-kolla | 07:17 | |
| *** fragatina has joined #openstack-kolla | 07:18 | |
| *** msimonin has joined #openstack-kolla | 07:20 | |
| *** msimonin has quit IRC | 07:22 | |
| *** sayantan_ has quit IRC | 07:22 | |
| *** JRobinson__afk is now known as JRobinson__ | 07:23 | |
| openstackgerrit | Zeyu Zhu proposed openstack/kolla-ansible: Optimize the code https://review.openstack.org/407815 | 07:24 |
| *** tovin07_ has joined #openstack-kolla | 07:25 | |
| openstackgerrit | Zeyu Zhu proposed openstack/kolla-ansible: Optimize the code https://review.openstack.org/407815 | 07:30 |
| *** alteriks has joined #openstack-kolla | 07:32 | |
| *** vilobhmm11 has joined #openstack-kolla | 07:41 | |
| *** pcaruana has joined #openstack-kolla | 07:41 | |
| *** shardy has joined #openstack-kolla | 07:46 | |
| *** msimonin has joined #openstack-kolla | 07:48 | |
| *** msimonin has quit IRC | 07:49 | |
| jmccarthy | Any ansible gurus about ? :) In this snippet here - would the serial be for both iscsid and tgtd, or is only the actions done against the 'tgtd' nodes done serially ? | 07:49 |
| jmccarthy | (I mean when not using the default of 0) | 07:49 |
| jmccarthy | - hosts: | 07:49 |
| jmccarthy | - iscsid | 07:49 |
| jmccarthy | - tgtd | 07:49 |
| jmccarthy | serial: '{{ serial|default("0") }}' | 07:49 |
| jmccarthy | ... | 07:49 |
| *** hfu has quit IRC | 07:50 | |
| *** hfu has joined #openstack-kolla | 07:51 | |
| *** sdake has joined #openstack-kolla | 07:55 | |
| *** kristian__ has quit IRC | 07:56 | |
| *** fragatina has quit IRC | 07:58 | |
| *** kristian__ has joined #openstack-kolla | 07:59 | |
| *** zhurong has quit IRC | 08:00 | |
| jascott1 | jmccarthy id guess both | 08:01 |
| *** pbourke_ has joined #openstack-kolla | 08:01 | |
| *** zhurong has joined #openstack-kolla | 08:01 | |
| *** pbourke has quit IRC | 08:01 | |
| jmccarthy | Yea I thought both also, but I'm second guessing myself :) | 08:02 |
| *** tonanhngo has quit IRC | 08:02 | |
| jascott1 | do both hosts <serial value> at a time | 08:02 |
| sdake | Jeffrey4l when will you be tagging 1.0.0.0b2 of kollla-kubernetes? | 08:02 |
| sdake | or rather all of the deliverables related to kolla? | 08:03 |
| *** wirehead_ has quit IRC | 08:03 | |
| sdake | say grats zhubingbing well deserved! | 08:03 |
| zhubingbing | hi | 08:04 |
| Jeffrey4l | sdake, depends on yours ;) | 08:04 |
| Jeffrey4l | i asked ryan when i tag b1, | 08:04 |
| sdake | Jeffrey4l i dont understand what you mean | 08:04 |
| jmccarthy | Ok, but do iscsid 'serial' and then do tgtd 'serial' is it ? | 08:04 |
| sdake | we didn't tag a b1 because we weren't sure if we would have a 1.0.0 deliverable | 08:04 |
| Jeffrey4l | he told me kolla-kubernetes will independent until 1.0. | 08:05 |
| Jeffrey4l | so i didn't make b1 for kolla-k8s. | 08:05 |
| *** wirehead_ has joined #openstack-kolla | 08:05 | |
| sdake | even independent should follow the standard release cycle | 08:05 |
| Jeffrey4l | ok. no idea about this. | 08:05 |
| sdake | have to take my word for it :) | 08:05 |
| sdake | doug hellman had indicated in some email i saw somewhere in the past that independent projects should strive to follow the standard cycle | 08:06 |
| sdake | doug hellman is ptl of the release team | 08:06 |
| Jeffrey4l | so based on the release scheduler, it should be Dec 15 to make the tag. | 08:06 |
| sdake | #openstack-release | 08:06 |
| Jeffrey4l | ok. but i don't see any rc tag for kolla-kubernetes now. | 08:07 |
| Jeffrey4l | milestone* ( not rc ) | 08:07 |
| Jeffrey4l | anyway, i will make the milestone2 tag for k8s around Dec 15. | 08:07 |
| *** yuanying has quit IRC | 08:08 | |
| sdake | Jeffrey4l we need to see if the core team feels confident we can hit the schedules | 08:08 |
| sdake | we need to hit two of the schedules to be considered part of the integrated release | 08:09 |
| Jeffrey4l | OK. i will add this in the meeting agenda. | 08:09 |
| sdake | thanks! | 08:09 |
| *** gfidente has joined #openstack-kolla | 08:09 | |
| *** gfidente has joined #openstack-kolla | 08:09 | |
| Jeffrey4l | ;) thanks for remind this | 08:09 |
| sdake | kolla and kolla-ansible of course will hit the schedules - can always just shed features | 08:10 |
| Jeffrey4l | yes. | 08:10 |
| sdake | 7 days is pretty tight to hit the helm-microservices blueprint | 08:11 |
| sdake | if we can hit that, i feel good about 1.0.0.0.b2 | 08:11 |
| Jeffrey4l | what's the status for that? | 08:11 |
| Jeffrey4l | can we start the copy&paste work? | 08:11 |
| sdake | https://blueprints.launchpad.net/kolla-kubernetes/+spec/helm-microservices | 08:12 |
| sdake | its a little more then copy and paste | 08:12 |
| sdake | it requires some brain power | 08:12 |
| sdake | but ya, sort of copy and paste | 08:12 |
| sdake | need the team to make the call, i've been out of the loop working on other things for the last 3 days while people get things done ;) | 08:12 |
| Jeffrey4l | cool. i will take some time to the new stuff about kolla-k8s. | 08:13 |
| Jeffrey4l | there isn't much TODO work item in the list. seem pretty good. | 08:14 |
| sdake | ya - although alot of inprogress stuff hasn't hit the review queue as of yet | 08:14 |
| sdake | if we can hit that blueprint on the 15th, the rest is doable | 08:15 |
| sdake | that blueprint is the fundamental building block on which everything else is bsaed | 08:15 |
| Jeffrey4l | cool | 08:16 |
| *** Jeffrey4l has quit IRC | 08:17 | |
| *** Jeffrey4l has joined #openstack-kolla | 08:17 | |
| sdake | Jeffrey4l for some reason i had been operating under the idea the 20th was the tag deadline | 08:18 |
| *** hfu has quit IRC | 08:18 | |
| Jeffrey4l | roger. np/ | 08:18 |
| sdake | no idea why i thought that | 08:18 |
| sdake | but looking at the schedule - 15th is the date | 08:18 |
| *** hfu has joined #openstack-kolla | 08:20 | |
| Jeffrey4l | personally, i think tail-cycle project should have more longger time than other projects for milestone tag. | 08:20 |
| sdake | expectation of trailing cycle projects is milestone tags are on time | 08:21 |
| sdake | and the final tag can lag | 08:21 |
| sdake | we did a poor job (as an openstack community) of defining the rcs and whatnot for the rcs during the last cycle (when trailing cycle was introduced) | 08:21 |
| Jeffrey4l | rcs means? | 08:22 |
| sdake | release candidates | 08:22 |
| sdake | rc1 | 08:22 |
| sdake | rc2 | 08:22 |
| sdake | etc | 08:22 |
| Jeffrey4l | roger. | 08:22 |
| sdake | what the release team wants is one rc1 | 08:22 |
| Jeffrey4l | yep. | 08:22 |
| sdake | and then one rc2 with all the bugs fixed | 08:22 |
| openstackgerrit | Duong Ha-Quang proposed openstack/kolla-ansible: [WIP] Apply neutron database migration https://review.openstack.org/407922 | 08:22 |
| sdake | and then the rc2 is retagged as final | 08:23 |
| Jeffrey4l | trail project has lone rc cycle, but we can not add new feature, which is sad :( | 08:23 |
| sdake | idea is we branch at rc1 | 08:23 |
| sdake | and then new features can go into pike | 08:23 |
| openstackgerrit | Duong Ha-Quang proposed openstack/kolla: [WIP] Apply neutron database migration https://review.openstack.org/407840 | 08:23 |
| Jeffrey4l | actually, between rc1 and finally release, no one works on new feature. | 08:24 |
| sdake | ya - thats because we have a bajillion bugs to fix :) | 08:24 |
| sdake | can't ship a pile of garb | 08:24 |
| Jeffrey4l | yup. that's one of reason | 08:26 |
| *** Pavo has quit IRC | 08:30 | |
| *** msimonin has joined #openstack-kolla | 08:31 | |
| *** kristia__ has joined #openstack-kolla | 08:34 | |
| *** Pavo has joined #openstack-kolla | 08:35 | |
| *** kristian__ has quit IRC | 08:35 | |
| sdake | duonghq will you be able to make team meeting tomorrow? | 08:36 |
| sdake | or rather today | 08:36 |
| duonghq | sdake, yup, iirc, out meeting is done in Wed? | 08:37 |
| sdake | yup - in channel topic | 08:38 |
| duonghq | sure | 08:38 |
| sdake | duonghq so in roughly 8 hrs or so ;) | 08:38 |
| duonghq | ah, today in my tz, just turn my inner clock to your tz so I can understand why you said it's tomorrow | 08:39 |
| sdake | it is actually today in my tz as well | 08:39 |
| sdake | which means, i need to go to bed soon :) | 08:39 |
| sdake | jascott1 will you be able to make it? | 08:39 |
| jascott1 | morning meeting? yes sir | 08:39 |
| jascott1 | ya i need to crash as well | 08:40 |
| duonghq | see you in the meeting, sdake or sooner when you wake up and I back from school | 08:41 |
| openstackgerrit | Justin Scott proposed openstack/kolla-kubernetes: WIP Helm-ize memcached https://review.openstack.org/405677 | 08:41 |
| duonghq | guess that we have many topic here | 08:41 |
| duonghq | especially o-2 is coming | 08:41 |
| sdake | jascott1 no need to call me sir :) | 08:42 |
| sdake | sdake works :) | 08:42 |
| *** mkoderer has joined #openstack-kolla | 08:42 | |
| jascott1 | yes sdake | 08:42 |
| jascott1 | :) | 08:42 |
| sdake | ok - need to crash before i burn | 08:42 |
| jascott1 | have a good one | 08:42 |
| jascott1 | night all | 08:42 |
| sdake | later peeps | 08:42 |
| duonghq | cya sdake | 08:42 |
| *** hfu has quit IRC | 08:45 | |
| *** hfu has joined #openstack-kolla | 08:46 | |
| *** jtriley has joined #openstack-kolla | 08:52 | |
| *** hfu has quit IRC | 08:57 | |
| *** egonzalez90 has joined #openstack-kolla | 08:58 | |
| egonzalez90 | morning | 08:58 |
| *** hfu has joined #openstack-kolla | 08:58 | |
| *** yuanying has joined #openstack-kolla | 08:59 | |
| *** athomas has joined #openstack-kolla | 08:59 | |
| *** jtriley has quit IRC | 09:02 | |
| *** sdake has quit IRC | 09:03 | |
| *** JRobinson__ has quit IRC | 09:03 | |
| duonghq | morning egonzalez90 | 09:05 |
| *** vilobhmm11 has quit IRC | 09:10 | |
| zhubingbing | morning | 09:22 |
| zhubingbing | sup egonzalez90 | 09:22 |
| *** kristia__ has quit IRC | 09:25 | |
| *** matrohon has joined #openstack-kolla | 09:33 | |
| *** mdnadeem has quit IRC | 09:33 | |
| *** yingjun_ has quit IRC | 09:42 | |
| *** rmart04 has joined #openstack-kolla | 09:42 | |
| duonghq | see you guys in about 4hrs | 09:43 |
| *** duonghq has quit IRC | 09:43 | |
| *** mdnadeem has joined #openstack-kolla | 09:50 | |
| zhubingbing | https://review.openstack.org/#/c/400122/ | 09:57 |
| zhubingbing | hello guys | 09:57 |
| zhubingbing | who can hel | 09:57 |
| zhubingbing | who can help me review it ? | 09:58 |
| openstackgerrit | Merged openstack/kolla: Update "searchlight-1.0.0.tar.gz" to "searchlight-master.tar.gz" https://review.openstack.org/407851 | 10:00 |
| *** tovin07_ has quit IRC | 10:02 | |
| *** tonanhngo has joined #openstack-kolla | 10:02 | |
| *** zhurong has quit IRC | 10:03 | |
| *** tonanhngo has quit IRC | 10:04 | |
| *** hfu has quit IRC | 10:05 | |
| *** hfu has joined #openstack-kolla | 10:06 | |
| *** huhaoran has quit IRC | 10:07 | |
| *** ccesario has joined #openstack-kolla | 10:13 | |
| *** prameswar has joined #openstack-kolla | 10:14 | |
| *** hfu has quit IRC | 10:16 | |
| *** hfu has joined #openstack-kolla | 10:16 | |
| *** hfu has quit IRC | 10:23 | |
| *** klindgren__ has joined #openstack-kolla | 10:29 | |
| *** Pavo has quit IRC | 10:30 | |
| *** klindgren_ has quit IRC | 10:32 | |
| *** DuncanT has quit IRC | 10:33 | |
| *** DuncanT has joined #openstack-kolla | 10:33 | |
| *** Pavo has joined #openstack-kolla | 10:34 | |
| *** david-lyle_ has joined #openstack-kolla | 10:35 | |
| *** prameswar has quit IRC | 10:37 | |
| *** prameswar has joined #openstack-kolla | 10:37 | |
| *** david-lyle has quit IRC | 10:37 | |
| *** prameswar has quit IRC | 10:38 | |
| *** Serlex has joined #openstack-kolla | 10:38 | |
| *** prameswar has joined #openstack-kolla | 10:38 | |
| *** kristian__ has joined #openstack-kolla | 10:38 | |
| openstackgerrit | chen.xing proposed openstack/kolla: Fix the format issue https://review.openstack.org/407343 | 10:39 |
| *** kristian__ has quit IRC | 10:39 | |
| *** kristia__ has joined #openstack-kolla | 10:40 | |
| *** kristia__ has quit IRC | 10:40 | |
| *** kristian__ has joined #openstack-kolla | 10:40 | |
| *** zhubingbing has quit IRC | 10:41 | |
| *** kristian__ has quit IRC | 10:41 | |
| *** kristian__ has joined #openstack-kolla | 10:41 | |
| *** zhubingbing has joined #openstack-kolla | 10:42 | |
| *** kristian__ has quit IRC | 10:42 | |
| *** kristian__ has joined #openstack-kolla | 10:42 | |
| *** kristian__ has quit IRC | 10:47 | |
| *** jtriley has joined #openstack-kolla | 10:53 | |
| *** kristian__ has joined #openstack-kolla | 10:54 | |
| *** sdake has joined #openstack-kolla | 10:55 | |
| *** kristia__ has joined #openstack-kolla | 10:55 | |
| *** jtriley has quit IRC | 10:57 | |
| *** jtriley has joined #openstack-kolla | 10:58 | |
| *** shardy has quit IRC | 10:58 | |
| *** kong_ has quit IRC | 10:58 | |
| *** kong_ has joined #openstack-kolla | 10:59 | |
| *** kristian__ has quit IRC | 10:59 | |
| *** akwasnie has quit IRC | 11:00 | |
| *** akwasnie has joined #openstack-kolla | 11:00 | |
| *** jtriley has quit IRC | 11:03 | |
| *** jascott1 has quit IRC | 11:06 | |
| *** EmilienM has left #openstack-kolla | 11:12 | |
| *** zhubingbing has quit IRC | 11:17 | |
| *** SaMnCo has quit IRC | 11:22 | |
| *** SaMnCo has joined #openstack-kolla | 11:22 | |
| *** kristia__ has quit IRC | 11:23 | |
| *** openstackgerrit has quit IRC | 11:32 | |
| *** turvey has quit IRC | 11:44 | |
| *** hyakuhei has quit IRC | 11:51 | |
| *** hyakuhei has joined #openstack-kolla | 11:51 | |
| *** hyakuhei has quit IRC | 11:51 | |
| *** hyakuhei has joined #openstack-kolla | 11:51 | |
| *** turvey has joined #openstack-kolla | 11:51 | |
| *** zhubingbing has joined #openstack-kolla | 11:53 | |
| *** yingjun has joined #openstack-kolla | 11:55 | |
| *** jtriley has joined #openstack-kolla | 11:59 | |
| *** jtriley has quit IRC | 12:03 | |
| *** dave-mccowan has joined #openstack-kolla | 12:10 | |
| *** openstackgerrit has joined #openstack-kolla | 12:16 | |
| openstackgerrit | prameswar proposed openstack/kolla: bifrost guide improved https://review.openstack.org/406177 | 12:16 |
| *** jtriley has joined #openstack-kolla | 12:18 | |
| openstackgerrit | prameswar proposed openstack/kolla: bifrost guide improved https://review.openstack.org/406177 | 12:21 |
