| *** goldyfruit has quit IRC | 00:02 | |
| *** goldyfruit has joined #openstack-kolla | 00:03 | |
| spsurya | portdirect: around ? | 00:04 |
|---|---|---|
| spsurya | reagarding this https://review.openstack.org/#/c/457042/ | 00:04 |
| *** sayantan_ has quit IRC | 00:06 | |
| *** manheim has joined #openstack-kolla | 00:07 | |
| *** kollian has quit IRC | 00:07 | |
| *** Guest68918 has joined #openstack-kolla | 00:09 | |
| portdirect | spsurya: will be around properly in an hour or so. Sup? | 00:17 |
| spsurya | *autocorrect * ? sorry not properly understand | 00:18 |
| spsurya | https://review.openstack.org/#/c/457042/ | 00:18 |
| portdirect | spsurya: wrote that on my phone, (as i am now). It wants to change your name to situation each time i type it :( | 00:19 |
| spsurya | :) :) | 00:20 |
| spsurya | np | 00:20 |
| spsurya | your pohone will be use to of this soon | 00:21 |
| portdirect | Have you been using the aio i wrote for openstack-helm? Be great to get feedback | 00:21 |
| spsurya | yes | 00:21 |
| spsurya | following the same pattern for kk8s | 00:22 |
| spsurya | dev part AIO | 00:22 |
| spsurya | portdirect: i was getting the understaing of that | 00:22 |
| *** jayahn has joined #openstack-kolla | 00:28 | |
| *** dave-mccowan has quit IRC | 00:39 | |
| *** manheim has quit IRC | 00:40 | |
| *** zhurong has joined #openstack-kolla | 00:41 | |
| *** tovin07_ has joined #openstack-kolla | 00:47 | |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible master: Remove the unnecessary tenant name https://review.openstack.org/457006 | 00:50 |
| *** jayahn has quit IRC | 00:56 | |
| openstackgerrit | Surya Prakash (spsurya) proposed openstack/kolla-kubernetes master: [WIP] Kubeadm-aio container for kolla-k8s dev and gates https://review.openstack.org/457042 | 00:56 |
| *** jayahn has joined #openstack-kolla | 00:58 | |
| *** tonanhngo has joined #openstack-kolla | 01:00 | |
| *** duonghq has joined #openstack-kolla | 01:01 | |
| *** tonanhngo has quit IRC | 01:01 | |
| *** Guest68918 has quit IRC | 01:03 | |
| openstackgerrit | Surya Prakash (spsurya) proposed openstack/kolla-kubernetes master: [WIP] Kubeadm-aio container for kolla-k8s dev and gates https://review.openstack.org/457042 | 01:04 |
| *** cuongnv has joined #openstack-kolla | 01:05 | |
| *** unicell has joined #openstack-kolla | 01:07 | |
| *** guoguo has joined #openstack-kolla | 01:09 | |
| *** tervancovan has joined #openstack-kolla | 01:28 | |
| *** zhurong has quit IRC | 01:30 | |
| openstackgerrit | Shunli Zhou proposed openstack/kolla-ansible master: Add kolla-ansible to operating-kolla.rst document https://review.openstack.org/456110 | 01:31 |
| *** hrw has quit IRC | 01:31 | |
| spsurya | sdake it would be great, if you review recent generated one. | 01:31 |
| spsurya | for now added only the required files | 01:32 |
| *** hrw has joined #openstack-kolla | 01:32 | |
| *** dixiaoli has joined #openstack-kolla | 01:34 | |
| *** zhuzeyu has joined #openstack-kolla | 01:35 | |
| *** StephenWang1991 has joined #openstack-kolla | 01:35 | |
| *** dixiaoli has quit IRC | 01:40 | |
| *** jayahn has quit IRC | 01:41 | |
| *** tonanhngo has joined #openstack-kolla | 01:42 | |
| *** tonanhngo has quit IRC | 01:42 | |
| tervancovan | hi all | 01:43 |
| *** jayahn has joined #openstack-kolla | 01:44 | |
| sdake | spsurya i think you can simplify it a bit by removing nfs for example | 01:46 |
| sdake | spsurya and implementing some code | 01:46 |
| sdake | and possibly just copying install-k8s.sh from that review i linked | 01:47 |
| sdake | spsurya inintial goai is for it to match the deployment document as much as possible (centos for e.g.) | 01:47 |
| spsurya | sdake: roger | 01:48 |
| *** Pavo has joined #openstack-kolla | 01:50 | |
| *** Pavo has quit IRC | 01:53 | |
| zhubingbing | sdake sup | 01:53 |
| sdake | hi zhubingbing | 01:53 |
| *** tovin07 has joined #openstack-kolla | 01:59 | |
| *** zhugaoxiao has quit IRC | 02:02 | |
| *** zhugaoxiao has joined #openstack-kolla | 02:02 | |
| *** zhurong has joined #openstack-kolla | 02:03 | |
| *** jayahn has quit IRC | 02:06 | |
| openstackgerrit | Steven Dake proposed openstack/kolla-kubernetes master: A start at a simplified devtool environment https://review.openstack.org/457075 | 02:12 |
