| *** shashank_t_ has joined #openstack-kolla | 00:05 | |
| *** shashank_t_ has quit IRC | 00:09 | |
| *** m1dev has quit IRC | 00:10 | |
| *** eglute_s has quit IRC | 00:14 | |
| *** cloudnull has quit IRC | 00:14 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes master: WIP Adding virtualbmc to the gate https://review.openstack.org/441550 | 00:14 |
|---|---|---|
| *** cloudnull has joined #openstack-kolla | 00:14 | |
| *** eglute has joined #openstack-kolla | 00:15 | |
| *** tonanhngo has joined #openstack-kolla | 00:19 | |
| *** tonanhngo has quit IRC | 00:22 | |
| *** papacz has quit IRC | 00:29 | |
| *** rwallner has joined #openstack-kolla | 00:32 | |
| *** rwallner has quit IRC | 00:36 | |
| *** duonghq has joined #openstack-kolla | 00:41 | |
| duonghq | morning guys | 00:41 |
| *** crushil has joined #openstack-kolla | 00:41 | |
| *** harlowja has joined #openstack-kolla | 00:42 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes master: WIP Adding virtualbmc to the gate https://review.openstack.org/441550 | 00:47 |
| *** MasterOfBugs has quit IRC | 00:48 | |
| *** tonanhngo has joined #openstack-kolla | 00:51 | |
| *** dave-mcc_ has joined #openstack-kolla | 00:52 | |
| *** tovin07 has quit IRC | 00:52 | |
| *** zhurong has joined #openstack-kolla | 00:53 | |
| *** dave-mccowan has quit IRC | 00:53 | |
| *** tonanhngo has quit IRC | 00:55 | |
| *** tovin07_ has joined #openstack-kolla | 00:58 | |
| spsurya | duonghq: morning | 01:03 |
| *** deadnull has quit IRC | 01:05 | |
| duonghq | morning spsurya | 01:10 |
| *** qiliang28 has joined #openstack-kolla | 01:12 | |
| *** qiliang27 has quit IRC | 01:13 | |
| *** qiliang28 is now known as qiliang27 | 01:13 | |
| *** tonanhngo has joined #openstack-kolla | 01:25 | |
| *** tonanhngo has quit IRC | 01:26 | |
| *** Margin has joined #openstack-kolla | 01:29 | |
| *** zhubingbing has joined #openstack-kolla | 01:35 | |
| masber | good morning | 01:44 |
| masber | I am installing kolla-ansible 4.0.0 using different network configurations, so far I can finish the installation and start using openstack, however my vms don't have access to the internet. I am assuming that the external traffic from the instances will go through neutron_external_interface because br-ex is mapped to that interface. Am I right? | 01:44 |
| Margin | right | 01:46 |
| *** qiliang28 has joined #openstack-kolla | 01:47 | |
| *** qiliang27 has quit IRC | 01:48 | |
| *** qiliang28 is now known as qiliang27 | 01:48 | |
| *** cuongnv has joined #openstack-kolla | 01:51 | |
| *** sayantani01 has quit IRC | 01:52 | |
| *** sayantani01 has joined #openstack-kolla | 01:52 | |
| *** sayantani01 has quit IRC | 01:53 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes master: WIP Adding virtualbmc to the gate https://review.openstack.org/441550 | 01:54 |
| *** tonanhngo has joined #openstack-kolla | 01:57 | |
| *** dave-mccowan has joined #openstack-kolla | 01:58 | |
| *** tonanhngo has quit IRC | 01:58 | |
| *** dave-mcc_ has quit IRC | 02:00 | |
| SamYaple | masber: it will go through br-ex on the node with the L3 router. not form the compute nodes (unless you are doing DVR or similiar) | 02:07 |
| masber | SamYaple, thank, you right now I am testing on a all-in-one environment to make things more simple and easier for me to understand | 02:09 |
| masber | right now, my server only has 1 nic, would it be possible to create a subinterface and map it to neutron_external_interface? | 02:13 |
| SamYaple | masber: kind of.... but its not easy to describe | 02:15 |
| SamYaple | you can't do it with a sub interface | 02:15 |
| SamYaple | it must be a veth pair | 02:15 |
| SamYaple | and its really not going to do what you want | 02:15 |
| SamYaple | best option is to setup the interface in openvswitch ahead of time | 02:15 |
| *** unicell has quit IRC | 02:16 | |
| *** caowei has joined #openstack-kolla | 02:18 | |
| *** Margin has quit IRC | 02:22 | |
| *** tonanhngo has joined #openstack-kolla | 02:28 | |
| *** tonanhngo has quit IRC | 02:31 | |
| *** zhubingbing has quit IRC | 02:41 | |
| *** zhubingbing has joined #openstack-kolla | 02:51 | |
| *** Margin has joined #openstack-kolla | 02:58 | |
| *** rwsu has quit IRC | 03:03 | |
| *** yuanying has quit IRC | 03:05 | |
| *** goldyfruit has joined #openstack-kolla | 03:05 | |
| *** klindgren has joined #openstack-kolla | 03:06 | |
| zhubingbing | when 's kolla-kubernetes meeting ? kfox1111 | 03:11 |
| *** tovin07 has joined #openstack-kolla | 03:14 | |
| *** david-lyle has joined #openstack-kolla | 03:30 | |
| *** harlowja has quit IRC | 03:35 | |
| *** david-lyle has quit IRC | 03:36 | |
| *** dave-mccowan has quit IRC | 03:51 | |
| openstackgerrit | Steve Baker proposed openstack/kolla master: Use version.release_version() https://review.openstack.org/448380 | 04:12 |
| *** yuanying has joined #openstack-kolla | 04:18 | |
| *** goldyfruit has quit IRC | 04:18 | |
| *** kbaegis has quit IRC | 04:21 | |
| *** fooliouno has left #openstack-kolla | 04:31 | |
| *** zhurong has quit IRC | 04:37 | |
| *** Administrator_ has joined #openstack-kolla | 04:38 | |
| *** crushil has quit IRC | 04:38 | |
| *** zhugaoxiao has quit IRC | 04:41 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla stable/newton: [TEST][NEWTON] test scenario2 in gate for newton branch https://review.openstack.org/417307 | 04:45 |
| *** unicell has joined #openstack-kolla | 04:48 | |
| sdake | zhubingbing kolla-kubernete meting is 16:00 UTC on wednesdays | 04:52 |
| sdake | Jeffrey4l_ is the gate still borked on kolla-ansible? | 04:52 |
| Jeffrey4l_ | sdake, kolla-ansible is OK. and kolla is broken. | 04:58 |
| sdake | Jeffrey4l_ mnaser was working on this problem - i believe he found the root cause in the implementation of bifrost | 04:58 |
| sdake | Jeffrey4l_ mnaser filed a bug which I don't have a reference for | 04:59 |
| sdake | Jeffrey4l_ we need to get the gate unblocked asap - perhaps a "ignore biforst build" would be a good short term answer? | 04:59 |
| sdake | if we dont fix it soon we will have a big backlog of patches causing all kinds of rebasing needed | 05:00 |
| sdake | 50,862 total steps | 05:01 |
| sdake | 2,392 avg. daily calorie burn | 05:01 |
| sdake | pretty sure that is superstar status for me with my fitbit :) | 05:01 |
| sdake | 1 week of my life on the road.. :( | 05:01 |
| sdake | no idea i burned up 2400 calories on the road | 05:01 |
| sdake | proably more like 1500 at home | 05:01 |
| sdake | explains why i lose weight when i travel even though i heat furiously | 05:02 |
| sdake | heat/eat | 05:02 |
| Jeffrey4l_ | i am checking this issue too. | 05:04 |
| Jeffrey4l_ | sees need add SUDO_USER envionements | 05:04 |
| *** l4yerffeJ has joined #openstack-kolla | 05:05 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla master: Fix bifrost failure https://review.openstack.org/448385 | 05:10 |
| *** shashank_t_ has joined #openstack-kolla | 05:10 | |
| *** skramaja has joined #openstack-kolla | 05:13 | |
| *** cuongnv has quit IRC | 05:15 | |
| *** l4yerffeJ has quit IRC | 05:15 | |
| *** shashank_t_ has quit IRC | 05:15 | |
| *** l4yerffeJ has joined #openstack-kolla | 05:16 | |
| openstackgerrit | Merged openstack/kolla-ansible master: Add an Ansible group for bifrost https://review.openstack.org/435001 | 05:16 |
| *** l4yerffeJ has quit IRC | 05:17 | |
| *** l4yerffeJ has joined #openstack-kolla | 05:18 | |
| *** l4yerffeJ has quit IRC | 05:18 | |
| *** l4yerffeJ has joined #openstack-kolla | 05:19 | |
| sdake | Jeffrey4l_ mnaser had a fix he submitted ti ironic | 05:23 |
| sdake | i am not sure if it merged | 05:23 |
| sdake | rather to bifrost | 05:23 |
| sdake | he filed a bug | 05:23 |
| sdake | let me check my mail and see if it hit my inbox | 05:24 |
| sdake | https://bugs.launchpad.net/kolla/+bug/1674483 | 05:24 |
| openstack | Launchpad bug 1674483 in kolla "Bifrost failing because of missing SUDO_USER" [Undecided,New] | 05:24 |
| l4yerffeJ | suffering network issue here. may lost some message. | 05:24 |
| l4yerffeJ | roger. seems he still do not have a fix. i will link this bug if my patch works. sdake | 05:25 |
| masber | are all images deployed on both tarballs and docker hub? which one is more reliable? | 05:25 |
| l4yerffeJ | masber, about this https://bugs.launchpad.net/kolla/+bug/1674483 | 05:26 |
| openstack | Launchpad bug 1674483 in kolla "Bifrost failing because of missing SUDO_USER" [Critical,Confirmed] | 05:26 |
| l4yerffeJ | the user is a simple version, i think . | 05:26 |
| l4yerffeJ | it should not related to becaome. | 05:26 |
| *** Jeffrey4l_ has quit IRC | 05:27 | |
| *** cuongnv has joined #openstack-kolla | 05:28 | |
| *** iceyao has joined #openstack-kolla | 05:29 | |
| *** pbourke has joined #openstack-kolla | 05:29 | |
| *** Jeffrey4l_ has joined #openstack-kolla | 05:29 | |
| *** zhurong has joined #openstack-kolla | 05:33 | |
| *** bmace has quit IRC | 05:35 | |
| *** bmace has joined #openstack-kolla | 05:48 | |
| *** kbaegis has joined #openstack-kolla | 05:48 | |
| openstackgerrit | Merged openstack/kolla-ansible master: Doc syntax error and expressions updates https://review.openstack.org/446882 | 05:52 |
| spsurya | caowei: ping...... | 05:58 |
| caowei | hi spsurya | 05:58 |
| sdake | Jeffrey4l_ https://bugs.launchpad.net/kolla/+bug/1674483 | 05:59 |
| openstack | Launchpad bug 1674483 in kolla "Bifrost failing because of missing SUDO_USER" [Critical,Confirmed] | 05:59 |
| sdake | Jeffrey4l_ mnaser I believe is the correct irc handle for that bug filer and the person who did some initial discovery work | 06:00 |
| spsurya | caowei: can i add my name as Co-Author to this https://blueprints.launchpad.net/kolla/+spec/dragonflow | 06:00 |
| sdake | Jeffrey4l_ mnaser and masber are two different people | 06:00 |
| spsurya | caowei: https://review.openstack.org/#/c/418193/ | 06:00 |
| Jeffrey4l_ | aha, i am not sure. | 06:01 |
| *** cuongnv has quit IRC | 06:01 | |
| Jeffrey4l_ | ok i pinged wrong person. | 06:01 |
| spsurya | caowei: ^^ | 06:02 |
| caowei | spsurya : I agree | 06:03 |
| spsurya | caowei: thanks | 06:03 |
| caowei | spsurya: cool | 06:04 |
| spsurya | caowei: https://review.openstack.org/#/c/418193/ is in merge conflict | 06:05 |
| spsurya | need to push a patch for this | 06:06 |
| *** cuongnv has joined #openstack-kolla | 06:08 | |
| openstackgerrit | Zeyu Zhu proposed openstack/kolla-ansible master: Update the kubernetes-deployment.rst file https://review.openstack.org/448035 | 06:11 |
| duonghq | sdake, seem that I missed some email again, how often do we have kolla-k8s meeting? and from where I can get info? | 06:20 |
| sdake | duonghthe k8s meeting is the kolla meeitng | 06:20 |
| sdake | duonghq it is weekly at 16:00 UTC on wednesday (see topic) | 06:20 |
| duonghq | It's our usual weekly meeting? | 06:21 |
| sdake | yup | 06:22 |
| sdake | just add agenda items for kolla, kolla-ansible, or kolla-kubernetes deliverable | 06:22 |
| spsurya | sdake: roger | 06:23 |
| duonghq | okay, thought that we have another meeting | 06:23 |
| *** masber has quit IRC | 06:24 | |
| *** masber has joined #openstack-kolla | 06:24 | |
| spsurya | sdake: are we going to set another version release target ?? As currently we have 2017-03-15 for 0.6.0 | 06:26 |
| spsurya | sdake: I am just confuse with releases | 06:26 |
| *** rwsu has joined #openstack-kolla | 06:27 | |
| spsurya | sdake: ^^ | 06:30 |
| sdake | spsurya i think we should move the 3/15 release to 4/15 possibly | 06:34 |
| sdake | as I think we haven' tmade much progress towards completion of anything in this sprint | 06:34 |
| sdake | spsurya although ironic has been mostly completed | 06:35 |
| sdake | duonghq only one meeting for all deliverables | 06:35 |
| duonghq | roger | 06:35 |
| spsurya | sdake: roger | 06:36 |
| *** jrobinson has quit IRC | 06:42 | |
| spsurya | sdake: bug that you were looking for IIRC for gate issue https://bugs.launchpad.net/kolla/+bug/1674483 | 06:45 |
| openstack | Launchpad bug 1674483 in kolla "Bifrost failing because of missing SUDO_USER" [Critical,Confirmed] | 06:45 |
| spsurya | in kolla deliverable | 06:45 |
| masber | "The requested image does not exist: 129.94.72.62:4000/lokolla/centos-source-kolla-toolbox:4.0.0" I tried both source and binary, I am using centos distro and I am using local images downloaded from tarballs.openstack.org | 06:50 |