| *** srwilkers_ has joined #openstack-kolla | 12:21 | |
| *** Pavo has quit IRC | 12:30 | |
| *** Pavo has joined #openstack-kolla | 12:34 | |
| openstackgerrit | Merged openstack/kolla-ansible: Enable designate dashboard when enable_designate is yes https://review.openstack.org/405942 | 12:35 |
| *** Bico_Fino has joined #openstack-kolla | 12:37 | |
| *** portdirect has joined #openstack-kolla | 12:42 | |
| *** Bico_Fino has quit IRC | 12:42 | |
| portdirect | morning o/ | 12:42 |
| *** Bico_Fino has joined #openstack-kolla | 12:50 | |
| *** jtriley has quit IRC | 12:50 | |
| openstackgerrit | Merged openstack/kolla: Fix typo error fix typo error "libary" -> "library" https://review.openstack.org/407876 | 12:53 |
| openstackgerrit | Merged openstack/kolla: Update Kibana documentation https://review.openstack.org/403729 | 12:53 |
| Bico_Fino | There is a order to restart all containers? | 12:55 |
| *** yingjun has quit IRC | 12:56 | |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Keystone Helm Packaging https://review.openstack.org/407780 | 12:58 |
| *** lamt has quit IRC | 13:01 | |
| *** mgiles has joined #openstack-kolla | 13:03 | |
| *** Bico_Fino has quit IRC | 13:08 | |
| *** schwicht has joined #openstack-kolla | 13:09 | |
| kfox1111 | morning. | 13:10 |
| *** duonghq has joined #openstack-kolla | 13:11 | |
| duonghq | evening guys | 13:11 |
| *** yingjun has joined #openstack-kolla | 13:11 | |
| *** kristian__ has joined #openstack-kolla | 13:13 | |
| zhubingbing | hi | 13:13 |
| Jeffrey4l | zhubingbing, congratulation ;) | 13:14 |
| zhubingbing | thank you very much | 13:16 |
| Jeffrey4l | since inc0 is on vocation. i will give some training as a core reviewer ;) | 13:16 |
| *** mdnadeem has quit IRC | 13:16 | |
| *** heroux has joined #openstack-kolla | 13:17 | |
| Jeffrey4l | i have added you into kolla-core and kolla-ansible group. now you can +2 and +w in kolla and kolla-ansible project. | 13:17 |
| *** Bico_Fino has joined #openstack-kolla | 13:17 | |
| *** kristian__ has quit IRC | 13:17 | |
| zhubingbing | I will do my best to contribute kolla | 13:17 |
| Jeffrey4l | nice ;) | 13:18 |
| Jeffrey4l | normally, if you are the first person +2, you shouldn't +w we need two at least two +2 | 13:18 |
| Jeffrey4l | if you are the second person +w , you are free to workflow +1, but remember you need +2 and +w at the same time. | 13:19 |
| zhubingbing | ;) understand | 13:19 |
| Jeffrey4l | do not just click the +workflow button. | 13:19 |
| Jeffrey4l | and please do not workflow+1 unless the gate is finished are gree. | 13:20 |
| Jeffrey4l | we do not require the core reviewer to test the patch. but you can ask the author that wether the code works. | 13:20 |
| Jeffrey4l | if no, you can -1 for it. | 13:21 |
| Jeffrey4l | if you are coauthor of the patch, you can not +2 for it. | 13:21 |
| *** matrohon has quit IRC | 13:21 | |
| Jeffrey4l | but you can +1 to indicate you think the code is OK to you. | 13:21 |
| Jeffrey4l | sometime, you can +1 for the patch, then you name will list in the reviewers list. it will tell others you want tracking this patch. | 13:22 |
| Jeffrey4l | and you will get some notification when the patch is updated. | 13:23 |
| zhubingbing | roger | 13:24 |
| Jeffrey4l | sometime you may need your -2 right. it means this patch is blocked. two possible reason, we can not merge this for release reason. or we won't merge this, it is not what kolla will do. | 13:24 |
| Jeffrey4l | for spec patch, it require majority vote. | 13:25 |
| Jeffrey4l | we have 14 core members right now. | 13:25 |
| Jeffrey4l | so it require at least 8 votes. | 13:25 |
| Jeffrey4l | this is all about review in gerrit | 13:26 |
| Jeffrey4l | zhubingbing, any question? | 13:26 |
| zhubingbing | no question | 13:26 |
| zhubingbing | thank you | 13:26 |
| Jeffrey4l | ok. | 13:26 |
| Jeffrey4l | as core reviewer, you also have responsibility to manage bugs in launchpad. | 13:27 |
| Jeffrey4l | I know you are familiar with that ;) | 13:27 |
| *** hfu has joined #openstack-kolla | 13:27 | |
| Jeffrey4l | take care when dealing the states of bug. other wise it will be got lost in the tracker . | 13:28 |
| Jeffrey4l | that all. | 13:28 |
| zhubingbing | understand | 13:29 |
| Jeffrey4l | cool | 13:29 |
| *** alteriks has quit IRC | 13:29 | |
| *** schwicht has quit IRC | 13:29 | |
| Jeffrey4l | and welcome join the kolla team ;) | 13:29 |
| zhubingbing | my pleasure | 13:30 |
| zhubingbing | ;) | 13:30 |
| duonghq | Jeffrey4l, as I see, the Kolla is updated after such a long time? | 13:30 |
| *** v1k0d3n has joined #openstack-kolla | 13:30 | |
| duonghq | Kolla meeting agenda | 13:31 |
| duonghq | sup v1k0d3n | 13:31 |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Keystone Helm Packaging https://review.openstack.org/407780 | 13:31 |
| *** v1k0d3n has quit IRC | 13:31 | |
| Jeffrey4l | duonghq, sorry? kolla is updated means? | 13:32 |
| duonghq | Jeffrey4l, missed typing, I mean Kolla meeting agenda | 13:32 |
| duonghq | sorry | 13:32 |
| Jeffrey4l | i made some change in the agenda. feel free to add some thing you want to talk in it. | 13:33 |
| portdirect | Jeffrey4l: do you have a link to the agenda? | 13:33 |
| Jeffrey4l | yep | 13:33 |
| Jeffrey4l | https://wiki.openstack.org/wiki/Meetings/Kolla | 13:33 |
| duonghq | Jeffrey4l, I added kolla-kubernetes helm versioning to the agenda, | 13:35 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Hel-ming Rabbitmq https://review.openstack.org/405422 | 13:35 |
| Jeffrey4l | cool | 13:35 |
| Jeffrey4l | thanks. | 13:35 |
| Jeffrey4l | we have the meeting relate info in the channel topic: you can check it if you forget: Kolla IRC meetngs on Wednesdays @ 16:00 UTC - see agenda @ https://goo.gl/OXB0DL | 13:36 |
| duonghq | zhubingbing, can you put your name on our Wiki, just for reference: https://wiki.openstack.org/wiki/Kolla :D | 13:36 |
| zhubingbing | ok | 13:37 |
| duonghq | not really many people read the wiki but it'd be nice if people can get you tz when somebody need to see you | 13:38 |
| portdirect | duonghq/zhubingbing: if either of you could add me to that list I'd really apreicate it - I'm unable to create an account to log into the wiki for some reason | 13:39 |
| sdake | morning peeps | 13:39 |
| sdake | is srwilkers around? | 13:39 |
| portdirect | morning sdake | 13:40 |
| sdake | aup portdirect | 13:40 |
| srwilkers_ | morning everyone | 13:40 |
| openstackgerrit | caoyuan proposed openstack/kolla: Remove the unnecessary "openstack-zaqar-ui" https://review.openstack.org/407576 | 13:40 |
| sdake | portdirect can you please email me a link to that article you send me the other day | 13:40 |
| duonghq | I go out for awhile, | 13:40 |
| sdake | hey srwilkers - wanted to know if you want to join the kolla-kubernetes-drivers team - this team does bug triage and generally maintains launchpad | 13:41 |
| srwilkers_ | yeah definitely. count me in | 13:41 |
| sdake | its is atypical to directly approve blueprints but other tings are ok with blueprints | 13:41 |
| sdake | which is your lauchpad id | 13:41 |
| zhubingbing | https://wiki.openstack.org/wiki/Kolla#Active_Contributors | 13:41 |
| zhubingbing | done | 13:41 |
| sdake | btw, this is a superset of the core reviewer team | 13:41 |
| *** lamt has joined #openstack-kolla | 13:41 | |
| srwilkers_ | sdake, srwilker | 13:41 |
| sdake | as in, it has more people in it then just the core reviewer team | 13:41 |
| srwilkers_ | yep, got it :) | 13:42 |
| sdake | it does giveyou full access to launchpad | 13:42 |
| sdake | the one thing i ask is not to approve blueprints and leave that to the core reviewers | 13:42 |
| srwilkers_ | yeah definitely. goes without saying :) | 13:42 |
| sdake | but feel free to getaquanted with launchpad | 13:42 |
| sdake | srwilkers_ nah its not really clear - actually very few people approve blueprints in the core reviewer team iether unforutntaely :( | 13:43 |
| sdake | give me a min to get my day going | 13:43 |
| prameswar | good morning All | 13:43 |
| srwilkers_ | sdake, i get that. makes sense to not have a bunch of people just approving things left and right, especially if they're not a core member to boot | 13:44 |
| srwilkers_ | heading into the office. will be back on in ~30 or so | 13:45 |
| *** srwilkers_ has quit IRC | 13:45 | |
| sdake | its not core member | 13:45 |
| sdake | its core reviewer | 13:45 |
| sdake | but roger :) | 13:46 |
| prameswar | can you please review : https://review.openstack.org/#/c/406177/ and https://review.openstack.org/#/c/407061/ again ? | 13:49 |
| prameswar | it is very difficult to write doc. :) | 13:49 |
| Jeffrey4l | klindgren__, around? | 13:50 |
| sdake | prameswar yes - writing documentation is harder then writing code | 13:51 |
| sdake | writing tests is harder then writing code too | 13:51 |
| sdake | not sure if tests or docs are harder ;) | 13:51 |
| prameswar | in my first patch there is one line that suggested to change then in second patch someone suggest change it to something diffent finally it become that i pushed in 1st patch :) | 13:51 |
| *** tyrola has quit IRC | 13:53 | |
| *** schwicht has joined #openstack-kolla | 13:54 | |
| *** rhallisey has joined #openstack-kolla | 13:56 | |
| egonzalez90 | Anyone know if there is a limit of volumes attached to an instance? | 13:58 |
| openstackgerrit | junbo proposed openstack/kolla-ansible: Integrate ceph-rgw with keystone https://review.openstack.org/406971 | 14:00 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Add qemu-img into glance api image https://review.openstack.org/408129 | 14:00 |
| *** jtriley has joined #openstack-kolla | 14:01 | |
| *** lamt has quit IRC | 14:03 | |
| *** schwicht has quit IRC | 14:04 | |
| *** tonanhngo has joined #openstack-kolla | 14:04 | |
| *** schwicht has joined #openstack-kolla | 14:05 | |
| *** sdake_ has joined #openstack-kolla | 14:06 | |
| sbezverk_ | good morning | 14:06 |
| *** tonanhngo has quit IRC | 14:06 | |
| portdirect | prameswar: yeah (that was me...) subjective things are the hardest to reach consensus on. Anything that is primarity rated on qualative vs quantative metrics is virtually impossible to draw a unviersially accepted consensus on, I used to teach engineering and hated marking students essays - as at least with pure mathematical work there was a definative answer you were looking for - though sometimes it was quite fun when I totally disagreed wit | 14:06 |
| portdirect | h their PoV but they presented a solid argument for it. | 14:06 |
| *** lamt has joined #openstack-kolla | 14:06 | |
| *** mdnadeem has joined #openstack-kolla | 14:07 | |
| kfox1111 | sbezverk_: morning | 14:08 |
| sbezverk_ | kfox1111: so rabbitmq is a go :-) | 14:09 |
| kfox1111 | yup. :) | 14:09 |
| *** sdake has quit IRC | 14:09 | |
| kfox1111 | its workflowed. I had to tweak the commit message to remove a work in progress. | 14:09 |
| kfox1111 | so its retesting. | 14:09 |
| prameswar | portdirect: :) | 14:09 |
| sbezverk_ | kfox1111: qq have you tried generate config after the repo split? I have not tried but one of guys I met at the summit told me that it is not working | 14:10 |
| kfox1111 | sbezverk_: yeah, the gate's still using trunk I believe. | 14:10 |
| sbezverk_ | kfox1111: I checked the gate code and you do some magic with zuul_clone or something, is it now required? | 14:11 |
| kfox1111 | it checks out kolla-ansible via zuul_clone, which is fater/more reliable then git clone in the gate. otherwise, its the same. | 14:11 |
| kfox1111 | I did check it out named kolla rather then kolla-ansible so I didn't have to go through and fix all the dir references. | 14:12 |
| kfox1111 | probably should fix that at some point. but wanted to fix a the gate asap when it broke. | 14:12 |
| sbezverk_ | kfox1111: ok I need to regen config for new test bed, hopefully gen config will just work.. | 14:13 |
| duonghq | just scroll back, sdake_, prameswar would it be better you shared volumes doc is put in seperated file? | 14:13 |
| portdirect | sbezverk_: I've been using the method that kfox1111 implemented in the gate - and its been working fine for me | 14:13 |
| sbezverk_ | portdirect: cool, thanks for confirming | 14:15 |
| openstackgerrit | Merged openstack/kolla-kubernetes: Hel-ming Rabbitmq https://review.openstack.org/405422 | 14:15 |
| portdirect | kfox1111: should we think about move genconfig into kolla? it seems counterintative for kolla-k8s to depend on kolla-ansible long term? Though I'm totally fine with being shot down on that :) | 14:15 |
| kfox1111 | portdirect: I agree. | 14:16 |
| *** saneax is now known as saneax-_-|AFK | 14:16 | |
| duonghq | portdirect, kfox1111 me too | 14:16 |
| sbezverk_ | kfox1111: portdirect: 100% agree | 14:16 |
| kfox1111 | there's been some talk about moving genconfig back into kolla repo and removing the ansible part of the config. | 14:16 |
| sbezverk_ | kfox1111: portdirect: there is one thing though, kolla-ansible generates openstack config for us | 14:17 |
| kfox1111 | sbezverk_: yeah. | 14:17 |
| sbezverk_ | I am not sure how feasible would it be to do just in kolla without kolla-ansible | 14:18 |
| kfox1111 | and genconfiging the openstack stuff is complicated enough, it would be nice to share that code if at all possible. | 14:18 |
| kfox1111 | we should have to have a really good reason to maintain a parallel config system. | 14:18 |
| portdirect | sbezverk_ this was my concern, could we discuss this at the meeting? (I cant edit the wii to add it to the agenda) | 14:19 |
| duonghq | config layering and merging is wfm | 14:19 |
| sbezverk_ | kfox1111: if genconfig can consume config for ansible and config for kolla-kubernetes then no reason :-) | 14:19 |
| sbezverk_ | portdirect: good idea, let's add it | 14:19 |
| kfox1111 | rabbit merged | 14:20 |
| portdirect | Nice :) I'm about to push a ps for keystone that will raise a few q's about db management (where we store params) - I've intentionally left that open for disscussion (ie incomplete) as it will set the tone for other api servces, | 14:23 |
| kfox1111 | portdirect: k. sounds good. | 14:23 |
| kfox1111 | portdirect: I'm still on the fence if we should push keystone/horizon through common_deployment. it basically requires turning off everything added in common_deployment. it may be cleaner to just use common lib. | 14:25 |
