*** jrist has joined #openstack-kolla | 00:04 | |
*** chas has joined #openstack-kolla | 00:04 | |
*** bmaltais has joined #openstack-kolla | 00:06 | |
*** chas has quit IRC | 00:09 | |
*** huzhengchuan has quit IRC | 00:09 | |
*** stefan_bo has joined #openstack-kolla | 00:17 | |
*** jamesPR has quit IRC | 00:19 | |
*** chas has joined #openstack-kolla | 00:20 | |
*** stefan_bo has quit IRC | 00:21 | |
*** chas has quit IRC | 00:25 | |
spsurya | morning | 00:26 |
---|---|---|
*** stefan_bo has joined #openstack-kolla | 00:26 | |
*** yuanying has quit IRC | 00:30 | |
jascott1 | good morning spsurya | 00:32 |
*** jtriley has joined #openstack-kolla | 00:33 | |
jascott1 | bb in a few (hrs) | 00:36 |
*** hongbin has quit IRC | 00:37 | |
*** yuanying has joined #openstack-kolla | 00:39 | |
duonghq | morning guys | 00:44 |
*** huzhengchuan has joined #openstack-kolla | 00:49 | |
*** dixiaoli has joined #openstack-kolla | 00:51 | |
*** hieulq has joined #openstack-kolla | 00:51 | |
*** cuongnv has joined #openstack-kolla | 00:53 | |
*** jamesPR has joined #openstack-kolla | 01:02 | |
*** daidv has joined #openstack-kolla | 01:02 | |
bmaltais | morning | 01:04 |
jamesPR | morning | 01:06 |
*** itlinux has quit IRC | 01:07 | |
*** harlowja has quit IRC | 01:08 | |
spsurya | duonghq: sup | 01:09 |
bmaltais | Been testing the AIO kolla-kubernetes. So far it is working well with the latest fix from yesterday | 01:11 |
bmaltais | But been noticing that rebooting the VM running everything result in a non functional environment. | 01:11 |
bmaltais | Anyone has experienced this? | 01:11 |
bmaltais | Is it possible that the PODs are not coming up in the right order and as a result it introduce issues in the system? | 01:12 |
bmaltais | They all seem to come up at the same time in a chaotic way | 01:13 |
*** itlinux has joined #openstack-kolla | 01:16 | |
*** rwallner has joined #openstack-kolla | 01:18 | |
jamesPR | bmaltais, sorry can not help you, have not tried kolla-kurbenetes | 01:21 |
jamesPR | yet | 01:21 |
jamesPR | ;) | 01:21 |
*** rwallner has quit IRC | 01:22 | |
*** YuYangWang has joined #openstack-kolla | 01:25 | |
*** salv-orlando has quit IRC | 01:25 | |
*** tovin07_ has joined #openstack-kolla | 01:26 | |
*** YuYangWang has quit IRC | 01:28 | |
*** YuYangWang has joined #openstack-kolla | 01:28 | |
*** dave-mccowan has quit IRC | 01:28 | |
bmaltais | jamesPR: Are you using kolla-ansible? | 01:29 |
bmaltais | jamesPR: Been testing it. Work pretty well... but I find kubernetes intriguing so I am investing time to test it | 01:30 |
*** dixiaoli has quit IRC | 01:30 | |
*** huzhengchuan has quit IRC | 01:43 | |
*** zhubingbing has joined #openstack-kolla | 01:43 | |
jamesPR | bmaltais, trying kolla-ansible plain docker | 01:44 |
*** caowei has joined #openstack-kolla | 01:44 | |
jamesPR | kurbenetes is too much for what I want | 01:44 |
*** itlinux has quit IRC | 01:51 | |
*** dave-mccowan has joined #openstack-kolla | 01:51 | |
*** huzhengchuan has joined #openstack-kolla | 01:59 | |
*** zhubingbing has quit IRC | 01:59 | |
*** itlinux has joined #openstack-kolla | 02:00 | |
rwellum | bmaltais: agreed, rebooting seems deadly to k8s. | 02:03 |
rwellum | BTW which AIO? | 02:04 |
rwellum | jamesPR: there's an ansible playbook for kolla-kubernetes :) | 02:04 |
*** unicell has quit IRC | 02:06 | |
*** zhubingbing has joined #openstack-kolla | 02:12 | |
jamesPR | rwellum, I know but too much for me for now ;) playing with plain docker for now | 02:18 |
rwellum | jamesPR: :) | 02:18 |
*** eanylin has quit IRC | 02:20 | |
*** chas has joined #openstack-kolla | 02:21 | |
*** chas has quit IRC | 02:25 | |
*** itlinux has quit IRC | 02:25 | |
*** qiliang27 has joined #openstack-kolla | 02:31 | |
*** itlinux has joined #openstack-kolla | 02:36 | |
openstackgerrit | Eric Lei proposed openstack/kolla-ansible master: [Trivial] Fix two typos in kolla-ansible. https://review.openstack.org/465213 | 02:37 |
*** dave-mccowan has quit IRC | 02:42 | |
*** huzhengchuan has quit IRC | 02:43 | |
*** YuYangWang has quit IRC | 02:45 | |
*** YuYangWang has joined #openstack-kolla | 02:45 | |
*** itlinux has quit IRC | 02:48 | |
*** eanylin has joined #openstack-kolla | 02:50 | |
*** qiliang28 has joined #openstack-kolla | 02:53 | |
*** xinliang has quit IRC | 02:54 | |
*** huzhengchuan has joined #openstack-kolla | 02:55 | |
*** qiliang27 has quit IRC | 02:57 | |
*** qiliang28 is now known as qiliang27 | 02:57 | |
*** itlinux has joined #openstack-kolla | 02:57 | |
openstackgerrit | pangliye proposed openstack/kolla master: Install python-pip https://review.openstack.org/465216 | 03:04 |
*** qiliang27 has quit IRC | 03:04 | |
*** qiliang27 has joined #openstack-kolla | 03:04 | |
*** xinliang has joined #openstack-kolla | 03:06 | |
*** qiliang28 has joined #openstack-kolla | 03:12 | |
openstackgerrit | pangliye proposed openstack/kolla master: Install the plugin of elasticsearch-kopf https://review.openstack.org/465220 | 03:12 |
*** qiliang27 has quit IRC | 03:13 | |
*** qiliang28 is now known as qiliang27 | 03:13 | |
*** itlinux has quit IRC | 03:24 | |
*** goldyfruit has joined #openstack-kolla | 03:28 | |
*** stefan_bo has quit IRC | 03:34 | |
*** zhubingbing has quit IRC | 03:42 | |
*** huzhengchuan has quit IRC | 03:47 | |
spsurya | duonghq: around ? | 03:50 |
spsurya | duonghq: Have you tried development environment setup with vagrant for kolla-ansible | 03:51 |
spsurya | ? | 03:51 |
spsurya | does the available docs for vagrant dev env setup works fine ? | 03:52 |
duonghq | spsurya, not yet | 03:54 |
duonghq | I use vm | 03:54 |
duonghq | directly | 03:54 |
spsurya | duonghq: ok, thanks | 03:54 |
spsurya | does anyone tried this ? | 03:55 |
*** huzhengchuan has joined #openstack-kolla | 03:59 | |
*** xinliang has quit IRC | 04:04 | |
*** xinliang has joined #openstack-kolla | 04:05 | |
*** xinliang has quit IRC | 04:05 | |
*** xinliang has joined #openstack-kolla | 04:05 | |
*** zhubingbing_ has joined #openstack-kolla | 04:06 | |
*** gomarivera_ has joined #openstack-kolla | 04:13 | |
*** huzhengchuan has quit IRC | 04:16 | |
*** goldyfruit has quit IRC | 04:21 | |
*** chas has joined #openstack-kolla | 04:22 | |
openstackgerrit | xuchao proposed openstack/kolla master: Change Kolla Kubernetes Developer Quickstart url. https://review.openstack.org/465396 | 04:23 |
*** chas has quit IRC | 04:27 | |
*** harlowja has joined #openstack-kolla | 04:33 | |
*** caowei has quit IRC | 04:34 | |
openstackgerrit | xuchao proposed openstack/kolla master: Change Kolla Kubernetes Developer Quickstart url fix the url 404 not found issue. https://review.openstack.org/465396 | 04:36 |
*** zhubingbing_ has quit IRC | 04:42 | |
*** skramaja has joined #openstack-kolla | 04:44 | |
*** unicell has joined #openstack-kolla | 04:45 | |
openstackgerrit | Anthony Lin proposed openstack/kolla-kubernetes master: Script to Deploy Kolla Kubernetes AIO https://review.openstack.org/464312 | 04:49 |
*** huzhengchuan has joined #openstack-kolla | 05:00 | |
*** rwellum has quit IRC | 05:11 | |
*** caowei has joined #openstack-kolla | 05:13 | |
*** huzhengchuan has quit IRC | 05:14 | |
*** janki has joined #openstack-kolla | 05:15 | |
openstackgerrit | pangliye proposed openstack/kolla master: Install the plugin of elasticsearch-kopf https://review.openstack.org/465220 | 05:16 |
*** huzhengchuan has joined #openstack-kolla | 05:27 | |
*** stefan_bo has joined #openstack-kolla | 05:31 | |
*** salv-orlando has joined #openstack-kolla | 05:32 | |
openstackgerrit | Li Yingjun proposed openstack/kolla-ansible master: Removing ovs bridge before destroying container https://review.openstack.org/464998 | 05:38 |
*** kollian has joined #openstack-kolla | 05:41 | |
*** huzhengchuan has quit IRC | 05:42 | |
kollian | Has anyone faced this http://paste.openstack.org/show/609732/ ? | 05:42 |
kollian | fatal: [localhost]: FAILED! => {"changed": true, "failed": true, "msg": "'Traceback (most recent call last):\\n File \"/tmp/ansible_Y_kqzG/ansible_module_kolla_docker.py\", line 781, in main\\n | 05:43 |
*** zhubingbing has joined #openstack-kolla | 05:43 | |
kollian | during deploy of kolla using kolla-ansible deploy | 05:43 |
kollian | zhubingbing http://paste.openstack.org/show/609732/ | 05:44 |
kollian | have you faced this ? | 05:44 |
zhubingbing | you should update ansible | 05:44 |
zhubingbing | and docker-py | 05:44 |
*** sgrasley has quit IRC | 05:45 | |
jascott1 | does this exist? kolla/centos-binary-memcached:4.0.1 | 05:45 |
*** zhugx has joined #openstack-kolla | 05:45 | |
kollian | zhubingbing: ansible --version ansible 2.3.0.0 config file = /etc/ansible/ansible.cfg configured module search path = Default w/o overrides python version = 2.7.5 (default, Nov 6 2016, 00:28:07) [GCC 4.8.5 20150623 (Red Hat 4.8.5-11)] | 05:46 |
zhubingbing | 2.2 | 05:46 |
zhubingbing | ansible 2.2 | 05:46 |
kollian | docker --version Docker version 17.05.0-ce, build 89658be | 05:47 |
zhubingbing | manifest for kolla/centos-binary-memcached:4.0.1 not found | 05:47 |
zhubingbing | you haven't this images | 05:47 |
zhubingbing | you check docker registry | 05:48 |
*** pcaruana has joined #openstack-kolla | 05:48 | |
*** unicell1 has joined #openstack-kolla | 05:51 | |
*** unicell has quit IRC | 05:52 | |
*** harlowja has quit IRC | 05:54 | |
*** huzhengchuan has joined #openstack-kolla | 05:54 | |
*** zhugx_ has joined #openstack-kolla | 05:55 | |
*** zhugx has quit IRC | 05:58 | |
*** rwallner has joined #openstack-kolla | 06:01 | |
*** rwallner has quit IRC | 06:06 | |
*** jtriley has quit IRC | 06:15 | |
Seyong | Hello guys Im gonna try kolla on multinode kubernetes cluster. Any body have experenced it? Appreciated for any information in advance :) | 06:18 |
jascott1 | hi Seyong what exactly are you interested in? | 06:22 |
*** rwallner has joined #openstack-kolla | 06:22 | |
*** chas has joined #openstack-kolla | 06:23 | |
*** mannamne has joined #openstack-kolla | 06:24 | |
*** zhubingbing has quit IRC | 06:25 | |
*** rwallner has quit IRC | 06:27 | |
*** zhubingbing has joined #openstack-kolla | 06:27 | |
*** chas has quit IRC | 06:28 | |
*** athomas has joined #openstack-kolla | 06:35 | |
*** mannamne has quit IRC | 06:36 | |
openstackgerrit | Duong Ha-Quang proposed openstack/kolla-ansible master: Implement Keystone zero-downtime upgrade https://review.openstack.org/425446 | 06:44 |
*** chas has joined #openstack-kolla | 06:45 | |
*** matrohon has joined #openstack-kolla | 06:47 | |
*** satyar has joined #openstack-kolla | 06:48 | |
satyar | Hi Jeffrey4l_ | 06:48 |