| openstackgerrit | Merged openstack/kolla-ansible master: Add new section "Log delivery" in CONTRIBUTING.rst https://review.openstack.org/446174 | 06:56 |
| *** unicell has quit IRC | 07:02 | |
| openstackgerrit | Merged openstack/kolla-ansible master: Fix the Nova SSH port prechecks missing https://review.openstack.org/447268 | 07:03 |
| openstackgerrit | Merged openstack/kolla-ansible master: Update Searchlight resource_plugin configuration https://review.openstack.org/442628 | 07:04 |
| openstackgerrit | Merged openstack/kolla-ansible master: Apply correct permissions to kolla_logs/ceilometer https://review.openstack.org/443066 | 07:04 |
| *** kbaegis has quit IRC | 07:04 | |
| *** unicell has joined #openstack-kolla | 07:07 | |
| openstackgerrit | Merged openstack/kolla-ansible master: Sahara: add oslo_messaging_notifications config https://review.openstack.org/444420 | 07:08 |
| *** zioproto has joined #openstack-kolla | 07:15 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla master: Get latest versions from openstack/releases repo https://review.openstack.org/432300 | 07:19 |
| *** jaosorior has joined #openstack-kolla | 07:21 | |
| Jeffrey4l_ | duonghq, congrats! | 07:26 |
| spsurya | duonghq: Congratulations !! | 07:30 |
| *** manheim has joined #openstack-kolla | 07:36 | |
| *** manheim has quit IRC | 07:36 | |
| *** manheim has joined #openstack-kolla | 07:38 | |
| *** manheim has quit IRC | 07:38 | |
| *** manheim has joined #openstack-kolla | 07:39 | |
| *** manheim has quit IRC | 07:39 | |
| *** pcaruana has joined #openstack-kolla | 07:39 | |
| *** manheim has joined #openstack-kolla | 07:40 | |
| *** manheim has quit IRC | 07:41 | |
| *** manheim has joined #openstack-kolla | 07:42 | |
| *** manheim has quit IRC | 07:42 | |
| *** manheim has joined #openstack-kolla | 07:43 | |
| *** manheim has quit IRC | 07:43 | |
| *** manheim has joined #openstack-kolla | 07:45 | |
| *** manheim has quit IRC | 07:45 | |
| *** iceyao has quit IRC | 07:53 | |
| duonghq | Jeffrey4l_, spsurya thank you | 07:53 |
| *** matrohon has joined #openstack-kolla | 07:53 | |
| openstackgerrit | Wei Cao proposed openstack/kolla master: Add dragonflow container https://review.openstack.org/418193 | 07:57 |
| *** jmccarthy has joined #openstack-kolla | 07:58 | |
| *** zhurong has quit IRC | 08:00 | |
| openstackgerrit | Wei Cao proposed openstack/kolla master: Add dragonflow container https://review.openstack.org/418193 | 08:03 |
| *** Serlex has joined #openstack-kolla | 08:04 | |
| openstackgerrit | Zeyu Zhu proposed openstack/kolla-ansible master: Remove the redundance extra-vars https://review.openstack.org/448433 | 08:07 |
| *** zhurong has joined #openstack-kolla | 08:11 | |
| openstackgerrit | Zeyu Zhu proposed openstack/kolla-ansible master: Update the kubernetes-deployment.rst file https://review.openstack.org/448035 | 08:16 |
| openstackgerrit | zhubingbing proposed openstack/kolla-ansible master: Add magnum_trustee_domain_id in magnum.conf https://review.openstack.org/448438 | 08:21 |
| *** scsnow_ has joined #openstack-kolla | 08:22 | |
| *** athomas has joined #openstack-kolla | 08:25 | |
| *** shashank_t_ has joined #openstack-kolla | 08:26 | |
| *** matrohon has quit IRC | 08:30 | |
| *** shashank_t_ has quit IRC | 08:30 | |
| *** mgoddard has joined #openstack-kolla | 08:32 | |
| *** matrohon has joined #openstack-kolla | 08:33 | |
| openstackgerrit | zhubingbing proposed openstack/kolla-ansible master: Fix can't find /usr/lib/libCryptoki2_64.so in barbican https://review.openstack.org/444561 | 08:36 |
| *** bjolo has joined #openstack-kolla | 08:37 | |
| *** targon has joined #openstack-kolla | 08:41 | |
| *** skramaja is now known as skramaja_afk | 08:41 | |
| *** cuongnv has quit IRC | 08:43 | |
| *** egonzalez has joined #openstack-kolla | 08:43 | |
| *** cuongnv has joined #openstack-kolla | 08:44 | |
| *** l4yerffeJ has quit IRC | 08:50 | |
| *** l4yerffeJ has joined #openstack-kolla | 08:50 | |
| *** shardy has joined #openstack-kolla | 08:53 | |
| openstackgerrit | Merged openstack/kolla-ansible master: Activate binary-logs in MariaDB/MySQL https://review.openstack.org/447265 | 08:53 |
| *** l4yerffeJ has quit IRC | 08:55 | |
| openstackgerrit | Merged openstack/kolla-ansible master: Fix error during Grafana container restart https://review.openstack.org/447191 | 08:55 |
| *** l4yerffeJ has joined #openstack-kolla | 08:55 | |
| *** athomas has quit IRC | 08:56 | |
| *** iceyao has joined #openstack-kolla | 08:58 | |
| bjolo | morning | 08:58 |
| *** aolwas has joined #openstack-kolla | 09:00 | |
| *** shardy has quit IRC | 09:03 | |
| *** shardy has joined #openstack-kolla | 09:03 | |
| *** iceyao has quit IRC | 09:05 | |
| *** iceyao has joined #openstack-kolla | 09:06 | |
| *** mgoddard has quit IRC | 09:08 | |
| *** iceyao has quit IRC | 09:08 | |
| *** iceyao has joined #openstack-kolla | 09:11 | |
| *** mcnanci has quit IRC | 09:12 | |
| zhubingbing | sup duonhq | 09:16 |
| *** iceyao has quit IRC | 09:17 | |
| egonzalez | duonghq, congrats dude! | 09:18 |
| *** rmart04 has joined #openstack-kolla | 09:18 | |
| openstackgerrit | Wei Cao proposed openstack/kolla master: Add dragonflow container https://review.openstack.org/418193 | 09:19 |
| *** iceyao has joined #openstack-kolla | 09:20 | |
| *** mcnanci has joined #openstack-kolla | 09:21 | |
| zhubingbing | sup egonzalez | 09:22 |
| zhubingbing | ;) | 09:22 |
| *** yuanying has quit IRC | 09:31 | |
| *** iceyao has quit IRC | 09:32 | |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible master: Update the "virt_driver" to "virt_type" https://review.openstack.org/447336 | 09:34 |
| *** mgoddard has joined #openstack-kolla | 09:34 | |
| *** l4yerffeJ has quit IRC | 09:38 | |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible master: Remove tenant_id from compute endpoint https://review.openstack.org/447262 | 09:38 |
| *** l4yerffeJ has joined #openstack-kolla | 09:38 | |
| masber | hi, | 09:39 |
| egonzalez | sup | 09:41 |
| *** papacz has joined #openstack-kolla | 09:43 | |
| openstackgerrit | Mark Goddard proposed openstack/kolla-ansible master: Improve developer documentation for Bifrost https://review.openstack.org/436462 | 09:44 |
| masber | I am testing kolla-ansible with different network configurations, so far I have been able to finish the installation and access horizon, create networks, etc. However it seems I don't have internet access from the router's namespace, I can see br-ex (ip a) but the neutron_external_interface is not hook up to it. I tried to check the br-ex configuration but ifcfg-br-ex is missing and ovs-vsctl command is not found on the host. Any idea what would | 09:48 |
| masber | be the next steps for troubleshooting? | 09:48 |
| *** jmccarthy1 has joined #openstack-kolla | 09:50 | |
| *** jmccarthy has quit IRC | 09:51 | |
| *** zhurong has quit IRC | 09:52 | |
| *** duonghq has quit IRC | 09:53 | |
| *** skramaja_afk is now known as skramaja | 09:56 | |
| *** cuongnv has quit IRC | 10:05 | |
| openstackgerrit | Bertrand Lallau proposed openstack/kolla-ansible master: Heat: add oslo_messaging_notifications config https://review.openstack.org/444515 | 10:06 |
| *** pbourke has quit IRC | 10:08 | |
| *** tovin07_ has quit IRC | 10:11 | |
| *** yuanying has joined #openstack-kolla | 10:11 | |
| openstackgerrit | Bertrand Lallau proposed openstack/kolla-ansible master: Cloudkitty: remove oslo_messaging_notifications config https://review.openstack.org/444405 | 10:14 |
| *** blallau has joined #openstack-kolla | 10:14 | |
| *** pbourke has joined #openstack-kolla | 10:14 | |
| *** athomas has joined #openstack-kolla | 10:27 | |
| *** l4yerffeJ has quit IRC | 10:29 | |
| *** l4yerffeJ has joined #openstack-kolla | 10:29 | |
| openstackgerrit | Paul Bourke (pbourke) proposed openstack/kolla-ansible master: Add an extra network node for HA https://review.openstack.org/448494 | 10:36 |
| openstackgerrit | Paul Bourke (pbourke) proposed openstack/kolla-ansible master: Add an extra network node for HA https://review.openstack.org/448494 | 10:44 |
| *** shardy is now known as shardy_dentist | 10:45 | |
| *** asettle has joined #openstack-kolla | 10:46 | |
| asettle | Hey! Who in kolla is euro? :) | 10:48 |
| asettle | I'm currently looking at the new kolla-ansible deployment guide, and it doesn't appear to link to a deployment guide, it links to the entire dev docs on kolla-ansible | 10:49 |
| Serlex | you mean this - https://docs.openstack.org/developer/kolla-ansible/quickstart.html ? | 10:52 |
| pbourke | asettle: can you link to what you mean? | 10:53 |
| asettle | pbourke: https://docs.openstack.org/project-deploy-guide/kolla-ansible/ocata/kolla-ansible-index.html | 10:54 |
| asettle | This, should link to the quickstart | 10:54 |
| asettle | :) | 10:54 |
| asettle | But it links to all the dev docs | 10:54 |
| asettle | For Kolla-Ansible | 10:54 |
| pbourke | asettle: I think right now that's all we have | 10:54 |
| asettle | THat's okay, lemme poke around and fix it. | 10:54 |
| pbourke | asettle: but they will be refactored as time goes on | 10:54 |
| asettle | pbourke: it should only link to the quickstart, not everything else :) | 10:54 |
| pbourke | everything there is end user docs | 10:55 |
| asettle | Yep, end user should be kept in your own tree, not in the 'deployment guide' | 10:55 |
| asettle | Deployment guide should only be installation material | 10:55 |
| pbourke | ah | 10:55 |
| asettle | Not configuration or troubleshooting | 10:55 |
| asettle | :) | 10:55 |
| openstackgerrit | Bertrand Lallau proposed openstack/kolla-ansible master: Set neutron RPC workers and state report workers https://review.openstack.org/444151 | 10:55 |
| asettle | Just gonna quickly poke around, uno momento. I wonder if the cookie cutter has accidentally wrapped the entire doc tree | 10:55 |
| asettle | If so, that's not so flash | 10:55 |
| pbourke | asettle: I think its just a misunderstanding, I remember reviewing the changes for the deploy guide to link everything | 10:56 |
| asettle | Yes okay, no problem :) | 10:56 |
| asettle | I have a slight concern though - https://github.com/openstack/kolla-ansible/blob/master/deploy-guide/source/quickstart.rst | 10:56 |
| pbourke | Sayantani Goswami is the person who did most of the work on this, I'd ping them but unsure of the irc name! | 10:57 |
| asettle | You're linking between your doc tree, and your deploy guide tree | 10:57 |
| asettle | pbourke: no sweat :) i've got Sayantani on slack somewhere | 10:57 |
| pbourke | great :) | 10:57 |
| pbourke | he/she should be able answer your questions more accurately | 10:57 |
| *** Deys has joined #openstack-kolla | 10:58 | |
| asettle | pbourke: coolio hoolio :) I might just play around a bit and get this lined up - i'll tag you as a review, if that's okay pbourke ? :) | 10:59 |
| asettle | Sorry ,tagged your name twice | 10:59 |
| *** Deys has left #openstack-kolla | 10:59 | |
| *** zhubingbing has quit IRC | 11:02 | |
| pbourke | asettle: no problem | 11:03 |
| *** zhubingbing_ has joined #openstack-kolla | 11:07 | |
| openstackgerrit | Alexandra Settle proposed openstack/kolla-ansible master: [deploy-guide] Moving the quick start to the deploy folder https://review.openstack.org/448509 | 11:08 |
| asettle | pbourke: ^^ | 11:09 |
| asettle | I just want to check that build | 11:09 |
| egonzalez | masber, ovs-vsctl command are inside openvswitch containers | 11:14 |
| openstackgerrit | Merged openstack/kolla-ansible master: Use service_name variable when defined in neutron role https://review.openstack.org/448269 | 11:23 |
| openstackgerrit | Paul Bourke (pbourke) proposed openstack/kolla master: Allowing internal pip install https://review.openstack.org/448515 | 11:30 |
| openstackgerrit | Paul Bourke (pbourke) proposed openstack/kolla master: Fix source install for vmtp https://review.openstack.org/448515 | 11:31 |
| openstackgerrit | Paul Bourke (pbourke) proposed openstack/kolla master: Fix source install for vmtp https://review.openstack.org/448515 | 11:32 |
| *** shardy_dentist is now known as shardy | 11:33 | |
| *** athomas has quit IRC | 11:33 | |
| *** magicboiz has quit IRC | 11:34 | |
| *** rwallner has joined #openstack-kolla | 11:39 | |
| *** rwallner has quit IRC | 11:39 | |
| *** rwallner has joined #openstack-kolla | 11:40 | |
| asettle | Thanks pbourke :) I've messaged Sayantani and let her know what changes I've made and why. | 11:44 |