| kfox1111 | calling it common_deployment might have been a bit of a misnomer, as its not intended for all cases a kubernetes deployment object could be used. | 14:26 |
| *** v1k0d3n has joined #openstack-kolla | 14:26 | |
| *** v1k0d3n has quit IRC | 14:26 | |
| kfox1111 | but handles just about all of the openstack api like services that are not in apache. | 14:26 |
| *** v1k0d3n has joined #openstack-kolla | 14:26 | |
| *** coolsvap has quit IRC | 14:27 | |
| *** krtaylor has quit IRC | 14:28 | |
| *** severion has joined #openstack-kolla | 14:28 | |
| portdirect | kfox1111: I'm open to that, more openstack projects are beginning to support apache as a deployment mechanism - I could rename common_deployment to common_python_deployment, and create a common_apache_deployment? | 14:29 |
| *** krtaylor has joined #openstack-kolla | 14:29 | |
| kfox1111 | portdirect: that sounds good to me. | 14:30 |
| *** Pavo has quit IRC | 14:30 | |
| kfox1111 | then we can easily share the code between keystone and horizon. | 14:30 |
| *** Jeffrey4l has quit IRC | 14:30 | |
| *** srwilkers_ has joined #openstack-kolla | 14:31 | |
| *** Jeffrey4l has joined #openstack-kolla | 14:31 | |
| *** matrohon has joined #openstack-kolla | 14:31 | |
| portdirect | np - will do, I'll create a sperate ps for that - to clearly seperate the tasks and also allow us to get it in before keystone is done - whihc will help people working on other api's | 14:31 |
| *** v1k0d3n has quit IRC | 14:32 | |
| *** schwicht has quit IRC | 14:32 | |
| kfox1111 | sounds good. :) | 14:33 |
| *** Pavo has joined #openstack-kolla | 14:35 | |
| *** hfu has quit IRC | 14:40 | |
| *** matrohon has quit IRC | 14:41 | |
| *** hfu has joined #openstack-kolla | 14:43 | |
| *** schwicht has joined #openstack-kolla | 14:47 | |
| *** fguillot has joined #openstack-kolla | 14:50 | |
| openstackgerrit | junbo proposed openstack/kolla-ansible: Integrate ceph-rgw with keystone https://review.openstack.org/406971 | 14:52 |
| *** msimonin has quit IRC | 14:55 | |
| *** hfu has quit IRC | 14:57 | |
| *** yingjun has quit IRC | 14:57 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP Hel-ming mariadb components https://review.openstack.org/408160 | 14:57 |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Keystone Helm Packaging https://review.openstack.org/407780 | 14:58 |
| *** schwicht has quit IRC | 15:00 | |
| *** yingjun has joined #openstack-kolla | 15:03 | |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Change _common_deployment.yaml to _common_api_python_deployment.yaml https://review.openstack.org/408165 | 15:11 |
| *** ccesario has quit IRC | 15:16 | |
| *** msimonin has joined #openstack-kolla | 15:19 | |
| *** lrensing has joined #openstack-kolla | 15:20 | |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Change _common_deployment.yaml to _common_api_python_deployment.yaml https://review.openstack.org/408165 | 15:21 |
| *** dave-mccowan has quit IRC | 15:23 | |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Change _common_deployment.yaml to _common_api_python_deployment.yaml https://review.openstack.org/408165 | 15:25 |
| *** Jeffrey4l has quit IRC | 15:27 | |
| *** Jeffrey4l has joined #openstack-kolla | 15:27 | |
| *** krishnaposa has joined #openstack-kolla | 15:28 | |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Change _common_deployment.yaml to _common_api_python_deployment.yaml https://review.openstack.org/408165 | 15:29 |
| *** sdake_ has quit IRC | 15:29 | |
| kfox1111 | portdirect: alive? | 15:29 |
| portdirect | kfox1111: just working on too many ps at once - hence gerrit madness :) | 15:30 |
| *** mdnadeem has quit IRC | 15:30 | |
| kfox1111 | portdirect: yeah. :) | 15:30 |
| kfox1111 | figured it would be better to talk it out real quick. :) | 15:30 |
| portdirect | shoot :) | 15:30 |
| kfox1111 | so, do you think the parent itself would ever want to disaable haproxy overriding what the user specifies via values? | 15:30 |
| portdirect | not with the seperation between api_python and api_apache | 15:31 |
| kfox1111 | k. so, I thin kthe cleanest thing would be to: | 15:31 |
| *** magicboiz has joined #openstack-kolla | 15:32 | |
| *** sdake has joined #openstack-kolla | 15:32 | |
| kfox1111 | * have {{- $haproxy := .Values.haproxy | default true }} of the common file. | 15:32 |
| kfox1111 | * s/.haproxy/$haproxy/ in the common file | 15:32 |
| *** absubram has joined #openstack-kolla | 15:32 | |
| kfox1111 | * not define haproxy at all in neutron-server | 15:32 |
| kfox1111 | that should eliminate all dead code. | 15:33 |
| portdirect | ok - that sounds logical - Need to get my head round gotpl a but more i think (as I'm not sure how to implement that ohhand if the 2nd ps wouldnt hav achived that) - will proba have worked it out in 2 mins - but I'll hit you up if i get stuck :) | 15:34 |
| *** athomas has quit IRC | 15:35 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP Hel-ming mariadb components https://review.openstack.org/408160 | 15:35 |
| kfox1111 | k. sounds good. | 15:37 |
| kfox1111 | yeah. gotl is a little weird. | 15:37 |
| kfox1111 | . vars vs $ vars especially. :/ | 15:37 |
| kfox1111 | and no way to set a . var except with a 'with' is goofy and painful. :/ | 15:37 |
| sdake | duonghq busy - not sure why you pinged me - working on getting my kids to school | 15:38 |
| *** tonanhngo has joined #openstack-kolla | 15:38 | |
| *** portdirect is now known as portdirect_lunch | 15:39 | |
| duonghq | sdake, about this ps: https://review.openstack.org/#/c/407061/ | 15:40 |
| duonghq | I think we should not continue putting stuff like this to QSG? | 15:40 |
| sdake | busy | 15:41 |
| duonghq | can you join IRC meeting? | 15:41 |
| *** dave-mccowan has joined #openstack-kolla | 15:41 | |
| duonghq | sorry, just do not get your meaning | 15:43 |
| *** athomas has joined #openstack-kolla | 15:44 | |
| *** pomac_ is now known as pomac | 15:45 | |
| egonzalez90 | duonghq: have the same opinion as you with QSG | 15:48 |
| duonghq | thank egonzalez90 | 15:48 |
| *** heroux has quit IRC | 15:48 | |
| *** sayantan_ has joined #openstack-kolla | 15:49 | |
| *** wwriverrat has joined #openstack-kolla | 15:50 | |
| *** hogepodge has joined #openstack-kolla | 15:50 | |
| *** portdirect_lunch is now known as portdirect | 15:52 | |
| *** jheroux has joined #openstack-kolla | 15:52 | |
| *** sayantan_ has quit IRC | 15:54 | |
| Jeffrey4l | we can talk this in the meeting. | 15:55 |
| Jeffrey4l | doc is a essential thing we need improve in this cycle. there are lots of complain about it. | 15:56 |
| duonghq | Jeffrey4l, can you make a topic about doc? | 15:57 |
| Jeffrey4l | duonghq, yep. i added it into the agenda | 15:57 |
| *** jascott1 has joined #openstack-kolla | 15:58 | |
| duonghq | cool | 15:58 |
| Jeffrey4l | meeting time guys | 16:00 |
| wirehead_ | If I wander in to a meeting channel and the previous meeting hasn't stopped yet, do I retain my usual right to make bad puns? | 16:00 |
| portdirect | of course :) | 16:00 |
| *** jheroux has quit IRC | 16:00 | |
| *** yingjun has quit IRC | 16:01 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP Hel-ming mariadb components https://review.openstack.org/408160 | 16:01 |
| *** yingjun has joined #openstack-kolla | 16:01 | |
| *** msimonin has quit IRC | 16:02 | |
| *** msimonin has joined #openstack-kolla | 16:03 | |
| *** adrian_otto has joined #openstack-kolla | 16:04 | |
| *** yingjun has quit IRC | 16:05 | |
| *** schwicht has joined #openstack-kolla | 16:05 | |
| *** jheroux has joined #openstack-kolla | 16:06 | |
| *** msimonin has quit IRC | 16:08 | |
| *** jgriffith_away is now known as jgriffith | 16:09 | |
| *** rmart04 has quit IRC | 16:11 | |
| dmsimard | Jeffrey4l: which channel is the meeting in ? | 16:11 |
| srwilkers_ | openstack-meeting-4 | 16:11 |
| Jeffrey4l | dmsimard, #openstack-meeting-4 | 16:12 |
| *** yingjun has joined #openstack-kolla | 16:12 | |
| *** sayantan_ has joined #openstack-kolla | 16:17 | |
| *** sayantan_ has quit IRC | 16:17 | |
| *** kollabot2 has quit IRC | 16:18 | |
| *** kollabot has joined #openstack-kolla | 16:18 | |
| *** yingjun has quit IRC | 16:18 | |
| *** pcaruana has quit IRC | 16:18 | |
| *** yingjun has joined #openstack-kolla | 16:18 | |
| *** sayantan_ has joined #openstack-kolla | 16:19 | |
| *** gfidente has quit IRC | 16:24 | |
| *** fguillot_ has joined #openstack-kolla | 16:25 | |
| *** fguillot has quit IRC | 16:25 | |
| *** Pavo has quit IRC | 16:30 | |
| *** mdnadeem has joined #openstack-kolla | 16:33 | |
| *** Pavo has joined #openstack-kolla | 16:35 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP Hel-ming mariadb components https://review.openstack.org/408160 | 16:36 |
| *** yingjun has quit IRC | 16:39 | |
| *** yingjun has joined #openstack-kolla | 16:39 | |
| *** yingjun has quit IRC | 16:43 | |
| *** adrian_otto has quit IRC | 16:49 | |
| *** zhubingbing has quit IRC | 16:54 | |
| *** vhosakot has joined #openstack-kolla | 16:55 | |
| *** Bico_Fino has quit IRC | 16:59 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP Hel-ming mariadb components https://review.openstack.org/408160 | 17:00 |
| *** kfox1111 has quit IRC | 17:01 | |
| dmsimard | Sorry, got side tracked during the meeting. I had a question for the open floor. | 17:01 |
| *** kfox1111 has joined #openstack-kolla | 17:02 | |
| dmsimard | I'd like to know if you would be open to splitting the baremetal role out of kolla-ansible: https://github.com/openstack/kolla-ansible/tree/master/ansible/roles/baremetal (the one responsible for bootstrap-servers) | 17:02 |
| Jeffrey4l | dmsimard, split it to? | 17:02 |
| openstackgerrit | sayantani proposed openstack/kolla: Minor documentation fixes https://review.openstack.org/406534 | 17:02 |
| dmsimard | Context is kolla-ansible (the tool) does not really bring much value in this context, and really all I want is to leverage the bootstrap server role in order to use kolla-build without needing to do the setup myself | 17:03 |
| dmsimard | Jeffrey4l: a separate standalone role, outside of kolla-ansible | 17:03 |
| dmsimard | I'm trying to build kolla containers with ansible right now, I /could/ use the kolla-ansible command but that hides away all the tasks it does (it's like a bash command inside an ansible task), so I'd like to include the role properly instead | 17:04 |
| dmsimard | I'm not sure if you understand | 17:04 |
| Jeffrey4l | hrm but it is doing lots of work related to kolla. | 17:04 |
| Jeffrey4l | like make /run shared | 17:04 |
| Jeffrey4l | prepare /etc/hosts based on the api_interface | 17:05 |
| *** klindgren has joined #openstack-kolla | 17:05 | |
| rhallisey | dmsimard, that idea has been throw around. Maybe a kolla-host repo or something | 17:05 |
| dmsimard | Jeffrey4l: it's still related to kolla if it's split into git.o.o/openstack/ansible-role-kolla-bootstrap or something like that | 17:05 |
| Jeffrey4l | but it just installs docker-engine except that. | 17:05 |
| dmsimard | afk, getting food. | 17:05 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP Hel-ming mariadb components https://review.openstack.org/408160 | 17:06 |
| *** jemcevoy has joined #openstack-kolla | 17:06 | |
| Jeffrey4l | rhallisey, kolla-host is mange baremetal, imo. | 17:06 |
| rhallisey | it could also do config generation | 17:06 |
| Jeffrey4l | not mainly prepare the kolla running env. | 17:07 |
| klindgren | Jeffrey4l, pong | 17:07 |
| Jeffrey4l | sup klindgren | 17:07 |
| Jeffrey4l | klindgren, can we talk about https://review.openstack.org/#/c/405647/ | 17:07 |
| *** klindgren__ has quit IRC | 17:08 | |
| klindgren | yep - can I have ~15 minutes? Finishing a meeting - but I know you are not in the states | 17:08 |
| Jeffrey4l | np. | 17:09 |
| Jeffrey4l | ping me when you are free klindgren | 17:09 |
| duonghq | Jeffrey4l, if I understand correctly, your approach in reconfiguration still need the config is copied to every node (in their type) and then the Kolla know whether the config is changed and trigger the handler? | 17:09 |
| Jeffrey4l | i am still up in another 1 hours. | 17:10 |
| *** jemcevoy has quit IRC | 17:10 | |
| Jeffrey4l | duonghq, yep. | 17:10 |
| duonghq | ok | 17:10 |
| Jeffrey4l | so what u want? | 17:10 |
| duonghq | just confirm my understand | 17:10 |
| Jeffrey4l | exactly ;) | 17:10 |
| duonghq | I think it's better if we know which host need to be updates in advanced | 17:11 |
| duonghq | but it'll require huge effort | 17:11 |
| *** Bico_Fino has joined #openstack-kolla | 17:11 | |
| Jeffrey4l | duonghq, but how? | 17:11 |
| *** jascott1 has quit IRC | 17:11 | |
| Jeffrey4l | if u do not copy the configuration file, how could u know which host need update? | 17:11 |
| duonghq | Jeffrey4l, iirc my friend told me that some CM can tracking this | 17:11 |
| Jeffrey4l | no idea about CM. | 17:12 |
| Jeffrey4l | ;( | 17:12 |
| Jeffrey4l | is there any tool implement CM? | 17:12 |
| duonghq | I need to ask somebody and will bring this topic later | 17:12 |
| Jeffrey4l | cool. | 17:12 |
| Jeffrey4l | thaks | 17:12 |
| Jeffrey4l | i think you may be a DB to store everything else on other node. then you can get which node need to be updated without touch it. | 17:13 |
| Jeffrey4l | if so, i do not think ansible can implement this | 17:13 |
| egonzalez90 | duonghq: you meaning something like puppet-agents to check the status of everything? | 17:13 |
| duonghq | Jeffrey4l, yup, most trivial approach | 17:13 |
| duonghq | egonzalez90, yep | 17:13 |
| duonghq | my friend told me something like that | 17:13 |
| Jeffrey4l | egonzalez90, we have ansible-pull, just like agent. but no idea could it implement such. | 17:14 |
| egonzalez90 | personally I don't like that approach, leting a CM to modify whenever it wants something | 17:15 |
| egonzalez90 | I would prefer one step to check status and then other step to configure | 17:15 |
| severion | duonghq: hey man | 17:15 |
| severion | sorry late | 17:16 |
| severion | dang...one sec... | 17:16 |
| *** severion has quit IRC | 17:16 | |
| *** v1k0d3n has joined #openstack-kolla | 17:16 | |
| v1k0d3n | duonghq: hey...this is better | 17:16 |
| *** schwicht has quit IRC | 17:16 | |