satyar | Hi SamYaple | 06:49 |
satyar | I have quite a dumb question but need help on that... | 06:49 |
Jeffrey4l_ | satyar, yep? | 06:49 |
satyar | i have a newton deployment without the containers (openstack services are not containerized) | 06:50 |
satyar | is there any way i can upgrade to ocata with containerized kolla deployment? | 06:50 |
Jeffrey4l_ | there is no way, or no way from kolla community. | 06:51 |
satyar | Hmmm | 06:52 |
*** jtriley has joined #openstack-kolla | 06:53 | |
*** zhubingbing has quit IRC | 07:01 | |
*** jamesPR has quit IRC | 07:03 | |
*** zhubingbing_ has joined #openstack-kolla | 07:04 | |
masuberu | hi, kolla does not install ceph client is that right? | 07:05 |
masuberu | how can I run ceph health of ceph status? | 07:05 |
*** shardy has joined #openstack-kolla | 07:06 | |
masuberu | ok, I found out I just need to run the command on ceph_rgw container | 07:07 |
openstackgerrit | Bertrand Lallau proposed openstack/kolla-ansible master: Support osprofile usage https://review.openstack.org/455628 | 07:09 |
*** salv-orlando has quit IRC | 07:09 | |
*** bjolo has joined #openstack-kolla | 07:12 | |
*** bjolo_ has joined #openstack-kolla | 07:12 | |
*** hieulq_ has joined #openstack-kolla | 07:12 | |
*** hieulq_ has quit IRC | 07:13 | |
*** mgoddard__ has joined #openstack-kolla | 07:20 | |
openstackgerrit | Martin Mágr proposed openstack/kolla master: Sensu images https://review.openstack.org/463593 | 07:35 |
jascott1 | does anyone happen to know if openstack as a whole follows a consistent log format? | 07:37 |
jascott1 | nvmnd i *think* i found it | 07:39 |
spsurya | jascott1: not much idea about consistency of logging but yes this can help | 07:40 |
spsurya | https://specs.openstack.org/openstack/openstack-specs/specs/log-guidelines.html | 07:40 |
spsurya | ok ok | 07:40 |
*** zhubingbing_ has quit IRC | 07:40 | |
jascott1 | spsurya thanks, i was looking at https://wiki.openstack.org/wiki/LoggingStandards | 07:40 |
spsurya | jascott1: better follow mine | 07:41 |
spsurya | wiki may not be up to date | 07:41 |
jascott1 | for sure | 07:42 |
jascott1 | was going to mention mined looked ancient :) | 07:42 |
Seyong | jascott1: well any info. regarding multi node... I read the deployment guide is tested only for AIO environment, so... | 07:43 |
Seyong | May it work on multi node? I believe that | 07:43 |
*** huzhengchuan has quit IRC | 07:43 | |
jascott1 | Seyong I have figured out the multinode for that playbook but havent committed yet | 07:43 |
*** zhubingbing_ has joined #openstack-kolla | 07:44 | |
Seyong | jascott1: ooh can you share with me? mail@syum.net | 07:45 |
jascott1 | Seyong i think it more beneficial if I update PS with what ive got, i can just comment it out as its not super tested | 07:47 |
*** mgoddard__ has quit IRC | 07:47 | |
*** gfidente has joined #openstack-kolla | 07:47 | |
*** magicboiz has joined #openstack-kolla | 07:47 | |
jascott1 | essentially tho, its just adding nodes to k8s cluster and labeling them as kolla_compute | 07:47 |
*** huzhengchuan has joined #openstack-kolla | 07:47 | |
*** jaosorior_away is now known as jaosorior | 07:48 | |
Seyong | jascott1: yea thats my thought | 07:48 |
Seyong | I will try | 07:48 |
Seyong | However, when I join a minion to AIO kolla-kubernetes(so, master), master continuesly deletes pods and joining doesnt success.... Have you seen this..? | 07:50 |
jascott1 | hmm | 07:51 |
*** magicboiz has quit IRC | 07:52 | |
jascott1 | it might be labeling/taint related on master. The joining problem might be kubeadm init params, like the api-advertise-ip (or whatever) being correct | 07:52 |
*** manheim has joined #openstack-kolla | 07:53 | |
jascott1 | Seyong which pods are deleted/recreated? | 07:53 |
Seyong | jascott1: All of the kube-system namespace | 07:54 |
Seyong | and kolla namespace pods are backed to 'init' and wait forever | 07:54 |
*** lpetrut has joined #openstack-kolla | 07:54 | |
*** bjolo__ has joined #openstack-kolla | 07:55 | |
*** huzhengchuan has quit IRC | 07:55 | |
jascott1 | Seyong Tiller and kube-dns are failing or everything in kube-system is failing> | 07:56 |
jascott1 | ? | 07:56 |
Seyong | I can not see any failing (but could failed) and just deleted... everything in kube-system inclding kube-dns | 07:57 |
jascott1 | kolla-kubernetes is mostly contained in kolla so kube-system problems point to kubernetes | 07:57 |
jascott1 | config problem | 07:58 |
Seyong | sure ^^;; | 07:58 |
Seyong | I will check kubernetes point of view | 07:58 |
jascott1 | you are following kolla-k8s deploy guide? | 07:58 |
Seyong | yea | 07:58 |
*** stefan_bo has quit IRC | 07:59 | |
jascott1 | running on AWS or openstack or? | 07:59 |
Seyong | running on baremetal.. | 07:59 |
*** magicboiz has joined #openstack-kolla | 07:59 | |
Seyong | once kolla-kubernetes pods are running I tried join other baremetal as a minion then it happens | 07:59 |
jascott1 | what I dont understand is that kube-system pods are so affected | 08:00 |
Seyong | very strange, huh | 08:00 |
jascott1 | kolla namespaced pods will fail for various reasons but in my experience most of the kube-system is stable | 08:00 |
jascott1 | i have mostly seen kube-dns and tiller crashing | 08:01 |
Seyong | I agree that it is not related kolla-kube, but kubernetes issue | 08:01 |
jascott1 | probably taints or labels? | 08:01 |
Seyong | I did taint for the master to have pods | 08:01 |
Seyong | and also labeled control+compute to the master | 08:02 |
jascott1 | bare metal has good specs? I can tax out a t2.large VM at AWS sometimes with comput ekit | 08:02 |
jascott1 | etcd starts failing to respond usually when their is hardware resource problem | 08:03 |
jascott1 | oh sorry let me commit this, perhaps you can test :) | 08:03 |
Seyong | it has 64GB mem, 40cores of cpu It will be enough right? | 08:04 |
Seyong | oh thats good jascott1 | 08:04 |
openstackgerrit | pangliye proposed openstack/kolla-ansible master: Modify the configuration file of telegraf https://review.openstack.org/465446 | 08:06 |
*** zhugx has joined #openstack-kolla | 08:07 | |
*** huzhengchuan has joined #openstack-kolla | 08:08 | |
*** Serlex has joined #openstack-kolla | 08:09 | |
*** salv-orlando has joined #openstack-kolla | 08:10 | |
*** zhugx_ has quit IRC | 08:10 | |
*** zhubingbing_ has quit IRC | 08:11 | |
openstackgerrit | Merged openstack/kolla master: Fix logging for chrony https://review.openstack.org/462463 | 08:11 |
*** zhubingbing_ has joined #openstack-kolla | 08:12 | |
openstackgerrit | Bertrand Lallau proposed openstack/kolla-ansible master: Support osprofile usage https://review.openstack.org/455628 | 08:12 |
*** huzhengchuan has quit IRC | 08:13 | |
bjolo__ | morning | 08:14 |
*** salv-orlando has quit IRC | 08:14 | |
*** egonzalez has joined #openstack-kolla | 08:17 | |
jascott1 | Seyong looking at the diff, really Im just adding the node with sudo kubeadm join --token {{ token }} {{ api_advertise_ip }}:6443 | 08:19 |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible master: Implement Keystone zero-downtime upgrade https://review.openstack.org/425446 | 08:19 |
Seyong | you mean join works right? | 08:20 |
Seyong | which file did you commit? deployment-guide.rst? | 08:20 |
jascott1 | the ansible implementation of that guide | 08:22 |
jascott1 | https://review.openstack.org/#/c/457317/ | 08:22 |
Seyong | thanks for the link | 08:22 |
*** qiliang28 has joined #openstack-kolla | 08:22 | |
jascott1 | joined did work for me but I havent tested the resulting impact on the kolla pods | 08:22 |
Seyong | yea | 08:23 |
jascott1 | but I didnt see what you are describing afaik | 08:23 |
*** stefan_bo has joined #openstack-kolla | 08:24 | |
*** rwallner has joined #openstack-kolla | 08:24 | |
*** qiliang27 has quit IRC | 08:26 | |
*** qiliang28 is now known as qiliang27 | 08:26 | |
openstackgerrit | Bertrand Lallau proposed openstack/kolla master: Add Elasticsearch client package for OSprofile https://review.openstack.org/465457 | 08:27 |
*** duonghq has quit IRC | 08:27 | |
jascott1 | Seyong i missed earlier when you listed specs... nice! | 08:27 |
*** huzhengchuan has joined #openstack-kolla | 08:28 | |
openstackgerrit | Bertrand Lallau proposed openstack/kolla-ansible master: Support OSprofile usage https://review.openstack.org/455628 | 08:28 |
Seyong | jascott1: :D | 08:28 |
manheim | hi | 08:28 |
manheim | i have a question | 08:28 |
*** rwallner has quit IRC | 08:28 | |
*** qiliang27 has quit IRC | 08:28 | |
jascott1 | Seyong to be clear, everything works fine in kube-system and kolla until you join worker node? | 08:29 |
manheim | on reconfigure, i can see that all the steps in a roles will run (deploy, config, bootstrap), how does this with with idempotency? won't the steps be skipped? | 08:29 |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible master: Implement Keystone zero-downtime upgrade https://review.openstack.org/425446 | 08:30 |
manheim | and only the conigs copy will actually change things since the configs files are changed | 08:30 |
egonzalez | manheim, is idempotent since most task are skipped | 08:31 |
manheim | so most tasks will be skipped, right? | 08:31 |
manheim | expect for config.j2 copy over | 08:31 |
manheim | except* | 08:31 |
egonzalez | yep, db sync only is executed if dbs are created, endpoints will change if values differs from current, copy config only if changed, and restart containers in image or config changes | 08:32 |
manheim | aha | 08:32 |
manheim | thought so, just wanted to make sure | 08:32 |
*** duonghq has joined #openstack-kolla | 08:33 | |
jascott1 | that sounds like some wisdom that should be in a doc somewhere ;) | 08:33 |
Seyong | jascott1: Yes. everything was fine. So, I tried re-install centos / install kubernetes + canal and kubernetes node is getting ready / trying join from other baremetal / => Join not working | 08:33 |
manheim | yep, i was thinking on writing some docs and commiting them | 08:33 |
*** qiliang27 has joined #openstack-kolla | 08:34 | |
Seyong | so it is clear that kolla is not related of my issue | 08:34 |
jascott1 | Seyong but even with join not working, the attempt to join wrecks pods in kube-system correct? | 08:34 |
egonzalez | manheim, cool, a lot of work is needed in docs atm | 08:34 |
*** mgoddard__ has joined #openstack-kolla | 08:34 | |
Seyong | jascott1: yesyes | 08:34 |
manheim | yep, talked with inc and he gave me some ideas on this | 08:34 |