| *** athomas has joined #openstack-kolla | 11:51 | |
| egonzalez | zhubingbing_, aroung? | 11:53 |
| zhubingbing_ | hear | 11:53 |
| egonzalez | added you on this PS as reviewer, can you take a look? https://review.openstack.org/#/c/448479/ | 11:54 |
| egonzalez | fluentd related stuff | 11:54 |
| zhubingbing_ | 。。。 | 11:56 |
| zhubingbing_ | have some problmes | 11:56 |
| zhubingbing_ | because we have not support this project | 11:57 |
| *** srwilkers has joined #openstack-kolla | 11:59 | |
| *** srwilkers has left #openstack-kolla | 11:59 | |
| *** srwilkers has joined #openstack-kolla | 11:59 | |
| srwilkers | morning o/ | 11:59 |
| *** manheim has joined #openstack-kolla | 12:01 | |
| *** NachoDuck has quit IRC | 12:01 | |
| sbezverk | srwilkers: morning | 12:05 |
| *** mnasiadka has joined #openstack-kolla | 12:11 | |
| Serlex | hi sbezverk, have you booted ironic instance using kolla-ansible? | 12:12 |
| sbezverk | Serlex: no, I am working on kolla-kube side of house | 12:16 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/kolla-ansible master: Updated from global requirements https://review.openstack.org/446772 | 12:17 |
| nea1 | Hi, I have the error that there still is a wrong IP for e.g. in the iSCSI settings as I had to change it, how do I fix it? rerunning the deploy doesn't help | 12:19 |
| *** rhallisey has joined #openstack-kolla | 12:19 | |
| openstackgerrit | zhubingbing proposed openstack/kolla-ansible master: Fix can't find /usr/lib/libCryptoki2_64.so in barbican https://review.openstack.org/444561 | 12:23 |
| *** rwallner has quit IRC | 12:30 | |
| openstackgerrit | Merged openstack/kolla master: Fix README.rst visit link error https://review.openstack.org/447213 | 12:30 |
| egonzalez | nea1, what error are referring to? | 12:32 |
| *** jtriley has quit IRC | 12:34 | |
| *** shardy is now known as shardy_lunch | 12:34 | |
| nea1 | egonzalez: 'Could not login to any iSCSI portal.' | 12:35 |
| nea1 | becase it has somewhere the wrong IP still stored | 12:35 |
| *** rwallner has joined #openstack-kolla | 12:36 | |
| nea1 | I noticed that it has for e.g. the wrong values in the database, but I haven't found the config file containing the old IP (searched /var and /etc from the host | 12:37 |
| egonzalez | nea1, what's the IP in /etc/kolla/<tgtd or iscsi>/config.json ? | 12:38 |
| nea1 | except for "Binary file /var/lib/docker/devicemapper/devicemapper/data matches" the other matches aren't a problem | 12:38 |
| nea1 | egonzalez: the new one | 12:38 |
| masber | Jeffrey4l_, are you after this bug https://bugs.launchpad.net/kolla-ansible/+bug/1626339? | 12:38 |
| openstack | Launchpad bug 1626339 in kolla-ansible "Kolla mongodb error: can't find self in the replset config" [High,In progress] - Assigned to Jeffrey Zhang (jeffrey4l) | 12:38 |
| egonzalez | nea1, the service is bind in the correct IP:port `ss -antupo | grep 3260` | 12:40 |
| egonzalez | nea1, have you tried restarting tgtd container manually? | 12:40 |
| nea1 | tcp LISTEN 0 128 172.31.30.101:3260 *:* users:(("tgtd",pid=20854,fd=6)) | 12:41 |
| nea1 | that is fine | 12:41 |
| nea1 | no, I did a reboot of the server in between I think | 12:41 |
| zhubingbing_ | egonzalez can u review it ? egonzalez | 12:41 |
| zhubingbing_ | https://review.openstack.org/#/c/444561/ | 12:41 |
| zhubingbing_ | ;( | 12:41 |
| Jeffrey4l_ | masber, yep. and here is the fix. https://review.openstack.org/#/c/442999/ | 12:42 |
| egonzalez | zhubingbing_, change commit message as there is not any reference for safenet now ;) | 12:42 |
| zhubingbing_ | hmm | 12:42 |
| zhubingbing_ | ok | 12:42 |
| egonzalez | nea1, if the IP in ss is correct the container applied the new config on reconfigure. guess there is not any error or warn in container logs right? | 12:44 |
| openstackgerrit | zhubingbing proposed openstack/kolla-ansible master: Fix can't find /usr/lib/libCryptoki2_64.so in barbican https://review.openstack.org/444561 | 12:44 |
| nea1 | egonzalez: in the database if creating a new VM the connection_info in the block_device_mapping table of nova contains the old IP | 12:44 |
| masber | Jeffrey4l_, thank you, I am trying to find out why my br-ex is not working and will deploy your fix after that! | 12:44 |
| Jeffrey4l_ | got. thanks. | 12:45 |
| nea1 | egonzalez: "Failed to connect to iSCSI portal [old-ip]:3260" in nova-compute.log | 12:45 |
| zhubingbing_ | https://review.openstack.org/#/c/444561/ please reviews | 12:45 |
| *** eaguilar has joined #openstack-kolla | 12:48 | |
| egonzalez | nea1, cinder service-list have nodes updated? or still pointing to old cinder-volume nodes? | 12:48 |
| nea1 | egonzalez: http://paste.openstack.org/show/603755/ | 12:50 |
| nea1 | that gives a error, openstack volume service list works, and that has the dns name in there (not fqdn) | 12:50 |
| *** zhubingbing_ has quit IRC | 12:50 | |
| *** srwilkers is now known as srwilkers-coffee | 12:52 | |
| nea1 | my feeling is that the '/etc/hosts' file in /var/lib/docker/devicemapper/devicemapper/data could still have the old IP in it as that seams to be "SGI XFS filesystem data (blksz 4096, inosz 512, v2 dirs)" | 12:52 |
| nea1 | but on the other hand everything else seams to be working | 12:52 |
| openstackgerrit | Merged openstack/kolla-ansible master: Updated from global requirements https://review.openstack.org/446772 | 12:53 |
| *** fguillot has joined #openstack-kolla | 12:54 | |
| *** bjolo has quit IRC | 12:56 | |
| *** bjolo has joined #openstack-kolla | 12:56 | |
| nea1 | hm, no the hosts are in /var/lib/docker/containers/*/hosts and they are all fine | 12:59 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes master: WIP Adding virtualbmc to the gate https://review.openstack.org/441550 | 13:00 |
| egonzalez | nea1, have you checked https://lists.gt.net/openstack/operators/30097 ? | 13:00 |
| *** bjolo has quit IRC | 13:01 | |
| egonzalez | is quite old but maybe points you in the right direction | 13:01 |
| *** pbandark has joined #openstack-kolla | 13:03 | |
| *** srwilkers-coffee has quit IRC | 13:04 | |
| Jeffrey4l_ | masber, btw, any progress on bifrost issue in kolla gate? | 13:04 |
| *** srwilkers-coffee has joined #openstack-kolla | 13:04 | |
| masber | sorry I think you are confused with mnaser? | 13:05 |
| Jeffrey4l_ | aha, sorry. | 13:06 |
| mnaser | heh ^ | 13:06 |
| *** goldyfruit has joined #openstack-kolla | 13:06 | |
| mnaser | Jeffrey4l_ i had things to do in $work yesterday, and i have to finish up things today so i wont be able to get to it :( | 13:06 |
| mnaser | i mentioned yesterday if someone could pick up and see what the rest is happening that would be nice if not ill pick off my work later today or tomorrow | 13:06 |
| Jeffrey4l_ | mnaser, any new findings? | 13:07 |
| mnaser | Jeffrey4l_ i submitted a patch to fix the gate, well, what i thought was going to fix it, but that didnt seem to have cleared it out | 13:07 |
| Jeffrey4l_ | mind paste the link? | 13:07 |
| mnaser | i submitted an initial patch which i was sure would eliminate the issue but bifrost team asked to maintain the same way of doing things before | 13:07 |
| mnaser | sure 1 sec | 13:07 |
| sdake | morning | 13:07 |
| nea1 | egonzalez: ok, I added the IP manually and now it is working | 13:07 |
| mnaser | o/ sdake | 13:07 |
| Jeffrey4l_ | i can take a few hour on this issue now. | 13:07 |
| Jeffrey4l_ | sup sdake | 13:08 |
| nea1 | I added it to the [lvm-1] block and that fixed it | 13:08 |
| mnaser | Jeffrey4l_ https://review.openstack.org/#/c/447713/ | 13:08 |
| mnaser | check out the 1st patch i submitted | 13:08 |
| mnaser | im pretty sure if the 1st one merged, we wouldn't have that issue | 13:08 |
| Jeffrey4l_ | got. thanks. | 13:08 |
| egonzalez | nea1, seems cinder does not update entries once created :/ | 13:08 |
| mnaser | but the other problem is why SUDO_USER is not being set /shrug | 13:08 |
| mnaser | feel free to ask me anything if you want info :) | 13:08 |
| Jeffrey4l_ | seems SUDO_USER come from "sudo -i" | 13:09 |
| nea1 | which entries? you mean existing volumes? I hadn't created any volumes before changing the IP | 13:11 |
| *** eaguilar has quit IRC | 13:20 | |
| *** shardy_lunch is now known as shardy | 13:24 | |
| *** mbruzek has joined #openstack-kolla | 13:27 | |
| *** mbruzek has joined #openstack-kolla | 13:27 | |
| *** eaguilar has joined #openstack-kolla | 13:30 | |
| mnaser | silly question why doesnt kolla-ansible (or kolla) adopt existing images rather than build from scratch | 13:32 |
| mnaser | ex: https://hub.docker.com/_/haproxy/ | 13:32 |
| *** jtriley has joined #openstack-kolla | 13:33 | |
| Jeffrey4l_ | mnaser, this is talked recently. first of all, i think is these image is out of kolla's control. for configuration management. start up method. | 13:36 |
| Jeffrey4l_ | and COPY_ONCE COPY_ALWAYS feature etc. | 13:36 |
| Jeffrey4l_ | another consideration is, all image size. | 13:36 |
| Jeffrey4l_ | total image size. | 13:36 |
| Jeffrey4l_ | and other cpu arch support . | 13:36 |
| Jeffrey4l_ | build an image is easy. i do not see the advantage to use existent images. | 13:37 |
| nea1 | which pip packages are installed on the host? (if it comes to system update, not that some python packages get updated breaking the setup) | 13:38 |
| egonzalez | nea1, using kolla-ansible bootstrap-servers? | 13:40 |
| egonzalez | nea1, on target nodes only docker-py from pip | 13:42 |
| nea1 | ok perfect | 13:47 |
| *** crushil has joined #openstack-kolla | 13:48 | |
| *** zhubingbing has joined #openstack-kolla | 13:49 | |
| *** srwilkers-coffee is now known as srwilkers | 13:50 | |
| zhubingbing | hello guys | 13:54 |
| zhubingbing | ;) | 13:54 |
| *** zioproto has quit IRC | 14:01 | |
| openstackgerrit | Merged openstack/kolla-ansible master: Fix can't find /usr/lib/libCryptoki2_64.so in barbican https://review.openstack.org/444561 | 14:11 |
| *** zioproto has joined #openstack-kolla | 14:11 | |
| spsurya | zhubingbing: hi.. | 14:12 |
| zhubingbing | hi <spsurya | 14:12 |
| zhubingbing | ;) | 14:12 |
| spsurya | zhubingbing: seems like topics are not added in agenda of today meeting | 14:14 |
| Jeffrey4l_ | egonzalez, could u review this backport? https://review.openstack.org/#/c/447300/ | 14:15 |
| openstackgerrit | caoyuan proposed openstack/kolla master: Add the missing project link into README.rst https://review.openstack.org/447255 | 14:16 |
| *** skramaja is now known as skramaja_afk | 14:16 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes master: WIP Adding virtualbmc to the gate https://review.openstack.org/441550 | 14:19 |
| openstackgerrit | caoyuan proposed openstack/kolla master: Add the missing project link into README.rst https://review.openstack.org/447255 | 14:19 |
| zhubingbing | spsurya link？ | 14:22 |
| openstackgerrit | caoyuan proposed openstack/kolla master: Add the missing project link into README.rst https://review.openstack.org/447255 | 14:22 |
| spsurya | zhubingbing: https://wiki.openstack.org/wiki/Meetings/Kolla | 14:24 |
| zhubingbing | ok | 14:24 |
| zhubingbing | i don；t know ；( | 14:25 |
| zhubingbing | ;( | 14:25 |
| *** dave-mccowan has joined #openstack-kolla | 14:25 | |
| *** deadnull has joined #openstack-kolla | 14:27 | |
| *** shashank_t_ has joined #openstack-kolla | 14:29 | |
| spsurya | zhubingbing: agenda is same as last week | 14:34 |
| sdake | agenda can be updated by anyone | 14:35 |
| *** crushil has quit IRC | 14:38 | |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible master: Remove tenant_id from compute endpoint https://review.openstack.org/447262 | 14:39 |
| *** athomas has quit IRC | 14:41 | |
| spsurya | sdake: yes it can be | 14:44 |
| *** shashank_t_ has quit IRC | 14:45 | |
| *** fooliouno has joined #openstack-kolla | 14:49 | |
| *** athomas has joined #openstack-kolla | 14:50 | |
| zhubingbing | ；） | 14:50 |
| sdake | https://wiki.openstack.org/wiki/Meetings/Kolla#Agenda_for_next_meeting_.28Mar_22nd_2017.29 | 14:51 |
| sdake | i've updated the agenda with some items i'm interested in talking about - perhaps other people can add to it :) | 14:51 |
| sdake | note i've also updated the release schedule for 0.6.0 of kolla-kubernetes here to 4/15/17: https://launchpad.net/kolla-kubernetes/+milestone/0.6.0 | 14:52 |
| *** manheim has quit IRC | 14:56 | |
| *** sayantani01 has joined #openstack-kolla | 14:59 | |
| *** g3ek has quit IRC | 14:59 | |
| zhubingbing | ;( | 15:00 |