| duonghq | hey there | 17:17 |
| Jeffrey4l | egonzalez90, agreed it is simple and works. | 17:17 |
| *** schwicht has joined #openstack-kolla | 17:17 | |
| duonghq | agreed with egonzalez90 | 17:18 |
| duonghq | v1k0d3n, you are severion? | 17:18 |
| *** sacharya has joined #openstack-kolla | 17:20 | |
| *** egonzalez90 has quit IRC | 17:20 | |
| v1k0d3n | when my client drops...yes | 17:21 |
| v1k0d3n | :) | 17:21 |
| duonghq | what can I do for you? | 17:22 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible: Implement only restart the container which configuration changed https://review.openstack.org/408225 | 17:22 |
| Jeffrey4l | duonghq, make another improvement. it is kinda complicated, but better ^^ | 17:23 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible: Implement only restart the container which configuration changed https://review.openstack.org/408225 | 17:24 |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Change _common_deployment.yaml to _common_api_python_deployment.yaml https://review.openstack.org/408165 | 17:25 |
| Jeffrey4l | no idea whether it is worth, especially for big service, like nova and neutron. | 17:25 |
| klindgren | Jeffrey4l, ping - Free now | 17:26 |
| Jeffrey4l | klindgren, pong, cool | 17:26 |
| *** ccesario has joined #openstack-kolla | 17:26 | |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Change _common_deployment.yaml to _common_api_python_deployment.yaml https://review.openstack.org/408165 | 17:26 |
| Jeffrey4l | klindgren, i replied your comment in the ps. | 17:27 |
| duonghq | Jeffrey4l, If we rewrite it in other way, it'll be more compact | 17:27 |
| *** rhallisey has quit IRC | 17:27 | |
| duonghq | but for many components services, I don't sure | 17:27 |
| duonghq | better performance but harder for maintainance | 17:28 |
| *** rhallisey has joined #openstack-kolla | 17:28 | |
| Jeffrey4l | current implementation 1 do not hard duonghq | 17:28 |
| Jeffrey4l | i wrote the nova part, i do not think it is hard. | 17:29 |
| Jeffrey4l | duonghq, https://review.openstack.org/407416 | 17:29 |
| duonghq | it still ok until the components is not change in name or number | 17:29 |
| klindgren | Jeffrey4l, yep, I was reading though. So I was not aware of the Kolla philosophy about being as hands off as possible from the host OS. | 17:30 |
| Jeffrey4l | change name is bad, we should avoid this. it will cause upgrade issue. | 17:30 |
| *** kristia__ has joined #openstack-kolla | 17:30 | |
| Jeffrey4l | duonghq, change number is not a big issue. | 17:30 |
| duonghq | ya, it should not be in most case | 17:31 |
| portdirect | sbezverk: ping | 17:31 |
| Jeffrey4l | klindgren, create the same user on host is useless except good for view by human | 17:31 |
| klindgren | I guess if I wanted to have the uid/gid/ match in the host os, I could do that. | 17:31 |
| klindgren | well the reason why I wanted to do that is that initially I wanted to use the scripts/logic I already have for generating the config files | 17:32 |
| portdirect | sbezverk/kfoxx1111: any preference for where we store common db port/hostname for consumption by services? | 17:32 |
| Jeffrey4l | klindgren, you can create it in another script/roles. | 17:32 |
| Jeffrey4l | but i do not think add this into kolla is a good idea. | 17:32 |
| klindgren | if I set the uid/gid the same as whats in the kolla images, then I dont have to worry about chowning them in the container? | 17:32 |
| *** prameswar has quit IRC | 17:33 | |
| Jeffrey4l | klindgren, generation config files? why not use kolla's configure generator? | 17:33 |
| Jeffrey4l | klindgren, you can use chown 1000:1000 foobar ;) | 17:33 |
| kfox1111 | portdirect: not quite sure I follow. | 17:34 |
| klindgren | because we have a bunch of extra stuff added into our - plus we run cells :-) | 17:34 |
| *** JoseMello has joined #openstack-kolla | 17:34 | |
| kfox1111 | portdirect: you mean mariadb location info? | 17:34 |
| klindgren | added into our config* | 17:34 |
| *** mdnadeem has quit IRC | 17:34 | |
| Jeffrey4l | cool. do u ever thought add cell support in kolla lol | 17:34 |
| klindgren | basically trying to get kolla in our existing env/process, but we are trying to do it in as small of chunks as a time | 17:34 |
| portdirect | kfox1111: yeah - eg for: https://review.openstack.org/#/c/407780/7/helm/kolla-common/templates/_common_db_create_job.yaml Line 18/19 | 17:35 |
| Jeffrey4l | anyway, create the same user in host os is a edge case, it is useless for most of user. and it breaks kolla's philosophy. | 17:36 |
| *** kristian__ has joined #openstack-kolla | 17:36 | |
| *** kristia__ has quit IRC | 17:36 | |
| *** tonanhngo has quit IRC | 17:36 | |
| Jeffrey4l | but we do not stop u create the same user in you owner script, which is very simple. | 17:36 |
| klindgren | and from my perspective where we have a bunch of external, non openstack, not yet kolla'ized services that do things on the host. Having matching uid/gid between container and host would make things simpler, as we get more and more of things into kolla it makes less of a deal | 17:36 |
| Jeffrey4l | after we implement fix uid and gid. | 17:36 |
| klindgren | so - I can keep that part out of any patch up stream, and simply make a script to parse the mapping | 17:36 |
| kfox1111 | portdirect: looking | 17:36 |
| klindgren | and create the users locally | 17:36 |
| Jeffrey4l | cool. ( if we create the user, what will happen, if the uid is already taken by another sysstem user? ) | 17:37 |
| klindgren | Jeffrey4l, re: cells. Yes we are going to have to add nova-cells. But the issue is this is cellsv1 and from previous convo's you guys really only want to add support for cellsv2. Because they behave very differently | 17:38 |
| Jeffrey4l | klindgren, we have agreement on this, right? move on? | 17:38 |
| *** jascott1 has joined #openstack-kolla | 17:38 | |
| Jeffrey4l | yep. you r right. | 17:39 |
| *** prameswar has joined #openstack-kolla | 17:39 | |
| kfox1111 | portdirect: Ah./ I see. your finishing up the common create db stuff I hadn't gotten around to yet. | 17:39 |
| kfox1111 | thanks for working on that. :) | 17:39 |
| Jeffrey4l | next is fix uid mapping or dynamic, right? fk | 17:39 |
| Jeffrey4l | klindgren, ^ | 17:39 |
| klindgren | Jeffrey4l, I think the only thing that is still open. Is where to store the mapping | 17:40 |
| Jeffrey4l | forget it now. | 17:40 |
| portdirect | kfox1111: yeah - I've got keystone endpopint jobs and db (initial) migration about to drop as well :) | 17:40 |
| kfox1111 | doh. | 17:40 |
| kfox1111 | been working on common endpoint myself. | 17:40 |
| Jeffrey4l | we need confirm that, if the uid is changable, how to handle the upgrade issue we meet now? | 17:40 |
| kfox1111 | I'll have a look at that too. | 17:40 |
| Jeffrey4l | klindgren, ^^ | 17:41 |
| klindgren | My only point was I would rather have it in a config file to override, vs's in code so I dont have to maintain a code patch to kolla. | 17:41 |
| Jeffrey4l | store it is simple. i have idea for this. but we need confirm above issue. | 17:41 |
| kfox1111 | portdirect: ok. so your question is about allowing the user override the db info, so its easier to do a multimariadb setup? | 17:42 |
| Jeffrey4l | if the uid is changable/customable, does kolla handle the different during upgrade? klindgren | 17:42 |
| klindgren | So as I understand the issue. If someone changes the uid between releases. At some point in the future | 17:42 |
| v1k0d3n | duonghq: i was just responding to you earlier. | 17:42 |
| v1k0d3n | been head down lately. | 17:43 |
| kfox1111 | portdirect: oh. I don't see code for endpoint creation yet. have you been owkring on it, or going to next? | 17:43 |
| kfox1111 | cause I have patches. | 17:43 |
| klindgren | when the start up script in the container is removed. they will cause permisions issues that kolla wont fix for them | 17:43 |
| portdirect | kfox1111: yep - also to support mariadb being in a different k8s namespace from the service | 17:43 |
| portdirect | kfox1111: putting together a ps with them all in - not pushed yest but soon as i have this resolved I;ll get them up | 17:44 |
| kfox1111 | portdirect: ah. I see. yeah. +1 to that configurability. | 17:44 |
| klindgren | from my point of view having a a warning with a doc fix would be sufficient? | 17:44 |
| kfox1111 | portdirect: do you actually have code yet for endpoints though? | 17:44 |
| duonghq | get some sleep now, thank you, see ya guys | 17:44 |
| Jeffrey4l | klindgren, cool. i hope so. if u use customize uid, you should take care of you uid mapping configuration. kolla do not take care of this. | 17:44 |
| kfox1111 | trying to ffigure out how much we are actually duplicating work at this point. | 17:44 |
| portdirect | yep | 17:44 |
| Jeffrey4l | klindgren, agreed. | 17:44 |
| klindgren | IE if you change this, you will also need to change the files on the host to match the new mapping. Read nore here. | 17:44 |
| portdirect | I was takeleing the keystone parts as on the blueprint ? | 17:44 |
| Jeffrey4l | then we can remove permissions secions in config.json file. | 17:44 |
| *** duonghq has quit IRC | 17:44 | |
| kfox1111 | portdirect: yeah, but depends on common stuff, and I was doing the same for neutron's common bits. | 17:45 |
| kfox1111 | i didn't see any reviews up so I started workin on that part. | 17:45 |
| Jeffrey4l | klindgren, yes. mainly the file in docker named volume. | 17:45 |
| *** magicboiz has quit IRC | 17:45 | |
| Jeffrey4l | trust me, no one want do this. it is very complicated. | 17:45 |
| kfox1111 | can you post what you have asap? | 17:46 |
| Jeffrey4l | so keep your uid mapping if u change it, and use it to build container all the time. | 17:46 |
| *** strigazi is now known as strigazi_AFK | 17:46 | |
| Jeffrey4l | and add some warning to when someone trying to use such feature. | 17:46 |
| klindgren | yep - that was my thought as well. Someone would change it to help ease kolla into their env, but no reason to change it again once kolla is all integrated. | 17:47 |
| Jeffrey4l | cool. | 17:47 |
| Jeffrey4l | so we can continue | 17:47 |
| *** jascott1 has quit IRC | 17:47 | |
| Jeffrey4l | about the uid mapping storage. | 17:47 |
| Jeffrey4l | we can save a static mapping in common/config.py file. | 17:48 |
| *** msimonin has joined #openstack-kolla | 17:48 | |
| Jeffrey4l | and add a new parameter like --uid-mapping which is a oslo.dict format. you can use this override the static mapping. | 17:49 |
| Jeffrey4l | does it work? | 17:49 |
| portdirect | kfox1111: will post in 5 mins | 17:49 |
| kfox1111 | portdirect: k. thanks. | 17:49 |
| kfox1111 | I'll try and merge with mine. | 17:49 |
| portdirect | just need to move it from my ws | 17:50 |
| kfox1111 | I've been doing it a bit differently though. finally got to the point where I couldn't look at that hacky json escaping nightmare in there, | 17:50 |
| kfox1111 | ansible being used when not really needed, | 17:50 |
| kfox1111 | and there being a race condition because of how ansible does things, | 17:51 |
| kfox1111 | so was breaking out service creation into its own job. | 17:51 |
| Jeffrey4l | klindgren, / | 17:51 |
| *** msimonin has quit IRC | 17:53 | |
| klindgren | I think that would work, if we can pass a mapping in. As long as we dont have to modify code - I would be ok with it | 17:55 |
| *** sacharya_ has joined #openstack-kolla | 17:56 | |
| Jeffrey4l | klindgren, we can i will write a piece of code for this. | 17:56 |
| Jeffrey4l | now need go to bed ;) | 17:56 |
| klindgren | ok - thanks - I will update the spec with the information that we talked about | 17:56 |
| Jeffrey4l | thanks a lot. | 17:57 |
| *** sacharya has quit IRC | 17:57 | |
| Jeffrey4l | klindgren, just fyi, kolla will make b2 tag arouch dec 15. hope this ps can be merged before that ;) | 17:58 |
| klindgren | kk | 17:58 |
| Jeffrey4l | thanks. bye guys. | 17:58 |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Keystone Helm Packaging https://review.openstack.org/407780 | 17:59 |
| *** jascott1 has joined #openstack-kolla | 18:00 | |
| *** athomas has quit IRC | 18:01 | |
| portdirect | kfox1111: I'll move these into a new ps for common libs so we can work on them there | 18:01 |
| kfox1111 | portdirect: oh... | 18:02 |
| *** david-lyle_ is now known as david-lyle | 18:02 | |
| portdirect | ? | 18:02 |
| kfox1111 | you uploaded the keystone create endpoitn job as common. | 18:02 |
| kfox1111 | keystone's own endpoint one is fairly keystone specific. | 18:02 |
| portdirect | oh - good point! | 18:03 |
| kfox1111 | I'd move that one back as a regular job package, and then use the common one I'm owrking on. | 18:03 |
| portdirect | whoops :) | 18:03 |
| kfox1111 | then we're not stepping on each other too. :) | 18:03 |
| portdirect | ofc :) | 18:03 |
| *** fragatina has joined #openstack-kolla | 18:03 | |
| portdirect | I'll move it back and get the init in as well - still need to test it on my local gate clone. | 18:04 |
| kfox1111 | cool. | 18:04 |
| *** fragatin_ has joined #openstack-kolla | 18:04 | |
| kfox1111 | I'd consider breaking up that ps more too. parts could probably have been merged already but I'm going to have to rereview it all as it keeps changing. | 18:04 |
| portdirect | yeah - I know it's not ideal - I'll smash it with the big git hammer just now :) | 18:05 |
| kfox1111 | cool. thanks. | 18:05 |
| *** krtaylor has quit IRC | 18:06 | |
| *** sacharya_ has quit IRC | 18:07 | |
| *** fguillot_ is now known as fguillot | 18:08 | |
| portdirect | though re your comment - I much prefer <service>-db-create to <serivce>-create-db, as it reads better for other actions, esp when services have multiple depencaices to manipultate eg rabbit etc. Is there a reson fro your peference the other way round? | 18:08 |
| *** fragatina has quit IRC | 18:08 | |
| *** sacharya has joined #openstack-kolla | 18:08 | |
| kfox1111 | most of the other services are already folowing the other convention. | 18:08 |
| kfox1111 | we'd have to change them all. | 18:08 |
| kfox1111 | it also looks a little better to me due to all the similar jobs lining up: | 18:09 |
| kfox1111 | neutron-create-keystone-service | 18:09 |