manheim | egonzalez, how do config files end up in /etc/kolla/<service>? | 08:36 |
manheim | how can I do this for my role | 08:36 |
manheim | i want to implement reconfigure for my role | 08:36 |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible master: Optimize the reconfiguration for 'common' container https://review.openstack.org/435236 | 08:37 |
egonzalez | manheim, the new reconfigure method is much better, idempotent and faster than older, you can take a look at what changed from older to new method https://review.openstack.org/#/c/435236/ | 08:37 |
Seyong | kubelet.go:1524] Deleting mirror pod "kube-controller-manager-localhost.localdomain_kube-system(606a08f7-3a0d-11e7-9e6f-bc305bf48848)" because it is outdated | 08:37 |
Seyong | This is continuesly repeated for the all of kube-system pods | 08:38 |
manheim | cool, i'll take a look when i'm done with the docs for hyperv role | 08:38 |
egonzalez | manheim, this task copy files there https://github.com/openstack/kolla-ansible/blob/master/ansible/roles/aodh/tasks/config.yml#L27 | 08:38 |
*** mannamne has joined #openstack-kolla | 08:38 | |
egonzalez | manheim, and this one copy those files into the container https://github.com/openstack/kolla-ansible/blob/master/ansible/roles/aodh/templates/aodh-api.json.j2 | 08:39 |
manheim | ah, so it's with mege_configs module, i see | 08:39 |
jascott1 | Seyong oh | 08:39 |
jascott1 | i have seen similar but what about kubectl get pods -n kube-system ? | 08:39 |
jascott1 | do you have running pods for all the basic kubernetes and helm stuff? | 08:39 |
Seyong | all of pods are repeating : creating -> deleting | 08:40 |
manheim | that is runned from within the container i see | 08:40 |
jascott1 | Seyong what does `kubectl describe pod <pod>` get you? | 08:40 |
jascott1 | i think thats the only way to see why its restarting | 08:41 |
Seyong | thats impossible... when it alive it is just "creating", nothing is special.. and some time it is just deleted. I can not see anything :( | 08:41 |
jascott1 | oops yeah i remember that situation now | 08:42 |
Seyong | ow | 08:42 |
Seyong | and do you remember what was your reason? | 08:42 |
jascott1 | sorry i just meant the whole `cant describe cause its restarting` thing | 08:42 |
jascott1 | let me see if i get same behavior | 08:43 |
Seyong | ^^ | 08:43 |
egonzalez | hi guys, guess this is a ubuntu dependency issue and we cannot do anything to solve right? rally : Depends: python-rally (= 0.8.1-0ubuntu3~cloud0) but it is not going to be installed | 08:51 |
*** shardy has quit IRC | 08:53 | |
*** shardy has joined #openstack-kolla | 08:56 | |
*** salv-orlando has joined #openstack-kolla | 09:00 | |
openstackgerrit | Martin André proposed openstack/kolla master: Revert "Clean up base Dockerfile" https://review.openstack.org/465472 | 09:05 |
*** salv-orlando has quit IRC | 09:06 | |
*** satyar has quit IRC | 09:06 | |
mandre | egonzalez, Jeffrey4l_, inc0: can you guys have a look at https://review.openstack.org/#/c/465472/? The base image cleanup broke customization, so I'm proposing a revert until we find a better solution | 09:09 |
openstackgerrit | shaofeng cheng proposed openstack/kolla-ansible master: Fix create and associate error of resellerAdmin https://review.openstack.org/465473 | 09:09 |
openstackgerrit | Surya Prakash (spsurya) proposed openstack/kolla-ansible master: Updated the quick start guide https://review.openstack.org/465474 | 09:12 |
openstackgerrit | shaofeng cheng proposed openstack/kolla-ansible master: Add heat_stack_owner role to admin project https://review.openstack.org/464913 | 09:16 |
openstackgerrit | shaofeng cheng proposed openstack/kolla-ansible master: Fix create and associate error of resellerAdmin https://review.openstack.org/465473 | 09:17 |
*** huzhengchuan has quit IRC | 09:23 | |
*** rwallner has joined #openstack-kolla | 09:25 | |
*** zhugx has quit IRC | 09:30 | |
*** rwallner has quit IRC | 09:30 | |
*** jchhatbar has joined #openstack-kolla | 09:30 | |
*** janki has quit IRC | 09:31 | |
*** jchhatbar_ has joined #openstack-kolla | 09:33 | |
openstackgerrit | Merged openstack/kolla master: fix handling of delorean on non-x86 https://review.openstack.org/465098 | 09:34 |
openstackgerrit | Merged openstack/kolla master: Change Kolla Kubernetes Developer Quickstart url fix the url 404 not found issue. https://review.openstack.org/465396 | 09:34 |
*** jchhatbar has quit IRC | 09:36 | |
*** huzhengchuan has joined #openstack-kolla | 09:36 | |
*** yingjun has quit IRC | 09:38 | |
*** zhubingbing_ has quit IRC | 09:44 | |
*** zhubingbing has joined #openstack-kolla | 09:49 | |
openstackgerrit | Merged openstack/kolla-ansible master: Add heat_stack_owner role to admin project https://review.openstack.org/464913 | 09:52 |
*** stefan_bo has quit IRC | 09:54 | |
*** tovin07_ has quit IRC | 09:56 | |
*** yingjun has joined #openstack-kolla | 09:58 | |
*** jascott1 has quit IRC | 10:03 | |
Seyong | jascott1: It was happening ^^;;; my master and minion has same hostname and those arent clustering. I changed hostname one of, and it goes as well | 10:04 |
Seyong | ops he quits | 10:05 |
openstackgerrit | shaofeng cheng proposed openstack/kolla-ansible master: Fix create and associate error of resellerAdmin https://review.openstack.org/465473 | 10:05 |
*** salv-orlando has joined #openstack-kolla | 10:10 | |
*** jtriley has quit IRC | 10:12 | |
*** salv-orlando has quit IRC | 10:15 | |
*** Daviey_ is now known as Daviey | 10:15 | |
*** feisky has joined #openstack-kolla | 10:17 | |
*** caowei has quit IRC | 10:21 | |
*** cuongnv has quit IRC | 10:25 | |
*** powerd has joined #openstack-kolla | 10:28 | |
*** daidv has quit IRC | 10:30 | |
openstackgerrit | zhuangzejian proposed openstack/kolla master: Fix typo https://review.openstack.org/465497 | 10:31 |
openstackgerrit | Merged openstack/kolla-ansible master: Fix create and associate error of resellerAdmin https://review.openstack.org/465473 | 10:35 |
*** pester has joined #openstack-kolla | 10:38 | |
*** fxpester has quit IRC | 10:41 | |
egonzalez | manheim, added a bunch of comments into hyperv role | 10:42 |
manheim | sure | 10:42 |
manheim | i'll take a look at them | 10:42 |
egonzalez | let me know if need some clarification | 10:42 |
manheim | atm writing the docs | 10:42 |
manheim | sure thing, thanks | 10:43 |
*** yingjun has quit IRC | 10:43 | |
*** rhallisey has quit IRC | 10:48 | |
*** pc_m has quit IRC | 10:50 | |
*** pc_m has joined #openstack-kolla | 10:50 | |
-openstackstatus- NOTICE: gerrit is being restarted to help stuck git replication issues | 10:54 | |
*** huzhengchuan has quit IRC | 10:54 | |
openstackgerrit | Merged openstack/kolla master: Revert "Clean up base Dockerfile" https://review.openstack.org/465472 | 10:55 |
bmaltais | jascott1: Multinode does not work very well, unstable based on my testing | 10:56 |
bmaltais | jascott1: Just addind a kolla-compute node does not work | 10:56 |
bmaltais | jascott1: Making the 2nd node a cotroller as well sort of work but break the iscsi support for volumes | 10:57 |
*** feisky has quit IRC | 10:57 | |
*** huzhengchuan has joined #openstack-kolla | 10:57 | |
*** feisky has joined #openstack-kolla | 10:58 | |
openstackgerrit | Merged openstack/kolla master: Add Elasticsearch client package for OSprofile https://review.openstack.org/465457 | 11:00 |
*** YuYangWang has quit IRC | 11:01 | |
*** chandankumar is now known as chkumar|fire | 11:01 | |
*** gomarivera_ has quit IRC | 11:05 | |
*** feisky has quit IRC | 11:06 | |
pester | egonzalez: is there any examples or guides how to add kolla component (let say magnum) to existing non kolla openstack ? | 11:07 |
*** rwallner has joined #openstack-kolla | 11:09 | |
*** salv-orlando has joined #openstack-kolla | 11:11 | |
*** rwallner_ has joined #openstack-kolla | 11:12 | |
*** rwallner_ has quit IRC | 11:13 | |
*** rwallner has quit IRC | 11:13 | |
*** rwallner has joined #openstack-kolla | 11:13 | |
*** feisky has joined #openstack-kolla | 11:15 | |
*** salv-orlando has quit IRC | 11:16 | |
*** rwellum has joined #openstack-kolla | 11:18 | |
*** jtriley has joined #openstack-kolla | 11:19 | |
*** chkumar|fire is now known as chandankumar | 11:29 | |
openstackgerrit | Paul Bourke (pbourke) proposed openstack/kolla master: Clean up unused sanity check modules https://review.openstack.org/465515 | 11:30 |
openstackgerrit | Paul Bourke (pbourke) proposed openstack/kolla-ansible master: Update cinder sanity check to use built in shade https://review.openstack.org/437920 | 11:30 |
kollian | ERROR:kolla.image.build.neutron-sfc-agent:Error'd with the following message ERROR:kolla.image.build.neutron-sfc-agent:The command '/bin/sh -c echo 'binary not yet available for centos' && /bin/false' returned a non-zero code: 1 | 11:40 |
*** powerd has quit IRC | 11:40 | |
kollian | is this expected during kolla-build processing | 11:40 |
*** powerd has joined #openstack-kolla | 11:40 | |
kollian | ? | 11:40 |
kollian | pbourke: | 11:40 |
openstackgerrit | Bertrand Lallau proposed openstack/kolla-ansible master: Grafana post_config checks only need to run locally https://review.openstack.org/463810 | 11:49 |
*** jchhatbar_ is now known as janki | 11:51 | |
*** rhallisey has joined #openstack-kolla | 11:52 | |
*** klindgren_ has joined #openstack-kolla | 11:53 | |
*** klindgren has quit IRC | 11:55 | |
openstackgerrit | Sam Betts proposed openstack/kolla-kubernetes master: Run ironic-dnsmasq in privileged mode https://review.openstack.org/465529 | 11:58 |
*** jtriley has quit IRC | 11:59 | |
*** srwilkers has joined #openstack-kolla | 12:01 | |
*** rwallner has quit IRC | 12:02 | |
*** rwallner has joined #openstack-kolla | 12:02 | |
*** sean-k-mooney has joined #openstack-kolla | 12:02 | |
*** janki has quit IRC | 12:04 | |
*** janki has joined #openstack-kolla | 12:04 | |
*** rwallner has quit IRC | 12:07 | |
*** jtriley has joined #openstack-kolla | 12:07 | |
*** goldyfruit has joined #openstack-kolla | 12:07 | |
openstackgerrit | Vladislav Belogrudov proposed openstack/kolla-ansible master: Add possibility to configure tenant and provider VLANs https://review.openstack.org/464675 | 12:07 |
*** goldyfruit has quit IRC | 12:07 | |
*** rwallner has joined #openstack-kolla | 12:09 | |
*** srwilkers has quit IRC | 12:12 | |
*** srwilkers has joined #openstack-kolla | 12:22 | |
openstackgerrit | Sam Betts proposed openstack/kolla-kubernetes master: Remove more dependencies on kolla-ansible https://review.openstack.org/465089 | 12:27 |