| *** tervancovan has quit IRC | 02:13 | |
| *** japestinho has joined #openstack-kolla | 02:15 | |
| Jeffrey4l | sdake, ping. busy yesterday. But have you solved the issue? | 02:22 |
| sdake | Jeffrey4l i have not | 02:22 |
| sdake | Jeffrey4l arey ou available now to debug? | 02:22 |
| sdake | Jeffrey4l could really use the help | 02:22 |
| Jeffrey4l | yes. | 02:22 |
| Jeffrey4l | let me check | 02:23 |
| sdake | Jeffrey4l i need to get it back to the state it was in | 02:23 |
| sdake | Jeffrey4l it will take 5-10 mins | 02:23 |
| sdake | Jeffrey4l will you be able to look then? | 02:23 |
| sdake | i believe it is an ordering problem | 02:24 |
| Jeffrey4l | OK. | 02:24 |
| sdake | but I'm not really sure what the exactly issue is | 02:24 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible master: Add barbican_crypto_password key in passwords.yml file https://review.openstack.org/457117 | 02:28 |
| sdake | Jeffrey4l if your on the machine now - check out watch kubectl pods -n kolla | 02:31 |
| sdake | Jeffrey4l su - | 02:32 |
| sdake | sudo -i | 02:32 |
| sdake | rather | 02:32 |
| sdake | sudo -i | 02:32 |
| sdake | su - sdake | 02:32 |
| portdirect | spsurya: you about? | 02:32 |
| Jeffrey4l | yep, i am in. checking | 02:32 |
| spsurya | portdirect: your kubeadm AIO for opensatck | 02:34 |
| portdirect | You want to collaborate on that? | 02:34 |
| sdake | portdirect i think what we are really looking for is something more like install-k8s.sh | 02:35 |
| sdake | portdirect kubeadm-aio is pretty nicely done - although doesn't particularly match our workflow | 02:35 |
| Jeffrey4l | sdake, /home/sdake/keystonerc_admin do not work? | 02:35 |
| sdake | Jeffrey4l owkring on deplying dude :) | 02:36 |
| sdake | run watch kubeadm get pods -n kolla | 02:36 |
| Jeffrey4l | kubectl get pods -n kolla? | 02:37 |
| Jeffrey4l | get "No resources found." | 02:37 |
| sdake | Jeffrey4l hang tight | 02:38 |
| sdake | use the "watch" command | 02:38 |
| Jeffrey4l | works. | 02:38 |
| sdake | it is deployign now | 02:39 |
| Jeffrey4l | got. | 02:39 |
| sdake | about 150 microcharts are doing their thing | 02:39 |
| sdake | at the end of this process I register the cells jobs | 02:39 |
| portdirect | sdake: no problem, just thought that this was a great area for me to try and work on both kolla-k8s and openstack-helm. https://blueprints.launchpad.net/kolla-kubernetes/+spec/devenv-taketwo | 02:40 |
| portdirect | Feel free to ping me if you need a hand with anything | 02:40 |
| sdake | portdirect it could be - spsurya may ened some help - although I'd like whatever devenv we come up with to match our deployment documentation | 02:41 |
| sdake | portdirect so if your willing to help on two parallel implementations of the same hting, wfm :) | 02:41 |
| sdake | Jeffrey4l its almost done deploying | 02:42 |
| portdirect | Np, though ideally we'd just have a single base for k8s Dev deployment | 02:42 |
| *** goldyfruit has quit IRC | 02:42 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible master: [TEST][MASTER][KOLLA-ANSIBLE]TEST CEILOMETER RELATED SERVICE https://review.openstack.org/412278 | 02:43 |
| portdirect | But understand if you think we should be going custom with kolla k8s | 02:43 |
| sdake | portdirect we tried that with halcyon - no ciesta bien | 02:43 |
| *** tonanhngo has joined #openstack-kolla | 02:44 | |
| *** tonanhngo has quit IRC | 02:44 | |
| portdirect | No fiesta bein? New lingo for me | 02:44 |
| sdake | ;-) | 02:45 |
| sdake | it didn't work out | 02:45 |
| sdake | halycon not maintained | 02:45 |
| sdake | kolla-kbuernetes can not have a critical path dependency that is unmaintained | 02:45 |
| sdake | nor can we have a critical path dependency that we cannot personally maintain | 02:45 |
| sdake | Jeffrey4l libvirt and compute imploded on that go | 02:46 |
| portdirect | Totally get that, thats why I'm offering any assistance i can. | 02:47 |
| sdake | sounds good thanks :) | 02:48 |
| SamYaple | a | 02:49 |
| SamYaple | a typo apparently | 02:49 |
| portdirect | Cheers for the fb regarding that container, I'd be open to making one for Jolla with the customisation it needs | 02:49 |