| kfox1111 | neutron-create-keystone-endpoint-public | 18:09 |
| kfox1111 | neutron-create-db | 18:09 |
| kfox1111 | rather then: | 18:09 |
| kfox1111 | neutron-keystone-service-create | 18:10 |
| kfox1111 | neutron-keystone-endpoint-create | 18:10 |
| kfox1111 | neutron-db-create | 18:10 |
| kfox1111 | thought thats pretty minor. | 18:10 |
| portdirect | though the dont esist in helm yet - so is this not a perfect opportunity to do so? though from your messages i see we are just encoutering a difference in the way we visulaise a workflow for a service :) | 18:10 |
| *** kollabot3 has joined #openstack-kolla | 18:11 | |
| *** kollabot has quit IRC | 18:11 | |
| kfox1111 | its mostly convention I'm owrried about. theyu all shouodl be the same, whatever the convnention is. | 18:12 |
| kfox1111 | we'll, we're mostly doing a one to one convert. but, yeah, might be easier now then later. | 18:12 |
| kfox1111 | so, I prefer one way, you prefer the other. need a tie breaker. :) | 18:12 |
| kfox1111 | sbezverk_: you have an opinion? :) | 18:12 |
| portdirect | ^^ :D | 18:12 |
| kfox1111 | I don't care too much, so whatever others reach consensus around. but don't want extra work, so unless consensus is to change, lets just keep it the way it is. :) | 18:13 |
| *** tonanhngo has joined #openstack-kolla | 18:16 | |
| *** srwilkers_ is now known as srwilkers_lunch | 18:17 | |
| *** tonanhngo has quit IRC | 18:19 | |
| *** prameswar has quit IRC | 18:20 | |
| kfox1111 | arg... hate that about helm... so useful... :( | 18:23 |
| kfox1111 | Error: YAML parse error on neutron-create-keystone-service/templates/neutron-create-keystone-service.yaml: error converting YAML to JSON: yaml: mapping values are not allowed in this context | 18:24 |
| kfox1111 | I wonder if 2.0.2 is any better there... | 18:24 |
| portdirect | kfox1111: that means you have an error in your template somewhere | 18:24 |
| kfox1111 | heh. yeah. | 18:25 |
| kfox1111 | "general car fault light" | 18:25 |
| *** vilobhmm11 has joined #openstack-kolla | 18:25 | |
| portdirect | :D | 18:25 |
| *** vilobhmm111 has joined #openstack-kolla | 18:27 | |
| *** mgiles has quit IRC | 18:28 | |
| openstackgerrit | Justin Scott proposed openstack/kolla-kubernetes: WIP Helm-ize memcached https://review.openstack.org/405677 | 18:29 |
| *** vilobhmm11 has quit IRC | 18:29 | |
| *** Pavo has quit IRC | 18:30 | |
| kfox1111 | neutron-create-keystone-service/charts/kolla-common/templates/_common_create_keystone_service.yaml:15:20: executing "common_create_keystone_service" at <.Values.selector_key>: can't evaluate field selector_key in type interface {} | 18:32 |
| kfox1111 | uhhh.. thats really odd.... | 18:33 |
| qwang | kfox1111: I have a question about current helm packaging. eg. for keystone, why do we need to split keystone-api/keystone-endpoint/keystone-svc/etc into many charts? | 18:33 |
| kfox1111 | qwang: yeah. one per k8s object in microservices. | 18:33 |
| *** Pavo has joined #openstack-kolla | 18:34 | |
| jascott1 | qwang its for flexibility | 18:34 |
| qwang | not quite understand. could you give an example? | 18:35 |
| jascott1 | so they can be installed/upgraded individually | 18:35 |
| jascott1 | with helm | 18:35 |
| kfox1111 | its to allow for manual orchestration, and to allow the end user assemble custom archetectures. | 18:35 |
| jascott1 | ^ what he said ;) | 18:35 |
| kfox1111 | so, if they want to do one rabbit per major service, except share a rabbit between nova/neutron, they can do that. | 18:36 |
| kfox1111 | or cluster the nova/neutron rabbit, but make the others not. | 18:36 |
| kfox1111 | etc. | 18:36 |
| kfox1111 | they are building blocks of which a full system can be built out of. | 18:37 |
| qwang | I see. thanks jascott1 kfox1111 | 18:37 |
| kfox1111 | np | 18:37 |
| kfox1111 | ah. found the problem. | 18:38 |
| kfox1111 | man I really wish you could word wrap a with statement in gotl. :/ | 18:39 |
| kfox1111 | if you have a long dict definition, its hard to count them in pairs in that case. | 18:40 |
| openstackgerrit | sayantani proposed openstack/kolla: [WIP] Clean up kolla-ansible related files https://review.openstack.org/406429 | 18:41 |
| *** Serlex has quit IRC | 18:41 | |
| kfox1111 | "http://keystone:%!s(float64=35357)" | 18:42 |
| kfox1111 | heh.... thats not right... | 18:42 |
| *** mkoderer has quit IRC | 18:46 | |
| *** tonanhngo has joined #openstack-kolla | 18:46 | |
| jascott1 | kfox1111 ya im surprised there is no solution for that | 18:46 |
| *** tummy has quit IRC | 18:47 | |
| kfox1111 | value: "http://keystone:35357.000000" | 18:48 |
| kfox1111 | hmm... almost... :/ | 18:48 |
| kfox1111 | int maybe... | 18:49 |
| mliima_ | hey guys | 18:50 |
| sbezverk_ | portdirect: ping | 18:50 |
| mliima_ | when i ran showmount -e within cinder_volume container a "get clnt_create: RPC: Unknown host | 18:51 |
| mliima_ | " | 18:51 |
| kfox1111 | uh.... | 18:51 |
| kfox1111 | function "int" not defined | 18:51 |
| kfox1111 | thats weird.... | 18:51 |
| kfox1111 | its in the sprig reference... | 18:51 |
| sbezverk_ | kfox1111: was there any changes done for neutron? | 18:51 |
| portdirect | sbezverk_ pong | 18:52 |
| sbezverk_ | neutron-server? it got broken on my last run | 18:52 |
| sbezverk_ | Error: release neutron-server failed: unable to decode "": Object 'Kind' is missing in '{"template":"no template \"common_deployment\" associated with template \"gotpl\""}' | 18:52 |
| sbezverk_ | portdirect: is it yours? | 18:52 |
| portdirect | kfox1111: nothing I've done has been merged | 18:52 |
| kfox1111 | sbezverk_: I don't think so? | 18:52 |
| portdirect | sry sbezverk ^^ | 18:52 |
| kfox1111 | what broke? got a link? | 18:52 |
| kfox1111 | oh. | 18:53 |
| sbezverk_ | kfox1111: sure http://logs.openstack.org/65/408165/6/check/gate-kolla-kubernetes-deploy-centos-binary-external-ovs-nv/e04fc23/console.html | 18:53 |
| portdirect | sbezverk_: my stuff has failed at the smae place - I thought because of changes I'd made but perhaps there is something else going on? | 18:54 |
| kfox1111 | the last thing merged wias hel-ming Rabbitmq | 18:55 |
| *** sayantan_ has quit IRC | 18:56 | |
| *** sayantan_ has joined #openstack-kolla | 18:57 | |
| portdirect | sbezverk_: thats logs from my ps you posted? | 18:57 |
| kfox1111 | sbezverk_: that one is a bug in portdirect's ps. | 18:58 |
| kfox1111 | neutron-server needs to include the renamed macro. | 18:58 |
| portdirect | will fix - sorry didnt realise you were using that ps | 18:59 |
| kfox1111 | there we go. %1.f did the trick. | 19:00 |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Change _common_deployment.yaml to _common_api_python_deployment.yaml https://review.openstack.org/408165 | 19:05 |
| sbezverk_ | portdirect: what I do not get is how not merged PS imapcted mine not merged PS, it is amistery | 19:06 |
| openstackgerrit | sayantani proposed openstack/kolla: [WIP] Clean up kolla-ansible related files https://review.openstack.org/406429 | 19:07 |
| kfox1111 | sbezverk_: 100% sure you were looking at the right logs? as you pasted portdirect's. | 19:07 |
| sbezverk_ | portdirect: have you read my comment on common deployment PS? | 19:07 |
| kfox1111 | second issue could be a bug in zuuel. I believe it does some very clever things to merge patches for testing. | 19:07 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service creation job https://review.openstack.org/408263 | 19:08 |
| portdirect | sbezverk_: never underestimate the power of my bugs :) | 19:09 |
| portdirect | sbezverk_: you mean here? https://review.openstack.org/#/c/408165/ | 19:09 |
| sbezverk_ | portdirect: yep | 19:10 |
| portdirect | sbezverk_: I'm not quite sure I follow? the only template that uses this is neutron server whihc should be fixed in the latest ps. | 19:10 |
| portdirect | sbezverk_: though I'm probably missing something? | 19:10 |
| sbezverk_ | portdirect: it is nt clear for me too, your ps has not been merged, so nothing could have changed common_deployment in the master | 19:11 |
| sbezverk_ | why all of a sudden it started misbehaving?? | 19:12 |
| portdirect | you have a log of your ps thats failing in zuul? | 19:12 |
| kfox1111 | yeah. would like to see the log myself. | 19:12 |
| *** topol has left #openstack-kolla | 19:13 | |
| sbezverk_ | http://logs.openstack.org/65/408165/6/check/gate-kolla-kubernetes-deploy-centos-binary-external-ovs-nv/e04fc23/console.html | 19:13 |
| sbezverk_ | portdirect: ^^ | 19:14 |
| kfox1111 | sbezverk_: wrong ps. | 19:14 |
| portdirect | could you post a link the the ps thats generating that log - as thats the log from my last ps | 19:14 |
| kfox1111 | thats https://review.openstack.org/#/c/408165/ | 19:14 |
| sbezverk_ | oh man. maybe it is just me going crazy %() | 19:15 |
| kfox1111 | :) | 19:15 |
| kfox1111 | I've done it before myself. | 19:15 |
| kfox1111 | thats why I double check the changeset id from the logs against the review from time to time. | 19:15 |
| kfox1111 | really easy to get tabs mixed up in the browser if your doing multiple revies at once. | 19:16 |
| portdirect | np - I've been stuggeling with too many ongoing projects today as well :) | 19:16 |
| *** fragatin_ has quit IRC | 19:16 | |
| sbezverk_ | portdirect: appologties, too many damn windows in the browser | 19:16 |
| portdirect | sbezverk_: absoultily no problem dude :) | 19:17 |
| *** krtaylor has joined #openstack-kolla | 19:17 | |
| * portdirect marks sbezverk_ on the naughty list :) | 19:17 | |
| *** sayantan_ has quit IRC | 19:17 | |
| sbezverk_ | portdirect: yep, no christmas gifts for me!! | 19:18 |
| kfox1111 | portdirect: you better put me on the same list then. ;) | 19:18 |
| portdirect | nah your both cool :) | 19:19 |
| *** JoseMello has quit IRC | 19:19 | |
| * portdirect marks kevin on the *really* naughy list | 19:20 | |
| kfox1111 | hehe | 19:20 |
| kfox1111 | he's on to me.... | 19:20 |
| kfox1111 | portdirect: sbezverk_: going to make it to the ptg? | 19:20 |
| portdirect | dont think so :( though I'm trying to pull together time/funds | 19:22 |
| sbezverk_ | kfox1111: probably not, it would require sdake to my manager, so she would let me go. do not know if he agrees to do it ;-) | 19:22 |
| sbezverk_ | I meant to talk to my manager | 19:22 |
| kfox1111 | bummer. :/ | 19:22 |
| sdake | 2hq5/ | 19:22 |
| sdake | WHAT? | 19:22 |
| sdake | i am not a people manager | 19:22 |
| sdake | ok - sorry catching up | 19:23 |
| sdake | sbezverk_ lets take that internal | 19:23 |
| openstackgerrit | Joshua Harlow proposed openstack/kolla: Allow for externally managed configuration https://review.openstack.org/332590 | 19:24 |
| kfox1111 | sdake: +1 to trying to get sbezverk_ to the ptg. :) | 19:24 |
| *** srwilkers_lunch is now known as srwilkers_ | 19:24 | |
| kfox1111 | I'm stil trying to get arangments myself, but no roadblocks yet. so looking hopeful. | 19:25 |
| srwilkers_ | as long as nothing drastic changes before then, i think i should be attending | 19:25 |
| kfox1111 | srwilkers_: nice. :) | 19:25 |
| sbezverk_ | sdake: right but you can convience her ;-) | 19:27 |
| portdirect | without the commerical aspect of the summit its a lot harder for me to justify unfortunatly :/ | 19:27 |
| srwilkers_ | portdirect, i was curious if that was going to affect some individual's ability to attend | 19:27 |
| srwilkers_ | sad to see it being the case | 19:27 |
| *** fragatina has joined #openstack-kolla | 19:29 | |
| openstackgerrit | Joshua Harlow proposed openstack/kolla: Allow for externally managed configuration https://review.openstack.org/332590 | 19:29 |
| kfox1111 | portdirect: I was abje to argue that the commertial aspects of it are directly related to ongoing development. and if development isn't assisted with our needed feedback, we may suffer. | 19:29 |
| mliima_ | somebody knows why showmout -e fails when I ran within cinder_volume container? | 19:32 |
| sbezverk_ | kfox1111: I hit interesting issue. do you have a comple of minutes to chat? | 19:33 |
| kfox1111 | sure. whats up? | 19:33 |
| *** jmccarthy has quit IRC | 19:34 | |
| *** jmccarthy has joined #openstack-kolla | 19:34 | |
| sbezverk_ | kfox1111: we use env variable to pass password info when a service gets its DB created. works fine for all services since it is passed as $password var and bash can substitute it with right value | 19:35 |
| kfox1111 | right | 19:36 |
| sbezverk_ | but for mariadb we do not have command where we can use it | 19:36 |
| kfox1111 | one sec. | 19:36 |
| sbezverk_ | we just pass env var with password value | 19:36 |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Change _common_deployment.yaml to _common_api_python_deployment.yaml https://review.openstack.org/408165 | 19:37 |
| sbezverk_ | it seems when I try to instantiate this env variable from secret mapped to nev var directly. it does not pick it up | 19:37 |
| kfox1111 | someone showed up in my office. will read in 2 min. | 19:38 |
| sbezverk_ | http://paste.openstack.org/show/591713/ | 19:39 |
| *** jascott1 has quit IRC | 19:39 | |
| *** jascott1 has joined #openstack-kolla | 19:39 | |
| *** fragatina has quit IRC | 19:40 | |
| openstackgerrit | Justin Scott proposed openstack/kolla-kubernetes: WIP Helm-ize memcached https://review.openstack.org/405677 | 19:43 |
| kfox1111 | ok. back. | 19:44 |
| kfox1111 | sry. | 19:44 |
| kfox1111 | that part should work. I'm more wondering if its getting munged on the way. | 19:45 |
| kfox1111 | have the review thats a part of link handy? | 19:45 |
| kfox1111 | more context might help. | 19:46 |
| *** sayantan_ has joined #openstack-kolla | 19:46 | |
| *** Bico_Fino has quit IRC | 19:46 | |
| kfox1111 | I'm kind of in a similar space right now, doing the keystone endpoint creation stuff. | 19:46 |
| kfox1111 | so we're probably going to run into the same issue too. | 19:46 |
| *** sayantan_ has quit IRC | 19:47 | |
| sbezverk_ | one sec | 19:47 |
| sbezverk_ | https://review.openstack.org/408160 | 19:48 |
| *** lamt has quit IRC | 19:49 | |
| kfox1111 | thx | 19:50 |
| kfox1111 | ah. thats using the bootstrap kolla flag. | 19:51 |
| kfox1111 | I'd expect that to work as is. | 19:52 |
| sbezverk_ | yes me too | 19:52 |
| portdirect | sbezverk_: weird I'm using the same statgy for db creation, but there it's getting the value and dumping it into a shell fine | 19:52 |