*** chas has quit IRC | 12:29 | |
pbourke | kollian: yes | 12:33 |
openstackgerrit | Sam Betts proposed openstack/kolla-kubernetes master: Remove more dependencies on kolla-ansible https://review.openstack.org/465089 | 12:33 |
*** powerd has quit IRC | 12:36 | |
kollian | pbourke: if image is not available then it should skip | 12:37 |
*** chenn has quit IRC | 12:38 | |
kollian | pbourke: will it impact kolla-ansible deploy ? | 12:39 |
*** salv-orlando has joined #openstack-kolla | 12:39 | |
pbourke | kollian: only if you have neutron_plugin_agent set to 'sfc' | 12:40 |
*** salv-orlando has quit IRC | 12:40 | |
*** salv-orlando has joined #openstack-kolla | 12:41 | |
kollian | pbourke: not aware of sfc | 12:41 |
pbourke | kollian: then you dont need to worry | 12:41 |
kollian | does we set it before kolla-ansible run ? | 12:41 |
openstackgerrit | Sam Betts proposed openstack/kolla-kubernetes master: Remove more dependencies on kolla-ansible https://review.openstack.org/465089 | 12:42 |
pbourke | kollian: if you're not using sfc you dont need to change anything | 12:42 |
kollian | pbourke: thanks | 12:44 |
*** ansmith has joined #openstack-kolla | 12:46 | |
*** jrist has quit IRC | 12:47 | |
*** srwilkers has quit IRC | 12:51 | |
*** srwilkers has joined #openstack-kolla | 12:52 | |
*** goldyfruit has joined #openstack-kolla | 12:54 | |
*** fguillot has joined #openstack-kolla | 12:56 | |
*** gomarivera_ has joined #openstack-kolla | 13:02 | |
*** salv-orlando has quit IRC | 13:03 | |
*** salv-orlando has joined #openstack-kolla | 13:04 | |
*** rwallner has quit IRC | 13:04 | |
*** rwallner has joined #openstack-kolla | 13:04 | |
*** gomarivera_ has quit IRC | 13:07 | |
*** salv-orlando has quit IRC | 13:08 | |
*** rwallner has quit IRC | 13:09 | |
*** rwallner has joined #openstack-kolla | 13:10 | |
*** athomas has quit IRC | 13:12 | |
*** athomas has joined #openstack-kolla | 13:13 | |
*** jtriley has quit IRC | 13:14 | |
*** jrist has joined #openstack-kolla | 13:14 | |
*** huzhengchuan has quit IRC | 13:15 | |
*** jtriley has joined #openstack-kolla | 13:16 | |
*** mkoderer has joined #openstack-kolla | 13:17 | |
*** powerd has joined #openstack-kolla | 13:20 | |
*** jamesPR has joined #openstack-kolla | 13:20 | |
rwellum | sbezverk: morning! I am seeing an odd issue when deploying the Canal cni - like I've run out of ip addresses? https://www.irccloud.com/pastebin/cJXd0NFH/ | 13:20 |
*** bjolo has quit IRC | 13:21 | |
*** bjolo_ has quit IRC | 13:21 | |
*** sgrasley has joined #openstack-kolla | 13:22 | |
*** eanylin has quit IRC | 13:23 | |
*** scheuk has joined #openstack-kolla | 13:27 | |
scheuk | hello all | 13:27 |
scheuk | I'm running some test deployments of kolla-ansible 4.0.1 with ceph enabled | 13:27 |
scheuk | I am seeing some weirdness around starting the ceph-mons, it seems like all three are coming up too fast and they are unable to get quorum | 13:28 |
scheuk | if I restart the containers one by one, then they get into quorum | 13:28 |
*** sgrasley has quit IRC | 13:35 | |
*** eanylin has joined #openstack-kolla | 13:39 | |
*** srwilkers has quit IRC | 13:40 | |
sbezverk | rwellum: cannot say I saw something like that. I would start by double checking ranges assigned to pods and consistency between what controller-manager thinks and what canal thinks | 13:40 |
*** jtriley has quit IRC | 13:40 | |
*** srwilkers has joined #openstack-kolla | 13:41 | |
*** jascott1 has joined #openstack-kolla | 13:44 | |
*** ducttape_ has joined #openstack-kolla | 13:45 | |
rwellum | Thanks sbezverk. That level of debugging is beyond my current level :). | 13:47 |
*** ducttape_ has quit IRC | 13:48 | |
rwellum | Is there a chance that canal updated upstream? | 13:48 |
sbezverk | rwellum: busy now, will ping you in a bit, we can check that out together | 13:48 |
rwellum | Sure ty. | 13:48 |
*** jascott1 has quit IRC | 13:48 | |
*** ducttape_ has joined #openstack-kolla | 13:48 | |
*** stefan_bo has joined #openstack-kolla | 13:51 | |
pester | hi all, it is possible to use kolla with existing openstack installation ? let`s say I want kolla to deploy only magnum - how do I pass keystone credentials ? | 13:53 |
*** sgrasley has joined #openstack-kolla | 13:53 | |
*** ansmith has quit IRC | 13:55 | |
*** stefan_bo has quit IRC | 13:55 | |
Reepicheep | pester: AFAIK it theory you should be able to do that. possibly by setting enable_keystone = no and setting the mangnum configs to use an existing keystone | 13:58 |
*** rhallisey has quit IRC | 13:59 | |
*** salv-orlando has joined #openstack-kolla | 14:00 | |
gema | where is the meeting? | 14:01 |
*** zhubingbing has quit IRC | 14:01 | |
gema | meeting-4 is starting a watcher meeting :? | 14:01 |
pester | hm, sounds nice) I can`t see enable_keystone setting in default config is it really exist ? | 14:01 |
pester | or enable_nova | 14:02 |
Reepicheep | pester: is it kolla-ansible? | 14:02 |
gema | oh, it is in two hours | 14:02 |
kfox1111 | morning. | 14:02 |
gema | I am a bit early then xD | 14:02 |
pester | my bad, I can see keystone and horizon bu not other services like nova | 14:03 |
*** srwilkers_ has joined #openstack-kolla | 14:03 | |
pester | so, let me change question) "how do I disable nova ?" | 14:03 |
Reepicheep | if it is kolla-ansible the defaults are in "kolla-ansible/ansible/group_vars/all.yml" that location is dependent on how you installed it | 14:04 |
Reepicheep | you can over write them in the globals.yml (not everything is represented there) | 14:05 |
manheim | egonzalez you on? | 14:05 |
Reepicheep | just add stuff like enable_keystone or enable_nova if you need it | 14:05 |
pester | thx, yes in manual it rely on globals.yml - there my mistake | 14:06 |
*** srwilkers_ has quit IRC | 14:08 | |
*** mgoddard__ has quit IRC | 14:09 | |
*** ansmith has joined #openstack-kolla | 14:09 | |
*** janki has quit IRC | 14:09 | |
*** schwicht has joined #openstack-kolla | 14:12 | |
*** itlinux has joined #openstack-kolla | 14:12 | |
*** ducttape_ has quit IRC | 14:14 | |
*** blallau has joined #openstack-kolla | 14:15 | |
egonzalez | manheim, sup | 14:20 |
openstackgerrit | Anthony Lin proposed openstack/kolla-kubernetes master: Script to Deploy Kolla Kubernetes AIO https://review.openstack.org/464312 | 14:22 |
manheim | it's about the user/password for hyperv in globals | 14:25 |
manheim | thing is I would not want to separate the user from the password, for simplicity sake | 14:26 |
manheim | coz it's not a randomly generated password | 14:26 |
manheim | the user has to input it | 14:26 |
openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes master: Local registry in gate for trunk builds https://review.openstack.org/418550 | 14:27 |
sdake | @sbezverk is the gate functional atm? | 14:33 |
sdake | sbezverk i suspect it is, just wanted confirmation | 14:33 |
sbezverk | sdake: yes | 14:33 |
sdake | thanks | 14:33 |
sdake | I have a patch stream which nearly removes kolla-ansible as a dep | 14:33 |
sdake | nearly as in it doesn't work yet | 14:33 |
sdake | :) | 14:33 |
sdake | but it does remove it | 14:33 |
*** ducttape_ has joined #openstack-kolla | 14:38 | |
*** dave-mccowan has joined #openstack-kolla | 14:38 | |
sdake | rwellum is the upstream deployment guide functional atm? I know sbezverk had a specially build kube* binary to make 1.6.3 function | 14:38 |
*** itlinux has quit IRC | 14:39 | |
*** ansmith has quit IRC | 14:42 | |
*** sgrasley has quit IRC | 14:44 | |
*** ducttape_ has quit IRC | 14:46 | |
*** chandankumar is now known as chkumar|off | 14:49 | |
inc0 | good morning | 14:51 |
*** ducttape_ has joined #openstack-kolla | 14:52 | |
*** srwilkers_ has joined #openstack-kolla | 14:52 | |
*** ducttape_ has quit IRC | 14:54 | |
*** zhubingbing_ has joined #openstack-kolla | 14:54 | |
*** ducttape_ has joined #openstack-kolla | 14:55 | |
*** skramaja has quit IRC | 14:56 | |
manheim | hei inc | 14:58 |
*** manheim has quit IRC | 14:58 | |
*** manheim has joined #openstack-kolla | 14:59 | |
*** salv-orlando has quit IRC | 14:59 | |
*** salv-orlando has joined #openstack-kolla | 15:00 | |
egonzalez | manheim, roger, just looked at there are other password like this case in globals | 15:00 |
sdake | sup inc0 | 15:01 |
inc0 | omfg, Netflix is making the Witcher TV series \o/ | 15:01 |
*** duonghq_ has joined #openstack-kolla | 15:02 | |
duonghq_ | evening guys | 15:02 |
sdake | sup duonghq_ | 15:02 |
*** duonghq has quit IRC | 15:02 | |
sdake | to those using my server - I have it racked and just need to plug it in, althoug hthere are some complications | 15:02 |
*** duonghq_ is now known as duonghq | 15:02 | |
sdake | I am not sure how much longer the outage will last | 15:02 |
sdake | however once i'm done, there will be 3 servers instead of 1 ;) | 15:02 |
*** duonghq_ has joined #openstack-kolla | 15:03 | |
*** manheim has quit IRC | 15:03 | |
spsurya | sdake: sup | 15:04 |
rwellum | sdake: upstream doc needs a couple of revisions. I am working on it. | 15:04 |
*** gardlt has joined #openstack-kolla | 15:04 | |
*** salv-orlando has quit IRC | 15:04 | |
spsurya | rwellum: +1 | 15:04 |
rwellum | Anyone know what is the working current version of kubernetes-cni ? | 15:07 |
*** sgrasley has joined #openstack-kolla | 15:08 | |
*** n0isyn0ise has joined #openstack-kolla | 15:08 | |
*** rhallisey has joined #openstack-kolla | 15:09 | |
*** itlinux has joined #openstack-kolla | 15:09 | |
*** zhubingbing_ has quit IRC | 15:10 | |
*** rwallner has quit IRC | 15:11 | |
*** srwilkers_ has quit IRC | 15:11 | |
*** rhallisey has quit IRC | 15:11 | |
spsurya | rwellum: AFAIK 1.6.3 | 15:12 |
*** rhallisey has joined #openstack-kolla | 15:12 | |
*** ansmith has joined #openstack-kolla | 15:12 | |
*** salv-orlando has joined #openstack-kolla | 15:14 | |
*** rhallisey has quit IRC | 15:15 | |
*** duritong_ has quit IRC | 15:15 | |
*** rwallner has joined #openstack-kolla | 15:16 | |
rwellum | spsurya: '-cni' ? | 15:16 |
*** rhallisey has joined #openstack-kolla | 15:16 | |
rwellum | Seems to be '0.5.1-0 ' on my centos vm | 15:16 |
jamesPR | I think I found a bug regarding certificate, first I load a genuine CA it works fine, then now I redeploy with self signed certificate and haproxy keeps saying inconsistencies between private key and certificate | 15:19 |
spsurya | rwellum: https://github.com/containernetworking/cni/releases | 15:20 |
spsurya | v0.5.2 | 15:20 |
*** hongbin has joined #openstack-kolla | 15:22 | |
*** salv-orlando has quit IRC | 15:22 | |
sdake | rwellum rockin dude thanks :) | 15:22 |