| zhubingbing | sdake ;) | 15:00 |
| zhubingbing | good nice | 15:00 |
| nea1 | does kolla run with selinux enabled? - not sure if it is the image which had it disabled or the it was done in one of the playbooks | 15:00 |
| sdake | plz add your own areas :) | 15:00 |
| sdake | nea1 kolla will run with selinux enabled (it does so in the gate) | 15:00 |
| spsurya | sdake: i have also added some review items for discussion | 15:00 |
| nea1 | sdake: thx | 15:00 |
| sdake | nea1 getting your host os into the proper state is a bit of a bear | 15:01 |
| sdake | nea1 touch /.autorelabel seems to do the job if you had selinux off previously | 15:01 |
| sdake | we don't run the host-setup playbook your speaking of in the gate - so ymmv on whether that part works with selinux | 15:01 |
| *** adrian_otto has joined #openstack-kolla | 15:06 | |
| sdake | nea the touch needs to be followed up by a reboot | 15:07 |
| sdake | nea1^ | 15:07 |
| nea1 | yea, think that could be useful ^^ | 15:08 |
| *** g3ek has joined #openstack-kolla | 15:08 | |
| lazyPower | next kolla meeting is in ~ 50 minutes? | 15:09 |
| sdake | lazyPower yup | 15:09 |
| lazyPower | ty | 15:09 |
| lazyPower | just confirming, i was off by an hour last week and missed it :| | 15:09 |
| *** fooliouno has quit IRC | 15:10 | |
| sdake | lazyPower yup we schedule meetings in UTC not DST :) | 15:11 |
| sdake | lazyPower to avoid that problem | 15:11 |
| *** david-lyle has joined #openstack-kolla | 15:12 | |
| sdake | Jeffrey4l_ is the gat eblocked atm or unblocked? | 15:13 |
| Jeffrey4l_ | still blocked ;( | 15:13 |
| *** duonghq has joined #openstack-kolla | 15:14 | |
| duonghq | evening guys | 15:14 |
| sdake | mnaser will you be able to make the meeting in 45 mins? | 15:15 |
| mnaser | yes sdake | 15:15 |
| sdake | cool I added it ot hte agenda | 15:15 |
| sdake | it needs to be unblocked last friday ;) | 15:15 |
| mnaser | we unblocked it and it blocked again for something else unfortunately | 15:16 |
| mnaser | but yeah this sort of stuff should be the priority #1 | 15:16 |
| sdake | mnaser right - its #1 on the agenda :) | 15:16 |
| mnaser | :) | 15:16 |
| openstackgerrit | Bertrand Lallau proposed openstack/kolla-ansible master: Cleanup requirements.txt file https://review.openstack.org/448634 | 15:18 |
| *** manheim has joined #openstack-kolla | 15:19 | |
| nea1 | is it normal that mysql has 9.6G of RAM reserved? | 15:22 |
| sdake | nea1 could you expand on what you mean? | 15:26 |
| *** mkoderer has joined #openstack-kolla | 15:27 | |
| inc0 | good morning | 15:28 |
| inc0 | congrats duonghq :) | 15:28 |
| duonghq | inc0, morning, | 15:29 |
| duonghq | thank you | 15:29 |
| duonghq | sbezverk, are you there? | 15:29 |
| *** japestinho has quit IRC | 15:30 | |
| spsurya | inc0: morning | 15:30 |
| *** hrw has quit IRC | 15:30 | |
| openstackgerrit | Alicja Kwasniewska proposed openstack/kolla-ansible master: Improved documentation for central logging. https://review.openstack.org/444477 | 15:31 |
| sbezverk | duonghq: hey, what is up? | 15:32 |
| duonghq | sbezverk, in my k8s cluster, I'm able bring up keystone helm service, but I cannot query anything due to keystone cannot connect to the keystone-internal service | 15:33 |
| duonghq | do you have any suggestion? | 15:33 |
| kfox1111 | morning | 15:33 |
| duonghq | morning kfox1111 | 15:33 |
| nea1 | ah, I mean has a virtual memory of 9.6G | 15:33 |
| duonghq | sbezverk, it seems that it can resolve keystone-internal to IP, but not connect to this | 15:34 |
| sbezverk | duonghq: yeah you need to enable a couple of settings for keystone to be accessible from outside | 15:34 |
| nea1 | USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND | 15:35 |
| nea1 | 42434 20038 0.5 2.4 10074952 1203000 ? Sl Mar20 16:11 /usr/sbin/mysqld [...] | 15:35 |
| sbezverk | duonghq: check tests/bin/common_workflow_config.yaml | 15:35 |
| duonghq | sbezverk, I enabled external ip for keystone-public, so I can access keystone public, but in this, it cannot connect to keystone-internal | 15:35 |
| sbezverk | kfox1111: good morning, I finally get to the breaking point for ironic at the gate | 15:36 |
| *** hrw has joined #openstack-kolla | 15:36 | |
| sbezverk | kfox1111: check this out http://logs.openstack.org/50/441550/110/experimental/gate-kolla-kubernetes-deploy-centos-source-3-ironic-nv/ee2959f/logs/vm-1.log | 15:36 |
| sbezverk | duonghq: mimic what we have for keystone at the gate and you should be good to go :) | 15:37 |
| duonghq | sbezverk, thanks | 15:37 |
| kfox1111 | sbezverk: looks very similar to the issue we had with the bridge iptables thing. | 15:38 |
| inc0 | sbezverk: I can't find place where you set variable {{ memcached_servers }} | 15:38 |
| *** fooliouno has joined #openstack-kolla | 15:39 | |
| sbezverk | kfox1111: not sure, I am collecting tcpdump at the gate :) (this type of issue made you do crazy things) and I do see arp request and reply | 15:40 |
| openstackgerrit | Alexandra Settle proposed openstack/kolla-ansible master: [deploy-guide] Moving the quick start to the deploy folder https://review.openstack.org/448509 | 15:40 |
| sbezverk | inc0: could you be more specific? where you got that var? | 15:40 |
| sbezverk | kfox1111: check this out http://logs.openstack.org/50/441550/110/experimental/gate-kolla-kubernetes-deploy-centos-source-3-ironic-nv/ee2959f/logs/packets | 15:41 |
| *** skramaja_afk is now known as skramaja | 15:41 | |
| *** targon has quit IRC | 15:41 | |
| inc0 | sbezverk: https://github.com/openstack/kolla-ansible/blob/master/ansible/roles/neutron/templates/neutron.conf.j2#L94 | 15:42 |
| sbezverk | inc0: we set it up here : tests/conf/ceph-all-in-one/kolla_config:memcached_servers: "memcached" | 15:42 |
| inc0 | ok... | 15:42 |
| inc0 | :( | 15:42 |
| sbezverk | inc0: that is for the gate | 15:42 |
| inc0 | yeah I know | 15:42 |
| sbezverk | inc0: but you can set it up in your cloud yaml for example too | 15:42 |
| inc0 | no, we need to have default in code | 15:43 |
| *** mnasiadka has quit IRC | 15:43 | |
| kfox1111 | sbezverk: yeah, interesting... | 15:43 |
| sbezverk | inc0: not sure if it is a good aproach hardcoding things | 15:43 |
| inc0 | sbezverk: default != hardcode | 15:44 |
| sbezverk | inc0: then we can add it to helm/all_values.yaml | 15:44 |
| inc0 | then it won't be rendered if we use ansible | 15:44 |
| inc0 | so as long as we're using ansible, this has to be set in group_vars/all.yaml | 15:45 |
| openstackgerrit | Paul Bourke (pbourke) proposed openstack/kolla-ansible master: Only start one glance_api when using file backend https://review.openstack.org/448654 | 15:46 |
| sdake | sbezverk tcpdump in the gate - /me lols :) | 15:46 |
| sbezverk | inc0: not sure if it is a good approach bend code for ansible, it is ansible needs to be bent | 15:46 |
| sbezverk | kfox1111: it looks like kube proxy is not delivering packet to the service backend | 15:46 |
| inc0 | sbezverk: if ansible lives in kolla-k8s, no issues | 15:47 |
| kfox1111 | sbezverk: thats kinf of a weird dump. its asking over and over for arp. | 15:47 |
| kfox1111 | so its like its not making it all the way back to the vm. | 15:47 |
| inc0 | anyway I'm doing this until someone feels like moving everything to helm:P | 15:47 |
| sbezverk | kfox1111: but it shows arp reply in the trace | 15:47 |
| sbezverk | ?!?!? | 15:47 |
| kfox1111 | sbezverk: the server is replying, for sure. | 15:47 |
| sdake | inc0 sbezverk is starting on that shortly iiuc :) | 15:47 |
| kfox1111 | but the vm is acting like it didn't get it. | 15:48 |
| kfox1111 | so some how its not making it all the way back maybe. | 15:48 |
| sbezverk | kfox1111: hmmm so it gets dropped at tenants bridge | 15:48 |
| sdake | inc0 note the agenda has been updated to today's date | 15:48 |
| sbezverk | but after it shows up in tcpdump | 15:48 |
| sdake | inc0 i also added you in there for kolla-ansible ejection from kolla-kubernetes dep chain | 15:49 |
| inc0 | kk | 15:49 |
| inc0 | thanks | 15:49 |
| sbezverk | kfox1111: another thing which I do not get, it has already downloaded more tan 5 MB from the same ip, so it must have a valid mac address | 15:49 |
| kfox1111 | sbezverk: can you please add an iptables -F to setup_gate.sh near the top and try again? | 15:49 |
| kfox1111 | bridging between linuxbridge and ovs may be more fragile then we think. :/ | 15:50 |
| sbezverk | kfox1111: I have it already right before I start the instance | 15:50 |
| kfox1111 | we might try switching to ovs for the vm too if we can't figure it out. | 15:50 |
| kfox1111 | sbezverk: before starting the instance? | 15:50 |
| sdake | kfox1111 i've never been able to get ovs to work with a single nic | 15:50 |
| sdake | sbezverk you want the iptables flush beforey ou start the deployment | 15:51 |
| sbezverk | kfox1111: :( it is freaking working in my test bed... | 15:51 |
| kfox1111 | yeah. the iptables needs to be at the very begining. before neutron/calico/etc. | 15:51 |
| sdake | sbezverk e.g. right at the entrypont to the script | 15:51 |
| sbezverk | sdake: that is what I have, before provision baremetal VM I do flush table | 15:51 |
| sdake | sbezverk i see | 15:52 |
| sdake | sbezverk i think it should be "at start of gate, flush ip tables" | 15:52 |
| *** rwallner has quit IRC | 15:52 | |
| *** manheim has quit IRC | 15:52 | |
| kfox1111 | sbezverk: where's that in the code? I'm trying to find it. | 15:53 |
| duonghq | inc0, sorry, but you mean you are going to put ansible in k8s/helm? | 15:53 |
| sbezverk | kfox1111: tests/bin/ironic_deploy_tests.sh | 15:53 |
| *** Serlex has quit IRC | 15:53 | |
| inc0 | duonghq: https://review.openstack.org/#/c/448258/ | 15:54 |
| *** vhosakot has joined #openstack-kolla | 15:54 | |
| inc0 | just copy code from kolla-ansible, thin it out and leace jsut config | 15:54 |
| sbezverk | kfox1111: close to the bootom of the file | 15:54 |
| inc0 | that would give us quick win in getting rid of deps | 15:54 |
| duonghq | get your idea | 15:54 |
| kfox1111 | sbezverk: yeah, thats much too late. | 15:55 |
| kfox1111 | that may interfear with kube-proxy and some other canal things I think. | 15:55 |
| duonghq | when we have haproxy/keepalived in k8s, do we still need kube-proxy? | 15:57 |
| sbezverk | kfox1111: well it did not work without it either | 15:57 |
| sbezverk | I can move it to before kube cluster installation | 15:57 |
| kfox1111 | sbezverk: no, I mean, its required, but it needs to be at the very very bening of the gate job. | 15:57 |
| kfox1111 | yeah. that. :) | 15:57 |
| sbezverk | kfox1111: ok will ive it a try, how do you like my tcpdump thing ;) crazy.. | 15:58 |
| kfox1111 | yeah. thats pretty cool. :) | 15:58 |
| inc0 | meeting time | 15:58 |
| duonghq | roger | 15:58 |
| spsurya | roger | 15:59 |
| *** Guest52040 has joined #openstack-kolla | 16:00 | |
| *** harlowja has joined #openstack-kolla | 16:06 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes master: WIP Adding virtualbmc to the gate https://review.openstack.org/441550 | 16:07 |
| *** MasterOfBugs has joined #openstack-kolla | 16:07 | |
| *** kbaegis has joined #openstack-kolla | 16:12 | |
| *** zhubingbing__ has joined #openstack-kolla | 16:13 | |
| *** zhubingbing has quit IRC | 16:16 | |
| *** l4yerffeJ has quit IRC | 16:20 | |
| *** l4yerffeJ has joined #openstack-kolla | 16:21 | |
| openstackgerrit | Mohammed Naser proposed openstack/kolla master: Remove bifrost-base from checks to fix gate https://review.openstack.org/448666 | 16:21 |
| mnaser | inc0 sdake Jeffrey4l_ ^ | 16:22 |
| *** Jeffrey4l_ has quit IRC | 16:22 | |
| *** Jeffrey4l_ has joined #openstack-kolla | 16:22 | |
| *** Jeffrey4l_ has quit IRC | 16:30 | |
| *** Jeffrey4l_ has joined #openstack-kolla | 16:30 | |
| *** gfhellma has joined #openstack-kolla | 16:32 | |
| *** Jeffrey4l_ has quit IRC | 16:32 | |
| *** Jeffrey4l_ has joined #openstack-kolla | 16:32 | |
| nea1 | is there a way to get horizon running behind an reverse proxy, I can't really find anything useful via google | 16:37 |
| nea1 | or better said, it tries to redirect to http if requesting / and login failes | 16:37 |
| *** rmart04 has quit IRC | 16:38 | |
| *** matrohon has quit IRC | 16:39 | |
| *** matrohon has joined #openstack-kolla | 16:44 | |
| *** eaguilar has quit IRC | 16:45 | |
| sbezverk | kfox1111: when I tried to place iptables -F before setup_iptables, everything got broken | 16:46 |