| kfox1111 | is it failing in the gate? | 19:52 |
| sbezverk_ | for mariadb there is no shell | 19:52 |
| kfox1111 | its gota be going thorugh a shell, as its KOLLA_BOOTSTRAP="" | 19:53 |
| *** mliima_ has quit IRC | 19:53 | |
| *** kristian__ has quit IRC | 19:54 | |
| sbezverk_ | for ketstone you use $var but for mariadb these no var to pass to command line | 19:54 |
| kfox1111 | sbezverk_: its passed directly through the var DB_ROOT_PASSWORD | 19:55 |
| sbezverk_ | as there is no command line boot strap sits in the container | 19:55 |
| sbezverk_ | ok my be i am missing something else will dig more thanks | 19:55 |
| kfox1111 | sbezverk_: it goes through the default command in the container. | 19:55 |
| kfox1111 | which ends up going through this: | 19:55 |
| kfox1111 | https://github.com/openstack/kolla/blob/master/docker/mariadb/extend_start.sh | 19:55 |
| *** fragatina has joined #openstack-kolla | 19:58 | |
| qwang | sbezverk_: so the password is stored in a k8s "secret" object already? | 20:00 |
| *** msimonin has joined #openstack-kolla | 20:01 | |
| kfox1111 | qwang: yeah. | 20:02 |
| kfox1111 | https://github.com/openstack/kolla-kubernetes/blob/master/tools/setup_gate.sh#L121 | 20:04 |
| sbezverk_ | qwang: sorry needed to step out; yeas in jinja2 version all services except mariadb used kube secret objects | 20:06 |
| *** msimonin1 has joined #openstack-kolla | 20:06 | |
| *** msimonin has quit IRC | 20:08 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP Hel-ming mariadb components https://review.openstack.org/408160 | 20:09 |
| *** msimonin1 has quit IRC | 20:10 | |
| *** kollabot3 has quit IRC | 20:12 | |
| *** kollabot has joined #openstack-kolla | 20:12 | |
| kfox1111 | ugg.... | 20:16 |
| kfox1111 | is openstack unified cli broken in kolla toolbox? | 20:16 |
| kfox1111 | http://logs.openstack.org/63/408263/1/check/gate-kolla-kubernetes-deploy-ubuntu-binary-ceph-nv/ca765a6/logs/pods/kolla-neutron-create-keystone-service-1xw38-main.txt | 20:17 |
| *** mgiles has joined #openstack-kolla | 20:18 | |
| sbezverk_ | kfox1111: you do not need to quote description? | 20:18 |
| kfox1111 | sbezverk_: its quoted already. the bash logger unquotes before printing. very dumb. | 20:18 |
| sbezverk_ | kfox1111: hmm confusing a bit .. | 20:19 |
| kfox1111 | it looks like this though: https://ask.openstack.org/en/question/98071/openstack-help-command-throwing-error-argument-of-type-osc_config-is-not-iterable/ :/ | 20:19 |
| sbezverk_ | kfox1111: it looks like they need to update openstack client pip install --upgrade python-openstackclient | 20:20 |
| kfox1111 | may be. :/ | 20:21 |
| kfox1111 | I would have thought that openstack cli would have come from pacakages though, | 20:21 |
| sbezverk_ | kfox1111: you use 3.0.1 tag for toolbox right? | 20:21 |
| kfox1111 | and both ubuntu and centos are affected. | 20:21 |
| sbezverk_ | kfox1111: I saw not exactly the same but similar issue, after upgrade everything worked automagically | 20:22 |
| kfox1111 | yeah. 3.0.1 | 20:22 |
| kfox1111 | darn. we may need a 3.0.2 to fix it. :/ | 20:22 |
| kfox1111 | and, this is why I want revisions. :/ | 20:22 |
| kfox1111 | 3.0.1-1 | 20:23 |
| kfox1111 | 3.0.1-2 | 20:23 |
| kfox1111 | hmm.... /usr/bin/openstack is not from a package. | 20:24 |
| kfox1111 | arg.... and the container's set to run as user 'ansible' | 20:26 |
| kfox1111 | :/ | 20:26 |
| *** schwicht has quit IRC | 20:26 | |
| kfox1111 | and can't sudo. :/ | 20:26 |
| kfox1111 | the other alternative is to skip the toolbox and use a different container. :/ | 20:27 |
| kfox1111 | ok... upgrade did fix it. :/ | 20:28 |
| kfox1111 | so, I need to force the container to run as root, then do an upgrade in it, until we can get a fixed container. :/ | 20:28 |
| *** Pavo has quit IRC | 20:30 | |
| *** Pavo has joined #openstack-kolla | 20:34 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 20:34 |
| openstackgerrit | sayantani proposed openstack/kolla: [WIP] Clean up kolla-ansible related files https://review.openstack.org/406429 | 20:38 |
| *** harlowja has quit IRC | 20:41 | |
| sdake | ok kfox1111 | 20:42 |
| sdake | got clarity on the relesae model | 20:42 |
| sdake | Jeffrey4l ^^ | 20:42 |
| kfox1111 | yeah? whats the deal? :) | 20:42 |
| sdake | i'll send the log to the mailing list | 20:42 |
| sdake | with a TLDR | 20:42 |
| sdake | the log is for context | 20:42 |
| sdake | sending email now | 20:42 |
| kfox1111 | k. cool. | 20:42 |
| *** diogogmt has joined #openstack-kolla | 20:43 | |
| sdake | i misunderstood release independent and we have been tracking our schedule according to release-cycle-trailing | 20:43 |
| sdake | read the log plz - it will help you understand | 20:43 |
| sdake | actually I cant do it now | 20:43 |
| *** kbyrne has quit IRC | 20:43 | |
| sdake | i have 1;1 | 20:44 |
| kfox1111 | waiting for it to show. :) | 20:44 |
| sdake | will do after i pick up my kids | 20:44 |
| kfox1111 | k. | 20:44 |
| sdake | if you want a sneak peak check out the logs for openstack-release on eavesdrop.openstack.org | 20:44 |
| sdake | just serach for kolla | 20:44 |
| sdake | it just happened now | 20:44 |
| *** kbyrne has joined #openstack-kolla | 20:45 | |
| sbezverk_ | kfox1111: man, this release name is killing me :-( in my test bed, I deleted mariadb-pvc release name, helm show it as delete, but when I try to install new one using the same name it fails saying that release name is already used, it is freaking deleted!!! | 20:46 |
| jascott1 | sbezverk_ you have todo purge | 20:46 |
| *** lamt has joined #openstack-kolla | 20:46 | |
| sbezverk_ | jascott1: really ?!?!?! | 20:47 |
| jascott1 | helm delete --purge | 20:47 |
| sbezverk_ | it sounds like windows now, are you sure you want to delete it?? | 20:47 |
| portdirect | sbezverk_: as jascott1 says you need to purge a release, this is a nightmare for development but will be quite nice for a production system | 20:47 |
| jascott1 | lol | 20:47 |
| kfox1111 | sbezverk_: are you on 2.0.1+? | 20:48 |
| kfox1111 | I think 2.0.0 and below had a delete bug. | 20:48 |
| *** schwicht has joined #openstack-kolla | 20:48 | |
| kfox1111 | jascott1: purge? | 20:48 |
| kfox1111 | oh, really? ugg. | 20:48 |
| jascott1 | ya so if you delete a release, it stays in history and name is taken unless you use --purge | 20:48 |
| kfox1111 | I wonder why they did that. | 20:48 |
| jascott1 | i think they were really thinking more like docker auto gen'd names where you dont care about the name but idk | 20:49 |
| kfox1111 | is there an undo delete? :) | 20:49 |
| portdirect | it's because delete doesnt really delete it but simpley remove the dolyment from the cluster - so I think its possible to undelete afterwards (so again we are in the world of things not being quite what they say on the tin, but i digress...) | 20:49 |
| jascott1 | i havent tried undo | 20:49 |
| sbezverk_ | kfox1111: I am on 2.0.2, but man purge after delete, give me a break :-( | 20:50 |
| kfox1111 | sbezverk_: yeah. thats kind of lame. | 20:50 |
| kfox1111 | sbezverk_: something we should bring up with the helm devs. | 20:50 |
| jascott1 | seems to me it should be more like image naming where you have a :<version> on the end | 20:51 |
| portdirect | jascott1: I think thats the idea | 20:51 |
| kfox1111 | sdake: thanks for the link. thats the outcome I was hoping for. | 20:51 |
| sbezverk_ | kfox1111: I mean we can always just unconditionally add --purge keyword | 20:51 |
| kfox1111 | sdake: and ofr asking the questions. | 20:52 |
| kfox1111 | sbezverk_: yeah, thats what I'd recommmend for now. as thats the behavior I was expecting out of the box. | 20:52 |
| sbezverk_ | jascott1: hopefully you have answer and for this one. when pvc ghets deleted, it does not seem to get properly disconnected from used PV | 20:53 |
| sbezverk_ | so next pvc get stuck in pending state | 20:53 |
| sbezverk_ | have you seen something like this? | 20:53 |
| sbezverk_ | kfox1111: I will add --purge to all helm delete I used in ceph workflow | 20:54 |
| kfox1111 | sbezverk_: ok. sounds like a plan. I'll do the same. | 20:54 |
| *** kristian__ has joined #openstack-kolla | 20:56 | |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Change _common_deployment to _common_api_python_deployment https://review.openstack.org/408165 | 20:56 |
| portdirect | kfox1111/sbezverk_: out of interest why do we delete jobs after they have run? | 20:57 |
| kfox1111 | portdirect: no reason to leave them running. | 20:57 |
| jascott1 | sbezverk_ sorry nowhere near ceph competent yet :/ isnt that lock issue well known tho or this different? | 20:57 |
| *** kristian__ has quit IRC | 20:58 | |
| portdirect | kfox1111: but after they have run they are not running anymore - but we are deleting the record of them having run - which will break dependancy checking using kube-entrypoint (or whatever its called now) and make operator(vessels) harder to write i think. | 20:59 |
| kfox1111 | portdirect: yeah, leaveing them will be important for kube-entrypoint deps. | 21:00 |
| kfox1111 | I don't plan on using them, | 21:00 |
| kfox1111 | and I really don't want them on my system laying around. | 21:01 |
| kfox1111 | seems like a loaded gun. | 21:01 |
| kfox1111 | I plan to remove credentials needed by the job after they are run too. | 21:01 |
| kfox1111 | so maybe a little less of an issue. | 21:01 |
| portdirect | kfox1111: np, again just a difference of opinon :) | 21:01 |
| kfox1111 | yup. | 21:01 |
| kfox1111 | I don't really trust kube-entrypoint too far either. it requies a lot more things to have k8s credentials. | 21:02 |
| kfox1111 | the same issues can be solved in other ways without it. | 21:02 |
| kfox1111 | just my 2 cents. | 21:02 |
| *** sayantan_ has joined #openstack-kolla | 21:02 | |
| *** kristian__ has joined #openstack-kolla | 21:03 | |
| portdirect | removing the creds/screts will prevent them being able to run again - thats what I do with my k8s systems - so I maintain the visibility of what has happened on the cluster since its inception but remove the bullets from the guns :) | 21:03 |
| jascott1 | ugh "Error: transport is closing" | 21:03 |
| * jascott1 beats himself to death with an empty gun | 21:04 | |
| portdirect | yeah - as you know I'm not totaly sold on the kube-entrypoint either, but cleaning up like this in gate tests will prevent us using it there - so we'll prob need to review that when we get it added in | 21:04 |
| portdirect | jascott1: can you get logs from tiller as it dies? | 21:04 |
| kfox1111 | portdirect: I've always kind of assumed I'd be able to redeploy from a snapshot of the r unning k8s objects. | 21:04 |
| kfox1111 | jobs won't work that way though. | 21:04 |
| kfox1111 | portdirect: we can make deletion a testable thing. | 21:05 |
| jascott1 | portdirect its from the gate | 21:05 |
| klindgren | so coming back to this: https://review.openstack.org/#/c/332590/ We have ran into an issue where https://github.com/openstack/kolla-ansible/blob/master/ansible/roles/glance/templates/glance-api.json.j2 is actually missing files that need to be managed in our config. | 21:05 |
| portdirect | kfox1111: +1 on testing options :) | 21:06 |
| portdirect | jascott1: urgh - thats not good :/ | 21:06 |
| klindgren | like glance-swift.conf and glance-api-paste.ini | 21:06 |
| jascott1 | portdirect: ya if you get that you have to run recheck :( | 21:06 |
| *** kristian__ has quit IRC | 21:07 | |
| kfox1111 | jascott1: this started happening shortly after we started using helm. never saw it pre'helm. | 21:08 |
| portdirect | jascott1: weird that it fails for all of them | 21:08 |
| kfox1111 | jascott1: I'm still wondering if its actually helm related. still could be a race condition elsewhere though. hard to say. | 21:08 |
| jascott1 | i like me some golang, might take a stroll thru helm code again and see whats what | 21:09 |
| portdirect | though looking thorugh the logs its seems really weird that each one has failed with a different error: http://logs.openstack.org/77/405677/9/check/gate-kolla-kubernetes-deploy-ubuntu-binary-ceph-nv/6d15920/console.html#_2016-12-07_19_54_09_698264 | 21:11 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP Hel-ming mariadb components https://review.openstack.org/408160 | 21:12 |
| portdirect | http://logs.openstack.org/77/405677/9/check/gate-kolla-kubernetes-deploy-centos-binary-ceph-multi-nv/f436bc5/console.html#_2016-12-07_20_05_03_965062 | 21:12 |
| kfox1111 | this looks weird: | 21:12 |
| kfox1111 | http://logs.openstack.org/77/405677/9/check/gate-kolla-kubernetes-deploy-ubuntu-binary-ceph-nv/6d15920/console.html#_2016-12-07_19_54_09_547190 | 21:12 |
| kfox1111 | it tries creating mariadb and says it already exists. | 21:13 |
| kfox1111 | oh. this: | 21:13 |
| kfox1111 | http://logs.openstack.org/77/405677/9/check/gate-kolla-kubernetes-deploy-ubuntu-binary-ceph-nv/6d15920/console.html#_2016-12-07_19_54_03_182553 | 21:13 |
| jascott1 | maybe we can get helm upstream to give more info | 21:13 |
| sbezverk_ | kfox1111: we need to reboot the gate ;-) it is really really strange | 21:13 |
| kfox1111 | it is launching with both helm and kollakube | 21:13 |
| portdirect | kfoxx1111: that will do it | 21:14 |
| sbezverk_ | kfox1111: another possibility is the collateral from merge conflicts | 21:16 |
| sbezverk_ | if both versions were combined while resolving merge conflict then it could happen | 21:17 |
| sbezverk_ | to have both rabbitmq launched as helm and as old | 21:17 |
| *** kristia__ has joined #openstack-kolla | 21:18 | |
| kfox1111 | possible. though I think we've only been touching one set of objects and converting them at the same time. | 21:18 |
| srwilkers_ | hey kfox1111 sbezverk_: im going through the changes ive got watched. with helm, https://review.openstack.org/#/c/316850/5 isnt relevant anymore correct? | 21:19 |
| srwilkers_ | just wanted to be sure | 21:19 |
| kfox1111 | for every change to add a helm install, there should be a corisponding removing from kollakube | 21:19 |
| kfox1111 | sbezverk_: yeah, thats so old we probably just just abandon it and start over. | 21:19 |
| kfox1111 | sry. | 21:20 |
| srwilkers_ | your statement just cleared that up for me kfox1111. thanks | 21:20 |