sdake | rwellum i have a bunch of work that i need to submit and having trouble doing dev :) | 15:22 |
rwellum | spsurya: Ok but when I try to yum install that version does not exist. | 15:24 |
rwellum | sdake: Is that a hint? :) | 15:25 |
*** gardlt has quit IRC | 15:25 | |
*** duritong has joined #openstack-kolla | 15:32 | |
kfox1111 | morning again. | 15:33 |
*** jmguzmanc has joined #openstack-kolla | 15:33 | |
rwellum | Morninf kfox1111 - how's the back? | 15:33 |
kfox1111 | 95%. thanks for asking. :) | 15:34 |
kfox1111 | walkings a little slow yet, but doing well. off all the meds now. | 15:34 |
spsurya | kfox1111: sup | 15:35 |
kfox1111 | spsurya: hi. :) | 15:35 |
spsurya | rwellum: not faced yet | 15:35 |
*** AzazeL_ has joined #openstack-kolla | 15:35 | |
egonzalez | duonghq, nice talk in the summit | 15:37 |
*** hieulq_ has joined #openstack-kolla | 15:38 | |
duonghq | egonzalez, thanks hope that some days I can meet you in person | 15:38 |
*** daidv has joined #openstack-kolla | 15:39 | |
*** gardlt has joined #openstack-kolla | 15:40 | |
*** gomarivera_ has joined #openstack-kolla | 15:40 | |
*** blallau has quit IRC | 15:41 | |
*** vhosakot has joined #openstack-kolla | 15:41 | |
inc0 | meeting in 15 | 15:44 |
*** jascott1 has joined #openstack-kolla | 15:44 | |
sdake | kfox1111 good to hear dude | 15:45 |
*** mgoddard__ has joined #openstack-kolla | 15:45 | |
sdake | sup mgkwill | 15:45 |
sdake | su mgoddard__ :) | 15:45 |
sdake | sup jascott1 | 15:45 |
sdake | su pall :) | 15:45 |
mgoddard__ | morning sdake | 15:45 |
mgkwill | Yo sdake | 15:46 |
inc0 | mgoddard__: I've met yet another scientist interested in Kolla:) | 15:46 |
mgoddard__ | inc0: nice - who is it? | 15:47 |
inc0 | Timm, from Berlin | 15:47 |
vhosakot | inc0: if you're talking about Tim we met in Boston, he has a great research usecase with big data for kolla. | 15:48 |
vhosakot | Timm* right, sorry for typo | 15:48 |
inc0 | yeah, him:) | 15:48 |
inc0 | high energy physics from univerity in Berlin | 15:48 |
mgoddard__ | inc0: cool. is he using kolla-ansible or k8s? | 15:49 |
*** jascott1 has quit IRC | 15:49 | |
inc0 | don't know yet | 15:49 |
kfox1111 | I can't remember for sure, but I think he was just kicking the tires. | 15:49 |
*** unicell has joined #openstack-kolla | 15:49 | |
*** unicell1 has quit IRC | 15:50 | |
mgoddard__ | is he aware of kayobe? | 15:50 |
*** srwilkers has quit IRC | 15:51 | |
inc0 | mgoddard__: mind reminding me link? I'll send him email | 15:53 |
inc0 | also connect with you if that's ok | 15:54 |
*** daidv has quit IRC | 15:54 | |
*** gomarivera_ has quit IRC | 15:55 | |
hrw | morning | 15:58 |
hrw | meeting in 2 minutes? | 15:58 |
inc0 | yup | 15:59 |
*** jmguzmanc has quit IRC | 15:59 | |
*** lpetrut has quit IRC | 16:01 | |
*** magicboiz has quit IRC | 16:01 | |
*** daidv has joined #openstack-kolla | 16:06 | |
*** britthouser has joined #openstack-kolla | 16:08 | |
*** hieulq_ has quit IRC | 16:09 | |
*** hieulq_ has joined #openstack-kolla | 16:09 | |
*** hieulq_ has quit IRC | 16:10 | |
*** hieulq_ has joined #openstack-kolla | 16:10 | |
openstackgerrit | Merged openstack/kolla-kubernetes master: Run ironic-dnsmasq in privileged mode https://review.openstack.org/465529 | 16:11 |
*** pcaruana has quit IRC | 16:11 | |
*** hieulq__ has joined #openstack-kolla | 16:13 | |
*** jaosorior is now known as jaosorior_away | 16:17 | |
*** matrohon has quit IRC | 16:17 | |
*** hieulq__ has quit IRC | 16:18 | |
*** zhubingbing has joined #openstack-kolla | 16:23 | |
*** jascott1 has joined #openstack-kolla | 16:32 | |
*** harlowja has joined #openstack-kolla | 16:32 | |
*** srwilkers has joined #openstack-kolla | 16:36 | |
openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes master: WIP: Cache k8s https://review.openstack.org/463331 | 16:37 |
*** manheim has joined #openstack-kolla | 16:43 | |
*** srwilkers_ has joined #openstack-kolla | 16:44 | |
*** unicell has quit IRC | 16:47 | |
*** manheim has quit IRC | 16:48 | |
*** Manheim has joined #openstack-kolla | 16:48 | |
duonghq | inc0, about sliming image, does it better if we start with guide like docker best practice first? | 16:53 |
egonzalez | daidv, the errors probably came because db sync is skipped, bootstrap_service is only "executed" when db creation changed | 16:53 |
inc0 | duonghq: probably that's a good idea | 16:53 |
hieulq_ | https://dzone.com/articles/tensorflow-in-kubernetes-in-88-mb | 16:54 |
egonzalez | daidv, problem with external db is how we ensure idempotency(not execute db sync) on each run | 16:54 |
hieulq_ | they reduce the size of tensorflow from 4GB to 88MB | 16:55 |
hieulq_ | maybe we need to take a look at this work for sliming image | 16:55 |
inc0 | yesterday I've broken my xwindow because I wanted drivers for nvidia for cuda+tensorflow | 16:55 |
inc0 | today I'll go back into the fray | 16:56 |
Manheim | hah, yes inc, i'm a hube witcher fan ^_^ | 16:56 |
inc0 | Manheim: did you read books? | 16:56 |
Manheim | no :( | 16:56 |
inc0 | read them | 16:57 |
duonghq | some idea about this: https://blog.replicated.com/engineering/refactoring-a-dockerfile-for-image-size/, https://ypereirareis.github.io/blog/2016/02/15/docker-image-size-optimization/ | 16:57 |
daidv | egonzalez, Yep, I really have something like separate boostrap.yml task. for creating database, for create user, grant privelge.. | 16:57 |
inc0 | and then replay all 3 witchers | 16:57 |
inc0 | a lot of things will make more sense | 16:57 |
duonghq | iirc, we had refactoring image bp in the past? | 16:57 |
daidv | egonzalez, but one thing in my mind: should we need to have it in real | 16:57 |
Manheim | i see, coo, still thing w3 is the best game ever, even better than mass effect | 16:57 |
Manheim | egonzalez so it's ok if i leave the password in globals? | 16:58 |
egonzalez | Manheim, from my side yep | 16:58 |
Manheim | cool | 16:58 |
*** srwilkers has quit IRC | 16:59 | |
egonzalez | daidv, if the only thing that change in the db/user creation are privs, maybe is worth make current privs configurable https://github.com/openstack/kolla-ansible/blob/master/ansible/roles/aodh/tasks/bootstrap.yml#L26 | 16:59 |
duonghq | inc0, can we reboot the bp? which refactoring image? | 16:59 |
inc0 | yeah | 16:59 |
inc0 | duonghq: I'd start with cleaning them up | 17:00 |
duonghq | sure | 17:00 |
inc0 | as in, readibility | 17:00 |
duonghq | also, I want to finish the bp: https://blueprints.launchpad.net/kolla-ansible/+spec/ansible-specific-task-become in P, hope that I can get some help | 17:00 |
egonzalez | inc0, we reverted the clean up change, mandre found https://bugs.launchpad.net/kolla/+bug/1691413 | 17:01 |
openstack | Launchpad bug 1691413 in kolla "Customization not honored in included templates" [Undecided,New] | 17:01 |
*** manheim_ has joined #openstack-kolla | 17:01 | |
inc0 | ah damn | 17:01 |
inc0 | well | 17:01 |
inc0 | back to business then | 17:01 |
*** egonzalez has quit IRC | 17:02 | |
*** hieulq__ has joined #openstack-kolla | 17:03 | |
*** hieulq_ has quit IRC | 17:04 | |
*** britthouser has quit IRC | 17:06 | |
duonghq | inc0, ya, see you, maybe we can start refactoring image bp, and do it step by step, from simple first, every impprovement should count in final work | 17:06 |
*** zhubingbing has quit IRC | 17:06 | |
*** hieulq__ has quit IRC | 17:07 | |
*** manheim_ has quit IRC | 17:08 | |
openstackgerrit | Sam Betts proposed openstack/kolla-kubernetes master: Remove kolla-ansible dependencies https://review.openstack.org/465089 | 17:09 |
Manheim | egonzalez by release notes you mean docs, right? | 17:09 |
*** sambetts is now known as sambetts|afk | 17:09 | |
*** duonghq has quit IRC | 17:14 | |
*** hieulq_ has joined #openstack-kolla | 17:15 | |
*** unicell has joined #openstack-kolla | 17:18 | |
*** hieulq_ has quit IRC | 17:18 | |
*** chatter29 has joined #openstack-kolla | 17:19 | |
kfox1111 | uh... weird.... | 17:19 |
kfox1111 | ok. so the difference between the working 3.x gate and the broken 4.x gate, | 17:19 |
chatter29 | hey guys | 17:19 |
chatter29 | allah is doing | 17:19 |
chatter29 | sun is not doing allah is doing | 17:19 |
chatter29 | to accept Islam say that i bear witness that there is no deity worthy of worship except Allah and Muhammad peace be upon him is his slave and messenger | 17:19 |
kfox1111 | is nova-compute on 3.x is ignoring the username cinder is giving it and using "nova" | 17:19 |
kfox1111 | and 4 is honoring the request for "cinder" | 17:19 |
*** manheim_ has joined #openstack-kolla | 17:20 | |
*** chatter29 has quit IRC | 17:20 | |
*** eaguilar has joined #openstack-kolla | 17:20 | |
*** chatter29 has joined #openstack-kolla | 17:20 | |
*** chatter29 has quit IRC | 17:20 | |
inc0 | that was wierd | 17:21 |
vhosakot | yes | 17:21 |
*** mannamne has quit IRC | 17:23 | |
*** pbourke has quit IRC | 17:23 | |
*** pbourke has joined #openstack-kolla | 17:25 | |
*** harlowja has quit IRC | 17:26 | |
inc0 | kfox1111: around? | 17:28 |
kfox1111 | inc0: yeah. whats up? | 17:28 |
inc0 | I'm thinking of point you made on mail list | 17:29 |
inc0 | I agree with Thierry that we shouldn't have more than 1 | 17:29 |
inc0 | image | 17:29 |
inc0 | so we won't keep more | 17:29 |
inc0 | but what we can do is to add autoincrement | 17:29 |
inc0 | ocata-4 | 17:29 |
inc0 | if you're running nova-api:ocata-1 | 17:30 |
inc0 | and want to upgrade to ocata-4 | 17:30 |
inc0 | ocata-1 is still on the host so you can rollback | 17:30 |
inc0 | but you won't ever be able to download ocata-1 again | 17:30 |
kfox1111 | I'd prefer to have 2, just so rollbacks are a thing. (like, we havent been bit by k8s rolling forward and deleting the workign release.... ;) | 17:30 |
inc0 | but 2 separated by 1 day? 1 week? 1 month? | 17:31 |
kfox1111 | but if its a big fight, 1 would still go a very long way over where we are now. | 17:31 |
kfox1111 | we build daily. we push whenever fingerprints change. | 17:31 |
inc0 | having 1 latest are only useful for people upgrade daily | 17:31 |
inc0 | I think archiving should be up to users tbh | 17:31 |
inc0 | we can provide tools for that | 17:31 |
kfox1111 | I really don't prefer that, but could live with that. | 17:32 |
inc0 | sure | 17:32 |
kfox1111 | revision numbers are critical though. | 17:32 |
kfox1111 | not "latest" like they are suggesting. | 17:32 |
inc0 | that's fine | 17:32 |
Manheim | egonzalez for the missing auth_uri in [placement], there is no auth_uri in nova role there | 17:32 |