| sbezverk | kfox1111: now trying to do it almost right after | 16:46 |
| *** rwallner has joined #openstack-kolla | 16:47 | |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla-kubernetes master: WIP Move config generation to kolla-k8s https://review.openstack.org/448258 | 16:47 |
| *** Jeffrey4l has joined #openstack-kolla | 16:49 | |
| *** unicell has quit IRC | 16:49 | |
| *** Jeffrey4l is now known as test_user | 16:49 | |
| *** unicell has joined #openstack-kolla | 16:49 | |
| *** eaguilar has joined #openstack-kolla | 16:54 | |
| openstackgerrit | Bertrand Lallau proposed openstack/kolla-ansible master: Remove deprecated options nova-network related https://review.openstack.org/448685 | 16:58 |
| mnaser | ok | 17:00 |
| mnaser | fun time | 17:00 |
| mnaser | the build failed | 17:00 |
| mnaser | because of horizon now | 17:00 |
| mnaser | on xenial | 17:00 |
| *** unicell has quit IRC | 17:00 | |
| kfox1111 | sbezverk: hmm.... maybe its breaking on the iptables rewriting rules that are in the gate. | 17:00 |
| *** egonzalez has quit IRC | 17:01 | |
| openstackgerrit | Bertrand Lallau proposed openstack/kolla-ansible master: Remove nova-network deprecated options https://review.openstack.org/448685 | 17:01 |
| *** kencjohnston has joined #openstack-kolla | 17:01 | |
| mnaser | http://logs.openstack.org/66/448666/1/check/gate-kolla-dsvm-build-ubuntu-source-ubuntu-xenial/886b1ff/console.html#_2017-03-22_16_49_22_173095 | 17:02 |
| *** rwsu has quit IRC | 17:02 | |
| *** sayantani01 has quit IRC | 17:02 | |
| mnaser | http://stackoverflow.com/a/40185488 | 17:02 |
| inc0 | brb, need some food | 17:02 |
| mnaser | i think its because its missing the netbase package i guess | 17:02 |
| *** sayantani01 has joined #openstack-kolla | 17:03 | |
| spsurya | inc0: please hv some | 17:03 |
| duonghq | sdake, about recheck topic, it'll automatic rebase before recheck? | 17:03 |
| mnaser | https://bugs.launchpad.net/cloud-images/+bug/1673930 | 17:03 |
| openstack | Launchpad bug 1673930 in cloud-images "ubuntu 16.04 docker image should include netbase" [Undecided,Invalid] | 17:03 |
| lazyPower | inc0: sdake: Ping me when you're ready to head over to openstack-infra, i'm goign to step out for 20 to grab a quick bite to eat myself and will be here for the remainder of my workday. Thanks for the assist | 17:05 |
| openstackgerrit | Mohammed Naser proposed openstack/kolla master: Install netbase package for Horizon https://review.openstack.org/448707 | 17:08 |
| mnaser | fack i think i broke the ordering of | 17:08 |
| mnaser | any idea how we can fix this guys | 17:10 |
| mnaser | the horizon fix would never pass gate, but the gate fix would never pass without the horizon one | 17:11 |
| duonghq | need to go to bed, cya guys | 17:11 |
| *** matrohon has quit IRC | 17:11 | |
| duonghq | mnaser, can you explain? | 17:12 |
| Jeffrey4l_ | mnaser, fix the two issue in one patches. | 17:13 |
| mnaser | duonghq this patch is meant to unblock the gate by disabling bifrost - https://review.openstack.org/#/c/448666/, but its failing because horizon is missing a package - https://review.openstack.org/#/c/448707 | 17:13 |
| *** magicboiz has joined #openstack-kolla | 17:13 | |
| mnaser | ok, ill let the second one pass tests and once its confirmed ill put them all into 1 | 17:14 |
| Jeffrey4l_ | voting jobs sometime annoying. ;) | 17:14 |
| duonghq | mnaser, you mean that you discover horizon bug while trying to fix bifrost bug? | 17:14 |
| Jeffrey4l_ | duonghq, a ubuntu base image bug actually. | 17:15 |
| mnaser | yeah. once bifrost was unblocked, horizon was failing because /etc/protocol is missing | 17:15 |
| duonghq | Jeffrey4l_, roger | 17:15 |
| duonghq | mnaser, so if we want gate green, we should put it together? | 17:15 |
| mnaser | duonghq yep seems like it, so ill do that once the 2nd change is tested (i made it dependent) | 17:15 |
| *** blallau has quit IRC | 17:16 | |
| duonghq | saw it | 17:16 |
| duonghq | :) | 17:16 |
| duonghq | bye guys | 17:16 |
| *** duonghq has quit IRC | 17:17 | |
| Jeffrey4l_ | duonghq, bye. | 17:18 |
| Jeffrey4l_ | inc0, check release next z stream kolla mitaka and newton branch https://review.openstack.org/448711 https://review.openstack.org/448709 | 17:21 |
| *** zhubingbing__ has quit IRC | 17:23 | |
| *** gfhellma has quit IRC | 17:23 | |
| *** gfhellma has joined #openstack-kolla | 17:23 | |
| *** test_user has quit IRC | 17:23 | |
| *** Jeffrey4l has joined #openstack-kolla | 17:24 | |
| inc0 | kfox1111: sbezverk how to debug this? http://logs.openstack.org/58/448258/6/check/gate-kolla-kubernetes-deploy-centos-binary-2-ceph-nv/1f7ffac/console.html#_2017-03-22_17_10_50_226143 | 17:26 |
| *** Jeffrey4l has quit IRC | 17:26 | |
| nea1 | how is the workflow if you want to change some config options? is there a way to roleout changes? or is there somewhere a way to overwrite settings in for e.g. nova.conf and role them out? | 17:26 |
| *** test_user has joined #openstack-kolla | 17:27 | |
| sbezverk | inc0: you need to check logs/pods | 17:30 |
| sbezverk | and check which container is crashing and why | 17:30 |
| *** test_user has quit IRC | 17:30 | |
| *** mbruzek has quit IRC | 17:31 | |
| *** unicell has joined #openstack-kolla | 17:33 | |
| inc0 | sbezverk: where are actual logs? | 17:33 |
| inc0 | pods doesn't seem to be helpful | 17:34 |
| openstackgerrit | Mohammed Naser proposed openstack/kolla master: Gate fixes: disable bifrost-base, fix ubuntu horizon image https://review.openstack.org/448666 | 17:34 |
| inc0 | ok got it...no amqp server on rabbitmq:5672 | 17:35 |
| sbezverk | inc0: logs/openstack some logs are there | 17:36 |
| *** satyar has joined #openstack-kolla | 17:36 | |
| inc0 | http://logs.openstack.org/58/448258/6/check/gate-kolla-kubernetes-deploy-centos-binary-2-ceph-nv/1f7ffac/logs/openstack/kolla-neutron-server-1004042052-c65h5.txt yeah... | 17:37 |
| sbezverk | inc0: looks like a connectivity or password issue | 17:37 |
| sbezverk | when netron tries to connect to rabbitmq | 17:38 |
| inc0 | nvm got it | 17:39 |
| inc0 | (I hate gates..) | 17:39 |
| inc0 | I mena i love gates | 17:39 |
| inc0 | ...it's complicated relationship | 17:40 |
| *** skramaja has quit IRC | 17:40 | |
| *** jaosorior has quit IRC | 17:40 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes master: WIP Adding virtualbmc to the gate https://review.openstack.org/441550 | 17:41 |
| sbezverk | inc0: I share the same sentiment ;) | 17:41 |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla-kubernetes master: WIP Move config generation to kolla-k8s https://review.openstack.org/448258 | 17:43 |
| sdake | inc0 did lazyPower get the questions answered he needed on openstack-infra? | 17:46 |
| lazyPower | sdake: we have not engaged with them, however I'll pick this up tomorrow. I still need to confer with management | 17:47 |
| sdake | lazyPower roger | 17:47 |
| sdake | lazyPower sorry for the bad information - not sure who gave it to ya - this is why we have historically discouraged third p arty gating - because everyone wants to promote it ot voting and from my understanindg this is not possible | 17:48 |
| inc0 | what is potential alternative is to not have 3rd party ci, just use your methodology of deployment in regular CI | 17:48 |
| inc0 | that would allow gates to be voting | 17:49 |
| *** MasterOfBugs has quit IRC | 17:49 | |
| *** tonanhngo has joined #openstack-kolla | 17:49 | |
| *** pbandark has quit IRC | 17:49 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/kolla stable/mitaka: Updated from global requirements https://review.openstack.org/448723 | 17:50 |
| sdake | lazyPower if you want to get to voting ci, you could always integrate directly with the gate | 17:52 |
| sdake | lazyPower its a bit of a bear to sort out and personally I'd wait until after zuulv3 to begin the work | 17:52 |
| sdake | lazyPower however doing it now will give you the experience you need to make it work when zuulv3 is out | 17:52 |
| *** tonanhngo has quit IRC | 17:54 | |
| mnaser | oh | 17:54 |
| mnaser | this bug is bigger than i thought.. | 17:54 |
| *** tonanhngo has joined #openstack-kolla | 17:54 | |
| mnaser | it seems like everything affecting eventlet | 17:54 |
| mnaser | is getting socket.error: protocol not found | 17:54 |
| mnaser | aka keystone fails to deploy | 17:54 |
| mnaser | i will assume that we can all agree installing netbase to provide /etc/protocols is fairly critical and i'll tack it into base | 17:55 |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla-kubernetes master: WIP: Playbook to add configmaps and secrets https://review.openstack.org/447733 | 17:55 |
| inc0 | mnaser: strange that we didn't pick it up | 17:56 |
| mnaser | inc0 it must be a recent change | 17:56 |
| openstackgerrit | Mohammed Naser proposed openstack/kolla master: Gate fixes: disable bifrost-base, fix ubuntu base image https://review.openstack.org/448666 | 17:56 |
| inc0 | but yeah adding one more package to base is not a big deal and this is potentially critical | 17:57 |
| mnaser | i thought it was horizon, but then keystone deploys were failing, so its in base now | 17:57 |
| *** mgoddard has quit IRC | 17:57 | |
| inc0 | mnaser: well horizon and ks uses apache | 17:57 |
| sdake | mnaser it may be a change in master requirements | 17:57 |
| mnaser | the traceback seems to stem off from eventlet | 17:57 |
| mnaser | http://logs.openstack.org/66/448666/2/check/gate-kolla-dsvm-deploy-ubuntu-source-ubuntu-xenial-nv/3aaae1c/console.html#_2017-03-22_17_50_35_416589 | 17:57 |
| sdake | mnaser which tend to pull stuff in that isn't in base | 17:57 |
| mnaser | anyways im going to grab something to eat but hopefully that clears out whatever is left.. | 17:58 |
| inc0 | well thing is, more and more projects wants to use appache | 17:58 |
| inc0 | so let's just install it in base to be on safe side | 17:58 |
| sdake | there is a community goal to use uswgi | 17:58 |
| sdake | whatever that is :) | 17:58 |
| mnaser | having /etc/protocols there will probably clear out any potential issues | 17:59 |
| mnaser | many things probably expect it | 17:59 |
| sdake | either a proposal or its approved - not sure | 17:59 |
| mnaser | who knows why its not there /shrug | 17:59 |
| mnaser | sdake keystone is already there | 17:59 |
| mnaser | but then keystone have always been the first to dive into these :P | 17:59 |
| inc0 | one will lead to another | 17:59 |
| *** rwsu has joined #openstack-kolla | 18:02 | |
| mnaser | on a fun note | 18:05 |
| *** scsnow_ has quit IRC | 18:05 | |
| mnaser | the patch review id for the gate fix is 448666 | 18:05 |
| mnaser | gate is source of all evil: confirmed? | 18:05 |
| kfox1111 | mnaser: keystone's doing it for a completely different reason. | 18:05 |
| kfox1111 | mnaser: they did it because they didn't want o have to reimplement all the authentication plugins apache supports. | 18:06 |
| kfox1111 | none of the other projects care about that, as they just have to validate keystone tokens. | 18:06 |
| *** athomas has quit IRC | 18:09 | |
| *** mbruzek has joined #openstack-kolla | 18:10 | |
| *** mbruzek has quit IRC | 18:10 | |
| *** mbruzek has joined #openstack-kolla | 18:10 | |
| *** gfhellma has quit IRC | 18:10 | |
| sdake | mnaser my son says "illuminati confirmed" all the time - I respond with "YESS." | 18:12 |
| *** gfhellma has joined #openstack-kolla | 18:17 | |
| *** MasterOfBugs has joined #openstack-kolla | 18:23 | |
| sdake | to all our friends in the UK, hope everyone is safe | 18:24 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes master: WIP Adding virtualbmc to the gate https://review.openstack.org/441550 | 18:24 |
| sbezverk | sdake: what happened in uk? | 18:25 |
| sdake | http://www.foxnews.com/world/2017/03/22/uk-parliament-terror-attack-4-people-killed-including-cop-and-alleged-assailant-20-injured.html | 18:25 |
| *** rmart04 has joined #openstack-kolla | 18:25 | |
| sbezverk | sdake: oh did not know that, indeed hopefully everybody safe.. | 18:27 |
| sdake | sbezverk i just found out 5 minutes ago or so | 18:27 |
| sdake | sbezverk i rarely if ever wtch the news and not during the day | 18:27 |
| *** eaguilar_ has joined #openstack-kolla | 18:32 | |
| *** eaguilar has quit IRC | 18:32 | |
| *** rmart04 has quit IRC | 18:38 | |
| *** rmart04_ has joined #openstack-kolla | 18:38 | |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla-kubernetes master: WIP: Playbook to add configmaps and secrets https://review.openstack.org/447733 | 18:44 |