| kfox1111 | srwilkers_: I meant | 21:20 |
| srwilkers_ | yeah, was going to suggest abandoning it | 21:20 |
| portdirect | think sbezverk_'s probably hit the nail on the head - jascott1 did you rebase? | 21:20 |
| openstackgerrit | sayantani proposed openstack/kolla: [WIP] Clean up kolla-ansible related files https://review.openstack.org/406429 | 21:20 |
| jascott1 | portdirect yes | 21:22 |
| portdirect | jascott1: think I've found it - just commented in your ps | 21:22 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 21:23 |
| jascott1 | portdirect that shouldnt be interfering with others tho right? | 21:23 |
| portdirect | jascott1: no but it explains why your gates are failing - as the ceph_workflow.sh you have is a combination of diffrent generations | 21:24 |
| jascott1 | ok thanks | 21:25 |
| jascott1 | i mustve foobar merge conflict | 21:25 |
| portdirect | np - I usually mess them up as well :) | 21:25 |
| sbezverk_ | :-) | 21:26 |
| sbezverk_ | why do you think I suggested it ;-) | 21:27 |
| sbezverk_ | been there done that | 21:27 |
| *** jascott1_ has joined #openstack-kolla | 21:27 | |
| *** jascott1 has quit IRC | 21:27 | |
| *** Jeffrey4l has quit IRC | 21:35 | |
| openstackgerrit | Justin Scott proposed openstack/kolla-kubernetes: WIP Helm-ize memcached https://review.openstack.org/405677 | 21:35 |
| *** Jeffrey4l has joined #openstack-kolla | 21:35 | |
| v1k0d3n | sbezverk_: is it documented somewhere in kolla what the build options are exactly in kolla-build.conf | 21:35 |
| v1k0d3n | or anyone really | 21:37 |
| v1k0d3n | ? | 21:37 |
| openstackgerrit | Justin Scott proposed openstack/kolla-kubernetes: WIP Helm-ize memcached https://review.openstack.org/405677 | 21:37 |
| v1k0d3n | i have what i think should be in there...but...i think my flags are wrong | 21:37 |
| openstackgerrit | Justin Scott proposed openstack/kolla-kubernetes: WIP Helm-ize memcached https://review.openstack.org/405677 | 21:40 |
| jascott1_ | portdirect thanks for all the help | 21:41 |
| *** harlowja has joined #openstack-kolla | 21:41 | |
| jascott1_ | and srwilkers_ thanks! | 21:42 |
| srwilkers_ | jascott1_, happy to help, cheers :) | 21:42 |
| *** schwicht has quit IRC | 21:42 | |
| portdirect | jascott1_: np dude - fingers crossed it goes through this time :) | 21:46 |
| jascott1_ | ;) | 21:46 |
| kfox1111 | shouldn't keystone show up in the catalog? | 21:49 |
| jascott1_ | it should right? | 21:52 |
| kfox1111 | alittle weirded out by this one: | 21:52 |
| kfox1111 | https://review.openstack.org/#/c/408263/ | 21:52 |
| kfox1111 | http://logs.openstack.org/63/408263/3/check/gate-kolla-kubernetes-deploy-centos-binary-external-ovs-nv/ac98ceb/logs/ | 21:53 |
| kfox1111 | openstack catalog is showing up with nothing. | 21:53 |
| kfox1111 | and openstack service create is failing: http://logs.openstack.org/63/408263/3/check/gate-kolla-kubernetes-deploy-centos-binary-external-ovs-nv/ac98ceb/logs/pods/kolla-neutron-create-keystone-service-jtxsj-main.txt | 21:53 |
| kfox1111 | the keystone enndpoint creation job claims to have worked though. | 21:55 |
| *** yatin has quit IRC | 21:55 | |
| *** krtaylor has quit IRC | 21:58 | |
| *** rhallisey has quit IRC | 21:59 | |
| *** schwicht has joined #openstack-kolla | 22:01 | |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Keystone Helm Packaging https://review.openstack.org/407780 | 22:03 |
| portdirect | kfox1111: yeah keystone should def be in the catalog - I'll have a look and see if anything pops out | 22:04 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 22:05 |
| portdirect | kfox1111: running helm install with --debug may make debugging easier in the gate (I've been using it and 'kollakube template' to confim that helm is generating what I expect) | 22:05 |
| kfox1111 | portdirect: I put in a capture of the equiv previously. every pod running is in the pods.yaml file. | 22:07 |
| kfox1111 | and the kube display pod * is in pods/ | 22:08 |
| kfox1111 | I haven't needed debug then. | 22:08 |
| *** yatin has joined #openstack-kolla | 22:08 | |
| *** dave-mccowan has quit IRC | 22:09 | |
| kfox1111 | so.... complete tangent... | 22:10 |
| kfox1111 | it sounsd like it is possible to put zuul jobs in that cache things, and pull them from other jobs. | 22:10 |
| kfox1111 | trove looks to be doing it. | 22:10 |
| kfox1111 | so we can cache up stuf like pregen'ed /etc/kolla dirs. | 22:11 |
| portdirect | id be looking at python-openstackclient at a possible candiate here, verything else seems ok? :/ | 22:11 |
| kfox1111 | which should shave off a couple minutes at least off the gate. | 22:11 |
| kfox1111 | portdirect: yeah. I'm still suspecting it. | 22:11 |
| kfox1111 | portdirect: like, its itself throwing an error if it sees no non keystone entries. | 22:11 |
| kfox1111 | but: http://docs.openstack.org/liberty/install-guide-ubuntu/keystone-services.html | 22:11 |
| jascott1_ | yay memcached passed | 22:12 |
| kfox1111 | seems to make it sound like it should work. | 22:12 |
| *** kristia__ has quit IRC | 22:12 | |
| srwilkers_ | jascott1_, \o/ | 22:13 |
| kfox1111 | and the job does run after keystone-entry' job runs. | 22:13 |
| kfox1111 | so there should be some of those. | 22:13 |
| portdirect | kfox1111: i'd be temepted to try with a 'openstack service list' just to see how it reacts to that? | 22:13 |
| kfox1111 | as apaosed to catalog list? | 22:14 |
| kfox1111 | cause its doing the latter on gate https://github.com/openstack/kolla-kubernetes/blob/master/tests/bin/gate_capture_logs.sh#L59 | 22:15 |
| kfox1111 | and it came up blank. | 22:15 |
| portdirect | kfox1111: sry meant 'catalog list' | 22:15 |
| *** yatin has quit IRC | 22:16 | |
| portdirect | hmm | 22:16 |
| kfox1111 | hmmm.... | 22:17 |
| *** vilobhmm111 has left #openstack-kolla | 22:17 | |
| kfox1111 | one of the commands threw this though: 2016-12-07 21:43:35.408559 | Missing value auth-url required for auth plugin password | 22:18 |
| kfox1111 | not sure which though. | 22:18 |
| kfox1111 | so could be not caputring becuase of that. :/ | 22:18 |
| kfox1111 | yeah. looks like it came from there. | 22:19 |
| kfox1111 | probably because the rc file wasn't built yet... | 22:19 |
| *** bmace has quit IRC | 22:19 | |
| *** bmace has joined #openstack-kolla | 22:19 | |
| kfox1111 | yup... its built riht after enpoint creation... I should move that up.... | 22:20 |
| kfox1111 | oh... hmm... | 22:20 |
| kfox1111 | domains might be diffreent here.... | 22:21 |
| kfox1111 | local admin keystonerc has OS_USEDR_DOMAIN_NAME=Default. | 22:21 |
| kfox1111 | and that works. | 22:21 |
| kfox1111 | but I think the other value I saw was =default. | 22:21 |
| portdirect | thats what i suspected for 2 mins | 22:21 |
| portdirect | but usually keystone paps Default to default unless things have changed | 22:21 |
| portdirect | *maps | 22:21 |
| kfox1111 | well, I think its usually a little wierder then that. | 22:22 |
| kfox1111 | I think they use the id=default and the name=Default. | 22:23 |
| *** jheroux has quit IRC | 22:23 | |
| portdirect | the only real diff i can see is that the old one did not set the value for OS_PROJECT_DOMAIN_NAME/OS_USER_DOMAIN_NAME ? though I'm still parsing | 22:23 |
| kfox1111 | so if you don't specify you want specifically id or name, then either work. | 22:23 |
| portdirect | kfox1111: thats correct | 22:23 |
| portdirect | it was to help the migration from v2 to v3 | 22:23 |
| portdirect | so both should work unless python-openstackclient's being weird again | 22:24 |
| kfox1111 | arg... here's a helm crash... http://logs.openstack.org/63/408263/4/check/gate-kolla-kubernetes-deploy-centos-binary-ceph-multi-nv/4aaac3d/console.html | 22:24 |
| kfox1111 | ok. moved the catalog list up, and changed it to Default instead... lets see where this goes... | 22:26 |
| *** yatin has joined #openstack-kolla | 22:28 | |
| *** Pavo has quit IRC | 22:29 | |
| portdirect | no helm debugging without a working k8s api - thats no fun :( | 22:31 |
| portdirect | i just want to render a chart... | 22:31 |
| *** fguillot has quit IRC | 22:32 | |
| *** jtriley has quit IRC | 22:33 | |
| srwilkers_ | nothing more embarrassing than not refreshing the page and doing a review on a patch set thats hours old | 22:34 |
| srwilkers_ | oops | 22:34 |
| jascott1_ | yup done that a couple times this week but one case it was only minutes old | 22:34 |
| srwilkers_ | yeahhh | 22:34 |
| *** Pavo has joined #openstack-kolla | 22:34 | |
| v1k0d3n | hey portdirect and srwilkers_ soooo....ceph for halcyon. | 22:35 |
| portdirect | v1k0d3n: yep | 22:35 |
| v1k0d3n | that's working for you, right? | 22:35 |
| kfox1111 | portdirect: yeah. that bites me hard. I develop kolla-kubernetes on a secondary user. so I | 22:36 |
| kfox1111 | often have to build the package, then helm install --debug it from a different user. | 22:36 |
| kfox1111 | as it must have a live tiller to work against. :/ | 22:36 |
| portdirect | v1k0d3n: yeah i've had it working, both with your aic-helm stuff, ceph upstream and kollakube on centos based halycon | 22:37 |
| *** sdake has quit IRC | 22:37 | |
| portdirect | v1k0d3n: though not yet made the progress I'd like on kolla-helm | 22:37 |
| v1k0d3n | portdirect: please take whatever you want out of our repo | 22:37 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 22:37 |
| v1k0d3n | srwilkers_ is dedicated to doing this. prob could use help with how quickly things are going...and we mean for that all to just be in kolla at some point to | 22:38 |
| kfox1111 | portdirect: kolla-helm? | 22:38 |
| portdirect | kfox1111: you know what I mean :) | 22:38 |
| v1k0d3n | don't want to have a "forever" poc out there. we're telling everyone else to do in kolla-kube too. | 22:38 |
| v1k0d3n | not everyone has execs saying "NOW!" :) | 22:39 |
| kfox1111 | portdirect: ok. just handn't heard the term. was wondiering if it was something new. :) | 22:39 |
| kfox1111 | v1k0d3n: :) | 22:39 |
| jascott1_ | bb 15m | 22:39 |
| *** jascott1_ has quit IRC | 22:39 | |
| *** jascott1 has joined #openstack-kolla | 22:40 | |
| openstackgerrit | Joshua Harlow proposed openstack/kolla: Refactor copy so that it shows recursive files copied https://review.openstack.org/408336 | 22:40 |
| v1k0d3n | kolla-helm? did i miss something big. | 22:41 |
| kfox1111 | if we could precache the genconfig, the kubernetes packages and some of the containers, we might be able to shave off 8-10 minutes at least off the gate. | 22:41 |
| kfox1111 | v1k0d3n: I don't think so. | 22:41 |
| kfox1111 | I think he was just reffering to the kolla-kubernetes helm stuff. | 22:42 |
| *** krtaylor has joined #openstack-kolla | 22:42 | |
| openstackgerrit | sayantani proposed openstack/kolla: Clean up kolla-ansible related files from Kolla https://review.openstack.org/406429 | 22:42 |
| *** absubram has quit IRC | 22:43 | |
| *** jascott1_ has joined #openstack-kolla | 22:44 | |
| *** jascott1 has quit IRC | 22:44 | |
| srwilkers_ | off to the office holiday party -- be back later this evening :) | 22:45 |
| kfox1111 | have fun. :) | 22:45 |
| *** srwilkers_ is now known as srwilkers_away | 22:45 | |
| *** jascott1_ has quit IRC | 22:46 | |
| *** srwilkers_away has quit IRC | 22:47 | |
| *** krishnaposa has quit IRC | 22:47 | |
| portdirect | v1k0d3n: yeah was just abberviating what kfox111 just said - sorry if it created confusion | 22:48 |
| v1k0d3n | oh yeah | 22:48 |
| v1k0d3n | actually i like it! | 22:48 |
| v1k0d3n | lol | 22:48 |
| *** ayoung has quit IRC | 22:48 | |
| v1k0d3n | so question... | 22:48 |
| *** lamt has quit IRC | 22:48 | |
| v1k0d3n | (oh boy...here it comes...are you all sitting?) | 22:49 |
| v1k0d3n | you are... | 22:49 |
| v1k0d3n | all...sitting, right? | 22:49 |
| * portdirect stands up | 22:49 | |
| v1k0d3n | crap | 22:49 |
| * v1k0d3n waits | 22:49 | |
| * portdirect sits back down | 22:49 | |
| v1k0d3n | anyway....we're planning on including this upstream in kubernetes/charts at the end stage, right? | 22:49 |
| * v1k0d3n jokes too much | 22:49 | |
| v1k0d3n | BUT NOT ABOUT THAT /me referring to portdirect | 22:50 |
| kfox1111 | possibly I think. if kubernetes/chart's build system can mature a bit or helm grows enough support for it. | 22:50 |
| v1k0d3n | kubernetes/charts makes the most sense. that's what we talked about a while back | 22:50 |
| *** schwicht has quit IRC | 22:50 | |
| v1k0d3n | ok...cool. i just wanted to make sure. | 22:50 |
| v1k0d3n | what do you mean "grows enough"? | 22:51 |
| kfox1111 | curently, their build system assumes a single build at a time, and requires binaries to be pulled for deps, | 22:51 |
| kfox1111 | even if the source is right in the same repo. | 22:51 |
| kfox1111 | so makes it really hard to build the whole repo. :/ | 22:51 |
| v1k0d3n | you mean have a master chart that deploys, and still deploy each individual components right? the "helm of helms" if you will? | 22:51 |
| kfox1111 | no, like, | 22:51 |
| portdirect | v1k0d3n: yeah that would be the ideal - possibly we could create meta charts for there that pull in carts direct from here? | 22:51 |
| kfox1111 | the you have say a wiki chart. it depends on mariadb. | 22:51 |
| v1k0d3n | portdirect: so we're doing that. that's our goal. | 22:52 |
| kfox1111 | they have a dep on mariadb that points directly to google's binary helm repo with a prebuilt chart. | 22:52 |
| kfox1111 | so you can't update the mariadb chart, and then rebuild all the things that its embeded in easily. :/ | 22:52 |
| v1k0d3n | that's what we're currently solving fwiw. | 22:52 |
| kfox1111 | (think, mariadb security issue) | 22:52 |
| kfox1111 | yeah. | 22:52 |
| v1k0d3n | any outliers we want to start opening up issues, but this can be done currently today. | 22:53 |
| portdirect | i think this is where the operator/vessel will come into its own | 22:53 |
| *** lrensing has quit IRC | 22:53 | |
| kfox1111 | operators/vessels/orchestrators/whatever are a different thing. | 22:53 |
| kfox1111 | the thingy would be in this build repo as well. | 22:53 |
| v1k0d3n | don't use operator for that. | 22:53 |
| v1k0d3n | just sayng | 22:53 |
| v1k0d3n | operators are unicorns. | 22:53 |
| * v1k0d3n wishes i didn't bring up operators. | 22:54 | |