*** shardy has quit IRC | 17:32 | |
inc0 | I don't think that's gonna be questioned | 17:32 |
inc0 | frankly, it's implementational detail | 17:32 |
inc0 | one thing tho | 17:32 |
inc0 | that will be hard to figure out which tag is installable today | 17:33 |
inc0 | and if it's installable today, won't be installable tomorrow | 17:33 |
inc0 | so maybe we could add revision to metadata? | 17:33 |
inc0 | it's gonna be technically hard to make it easy and revisioned at the same time | 17:34 |
inc0 | but that's different discussion, technical, we'll figure it out | 17:34 |
*** powerd has quit IRC | 17:34 | |
kfox1111 | for kolla-kubernetes at least, thats backed in to the helm packages. | 17:34 |
jascott1 | if we are pushing to hub/quay why only have n and n-1? | 17:34 |
kfox1111 | we release an updated helm package pointing to the updated containers. | 17:35 |
kfox1111 | so helm upgrade nova-compute will rolling upgrade to the newest released containers. | 17:35 |
inc0 | kfox1111: what I'm saying is, when you run helm, default for nova-api will be ocata-1? | 17:35 |
inc0 | then default is broken tomorrow | 17:35 |
inc0 | and manual checking it every day is horrible | 17:35 |
inc0 | and not-gatable | 17:35 |
kfox1111 | inc0: the pipeline for pushing out updated containers to the hub also pushes updated helm packages to point to them. | 17:35 |
inc0 | which means unless you install your openstack in the very same day you download code | 17:36 |
inc0 | you're screwed | 17:36 |
inc0 | I don't think we want that | 17:36 |
kfox1111 | yeah, would mean you have to heml update before you helm upgrade. | 17:36 |
kfox1111 | yum works the same way though. | 17:36 |
kfox1111 | apt too. | 17:36 |
inc0 | well, not if they remove older versions | 17:37 |
inc0 | also, kolla-ansible won't have this easily | 17:37 |
inc0 | and we *can't* break kolla-ansible | 17:37 |
kfox1111 | so long as we always release a new helm package to our repo that matches the hub container revisions, it should work fine. | 17:37 |
kfox1111 | kolla-ansible can continue to use 'latest'. | 17:37 |
kfox1111 | until it gets better support for rolling upgrades itself. | 17:37 |
inc0 | anyway, that's purely technical discussion which I'm not that concerned with | 17:38 |
kfox1111 | yeah. | 17:38 |
inc0 | we have several avenues we can pursue | 17:38 |
inc0 | we can have ocata and ocata-1 pointing to same image | 17:38 |
openstackgerrit | Andy Smith proposed openstack/kolla master: Add qdrouterd container https://review.openstack.org/465145 | 17:38 |
inc0 | keep defaults on ocata | 17:38 |
inc0 | and make using ocata-1 easy | 17:38 |
inc0 | regardless, we'll figure it out | 17:38 |
*** manheim__ has joined #openstack-kolla | 17:40 | |
kfox1111 | yeah. kolla-ansible can use ocata, and kolla-kubernetes can use 4.0.1-1, etc. | 17:40 |
openstackgerrit | Dan Ardelean proposed openstack/kolla-ansible master: Add Hyper-V role https://review.openstack.org/455684 | 17:42 |
*** manheim__ has quit IRC | 17:42 | |
*** salv-orlando has joined #openstack-kolla | 17:43 | |
*** salv-orlando has quit IRC | 17:48 | |
*** itlinux has quit IRC | 17:50 | |
*** scheuk has quit IRC | 17:51 | |
*** daidv has quit IRC | 17:51 | |
*** pritesh has joined #openstack-kolla | 17:53 | |
*** scheuk has joined #openstack-kolla | 17:54 | |
*** zhubingbing has joined #openstack-kolla | 17:57 | |
openstackgerrit | Rich Wellum proposed openstack/kolla-kubernetes master: Deployment guide: Added the work-around for kubelet 1.6.3 https://review.openstack.org/465692 | 18:00 |
sdake | rwellum ^^ nice work :) | 18:03 |
rwellum | sdake: thanks :) | 18:04 |
*** feisky has quit IRC | 18:05 | |
openstackgerrit | Merged openstack/kolla master: Install python-pip https://review.openstack.org/465216 | 18:05 |
sdake | inc0 doomsday just hit the mailing list ! | 18:08 |
rwellum | sdake: quite a few people at the summit told me they were going to try it out - and with all the criticism of lack of docs - was great to have something to point to that was recent and approved. | 18:09 |
sdake | good to hear :0 | 18:09 |
sdake | ;) I mean | 18:09 |
jascott1 | rwellum not to rain on parade but imo we'd stick a note in "1.6.3 is broken atm" instead of sending them down the sbezverk binary path | 18:14 |
*** jemcevoy has joined #openstack-kolla | 18:15 | |
jemcevoy | inc0: Are you going to tonight's meetup? | 18:16 |
inc0 | not sure, I have tons to do:/ | 18:17 |
*** sgrasley has quit IRC | 18:17 | |
*** bmace__ has quit IRC | 18:17 | |
*** bmace__ has joined #openstack-kolla | 18:17 | |
jascott1 | maybe I will start a golang meetup while im in MS ;) | 18:17 |
sdake | jascott1 i don't htink its prudent to block on 1.6.4 | 18:17 |
jascott1 | shouldnt we be recommending whatever is latest stable? | 18:18 |
jascott1 | lately they are breaking on patch releases | 18:18 |
rwellum | jascott1: issue is 1.6.2 was removed from the repo - so afaik there is no green path without that binary. | 18:19 |
jascott1 | i guess for me its release to release... a workaround is one thing but sending them to userX's repo for a binary idk | 18:19 |
jascott1 | 1.6.2 is still in repo | 18:19 |
rwellum | Maybe it was transitional? Lots complaining yesterday. | 18:20 |
jascott1 | i installed just this minute with playbook and its pinned to 1.6.2 | 18:20 |
jascott1 | they really need to stop pulling n-1 packages | 18:20 |
jascott1 | wtf is up with that | 18:20 |
*** imcsk8_ has quit IRC | 18:20 | |
rwellum | jascott1: possibly cached? Not - sure. | 18:20 |
rwellum | jascott1: agreed. | 18:21 |
jascott1 | possibly not propagated to all mirrors... im not caching anything | 18:21 |
rwellum | jascott1: I'm on the fence - if you'd rather a note than the binary that is fine too. sdake + or - ? | 18:22 |
sdake | rwellum whatever works WFM | 18:23 |
*** powerd has joined #openstack-kolla | 18:23 | |
jascott1 | just my .02 | 18:23 |
sdake | rwellum if 1.6.2 is in the repos, we should probably be recommending that path | 18:23 |
rwellum | ack - let me confirm it's there for me too | 18:24 |
jascott1 | damn going to have to find/write some mirror scanner that can tell us the state of things :/ | 18:24 |
*** zhubingbing has quit IRC | 18:24 | |
rwellum | jascott1: +1 | 18:24 |
*** bmaltais_ has joined #openstack-kolla | 18:24 | |
jascott1 | or send snarky msg to k8s team saying "stop doing this!" | 18:24 |
rwellum | jascott1: they were asking for volunteers at the summit.. | 18:24 |
* rwellum giggles a little | 18:25 | |
bmaltais_ | eanylin: Tried patch 6. Deployed just fine but the new keepalived VIP does not appear to work. | 18:25 |
bmaltais_ | eanylin: Is it possible it does not know how to talk to my IP across a gateway? | 18:25 |
jascott1 | stop it https://www.youtube.com/watch?v=Ow0lr63y4Mw | 18:25 |
bmaltais_ | I did not see a way to provide the VIP with a GW IP. | 18:25 |
sdake | jascott1 i dont think that would work :) | 18:27 |
sdake | jascott1 however perhaps your southern charm could talk em into it :) | 18:27 |
jascott1 | haha | 18:27 |
jascott1 | maybe my wife with her TN accent ;) | 18:27 |
*** jamesPR has quit IRC | 18:30 | |
*** salv-orlando has joined #openstack-kolla | 18:31 | |
openstackgerrit | Merged openstack/kolla-kubernetes master: Deployment guide: Added the work-around for kubelet 1.6.3 https://review.openstack.org/465692 | 18:32 |
*** Guest54241 has joined #openstack-kolla | 18:33 | |
jascott1 | one thing about that doc changes, not really worth fighting for either way since it will prob change in a few days :/ | 18:33 |
rwellum | jascott1: ack. Not sure how to handle that - just stay on top of it as much as possible. But you have the same issue with the playbook? | 18:35 |
bmaltais_ | eanylin: Also, trying to ping the VIP from the AIO VM does not work... so it look like it is not working atm. | 18:36 |
jascott1 | 1.6.2 is still working for me | 18:36 |
jascott1 | im using apt tho | 18:36 |
*** itlinux has joined #openstack-kolla | 18:36 | |
*** srwilkers_ has quit IRC | 18:38 | |
*** salv-orlando has quit IRC | 18:39 | |
rwellum | jascott1: well I meant in general - not just this issue | 18:39 |
jascott1 | think I will start "kube-klub" where you pay $1 and get to download all the old versions ;) | 18:40 |
sdake | jascott1 "We definately don't go there.." :) | 18:40 |
jascott1 | lol | 18:41 |
sdake | the punchline was epic | 18:42 |
sdake | jascott1 how old are you to remember that :) I was just a young tike when that was on the toobes | 18:42 |
jascott1 | oh yeh I didnt like that when I was a kid, went right over my head and was slow play | 18:43 |
jascott1 | i just saw that as an "adult" for the first time a couple months ago :) | 18:44 |
rwellum | Hey bmaltais_ QQ? | 18:45 |
sdake | this one is epic : ) https://www.youtube.com/watch?v=g2H-9OmErYQ | 18:45 |
bmaltais_ | rwellum: What's up | 18:45 |
sbezverk | rwellum: have you sorted out your canal issue? | 18:45 |
*** Guest54241 has quit IRC | 18:45 | |
bmaltais_ | rwellum: QQ? Not sure I understand the meaning ;-) | 18:45 |
bmaltais_ | rwellum: Quick Question? | 18:46 |
rwellum | QQ=Quick Questin - sorry! The script by eanylin doesn't do a dnslookup kubernetes check. Which is in the deployment guide. | 18:46 |
bmaltais_ | rwellum: Correct as it assumes all is right and does not need to prove it. | 18:46 |
bmaltais_ | rwellum: The guide just does this to make sure everything is working before moving to helm. | 18:47 |
rwellum | sbezverk: I had the same issue on two vm's. I think it may have been because I setup ntp... Anyway couldn't recover so pulled a new VM and now it's got past that. I do have the openvswitch pod in an infinite crash loop though now.. | 18:47 |
bmaltais_ | rwellum: With the script you know it will work so no need to check with nslookup. | 18:47 |
*** salv-orlando has joined #openstack-kolla | 18:48 | |
rwellum | bmaltais_: not sure that's true. What if you've pulled in a new version of canal etc and it's broken? | 18:48 |
bmaltais_ | rwellum: I think the script is prescriptive about the version so a newer version should not cause an issue since it won't get used unless you modify the script to do so. | 18:49 |
sbezverk | rwellum: any logs? | 18:49 |
jascott1 | rwellum I had a todo to automate that check but its just not a priority imo | 18:49 |
sbezverk | rwellum: docker log from crashing container might be helpfull | 18:49 |
rwellum | sbezverk: in about 10m, I retried and I've got about 50 pods still to come up. | 18:50 |