| inc0 | kfox1111: sbezverk sooo....http://logs.openstack.org/74/446274/18/check/gate-kolla-kubernetes-deploy-centos-binary-2-external-ovs-nv/fef4f61/logs/configmaps.yaml | 18:50 |
| inc0 | (this is fernet change) | 18:51 |
| inc0 | ctrl+f "ml2-conf.ini" | 18:51 |
| inc0 | real file is ml2_conf.ini | 18:51 |
| inc0 | ...any reason why it's changed? | 18:51 |
| inc0 | let me guess...it doesnt?;) | 18:54 |
| inc0 | doesn't work* | 18:54 |
| *** rwallner has quit IRC | 18:55 | |
| *** rwallner has joined #openstack-kolla | 18:59 | |
| *** harlowja has quit IRC | 18:59 | |
| inc0 | bbiaf lunch | 18:59 |
| *** eaguilar has joined #openstack-kolla | 18:59 | |
| *** harlowja has joined #openstack-kolla | 19:00 | |
| *** harlowja has quit IRC | 19:01 | |
| *** jmccarthy1 has quit IRC | 19:01 | |
| *** eaguilar_ has quit IRC | 19:03 | |
| *** adrian_otto has quit IRC | 19:07 | |
| *** Guest52040 has quit IRC | 19:08 | |
| *** rwsu has quit IRC | 19:15 | |
| sdake | inc0 http://lists.openstack.org/pipermail/user-committee/2017-March/001856.html | 19:16 |
| *** shardy has quit IRC | 19:21 | |
| vhosakot | kfox1111: sbezverk: making slow progess.. so, is the k8s dashboard needed? I'm seeing issues setting it up? have you set up the dashboard on kubeadm? | 19:21 |
| *** adrian_otto has joined #openstack-kolla | 19:34 | |
| *** krtaylor has quit IRC | 19:36 | |
| openstackgerrit | Merged openstack/kolla-ansible master: Cleanup requirements.txt file https://review.openstack.org/448634 | 19:40 |
| *** harlowja has joined #openstack-kolla | 19:45 | |
| *** deadnull has quit IRC | 19:46 | |
| *** shashank_t_ has joined #openstack-kolla | 19:49 | |
| *** fguillot has quit IRC | 19:50 | |
| *** rwallner has quit IRC | 19:53 | |
| inc0 | sdake: I saw that, on my todo list | 19:53 |
| inc0 | thanks | 19:53 |
| sdake | inc0 cool | 19:53 |
| sdake | inc0 i just saw it | 19:53 |
| sdake | i've got filters for my filters - thus finding information is difficult :) | 19:54 |
| *** rmart04_ has quit IRC | 19:55 | |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla-kubernetes master: WIP: Playbook to add configmaps and secrets https://review.openstack.org/447733 | 19:57 |
| stevebaker | inc0, sdake: hey, could you please take a look at this? https://review.openstack.org/#/c/448380/ | 19:57 |
| sdake | stevebaker our voting gate is broken | 19:58 |
| sdake | stevebaker we can look - however - it wont merge :( | 19:58 |
| stevebaker | sdake: just for discussion | 19:58 |
| sdake | stevebaker looking moment | 19:58 |
| inc0 | stevebaker: this is good point, I'll put it into next weeks agenda | 19:59 |
| inc0 | so far automated version caching causes a lot of problems:/ | 20:00 |
| stevebaker | sdake, inc0: while we're on the subject of contentious changes ;) https://review.openstack.org/#/c/442923/ | 20:00 |
| inc0 | uhh moving from docker-py to cli is...not ideal | 20:02 |
| inc0 | lots of screen scrapping | 20:02 |
| inc0 | I'd rather implement --volume to docker-py | 20:03 |
| stevebaker | inc0: but that would never be added to 1.x, and moving to 2.x is impractical for years due to API breakage and downstream packaging | 20:04 |
| *** gfhellma has quit IRC | 20:05 | |
| inc0 | yeah...but but...screenscraping | 20:05 |
| inc0 | I still sometimes have nightmares from time when we used screenscraping to detect disks from ceph | 20:06 |
| inc0 | for ceph* | 20:06 |
| stevebaker | inc0: by screenscraping I assume you mean parsing the response from the docker call. There is none of that required. Only checking the returncode for failure or success. | 20:07 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes master: WIP: Fernet Token Support https://review.openstack.org/446274 | 20:08 |
| sdake | stevebaker this one has been responded to: https://review.openstack.org/#/c/448380/1 | 20:08 |
| stevebaker | inc0: but even if response parsing is needed, the docker command has nice format templating | 20:08 |
| *** satyar has quit IRC | 20:10 | |
| inc0 | stevebaker: out of curiosity...why --volume on build? | 20:11 |
| stevebaker | inc0: when I thought that the volumes could be read-write I wanted to try mounting in a yum cache, but they're read-only | 20:12 |
| stevebaker | inc0: it could still be useful to mount in the host's /etc/yum.repos.d/ for more complex repo setups though, but I'll admit I don't have a solid use case for --volume on image build right now | 20:14 |
| sdake | stevebaker both reviews have been responded to | 20:14 |
| sdake | enjoy | 20:14 |
| stevebaker | sdake: thanks | 20:14 |
| inc0 | stevebaker: so I'd encourate to use overrides instead of this | 20:14 |
| inc0 | easier to reproduce builds and such | 20:15 |
| inc0 | stevebaker: https://docs.openstack.org/developer/kolla/image-building.html#dockerfile-customisation | 20:15 |
| stevebaker | inc0: yeah, overrides seems flexible enough | 20:16 |
| inc0 | repo changes was one of major use cases they were designed for | 20:16 |
| inc0 | so if you run into issue that overrides can't be used, I'd love to hear about it | 20:16 |
| sdake | elrog 845s in the mail | 20:17 |
| sdake | about time | 20:17 |
| sdake | my tube blew 3 months ago iirc | 20:18 |
| * sdake sadfaces | 20:18 | |
| *** rwsu has joined #openstack-kolla | 20:18 | |
| *** harlowja has quit IRC | 20:18 | |
| stevebaker | inc0: do you think its worth attempting to speed up package download times during image build? There is a lot of repeated downloads when building multiple images | 20:21 |
| inc0 | stevebaker: I'd do it with creating repo cache | 20:21 |
| inc0 | as in, https://help.ubuntu.com/community/Apt-Cacher-Server | 20:22 |
| sdake | stevebaker what did you ahve in mind | 20:22 |
| stevebaker | inc0: yes, I'm wondering about an ephemeral container which implements a caching proxy that runs for the length of the image build | 20:22 |
| inc0 | well, why ephemeral?:) | 20:23 |
| stevebaker | inc0: then kolla-build can manage its lifecycle, no external management of yet another long-lived thing required. It could be backed by a volume if you want the cache to persiste between runs | 20:24 |
| *** adrian_otto has quit IRC | 20:24 | |
| inc0 | well, I'm trying to limit amount of stuff in Kolla tbh. To me it would be great to write a doc piece or sth on how to run cache server to speed up | 20:25 |
| inc0 | https://docs.docker.com/engine/examples/apt-cacher-ng/ | 20:25 |
| *** srwilkers has quit IRC | 20:26 | |
| inc0 | then whole thing would be as simple as "find a good cache dockerfile, docker run path-to-this-image" | 20:26 |
| stevebaker | inc0: but downloading a package once is better than downloading it hundreds of times, so starting with an empty cache on every run would be a reasonable tradeoff between speeding runs and issues caused by stale caches | 20:26 |
| inc0 | ofc, but we don't have *that* much of overlap, I mean whole build process takes circa 15min, and that's once per cluster | 20:27 |
| sdake | inc0 not if your behind a proxy | 20:27 |
| inc0 | once you build it, you run local registry with it | 20:27 |
| sdake | then it takes 3 hours | 20:27 |
| inc0 | if you're behind proxy...run cache server, not just for your docke rimages...for your mental health too | 20:28 |
| sdake | you assume I have any :) | 20:28 |
| inc0 | I mean stevebaker I don't argue whether or not it's sensible | 20:28 |
| inc0 | I totally agree it is | 20:28 |
| inc0 | I'm just not sure if this should be in Kolla. But as part of contrib dockerfiles, I won't have issue at all | 20:29 |
| stevebaker | inc0: just the extent to which its integrated with kolla-build | 20:29 |
| sdake | stevebaker what is the interest in speeding up builds - out of curiosity | 20:29 |
| sdake | stevebaker on NVME750 SSD, build everything takes 15-20 mins - in the gate about 30 mins | 20:30 |
| inc0 | sdake: unless you're behind proxy | 20:30 |
| inc0 | that's the point | 20:30 |
| sdake | stevebaker what you propose is increased complexity for what tradeoff? | 20:30 |
| inc0 | stevebaker: seen this? http://pulpproject.org/demo/ | 20:31 |
| inc0 | if I were you, I'd run server like that somewhere with nightly updates | 20:31 |
| inc0 | and install packages from it | 20:31 |
| inc0 | that's what we did with my friends back when I didn't have access to stuff outside of proxy | 20:32 |
| stevebaker | sdake: personally, I'm at the wrong end of the world for fast downloads. But for tripleo CI we'd like to get to the point where we're rebuilding images on every CI run, *then* provisioning an overcloud and running tests. 30 minutes for a CI image build is a big hit for jobs which are already really long | 20:32 |
| inc0 | and we all used same caching repo | 20:32 |
| jascott1 | inc0 meeting | 20:32 |
| sdake | stevebaker - i'd recommend pulling the built images from tarballs.oo | 20:33 |
| sdake | stevebaker we could probably get master in tarballs.oo i'd think | 20:34 |
| stevebaker | sdake: we will for most jobs, but that doesn't help when the image needs to contain a change to test | 20:34 |
| sdake | i see | 20:34 |
| sdake | stevebaker right re NZ - internet there is pokey | 20:35 |
| sdake | stevebaker i guess it depends on how much it speeds things up - most of the docker usage is disk IO bound | 20:35 |
| sdake | not really networking bound | 20:35 |
| stevebaker | inc0: really all that is needed is a squid cache - then it doesn't need to be apt/yum/pip aware | 20:35 |
| sdake | have you tried a cache at all? | 20:35 |
| stevebaker | sdake: not yet, I agree it might not be a significant speedup | 20:36 |
| sdake | i have often wanted to set it up and see if i t makes a difference for me - with 1gb internet and 750 NVME SSDs | 20:36 |
| *** rhallisey has quit IRC | 20:36 | |
| sdake | stevebaker i think the first step there is to see if it actually makes a difference and under which circumstances | 20:37 |
| inc0 | stevebaker: for people with normal download speeds it's not really problem | 20:37 |
| stevebaker | yep | 20:37 |
| sdake | its a problem for china - Jeffrey4l_ complains about it all the time | 20:37 |
| inc0 | squid cache would solve it | 20:37 |
| inc0 | and again, we might justr document how to set it up | 20:37 |
| *** eaguilar has quit IRC | 20:38 | |
| inc0 | since there are tons of squid images in dockerhub | 20:38 |
| sdake | ya docs seem like a good solution here - although don't tap me to write em :) | 20:38 |
| *** crushil has joined #openstack-kolla | 20:40 | |
| *** crushil has left #openstack-kolla | 20:43 | |
| *** eaguilar has joined #openstack-kolla | 20:46 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes master: WIP Adding virtualbmc to the gate https://review.openstack.org/441550 | 20:47 |
| *** mkoderer has quit IRC | 20:47 | |
| inc0 | sbezverk: kfox1111 http://logs.openstack.org/33/447733/12/check/gate-kolla-kubernetes-deploy-centos-binary-2-ceph-nv/99edf6a/logs/pods/kolla-neutron-manage-db-1qm96.txt need help here | 20:48 |
| sbezverk | inc0: looking | 20:48 |
| inc0 | seems like missing configmap, but strange one | 20:48 |
| sbezverk | inc0: secret name looks wrong SecretName:default-token-bf711 | 20:49 |
| *** goldyfruit has quit IRC | 20:50 | |
| inc0 | sbezverk: I think this is k8s-prepared token to communicate in cluster | 20:50 |
| sbezverk | inc0: do not think so, it is name db job neutron-manage-db | 20:51 |
| inc0 | yeah, what I'm saying is k8s adds it to every pod | 20:51 |
| sbezverk | we pass neutron's db password so it could connect to mariadb | 20:51 |
| inc0 | http://logs.openstack.org/33/447733/12/check/gate-kolla-kubernetes-deploy-centos-binary-2-ceph-nv/99edf6a/logs/pods/kolla-mariadb-0.txt | 20:51 |
| inc0 | this is working mariadb pod, also has this secet | 20:52 |
| inc0 | but what is service-configmap? | 20:52 |
| inc0 | just a configmap with name neutron-server? | 20:53 |
| sbezverk | inc0: service-configmap is just volume name | 20:54 |
| sbezverk | but the configmap name should be neutron-server | 20:54 |
| sbezverk | it looks like it does not find it | 20:54 |
| inc0 | but it's right there...:( http://logs.openstack.org/33/447733/12/check/gate-kolla-kubernetes-deploy-centos-binary-2-ceph-nv/99edf6a/logs/configmaps.yaml | 20:55 |
| sbezverk | inc0: hm, strange indeed it is there.. | 20:58 |
| *** matrohon has joined #openstack-kolla | 21:02 | |
| inc0 | "references non existant key" | 21:03 |