| kfox1111 | operators shouldn't be building new packages. only laucnhing existing/tested ones Ithink. | 22:54 |
| kfox1111 | (with a possible exception of a config pachage... haven't really throught through that one) | 22:54 |
| v1k0d3n | operators should repair things if/when they go down...in place of a human operator. | 22:54 |
| v1k0d3n | ex... | 22:54 |
| kfox1111 | package | 22:54 |
| portdirect | i was meaning to manage the issue that kfox1111 reased re package management - I would not want them building new ones | 22:54 |
| v1k0d3n | galera cluster member goes down. there are predefined steps that need to occur, and can be leveraged to typically bring back a member. | 22:55 |
| kfox1111 | yeah. | 22:55 |
| kfox1111 | right. | 22:55 |
| kfox1111 | I think helm needs ralitive dependency support, which would go a long way in solving the issue. | 22:56 |
| kfox1111 | some kind of value dependency handling too. | 22:57 |
| kfox1111 | "merge child's values into my values, instead of hanging them of of globals or a child key" | 22:57 |
| v1k0d3n | kfox1111 you've been keeping up with alan? have you run the pic? | 22:57 |
| v1k0d3n | poc | 22:58 |
| v1k0d3n | just want to stay close | 22:58 |
| v1k0d3n | tbh...i do NOT want to stray or have to stray. | 22:58 |
| kfox1111 | havent seen him in a while. | 22:58 |
| v1k0d3n | please don't make us stray... | 22:58 |
| v1k0d3n | well, we're busy. | 22:58 |
| kfox1111 | yeah. I get that. | 22:58 |
| v1k0d3n | we need srwilkers to keep eyes on it :) | 22:58 |
| kfox1111 | I wana make sure you have what you need. | 22:58 |
| v1k0d3n | and ditto | 22:58 |
| * kfox1111 nods | 22:58 | |
| * v1k0d3n feels the love among this group | 22:59 | |
| v1k0d3n | ok cool. just keep checking (if you can/have time) to exchange | 22:59 |
| v1k0d3n | i just put our PM to create a kanban board out on github projects so the mission, goals, work...is all clearer. | 22:59 |
| kfox1111 | working like mad right now to just get the existing stuff over to helm asap. | 22:59 |
| v1k0d3n | so any eyes can seriously pluck work from one to the other | 22:59 |
| *** yatin has quit IRC | 23:00 | |
| kfox1111 | cool. | 23:00 |
| *** jascott1 has joined #openstack-kolla | 23:00 | |
| v1k0d3n | yeah, i get that. let srwilkers know where we can help. | 23:00 |
| * v1k0d3n shamelessly pluggin' steve in for more work | 23:00 | |
| kfox1111 | k | 23:01 |
| kfox1111 | basically thats where we're at though. just churning through the jinja2 -> helm stuff. so anything not currently in progress is fair game. | 23:02 |
| *** unicell has joined #openstack-kolla | 23:03 | |
| kfox1111 | weird... | 23:03 |
| kfox1111 | http://logs.openstack.org/63/408263/5/check/gate-kolla-kubernetes-deploy-ubuntu-binary-ceph-nv/ef2a866/logs/pods/kolla-neutron-create-keystone-endpoint-public-8tqdc-main.txt | 23:03 |
| kfox1111 | I think that one may lead some credence to the, dns is not ready when it says it is theory. | 23:04 |
| *** schwicht has joined #openstack-kolla | 23:06 | |
| kfox1111 | ok... so, got the end point catalog back. there's stuff in it... so openstack cli in the container's not quite right... seems like config probably... | 23:07 |
| jascott1 | kfox1111 wrt "{{- include "common_containers" . | indent 8 }}" and $env, it does exist in tree like in rabbit | 23:08 |
| kfox1111 | jascott1: k. we should fix rabbit then too. I think its wrong too. | 23:09 |
| jascott1 | should they all be $env and no "." ? | 23:09 |
| kfox1111 | if the with is defining $env, then we should pass $env. | 23:09 |
| jascott1 | got it thanks | 23:09 |
| openstackgerrit | Justin Scott proposed openstack/kolla-kubernetes: WIP Helm-ize memcached https://review.openstack.org/405677 | 23:11 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 23:11 |
| jascott1 | kfox1111 i think I got what you wanted back ^ | 23:11 |
| kfox1111 | yeah. looks better. | 23:12 |
| *** yatin has joined #openstack-kolla | 23:12 | |
| kfox1111 | can you please have a look at what resourceName is used for in common lib too? | 23:12 |
| kfox1111 | not sure if it needs to be unique per element. | 23:12 |
| jascott1 | ah crap forgot that one thanks | 23:12 |
| kfox1111 | and slide up a line 2 lines up. | 23:14 |
| kfox1111 | then I think its mergable. :) | 23:14 |
| *** harlowja has quit IRC | 23:16 | |
| kfox1111 | jascott1: looks like resourceName is used for the logger config name. | 23:16 |
| jascott1 | so make it element name? | 23:16 |
| kfox1111 | which, I don't know. coudl go either way I guess. | 23:16 |
| *** sdake has joined #openstack-kolla | 23:17 | |
| kfox1111 | would we want a different set of logging rules per the same type? I"m guessing not. | 23:17 |
| *** yatin has quit IRC | 23:17 | |
| sdake | kfox1111 no problem | 23:17 |
| *** schwicht has quit IRC | 23:17 | |
| kfox1111 | sdake: wb | 23:17 |
| sdake | kfox1111 lagging a bit here - wife out of town traveling - have kids on my own :) | 23:17 |
| kfox1111 | sdake: yeah. I know how that can be. :) | 23:17 |
| kfox1111 | no wories. | 23:18 |
| kfox1111 | jascott1: I think we're probably fine with resourceName. we can revisit it later I think. | 23:18 |
| sdake | and pretty soon have to jet to feed em and fix my father nad mother's computer | 23:18 |
| jascott1 | ok | 23:18 |
| kfox1111 | jascott1: so, just that one line move, and we're good Ithink. | 23:18 |
| openstackgerrit | Justin Scott proposed openstack/kolla-kubernetes: WIP Helm-ize memcached https://review.openstack.org/405677 | 23:19 |
| kfox1111 | oh. and an updated commit message. | 23:19 |
| kfox1111 | :) | 23:19 |
| jascott1 | drats! | 23:19 |
| kfox1111 | hehe | 23:19 |
| sdake | ok - gotta jet to feed the kids | 23:19 |
| sdake | bbiaf | 23:19 |
| pron | anyone familiar with maas? atm i am trying to figure out how to prepare disks for kolla ceph with maas ( gpt with labels ) | 23:19 |
| openstackgerrit | Justin Scott proposed openstack/kolla-kubernetes: Helm-ize memcached https://review.openstack.org/405677 | 23:19 |
| *** david-lyle_ has joined #openstack-kolla | 23:20 | |
| kfox1111 | pron: no, sorry. been actively avoiding it. | 23:20 |
| *** david-lyle_ has quit IRC | 23:20 | |
| pron | kfox1111: what do you use ? | 23:20 |
| *** sdake has quit IRC | 23:20 | |
| kfox1111 | just using cobbler for now, but would really like to get to ironic. | 23:20 |
| kfox1111 | as then I can make bare metal services more easily available to my users. | 23:21 |
| pron | oh my blood pressure went off instantly when i saw gui of cobbler | 23:21 |
| kfox1111 | and I don't have to switch between different software stacks. | 23:21 |
| kfox1111 | pron: heh. thats probably true. I don't use gui's much. | 23:22 |
| kfox1111 | I want to be able to "for x in seq 1 1000`; do cobbler system edit --name=cn$x ...." | 23:22 |
| pron | cant say cli worked for me ;D | 23:23 |
| pron | was erroring out about random shit | 23:23 |
| pron | anyways | 23:23 |
| kfox1111 | gui's are natirousyly bad at managing stuff at scale (usually :) | 23:23 |
| *** sacharya has quit IRC | 23:23 | |
| *** sacharya has joined #openstack-kolla | 23:24 | |
| pron | then again there are coworkers i would be able to deleagte some responsibilites if i had gui | 23:24 |
| kfox1111 | yeah. I get that. | 23:24 |
| kfox1111 | jascott1: workflowed. | 23:25 |
| *** JRobinson__ has joined #openstack-kolla | 23:26 | |
| jascott1 | woot i say! woot. | 23:26 |
| wirehead_ | Ah, there’s one problem I was about to mention with the memcached, kfox1111 / jascott1 | 23:27 |
| kfox1111 | wirehead_: oh, yeah? | 23:27 |
| * jascott1 shakes fist at sky | 23:28 | |
| kfox1111 | hehe | 23:28 |
| wirehead_ | We probably need to encourage the user to be careful with the memcached. | 23:28 |
| kfox1111 | wirehead_: generally true. :) | 23:28 |
| kfox1111 | but in which partuclar way are you concerned? | 23:28 |
| wirehead_ | Oh, so we’re still exposing memcached as a kube service, right? | 23:28 |
| kfox1111 | yeah. | 23:29 |
| wirehead_ | Okay, you can only have one memcached instance. :/ | 23:29 |
| *** yatin has joined #openstack-kolla | 23:29 | |
| kfox1111 | right. | 23:29 |
| kfox1111 | well, | 23:29 |
| wirehead_ | There’s a comment with no explanation in the etc/ dir | 23:29 |
| kfox1111 | one instance per service. | 23:29 |
| wirehead_ | Yah. | 23:29 |
| kfox1111 | jascott1: yeah, maybe a followon patch that just hardcodes the replica's to 1? | 23:30 |
| jascott1 | ok | 23:30 |
| kfox1111 | I can't think of a case where != would work or be desirable. | 23:31 |
| wirehead_ | Sorry ‘bout that. It was sticking in my head, but I’d wanted to make sure I knew the context and took a brief period of time today to do so… and then realized you’d literally just workflow+1’d it. | 23:31 |
| kfox1111 | the user can always launch multiple elements to gain that functionality. | 23:31 |
| kfox1111 | wirehead_: no worries. :) | 23:31 |
| kfox1111 | thanks for bringing it up. | 23:31 |
| wirehead_ | Left to my own devices, I’d either look into memcached proxies or just make the supported config use something that’s more amenable to kube usage. | 23:31 |
| kfox1111 | wirehead_: yeah. memcached's always made me a little unsettled. | 23:32 |
| kfox1111 | its has some big *'s in its documentation, left up as an excersize to the reader. | 23:32 |
| kfox1111 | and then most people just ignore it. | 23:32 |
| wirehead_ | Well, there was a day where memcached was the only game in town that wasn’t something secret and proprietary. | 23:32 |
| kfox1111 | yeah. | 23:33 |
| kfox1111 | even with the *'s, its quite usable. | 23:33 |
| kfox1111 | but I'd like to see each project that uses it, explain how they are meeting the requirements. | 23:33 |
| kfox1111 | cause if they get one little thing wrong.... :) | 23:33 |
| wirehead_ | There’s a few teams where I want to sit them down, ascertain if they understand what the word “linearizable” means, and then make me explain their concurrency model in terms of such. | 23:34 |
| kfox1111 | hmm... much further this time... almost there... | 23:34 |
| kfox1111 | wirehead_: hehe. yeah. | 23:34 |
| kfox1111 | but... but... you just shuvel stuf into memcached, and... fast! :) | 23:35 |
| wirehead_ | But, yah, as the maniac who crowbarred memcached to fit inside of a kube service the first time, well done, jascott1 :) | 23:36 |
| kfox1111 | I just threw up in my mouth a little. :) | 23:36 |
| jascott1 | lol | 23:36 |
| kfox1111 | helm package. | 23:36 |
| kfox1111 | its been in kolla-kubernetes for a while. | 23:37 |
| kfox1111 | in fact, I found a race condition in keystone with it on. :/ | 23:37 |
| kfox1111 | it caches endpoints and doesn't invalidate properly. :/ | 23:37 |
| wirehead_ | Yah. Dims made the service, I made it work as a kube service as part of the mannic hackery some months ago, and now jascott1 made it work as a helm package. :) | 23:37 |
| kfox1111 | so if you create endpoints in parallel too fast, it can miss some in subsequent commands. :/ | 23:37 |
| wirehead_ | Heh. | 23:38 |
| jascott1 | wirehead_ shoulders of giants over here but thanks | 23:38 |
| kfox1111 | I had to explicitly disable memcached in the gate. :/ | 23:38 |
| kfox1111 | memcached for keystone in the gate that is. | 23:38 |
| wirehead_ | Don’t think of yourself as sitting on the shoulders of giants. Think of yourself as being the viking-style helmet with horns that completes the badass look. | 23:38 |
| wirehead_ | :D | 23:38 |
| jascott1 | haha awesome | 23:39 |
| jascott1 | where would we be without our viking *helm* ? | 23:39 |
| kfox1111 | hehe | 23:40 |
| wirehead_ | I think I’m not a fan of blindly caching things in the data layer. | 23:40 |
| kfox1111 | oh.... so close: http://logs.openstack.org/63/408263/6/check/gate-kolla-kubernetes-deploy-centos-binary-ceph-nv/a79234b/logs/openstack-catalog.txt | 23:40 |
| kfox1111 | somehow region didn't make it. | 23:40 |
| kfox1111 | probably because I didn't specify it... but I would have expected it to use OS_REGION_NAME. | 23:41 |
| kfox1111 | hmm... | 23:41 |
| wirehead_ | Plus, with Redis the way it is, you can probably use some of the data structures instead of using a blind k-v cache anyway. | 23:41 |
| jascott1 | redis rocks | 23:42 |
| kfox1111 | redis is on my to look at list at some point. | 23:42 |
| jascott1 | hmm should prob just make the bot say that since someone always seems to say it | 23:42 |
| kfox1111 | killed a lot of time on various nosql stuff though. | 23:42 |
| jascott1 | ugh r | 23:42 |
| kfox1111 | mongodb turned out less useful then I had hoped. :/ | 23:42 |
| jascott1 | was interested and then all the nightmare came out about it | 23:43 |
| jascott1 | table locks | 23:43 |
| jascott1 | nice api tho | 23:43 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 23:43 |
| jascott1 | couch was cool but seem they couldnt get the x data center replication worked out | 23:44 |
| wirehead_ | The CMS I wrote for my website because I’m too obsessive to use Wordpress uses PostgreSQL + Redis. | 23:44 |
| kfox1111 | if this works, then it shoudl clean up the codee base a lot. :) | 23:44 |
| wirehead_ | For the most part, the requirement is that the Redis is either transient storage that it shouldn’t matter if it gets zapped or a cache atop postgresql. | 23:45 |
| jascott1 | great for tokens and such with the TTL built in | 23:46 |
| jascott1 | hey anyone got any ideas to make the bot useful? couple ppl mentioned it in the meeting this morning | 23:48 |
| *** sdake has joined #openstack-kolla | 23:50 | |
| jascott1 | i think it was srwilkers and ryan | 23:50 |
| openstackgerrit | Merged openstack/kolla-kubernetes: Helm-ize memcached https://review.openstack.org/405677 | 23:52 |
| jascott1 | yay! | 23:53 |
| wirehead_ | YAY! | 23:54 |
| jascott1 | Thank you everyone for all the help getting my first patch into the kolla world! | 23:54 |
| jascott1 | For real, this is a great community. Im very proud to be here. | 23:55 |
| *** schwicht has joined #openstack-kolla | 23:56 | |
| *** sdake has quit IRC | 23:57 | |
| *** sdake has joined #openstack-kolla | 23:57 | |
| kfox1111 | jascott1: congrats. :) | 23:59 |
| kfox1111 | thanks for sticking with it. :) | 23:59 |
| *** vhosakot has quit IRC | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!