bmaltais_ | rwellum: ## Version Pinning\n kubelet_version=1.6.2-0 canal_version=1.6 HELM_VERSION="2.3.0" | 18:50 |
jascott1 | sdake "at the risk of sounding realistic" haha | 18:50 |
*** gfidente is now known as gfidente|afk | 18:51 | |
rwellum | jascott1: Yeah I think you could create the busybox pod, then remote access to run the command? | 18:51 |
rwellum | bmaltais_: point taken. | 18:51 |
jascott1 | think so | 18:51 |
*** rwallner has quit IRC | 18:52 | |
rwellum | bmaltais_:so in my code review I said that these things would be good to be passed in by the user (there'd be a default). If it goes that route then I think we'd need this logic. Otherwise I agree with sane defaults not that bad. | 18:52 |
rwellum | sbezverk: can I create a pod, access it and run a command - retrieving the output? | 18:53 |
bmaltais_ | rwellum: That would make sense. Not sure if bash is the best choice. Maybe moving to python or something to automate the build might be more scalable. | 18:53 |
jascott1 | rwellum kubectl can exec into pods | 18:54 |
bmaltais_ | rwellum: Bash is a nice prototype I think and work remarkably well so far. | 18:54 |
rwellum | bmaltais_: https://github.com/RichWellum/k8s/blob/master/k8s.py :) | 18:54 |
rwellum | jascott1: great - makes sense of course. Will look into that. | 18:55 |
jascott1 | ah secret to multinode is to join nodes after all the kube-system pods have surely landed | 18:55 |
bmaltais_ | rwellum: Correct, something similar to it. I wonder if there is an easy bash2py util ;-) | 18:55 |
sbezverk | rwellum: not if it is crashing | 18:55 |
rwellum | bmaltais_: well the above follows the deployment guide. I was going to abandon it for the playbook in the future. But for now capturing my progress in it. | 18:56 |
*** rwallner has joined #openstack-kolla | 18:56 | |
rwellum | sbezverk: that is true :) | 18:56 |
bmaltais_ | rwellum: Any link to the playbook wip? | 18:56 |
jascott1 | bmaltais_ this one? https://review.openstack.org/#/c/457317/ | 18:57 |
rwellum | bmaltais_: ^^^^ | 18:57 |
rwellum | Very cool. | 18:58 |
* rwellum (I like my python though) | 18:58 | |
sbezverk | rwellum: sometime you might have a better chances to get logs using kubectl command | 18:58 |
sbezverk | rwellum: kubectl logs {pod name} -n kolla | 18:58 |
*** salv-orlando has quit IRC | 18:59 | |
rwellum | sbezverk: of course now it's come up just fine, first time, no crashing and I can access horizon with no problem. | 18:59 |
rwellum | How can it be so unstable at times? | 18:59 |
jascott1 | i blame wave potentiality and super position | 19:00 |
bmaltais_ | jascott1: Wow, this look nice. Is it functional? I should give it a try. Been manually creating my k8t multinode from AIO but it is not functional yet. | 19:00 |
jascott1 | yes its completely functional | 19:00 |
jascott1 | im working on multinode piece now and it seems to be working | 19:00 |
bmaltais_ | jascott1: Wow, why isn't it more front and center in the kolla-kubernetes guide? | 19:01 |
jascott1 | one thing in global vars, currenty have to manually set the api_advertise_ip | 19:01 |
jascott1 | its new | 19:01 |
jascott1 | still needs testing | 19:01 |
jascott1 | sorry function yes, well tested no | 19:01 |
jascott1 | but it works for me on centos and ubuntu | 19:01 |
bmaltais_ | jascott1: But still... a pointer to it would help folks discover the greatness ;-) | 19:01 |
jascott1 | sure I think that could happen soon | 19:02 |
jascott1 | i need to add autodetect api_advertise_ip from interface with optional override in vars | 19:02 |
bmaltais_ | jascott1: I will most certainly give it a try. Is there a short quickstart page on how to make use of it? Like a basic requirements and launch command? | 19:03 |
jascott1 | its not merged yet and sorry no docs but it is essentially a somewhat faithful reproduction of the deployment guide | 19:03 |
*** lpetrut has joined #openstack-kolla | 19:03 | |
rwellum | I think we'll be able to add a pointer in the deployment guide, with a description of how to use etc. | 19:04 |
rwellum | When stable | 19:04 |
bmaltais_ | rwellum: Nice. | 19:04 |
jascott1 | bmaltais_ I will annotate the vars asap but FYI you need to change global_vars/all.yml to have your VMs interface names (eth0, eth1) etc | 19:05 |
*** ansmith has quit IRC | 19:05 | |
rwellum | jascott1: how are you doing multi-node? Multiple vm's? | 19:05 |
jascott1 | and also change the api_advertise_ip as I mentioned earlier.. usually thats the IP of eth0 | 19:05 |
jascott1 | ya | 19:05 |
rwellum | Do you have to have the VM's up first? | 19:06 |
bmaltais_ | jascott1: Making notes. | 19:06 |
jascott1 | rwellum | 19:06 |
*** mgoddard__ has quit IRC | 19:06 | |
jascott1 | rwellum yes you need the node | 19:06 |
jascott1 | but then you just kubeadm join | 19:06 |
rwellum | ok cool - very cool actually. | 19:06 |
jascott1 | yeah i have it working as of yesterday with multi.. let me get this updated and pushed. I will add the autodetect api_advertise_ip later | 19:07 |
*** imcsk8 has joined #openstack-kolla | 19:08 | |
rwellum | Won't get it done if we don't stop bugging you.. | 19:08 |
bmaltais_ | jascott1: Might be a stupid one but what command do you type at your end to run the playbook? Just show how new I am at this ;-) | 19:09 |
bmaltais_ | jascott1: Once the vars and IP are set? | 19:09 |
bmaltais_ | I used kubeadm to build my multinode k8t | 19:10 |
bmaltais_ | jascott1: But never ran a yml playbook... | 19:10 |
jascott1 | ansible-playbook -i <path to inventory file> k8s-deply.yml | 19:10 |
bmaltais_ | jascott1: OK, ansible. Make sense. That is just enough info to make me dangerous ;-) | 19:11 |
*** bmaltais has quit IRC | 19:11 | |
*** egonzalez has joined #openstack-kolla | 19:12 | |
jascott1 | it may just be glorified ssh but it is glorified :) | 19:12 |
bmaltais_ | jascott1: ;-) Like glorification. | 19:13 |
*** stvnoyes has joined #openstack-kolla | 19:14 | |
*** stvnoyes has left #openstack-kolla | 19:14 | |
sdake | jascott1 did that review land? | 19:14 |
jascott1 | which? | 19:15 |
sdake | playbooks | 19:15 |
egonzalez | Manheim still around? | 19:15 |
jascott1 | about to push an update | 19:16 |
Manheim | yep | 19:19 |
egonzalez | Manheim re release note https://docs.openstack.org/developer/reno/usage.html | 19:19 |
*** ansmith has joined #openstack-kolla | 19:20 | |
jascott1 | what is everyone using to setup/tearsdown VMs ? | 19:21 |
bmaltais_ | jascott1: Errr... vCenter cloning ;-) I know, I should use some API ;-) | 19:22 |
*** gomarivera_ has joined #openstack-kolla | 19:22 | |
jascott1 | ok one vote for that :p | 19:23 |
bmaltais_ | jascott1: Make a base centos7 VM and happily clone ever after | 19:23 |
jascott1 | im using terraform | 19:23 |
jascott1 | it can write the ansible inventory file for you | 19:23 |
jascott1 | so everytime I create/destroy i can inventory automatically | 19:24 |
bmaltais_ | jascott1: terraform look cool. I should liik into it | 19:24 |
jascott1 | vagrant will also do this but kolla-kubenetes is too big for vagrant (on my MBP anyway) | 19:25 |
rwellum | Been using virsh-manager... | 19:26 |
*** gomarivera_ has quit IRC | 19:27 | |
*** mkoderer has quit IRC | 19:27 | |
jascott1 | terraform lets you also specify networks, subnets, floating ips etc | 19:29 |
*** rwallner has quit IRC | 19:29 | |
jascott1 | as in create them | 19:29 |
bmaltais_ | jascott1: And it does interface with vCenter ;-) Yeah! Like learning new tricks. | 19:30 |
jascott1 | like CloudFormation on AWS but terraform works on AWS, openstack and others (that Ive never used) | 19:30 |
rwellum | Is it automated or GUI based etc? | 19:30 |
jascott1 | ah cool | 19:30 |
*** rwallner has joined #openstack-kolla | 19:32 | |
jascott1 | terraform is cli | 19:32 |
jascott1 | declare infrastructure in a file and it makes it happen | 19:33 |
jascott1 | one nit I have tho is that an aws_instance does not equal openstack_instance | 19:33 |
jascott1 | so you cant make a generic setup and then point it to aws or openstack all willy nilly | 19:34 |
jascott1 | but those platforms are different so I kinda get it | 19:34 |
jascott1 | i have files for both aws and openstack providers but am trying to determine if anyone uses it or how we should handle these various deployment systems | 19:35 |
*** bmace__ has quit IRC | 19:35 | |
*** rwallner has quit IRC | 19:36 | |
rwellum | Sounds like a must-try | 19:37 |
*** mannamne has joined #openstack-kolla | 19:38 | |
*** harlowja has joined #openstack-kolla | 19:39 | |
jascott1 | its nice in that i run `terraform apply` then `ansible-playbook blah blah` then `terraform destroy` rinse and repeat | 19:39 |
bmaltais_ | jascott1: Cool. I will try it in the next few days and I will have a better opinion on the subject. | 19:43 |
bmaltais_ | jascott1: So you drive everything from your MBP OSX bash cli. | 19:43 |
*** mgoddard__ has joined #openstack-kolla | 19:47 | |
*** lpetrut has quit IRC | 19:47 | |
*** bmaltais_ has quit IRC | 19:47 | |
*** gfidente|afk has quit IRC | 19:48 | |
*** harlowja has quit IRC | 19:53 | |
*** stefan_bo has joined #openstack-kolla | 19:53 | |
*** myoung is now known as myoung|biab | 19:55 | |
*** myoung|biab is now known as myoung | 19:55 | |
*** scheuk has quit IRC | 19:56 | |
*** harlowja has joined #openstack-kolla | 19:57 | |
*** matrohon has joined #openstack-kolla | 19:57 | |
*** stefan_bo has quit IRC | 19:58 | |
jascott1 | bmaltais_ yea | 19:59 |
jascott1 | meeting, brb | 19:59 |
*** jascott1 has quit IRC | 19:59 | |
*** harlowja has quit IRC | 20:00 | |
*** ansmith has quit IRC | 20:08 | |
*** jascott1 has joined #openstack-kolla | 20:12 | |
kfox1111 | weird.... | 20:13 |
kfox1111 | nova 4+ is ignoring 'rbd_user' | 20:13 |
*** unicell has quit IRC | 20:14 | |
*** unicell has joined #openstack-kolla | 20:14 | |
openstackgerrit | Bertrand Lallau proposed openstack/kolla-ansible master: Support osprofile usage https://review.openstack.org/455628 | 20:14 |
openstackgerrit | Bertrand Lallau proposed openstack/kolla-ansible master: Support OSprofile usage https://review.openstack.org/455628 | 20:15 |
*** mgoddard__ has quit IRC | 20:16 | |
openstackgerrit | Justin Scott proposed openstack/kolla-kubernetes master: Playbook implements Deployment Guide https://review.openstack.org/457317 | 20:16 |
openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes master: WIP: v4 gates. https://review.openstack.org/454841 | 20:17 |
rwellum | jascott1: been trying for an hour+ to get the busybox pod to create and then exec against it, no luck getting this to work; any ideas here? | 20:19 |