| inc0 | would be helpful to see what's the key | 21:03 |
| *** jrobinson has joined #openstack-kolla | 21:03 | |
| sbezverk | please add --debug option to neutron-manage-db job | 21:04 |
| inc0 | where you add it | 21:05 |
| inc0 | ? | 21:05 |
| sbezverk | in the ceph workflow | 21:05 |
| sbezverk | one sec, let me find it | 21:05 |
| inc0 | thank you | 21:05 |
| kfox1111 | inc0: "references non existant key?" | 21:05 |
| kfox1111 | that sounds like a yaml error, not a missing cnfigmap? | 21:05 |
| inc0 | http://logs.openstack.org/33/447733/12/check/gate-kolla-kubernetes-deploy-centos-binary-2-ceph-nv/99edf6a/logs/pods/kolla-neutron-manage-db-1qm96.txt kfox1111 | 21:06 |
| kfox1111 | looking.... | 21:06 |
| *** krtaylor has joined #openstack-kolla | 21:06 | |
| sbezverk | inc0: tests/bin/ceph_workflow.sh | 21:07 |
| kfox1111 | hmm.... | 21:07 |
| sbezverk | inc0: there is loop to create manage-db jobs | 21:07 |
| sbezverk | inc0: let's add --debug to helm install command line | 21:07 |
| kfox1111 | so, first, it might help debugging to not be making so many changes at once. maybe pick one configmap and do that one thing first? | 21:09 |
| inc0 | well I fail one pod at once;) | 21:10 |
| kfox1111 | oh... | 21:10 |
| kfox1111 | so.... | 21:10 |
| kfox1111 | did you match up the item names in the generated configmap for neutron-server? | 21:10 |
| inc0 | well, fun fact, names don't match files | 21:10 |
| sbezverk | inc0: you miss + sign | 21:10 |
| sbezverk | between files in your configmap I think | 21:11 |
| inc0 | we had ml2-conf instead of ml2_conf | 21:11 |
| kfox1111 | inc0: right. the filenames contain characters that are invalid in k8s configmaps. | 21:11 |
| inc0 | ahh | 21:11 |
| kfox1111 | so you gotta use valid ones. | 21:11 |
| sbezverk | ml2-conf.ini: |+ | 21:11 |
| inc0 | _ is invalid in configmap? | 21:11 |
| sbezverk | this is how it should be | 21:11 |
| inc0 | sbezverk: yeah I'll add this too | 21:12 |
| sbezverk | inc0: but you have ml2-conf.ini: | | 21:12 |
| *** jtriley has quit IRC | 21:12 | |
| inc0 | question, will that create ml2-conf.ini file | 21:12 |
| inc0 | ?? | 21:12 |
| inc0 | will that even work with neutron? | 21:12 |
| *** adrian_otto has joined #openstack-kolla | 21:12 | |
| kfox1111 | inc0: https://github.com/openstack/kolla-kubernetes/blob/master/helm/microservice/neutron-server-deployment/templates/neutron-server.yaml#L21 | 21:12 |
| inc0 | ahh | 21:12 |
| inc0 | that explains it | 21:12 |
| inc0 | duh | 21:13 |
| kfox1111 | yeah, kubectl create configmap..... will change _'s to -'s while building. | 21:13 |
| inc0 | unless you specify _ in your yaml;) | 21:13 |
| kfox1111 | but I don't think k8s will like that. | 21:14 |
| inc0 | http://logs.openstack.org/33/447733/12/check/gate-kolla-kubernetes-deploy-centos-binary-2-ceph-nv/99edf6a/logs/configmaps.yaml _ is there | 21:14 |
| inc0 | and configmap gets created | 21:14 |
| inc0 | let me check locally if it likes that or not | 21:14 |
| kfox1111 | yeah, it doesnt' validate it. | 21:14 |
| kfox1111 | not sure it works though. | 21:14 |
| kfox1111 | on the volume side. | 21:14 |
| kfox1111 | or what other sideeffects it may have. | 21:14 |
| inc0 | let me try to change neutron server too and see if it works | 21:14 |
| kfox1111 | the k8s object maps it back to the right file names though. | 21:15 |
| kfox1111 | you can try dropping the whole items chunk. | 21:15 |
| kfox1111 | but I would be against doing that perminantly. | 21:15 |
| kfox1111 | k8s did it for a reason. I don't think its a good idea to second guess them, not knowing why. | 21:15 |
| inc0 | I'll read through that | 21:16 |
| kfox1111 | your config gen code shoudl probably do the same thing kubectl create configmap does. | 21:16 |
| inc0 | I mean...switching - with _ and around is strange | 21:16 |
| * kfox1111 shrugs | 21:16 | |
| inc0 | easy fix tho | 21:16 |
| kfox1111 | thats the tool we're using. | 21:16 |
| kfox1111 | we shouild use it as designed. | 21:16 |
| inc0 | np | 21:17 |
| inc0 | I'll switch _ with - ;) | 21:17 |
| kfox1111 | k | 21:17 |
| mnaser | hi | 21:17 |
| mnaser | gate failed | 21:17 |
| mnaser | a | 21:17 |
| mnaser | g | 21:17 |
| mnaser | ain | 21:17 |
| mnaser | wait | 21:17 |
| mnaser | 2017-03-22 18:39:52.488367 | Length too long: 4798716 | 21:17 |
| mnaser | http://logs.openstack.org/66/448666/3/check/gate-kolla-dsvm-build-centos-source-centos-7/e745bb1/console.html#_2017-03-22_18_39_52_488367 anyone ever seen it? | 21:17 |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla-kubernetes master: WIP: Playbook to add configmaps and secrets https://review.openstack.org/447733 | 21:19 |
| inc0 | mnaser: eee.. | 21:19 |
| mnaser | and after 45 minutes it just bows up like that | 21:20 |
| mnaser | with no logs | 21:20 |
| mnaser | i think we have a recursion bug somewhere here | 21:21 |
| mnaser | ugh | 21:26 |
| mnaser | why is there no traceback | 21:26 |
| openstackgerrit | Bertrand Lallau proposed openstack/kolla-ansible master: Remove Ubuntu 14.04 related code https://review.openstack.org/448792 | 21:26 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes master: WIP: Fernet Token Support https://review.openstack.org/446274 | 21:27 |
| sbezverk | kfox1111: I HATE LINUX BRIDGE!!!!! | 21:27 |
| inc0 | :D | 21:27 |
| sbezverk | After plugging VM into br-tenants it started working | 21:27 |
| sbezverk | kfox1111: baremetal vm + 2 normal vms :) | 21:29 |
| kfox1111 | sbezverk: +1 | 21:29 |
| kfox1111 | sbezverk: nice. :) | 21:29 |
| kfox1111 | I use to prefer linux bridge, as it is somewhat simpler then ovs. | 21:29 |
| kfox1111 | but ovs seems much more reliable. | 21:29 |
| kfox1111 | so I use it for everything now. | 21:30 |
| sbezverk | kfox1111: yeah.. Ubuntu though is still problematic.. need to find ubuntu/libvirt expert | 21:32 |
| *** Pavo has joined #openstack-kolla | 21:32 | |
| kfox1111 | heh. yeah. | 21:32 |
| kfox1111 | I was having problems with ubuntu/libvirt in 4.x too? | 21:32 |
| *** harlowja has joined #openstack-kolla | 21:33 | |
| *** l4yerffeJ has quit IRC | 21:34 | |
| *** Jeffrey4l_ has quit IRC | 21:34 | |
| *** mbruzek has quit IRC | 21:35 | |
| *** jascott1 has quit IRC | 21:37 | |
| *** inc0 has quit IRC | 21:38 | |
| *** manjeets has quit IRC | 21:38 | |
| *** pcaruana has quit IRC | 21:47 | |
| *** eaguilar has quit IRC | 21:47 | |
| *** Jeffrey4l_ has joined #openstack-kolla | 21:48 | |
| *** sayantani01 has quit IRC | 21:48 | |
| *** sayantani01 has joined #openstack-kolla | 21:49 | |
| *** Jeffrey4l_ has quit IRC | 21:50 | |
| *** Jeffrey4l_ has joined #openstack-kolla | 21:50 | |
| *** david-lyle has quit IRC | 21:52 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes master: Virtualbmc and deployment of baremetal VM using ironic https://review.openstack.org/441550 | 21:53 |
| *** masber has quit IRC | 21:56 | |
| *** inc0 has joined #openstack-kolla | 21:57 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes master: WIP: Fernet Token Support https://review.openstack.org/446274 | 21:57 |
| inc0 | network on vm died: | 21:57 |
| inc0 | :( | 21:57 |
| kfox1111 | inc0: which vm? | 21:58 |
| sbezverk | kfox1111: if you have time, please take a look at PS, need your review :) | 21:59 |
| inc0 | kfox1111: my vm I work from | 22:01 |
| inc0 | sbezverk: link plz | 22:01 |
| inc0 | ahh that was to kfox | 22:01 |
| inc0 | vm died, my brain followed | 22:01 |
| kfox1111 | sbezverk: k. will do. | 22:02 |
| kfox1111 | inc0: ah. ok. | 22:02 |
| sbezverk | https://review.openstack.org/441550 | 22:05 |
| sbezverk | need to start cleaning it up.. | 22:06 |
| *** turvey has quit IRC | 22:08 | |
| *** rwsu has quit IRC | 22:10 | |
| kfox1111 | sbezverk: nova-cell0-create-db-job? | 22:11 |
| inc0 | brb | 22:12 |
| inc0 | gonna reboot | 22:12 |
| inc0 | 22:12:29 up 50 days, 23:17, 11 users, load average: 0.14, 0.05, 0.01 | 22:12 |
| sbezverk | kfox1111: yeah with cell in 4.0.0 nothing starts up | 22:12 |
| inc0 | good bye sweet prince | 22:12 |
| sbezverk | I meant without | 22:13 |
| kfox1111 | sbezverk: so cells are required? | 22:13 |
| sbezverk | kfox1111: yep :( | 22:13 |
| sbezverk | in ocata | 22:13 |
| kfox1111 | should we templatize that instead of hardcoding it? | 22:13 |
| kfox1111 | --set cell_id=0 or something? | 22:13 |
| sbezverk | kfox1111: there is a cell0 which is default | 22:13 |
| kfox1111 | then it can be used to create other cells. | 22:13 |
| sbezverk | then if you need, you create additional cells | 22:13 |
| *** inc0 has quit IRC | 22:13 | |
| kfox1111 | is the code different for different cells? | 22:13 |
| sbezverk | it is like default context if you do not really need cells | 22:14 |
| sbezverk | btw kolla ansible also create defauly cell 0 | 22:14 |
| *** matrohon has quit IRC | 22:15 | |
| kfox1111 | does creating cell1 require anything different then crating cell0 other then id? | 22:15 |
| *** david-lyle has joined #openstack-kolla | 22:15 | |
| SamYaple | kfox1111: no | 22:16 |
| SamYaple | but cell0 must be named cell0, the other cells can be named other things | 22:16 |
| SamYaple | cell0 is a hardcoded throughout nova default | 22:16 |
| SamYaple | just fyi | 22:16 |
| *** krtaylor has quit IRC | 22:16 | |
| SamYaple | as fair as I know, it is the only special cell | 22:16 |
| *** shashank_t_ has quit IRC | 22:17 | |
| kfox1111 | k. | 22:17 |
| kfox1111 | just feels a little weird to have a package with cell0 in its name rather then template it out.... | 22:17 |
| *** inc0 has joined #openstack-kolla | 22:19 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/kolla master: Updated from global requirements https://review.openstack.org/446771 | 22:20 |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla-kubernetes master: WIP Move config generation to kolla-k8s https://review.openstack.org/448258 | 22:20 |
| *** masber has joined #openstack-kolla | 22:22 | |
| sbezverk | kfox1111: yeah, I read somewhere on nova page that they suggest to build default installation with cell0, just to be backward compatible with old way.. and then start uising new cells as needed | 22:24 |
| *** fooliouno has quit IRC | 22:24 | |
| sbezverk | at one point we will need to build a job to add cells, I just would prefer to do it separately | 22:24 |
| *** Pavo has quit IRC | 22:25 | |
| *** jascott1 has joined #openstack-kolla | 22:28 | |
| *** rwsu has joined #openstack-kolla | 22:41 | |
| *** rwsu has quit IRC | 22:41 | |
| kfox1111 | if its different code, we probably should do it differently. | 22:44 |
| kfox1111 | but if its exactly the same, probably should just share the code? | 22:44 |
| sbezverk | kfox1111: I agree, but it would require a bit of research.. I can take care of that after ironic is completed | 22:45 |
| sbezverk | kfox1111: unless somebody else decides to do it ;) | 22:46 |
| kfox1111 | k | 22:46 |
| kfox1111 | still tryign to get fernet rotation working.... its proving much more stubbern then I would have thought. | 22:46 |
| kfox1111 | its really really close. :/ | 22:46 |
| sbezverk | kfox1111: I briefly looked at your PS, man you do some dancing moves in there ;) | 22:47 |
| kfox1111 | heh. yeah. | 22:47 |
| kfox1111 | safe atomic updating of secrets is not the easiest thing todo today. | 22:48 |
| kfox1111 | but when converted to a scheduledjob, will be very slick. :) | 22:48 |
| *** fooliouno has joined #openstack-kolla | 22:49 | |
| *** shashank_t_ has joined #openstack-kolla | 22:50 | |
| *** goldyfruit has joined #openstack-kolla | 22:53 | |
| *** manheim has joined #openstack-kolla | 23:01 | |
| kfox1111 | sbezverk: does placement not get a public api? | 23:01 |
| *** shashank_t_ has quit IRC | 23:01 | |
| kfox1111 | nm. it just wasn't changed. | 23:02 |
| sbezverk | kfox1111: unfortunately I needed to touch a few things when I hit problems.. placement was one of them.. | 23:03 |
| kfox1111 | no worries. | 23:04 |
| kfox1111 | almost done reviewing. its a large ps. :) | 23:04 |
| kfox1111 | but in general, a fantastic piece of work. :) | 23:04 |
| sbezverk | kfox1111: thanks, it is big, that is why I am a bit hesitant to add there more.. | 23:07 |
| kfox1111 | so, simple_cell_setup only does cell0? | 23:07 |
| sbezverk | kfox1111: yep | 23:07 |
| kfox1111 | is cell0 the parent cell for all other cells? | 23:08 |
| sbezverk | kfox1111: I am not sure there is parent/child relation between cell0 and other cells | 23:08 |
| kfox1111 | there's a master cell somewhere I think. | 23:09 |
| sbezverk | I think they are all parallel and you assign compute resources between them | 23:09 |
| sbezverk | maybe SamYaple can comment | 23:09 |
| kfox1111 | it gets the request, picks then forwards it to the other cells I think. | 23:09 |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla-kubernetes master: WIP: Playbook to add configmaps and secrets https://review.openstack.org/447733 | 23:09 |
| inc0 | yay gates are green https://review.openstack.org/#/c/448258/ | 23:10 |
| inc0 | kfox1111 sbezverk ^ want me to do rest of services? | 23:10 |
| sbezverk | inc0: for sure but I think we should do it in a separate ps? after this one gets merged | 23:13 |
| inc0 | yeah | 23:13 |
| inc0 | that was my thinking | 23:13 |
| inc0 | after that I can get this one in too and turn configs into k8s resources | 23:15 |
| inc0 | so whole configmap/secret generation will be simple ansible-playbook -> kubectl create | 23:15 |
| kfox1111 | inc0: was most of this coppied verbatim from kolla-ansible? | 23:18 |
| inc0 | kfox1111: yeah, well, I did remove a bunch of stuff | 23:18 |
| inc0 | but config generation is generally the same | 23:19 |
| inc0 | I just changed templates to remove this if KUBERNETES stuff | 23:19 |
| kfox1111 | inc0: yeah. jsut wondering. cause theres still a bunch of stuff that could be removed too. | 23:19 |
| inc0 | yeah I can imagine | 23:19 |
| kfox1111 | if it was just a copy per file over though, its probably fine. | 23:19 |
| inc0 | cleanups is going to take a while | 23:19 |
| kfox1111 | yeah. | 23:19 |
| kfox1111 | and may not be worth cleaning. | 23:19 |
| inc0 | yup, depends what we want to do with it | 23:20 |
| kfox1111 | only worth spending a bunch of time cleaning, if it stays. | 23:20 |
| kfox1111 | yeah. | 23:20 |
| inc0 | sbezverk: we | 23:21 |
| inc0 | we're not using post-deploy to create admin-openrc? | 23:21 |
| inc0 | for gates? | 23:21 |
| kfox1111 | inc0: no. | 23:22 |
| kfox1111 | just genconfig. | 23:22 |
| inc0 | kk | 23:22 |
| inc0 | bottom line, both patchsets https://review.openstack.org/#/c/447733/ and https://review.openstack.org/#/c/448258/ will ultimately lead to removal of kolla-ansible dependency and kollakube tool | 23:23 |
| kfox1111 | inc0: looks really good to me. how hard do you think it would be to tweak the ps to get rid of kolla-ansible genconfig entirely? | 23:23 |
| inc0 | not hard at all | 23:23 |
| inc0 | I just wanted to do it one role at the time | 23:23 |
| kfox1111 | 448258 looks very close to merge able. | 23:24 |
| kfox1111 | 447733 has a lot of work todo still. | 23:24 |
| inc0 | yeah all the secrets | 23:24 |
| kfox1111 | 448258 is where most of the headache is though. | 23:24 |
| inc0 | and first it needs to pass gates;) | 23:24 |
| kfox1111 | secrets, haproxy config, keepalived/set-ip packages, etc. | 23:24 |
| inc0 | I was thinking of splitting it into multiple patchsets | 23:25 |
| inc0 | but yeah | 23:25 |
| inc0 | not sure how to generate stuff like haproxy conf tho, will need to dig through kollakube | 23:25 |
| kfox1111 | inc0:+1 to multiple ps's, but I thin the tool should replace flat out kollakube, so that part goes last. | 23:25 |
| inc0 | and probably make separate playbook to generate -haproxy configs in same style as regular ones | 23:25 |
| kfox1111 | but having some prep work ps's like moving the haproxy configmap generation code to helm packages, libvirt seceret to helm package, keepalived to helm package, etc | 23:26 |
| kfox1111 | can be done in seperate ps's. | 23:26 |
| kfox1111 | -1 for bringing that into ansible. | 23:26 |
| inc0 | that might be good idea | 23:26 |
| inc0 | right, if you say so:) | 23:26 |
| kfox1111 | its one thing to copy existing templates from one repo to another. | 23:26 |
| inc0 | yeah I guess | 23:26 |
| inc0 | also I agree that it all should land in helm at some point | 23:27 |
| kfox1111 | its another to actively port it to something we're all pretty sure wont be sticking around. | 23:27 |
| inc0 | right | 23:27 |
| inc0 | k, I'll get to it tomorrow then | 23:27 |
| kfox1111 | ok. cool. | 23:27 |
| inc0 | I really want to smooth things out in terms of steps to deploy kolla-k8s | 23:28 |
| kfox1111 | big +1 from me. :) | 23:28 |
| sbezverk | inc0: that was a big step forward, nice! | 23:28 |
| vhosakot | inc0: yes, I have been doing it since weeks and want to help improve the deployment and its docs for kolla-k8s... | 23:29 |
| sbezverk | inc0: hopefully people will stop complaining about creating soft links to get genconfig running | 23:29 |
| sbezverk | ;) | 23:29 |
| vhosakot | I'll help too, but my kolla-k8s must be up first :) | 23:29 |
| vhosakot | RAM issues, vagrant issues, now dashboard issues related to service accounts :) | 23:30 |
| sbezverk | vhosakot: man you are still not up?!? wow, you take time ;) | 23:30 |
| inc0 | I think process > docs right now | 23:30 |
| vhosakot | sbezverk: yes, I'm kaing slow progress.. now stuck with both "helm init" and dashboard not working. | 23:30 |
| inc0 | I mean if it's super complex, no amount of docs will fix it | 23:30 |
| inc0 | vhosakot: maybe you want to help me with removing genconfig? | 23:31 |
| inc0 | once first patch merges we can split work between multiple roles | 23:31 |
| inc0 | it's really simple | 23:31 |
| sbezverk | inc0: I think the problem is people are expecting to have a magical button, you push once and everything is working | 23:31 |
| kfox1111 | inc0: can you merge in the defaults in tests/conf/ceph-all-in-one? | 23:31 |
| vhosakot | looks like sbezverk and kfox1111 are the only kolls-k8s devs with workins kolla-k8s cluster.... :) they helping great, but my cluster has issues too | 23:31 |
| kfox1111 | no reason to not set them now and would simplify things. | 23:31 |
| vhosakot | inc0: yes, let us remove genconfig thing.. cool | 23:32 |
| inc0 | kfox1111: sure, as long as they're sensible defaults | 23:32 |
| kfox1111 | inc0: they are required defaults. :) | 23:32 |
| inc0 | (by sensible I mean...let's not mix and match image tags for example) | 23:32 |
| mnaser | did anyone get a chance to look at this weird gate issue | 23:32 |
| kfox1111 | gotta go with wahts tested. then break out from there. | 23:33 |
| mnaser | where it seems to fail at Length too long: 4798716 | 23:33 |
| kfox1111 | if it doesnt work in gate, how do we know users can deploy it. | 23:33 |
| inc0 | but using 2.0.2 images? | 23:33 |
| inc0 | I mean...mitaka? | 23:33 |
| inc0 | in pike project? | 23:34 |
| kfox1111 | does 4.x work? | 23:34 |
| inc0 | I'd sure hope so | 23:34 |
| sbezverk | kfox1111: yep | 23:34 |
| kfox1111 | sbezverk: all of it, or just nova/ironic? | 23:34 |
| vhosakot | inc0: goot go now.. I'll ping tomorrow once my "helm init" works.. good night all! | 23:34 |
| sbezverk | ironic jobs are running all 4.0.0 images | 23:34 |
| inc0 | if 4.0.0 doesnt work, we need bugs and patches submited | 23:34 |
| kfox1111 | last I tried, it faild to do volume attaches. | 23:34 |
| inc0 | bye vhosakot | 23:34 |
| vhosakot | o/ | 23:34 |
| sbezverk | kfox1111: works like a charm | 23:35 |
| *** vhosakot has quit IRC | 23:35 | |
| sbezverk | but I have not tried ceph though | 23:35 |
| kfox1111 | sbezverk: ceph works? | 23:35 |
| kfox1111 | yeah... exactly. | 23:35 |
| inc0 | what's bad about ceph? | 23:35 |
| sbezverk | I am using iscsi backend | 23:35 |
| inc0 | I mean I run ceph with 4.0.0 and it's fine | 23:35 |
| inc0 | (on ansible tho) | 23:35 |
| kfox1111 | sbezverk: yeah. but that doesn't mean it all works. just iscsi. so not a good default. | 23:35 |
| kfox1111 | inc0: ok. | 23:36 |
| inc0 | have you experienced any issues? | 23:36 |
| kfox1111 | so, lets enable all the gates 4.x and see what breaks. | 23:36 |
| inc0 | big difference is we run jewel from 3.0 on | 23:36 |
| inc0 | kk | 23:36 |
| kfox1111 | if its stable, then we promote it to default. | 23:36 |
| *** adrian_otto has quit IRC | 23:36 | |
| sbezverk | kfox1111: good idea, but we need some piceces from my ps | 23:36 |
| kfox1111 | sbezverk: right. | 23:36 |
| sbezverk | kfox1111: like cells | 23:36 |
| inc0 | I'd suggest keeping image tag out of defaults | 23:37 |
| kfox1111 | and maybe this ps: https://review.openstack.org/#/c/418550/ | 23:37 |
| kfox1111 | inc0: if its not defaulted to something sane though, how do we expect users to deploy it? | 23:37 |
| kfox1111 | sane = tested. | 23:37 |
| kfox1111 | they shouldn't have to git checkout, dump a bunch of stuff into config, then it works. | 23:38 |
| inc0 | kfox1111: what I'm saying is let's test it | 23:38 |
| kfox1111 | +1 | 23:38 |
| inc0 | as right now | 23:38 |
| inc0 | they're using 4.0 anyway | 23:38 |
| inc0 | because genconfig | 23:38 |
| kfox1111 | the istructinos say copy ceph-all-in-one in I think. | 23:38 |
| kfox1111 | so, not 4.x. | 23:38 |
| inc0 | we can keep saying that;) | 23:39 |
| inc0 | for now | 23:39 |
| kfox1111 | lets go with what is known working, then a step at a time, move over the defaults all to 4.x | 23:39 |
| inc0 | anyway, that's different conversation | 23:39 |
| kfox1111 | without breaking anything. :) | 23:39 |
| inc0 | we first need to finish getting rid of genconfig | 23:39 |
| kfox1111 | +1 | 23:40 |
| mnaser | okay | 23:40 |
| mnaser | so it looks like | 23:40 |
| mnaser | subunit recieved a log entry that was tooo big | 23:40 |
| *** eaguilar has joined #openstack-kolla | 23:41 | |
| mnaser | that was the reason why it failed | 23:41 |
| sbezverk | kfox1111: do you think you will be able to send your review today? | 23:42 |
| kfox1111 | sbezverk: sorry. got distracted. | 23:42 |
| inc0 | k guys, I'll get back to it tomorrow | 23:42 |
| sbezverk | kfox1111: no worries | 23:42 |
| sbezverk | inc0: have a good one | 23:42 |
| inc0 | bye | 23:42 |
| mnaser | later inc0 | 23:43 |
| mnaser | im going to get to the bottom of this gate stuff today | 23:43 |
| *** manheim has quit IRC | 23:45 | |
| kfox1111 | sbezverk: https://www.openstack.org/videos/video/nova-cells-v2-whats-going-on a little dated, but interesting. | 23:49 |
| kfox1111 | it does sound like cell0 isn't special though. | 23:51 |
| kfox1111 | at least in newton | 23:52 |
| sbezverk | kfox1111: yeah but on their first slide they have cell0 ;) | 23:52 |
| kfox1111 | cell0 is a default name if you don't specify one. | 23:52 |
| kfox1111 | but it sounds like its possible to delete it if you have other cells. | 23:53 |
| sbezverk | kfox1111: in newton is not required, it is starting from ocata | 23:53 |
| kfox1111 | allittle more specific please. a cell is required, or a cell named cell0? cause I think its a cell is required. | 23:53 |
| kfox1111 | https://docs.openstack.org/developer/nova/cells.html says cell0 explicitly is required... | 23:54 |
| sbezverk | kfox1111: when I first hit the problem I checked with nova people and they told me to create cell0, then Jeffrey4l confimred it | 23:54 |
| kfox1111 | Thus, a small nova deployment will have an API database, a cell0, and what we will call here a “cell1” database. | 23:55 |
| kfox1111 | fun... | 23:55 |
| *** kbaegis has quit IRC | 23:55 | |
| kfox1111 | heh. | 23:55 |
| kfox1111 | " Instances that are never scheduled are relegated to the cell0 database, which is effectively a graveyard of instances that failed to start. All successful/running instances are stored in “cell1”." | 23:55 |
| *** kbaegis has joined #openstack-kolla | 23:55 | |
| kfox1111 | so, how is it working without cell1? | 23:56 |
| sbezverk | kfox1111: I think this model work only when you have multiple cells cell0 and cell1 | 23:56 |
| *** mnaser has quit IRC | 23:56 | |
| kfox1111 | sbezverk: the doc says cell0 and cell1 are required. | 23:56 |
| sbezverk | if you have cell0 only then it is like a default gateway ;) | 23:56 |
| *** mnaser has joined #openstack-kolla | 23:57 | |
| kfox1111 | then multiple past cell1 is optional. | 23:57 |
| kfox1111 | thats not what the docs say. | 23:57 |
| sbezverk | kfox1111: the doc is obivously wrong | 23:57 |
| kfox1111 | maybe. | 23:57 |
| kfox1111 | or we have a problem we havent identified yet. :/ | 23:57 |
| sbezverk | as I have a single cell0 and all instances are coming up | 23:57 |
| sbezverk | kfox1111: do not think so | 23:57 |
| sbezverk | kolla-ansible use cell0 only too | 23:58 |
| sbezverk | and they do not have any issues | 23:58 |
| kfox1111 | how many 4.x deployments has kolla-ansble seen? | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!