jascott1 | hmm one sec | 20:20 |
*** fguillot has quit IRC | 20:27 | |
jascott1 | rwellum yeah Im having trouble gettig basic commands to run in busybox with kubectl exec | 20:33 |
kfox1111 | hmmmm.... the ceph code in nova-compute did change a fair amount... | 20:33 |
*** itlinux_ has joined #openstack-kolla | 20:34 | |
rwellum | jascott1: this finally worked https://www.irccloud.com/pastebin/8ai4uff3/ | 20:36 |
jascott1 | rwellum good job | 20:37 |
*** itlinux has quit IRC | 20:37 | |
jascott1 | although there has got to be a way to `kubectl run busybox ... && kubectl exec nslookup kubernetes | 20:37 |
rwellum | Right but I couldn't get that box to stay up | 20:38 |
jascott1 | does busybox have /sbin/init? cant ever remember | 20:38 |
openstackgerrit | Mathieu Rohon proposed openstack/kolla master: Add skydive service https://review.openstack.org/459995 | 20:39 |
rwellum | not sure tbh | 20:40 |
rwellum | The above created the busybox pod in namespace default - you think that's ok jascott1 ? | 20:41 |
*** ducttape_ has quit IRC | 20:45 | |
*** ducttape_ has joined #openstack-kolla | 20:45 | |
*** Serlex has quit IRC | 20:45 | |
jascott1 | rwellum ultimately the pod should be cleaned up anyway so doesnt matter afaik | 20:46 |
jascott1 | putting it in kolla might be cleaner tho | 20:47 |
*** rwallner has joined #openstack-kolla | 20:47 | |
jascott1 | then we can say *everything installed is in kolla namespace | 20:48 |
rwellum | good point. | 20:48 |
rwellum | Next question is how... | 20:48 |
jascott1 | in ansible? | 20:49 |
rwellum | no in the yaml file I created :) | 20:49 |
jascott1 | oh how to change namespace | 20:50 |
jascott1 | ? | 20:50 |
jascott1 | sorry not getting exact question | 20:50 |
rwellum | Yeah my bad.. I think all I do is add: --namespace=kolla to the create command. | 20:50 |
jascott1 | sounds legit :) | 20:51 |
*** rwallner has quit IRC | 20:52 | |
*** manheim__ has joined #openstack-kolla | 20:52 | |
jascott1 | if nslookup gives non-zero on fail to lookup then in ansible can take output of command and check <name>.rc != 0 to fail it | 20:53 |
jascott1 | instead of having to parse and make sense of output | 20:55 |
rwellum | Nice | 20:55 |
*** kbaegis has joined #openstack-kolla | 20:56 | |
*** manheim__ has quit IRC | 20:56 | |
jascott1 | im flying united tomorrow... was just thinking I hope no one gets sick | 20:59 |
jascott1 | "Is there a doctor on the plane?" crickets. | 20:59 |
vhosakot | jascott1: s/no one gets sick/I won't be thrown out of the plane ;) | 21:00 |
jascott1 | im travleling for pleasure so Id prob take the $800 | 21:00 |
jascott1 | hello new tube amp | 21:00 |
jascott1 | well er old tube amp | 21:01 |
vhosakot | yeah, I don;t mind taking the money and sleeing at the airport.. lol | 21:01 |
jascott1 | sleeping/drinking | 21:01 |
vhosakot | ^^^ | 21:01 |
jascott1 | so overpriced could prob burn thru $800 | 21:01 |
vhosakot | lol | 21:02 |
*** bmace__ has joined #openstack-kolla | 21:03 | |
*** jrist has quit IRC | 21:04 | |
jascott1 | how long til the drug us put us in boxes and density pack us for max efficiency | 21:04 |
jascott1 | frontier is almost there now ;) | 21:05 |
*** unicell1 has joined #openstack-kolla | 21:05 | |
*** unicell has quit IRC | 21:05 | |
*** fooliouno has joined #openstack-kolla | 21:06 | |
vhosakot | there will be a new class of airline seats soon along with business class and economy - person-on-lap lol | 21:07 |
jascott1 | the `turtles all the way down` class | 21:07 |
*** hongbin has quit IRC | 21:08 | |
vhosakot | lol | 21:08 |
jascott1 | rwellum might change name of container too, seomthing like kolla-dns-test or something | 21:09 |
jascott1 | sorry imeant pod | 21:10 |
jascott1 | hmm maybe a job? | 21:10 |
rwellum | good idea | 21:10 |
jascott1 | the fun never ends :) | 21:10 |
rwellum | I'm seeing: 'error: unable to upgrade connection: container not found ("busybox")'... | 21:10 |
*** gomarivera_ has joined #openstack-kolla | 21:11 | |
rwellum | But when I run the command manually it works.. | 21:11 |
jascott1 | is it still downloading/starting when second command runs? | 21:12 |
rwellum | Maybe - might need to loop checking for it. | 21:14 |
*** itlinux_ has quit IRC | 21:14 | |
*** gomarivera_ has quit IRC | 21:15 | |
*** hongbin has joined #openstack-kolla | 21:16 | |
*** gardlt has quit IRC | 21:17 | |
rwellum | jascott1: yup that worked. | 21:22 |
jascott1 | i cant get my busybox to do right today.. idk what is up | 21:23 |
rwellum | Now what's interesting is that this pod returns 'nslookup: can't resolve 'kubernetes'' - but if I create the one manually it's fine. | 21:23 |
*** matrohon has quit IRC | 21:24 | |
rwellum | Only thing I added was the namespace | 21:24 |
rwellum | bah | 21:24 |
jascott1 | hmm | 21:27 |
jascott1 | tried without namespace? | 21:27 |
jascott1 | the guide is not in kolla iirc sorry | 21:27 |
rwellum | yeah true. | 21:27 |
rwellum | Yup I'll try without namespace | 21:28 |
*** athomas has quit IRC | 21:31 | |
*** schwicht has quit IRC | 21:37 | |
*** mannamne has quit IRC | 21:39 | |
*** mannamne has joined #openstack-kolla | 21:40 | |
jascott1 | rwellum i can see how one can blow an hour messing with kubectl run and exec | 21:42 |
rwellum | jascott1: I was assuming that it was, as usual, my own ignorance. | 21:46 |
rwellum | BTW no namesapce worked. Wants to be in 'default'. Go figure. | 21:46 |
openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes master: WIP: v4 gates. https://review.openstack.org/454841 | 21:46 |
*** manheim__ has joined #openstack-kolla | 21:47 | |
jascott1 | im sure there is a very good reason for that :) | 21:47 |
rwellum | Well if it helps for your ansible playbook, feel free to take a look at: https://github.com/RichWellum/k8s/blob/master/k8s.py#L802 | 21:48 |
rwellum | Sure I could look for the return code - instead of counting addresses - will look tomorrow. | 21:49 |
*** egonzalez has quit IRC | 21:50 | |
*** aagate has joined #openstack-kolla | 21:54 | |
jascott1 | thanks | 21:54 |
*** kbaegis has quit IRC | 21:56 | |
*** esmute_ has quit IRC | 22:00 | |
Seyong | jascott1: hi :) | 22:01 |
kolla-slack1 | <jascott1> Seyong hey! | 22:02 |
kolla-slack1 | <jascott1> so I got multinode working and it didnt hose all my kube-system stuff | 22:02 |
Seyong | jascott1: yesterday my issue was just happening, my kubernetes master and minion have same hostname! and clustering does not be made | 22:02 |
jascott1 | oh! | 22:02 |
*** harlowja has joined #openstack-kolla | 22:02 | |
jascott1 | well the only thing I have figured out since we chatted was that workers should join after everything in kube-system has been scheduled | 22:03 |
*** schwicht has joined #openstack-kolla | 22:04 | |
Seyong | yes right | 22:04 |
jascott1 | otherwise dns pod gets scheduled on a worker node for me and dns and tiller break perpetually | 22:05 |
jascott1 | Seyong so everything is running now? | 22:05 |
*** aagate has quit IRC | 22:06 | |
*** ducttape_ has quit IRC | 22:07 | |
*** ducttape_ has joined #openstack-kolla | 22:11 | |
*** eaguilar has quit IRC | 22:13 | |
*** aagate has joined #openstack-kolla | 22:17 | |
*** rwsu has quit IRC | 22:17 | |
*** fooliouno has quit IRC | 22:18 | |
*** harlowja has quit IRC | 22:18 | |
*** aagate has quit IRC | 22:18 | |
*** vhosakot has quit IRC | 22:22 | |
sbezverk | kfox1111: ping | 22:23 |
kfox1111 | hi | 22:24 |
*** aagate has joined #openstack-kolla | 22:27 | |
*** Manheim has quit IRC | 22:27 | |
*** manheim_ is now known as manheim | 22:27 | |
kolla-slack1 | <margin> hi | 22:29 |
*** rhallisey has quit IRC | 22:31 | |
openstackgerrit | Seyong Um proposed openstack/kolla-kubernetes master: Correct canal pod network band https://review.openstack.org/465754 | 22:33 |
*** itlinux has joined #openstack-kolla | 22:33 | |
sbezverk | kfox1111: I was just wondering if you knew a way to ssh to the host and provide a password in the command line | 22:34 |
sbezverk | I cannot use ssh keys | 22:34 |
sbezverk | but I think I found a solution, to call expect script from bash script | 22:34 |
kfox1111 | the host of what, from where? | 22:35 |
kfox1111 | expect does work, but generally the tool of last resort. :) | 22:35 |
sbezverk | kfox1111: from unix box to cisco router | 22:36 |
sbezverk | I have very limited capacity in the router as it is not real shell | 22:36 |
sbezverk | I guess without expect there is no way | 22:36 |
*** schwicht has quit IRC | 22:37 | |
kfox1111 | ah. | 22:37 |
kfox1111 | soemtimes you can load certs into routers. | 22:37 |
kfox1111 | what are you trying to automate? | 22:37 |
sbezverk | kfox1111: see pm | 22:41 |
Seyong | jascott1: well I met a problem of neutron doesnt initializing, I will take a look at it why it doesnt. | 22:46 |
jascott1 | Seyong iirc for me it was volume related *i think* | 22:49 |
*** itlinux has quit IRC | 22:49 | |
*** bmaltais has joined #openstack-kolla | 22:54 | |
*** bmaltais has quit IRC | 22:54 | |
*** bmaltais has joined #openstack-kolla | 22:55 | |
*** powerd has quit IRC | 22:58 | |
*** itlinux has joined #openstack-kolla | 22:59 | |
*** eaguilar has joined #openstack-kolla | 23:01 | |
*** powerd has joined #openstack-kolla | 23:01 | |
Seyong | jascott1: let me check volume stuff either :) | 23:08 |
jascott1 | iirc it was syslog | 23:08 |
*** rwsu has joined #openstack-kolla | 23:10 | |
Seyong | ok | 23:10 |
*** goldyfruit has quit IRC | 23:10 | |
openstackgerrit | Marcus Williams proposed openstack/kolla-ansible master: Add OpenDaylight role https://review.openstack.org/416367 | 23:10 |
*** ducttape_ has quit IRC | 23:25 | |
*** mannamne has quit IRC | 23:26 | |
*** itlinux has quit IRC | 23:27 | |
*** rhallisey has joined #openstack-kolla | 23:28 | |
*** kbaegis has joined #openstack-kolla | 23:31 | |
*** powerd has quit IRC | 23:33 | |
*** eaguilar has quit IRC | 23:39 | |
*** eaguilar has joined #openstack-kolla | 23:42 | |
*** pritesh has left #openstack-kolla | 23:46 | |
*** zul has joined #openstack-kolla | 23:47 | |
*** manheim__ has quit IRC | 23:49 | |
openstackgerrit | Marcus Williams proposed openstack/kolla-ansible master: Add OpenDaylight role https://review.openstack.org/416367 | 23:51 |
jascott1 | anyone know if you can set a global var in ansible during pb execution? seems mostly host based | 23:54 |
*** stefan_bo has joined #openstack-kolla | 23:55 | |
*** hongbin has quit IRC | 23:56 | |
*** rwellum has quit IRC | 23:57 | |
*** stefan_bo has quit IRC | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!