| portdirect | *Kolla | 02:49 |
| sdake | whats a fb :) | 02:49 |
| SamYaple | yea im interested in helping with that too | 02:49 |
| portdirect | Feedback :) | 02:50 |
| SamYaple | portdirect: im very interested in Jolla. tell me more | 02:50 |
| sdake | what is this jolla thing you speak of :) | 02:50 |
| portdirect | It's like kolla, but with added autocorrect | 02:50 |
| portdirect | Next level stuff | 02:51 |
| portdirect | It makes day 2 opps easy | 02:52 |
| Jeffrey4l | seems so. | 02:52 |
| SamYaple | day two oops maybe | 02:52 |
| portdirect | Day one is oops SamYaple | 02:53 |
| portdirect | Day 2, well let's just say we don't talk about day 2 | 02:53 |
| *** Jeffrey4l has quit IRC | 02:58 | |
| openstackgerrit | jimmygc proposed openstack/kolla-ansible master: Remove useless variable glance_backend_file https://review.openstack.org/457120 | 03:05 |
| *** iceyao has joined #openstack-kolla | 03:06 | |
| *** Jeffrey4l has joined #openstack-kolla | 03:07 | |
| openstackgerrit | Chen proposed openstack/kolla master: Take a special key customization into account Dockerfile.j2 https://review.openstack.org/457121 | 03:08 |
| *** caowei has joined #openstack-kolla | 03:14 | |
| *** jayahn has joined #openstack-kolla | 03:15 | |
| sdake | Jeffrey4l did you isntall libvirt on beast? | 03:25 |
| openstackgerrit | Merged openstack/kolla-ansible master: Remove neutron_metadata_socket volume from agents https://review.openstack.org/454167 | 03:27 |
| sdake | Jeffrey4l someone had installed libvirt | 03:38 |
| sdake | so compute kitwas ompliding | 03:38 |
| sdake | imploding | 03:38 |
| Jeffrey4l | ops.. | 03:39 |
| sdake | i have fixed that | 03:39 |
| sdake | the deploy should be up in a min | 03:39 |
| Jeffrey4l | got. | 03:39 |
| *** david-lyle has quit IRC | 03:42 | |
| sdake | Jeffrey4l ok - machine ready for an init-runonce as my uid | 03:44 |
| *** gkadam has joined #openstack-kolla | 03:44 | |
| sdake | https://docs.openstack.org/developer/kolla-kubernetes/deployment-guide.html | 03:44 |
| *** tonanhngo has joined #openstack-kolla | 03:45 | |
| *** tonanhngo has quit IRC | 03:46 | |
| Jeffrey4l | ok. let me check | 03:46 |
| Jeffrey4l | sdake, where is init_once file? | 03:50 |
| sdake | RTD | 03:50 |
| sdake | kolla-bringup | 03:50 |
| sdake | kolla-ansible/tools/init-runonce | 03:50 |
| *** tovin07 has quit IRC | 03:50 | |
| sdake | make sure to source the openrc file first | 03:51 |
| openstackgerrit | Surya Prakash (spsurya) proposed openstack/kolla-kubernetes master: [WIP] Kubeadm-aio container for kolla-k8s dev and gates https://review.openstack.org/457042 | 03:51 |
| *** zhurong has quit IRC | 03:59 | |
| sdake | Jeffrey4l anything? | 04:01 |
| sdake | Jeffrey4l i see the cloud has not been initialized | 04:02 |
| Jeffrey4l | no right now. | 04:02 |
| sdake | are you looking at logs ? | 04:02 |
| Jeffrey4l | i thought you will run "runonce". never mind, let me try | 04:03 |
| sdake | run as my UID | 04:03 |
| sdake | sudo -i | 04:03 |
| sdake | su - sdake | 04:03 |
| sdake | source *keystone* | 04:03 |
| sdake | cd kolla-bringup | 04:03 |
| Jeffrey4l | running | 04:03 |
| sdake | ok | 04:03 |
| sdake | good thing we talked about it :) | 04:04 |
| sdake | i thoguht it would help you to do it so you coudl run the proper command | 04:04 |
| *** lamt has joined #openstack-kolla | 04:05 | |
| sdake | ya error state | 04:05 |
| Jeffrey4l | yes. checking the logs | 04:07 |
| *** newmember has quit IRC | 04:08 | |
| *** newmember has joined #openstack-kolla | 04:08 | |
| sdake | Jeffrey4l note this is a easy way to access logs: [sdake@beast-01 ~(keystone_admin)]$ kubectl exec -it nova-api-3047009415-7n1w2 bash | 04:16 |
| Jeffrey4l | sdake, got this in placement api log: 2017-04-16 14:15:57.311263 Target WSGI script not found or unable to stat: /usr/bin/nova-placement-api | 04:16 |
| sdake | ok | 04:16 |
| Jeffrey4l | correct path should be: /var/lib/kolla/venv/bin/nova-placement-api | 04:17 |
| sdake | i know how to fix | 04:17 |
| sdake | moment let me redeploy | 04:17 |
| sdake | exit containers plz | 04:18 |
| Jeffrey4l | exited | 04:18 |
| *** zhurong has joined #openstack-kolla | 04:19 | |
| sdake | Jeffrey4l did you start libvirt earlier ? | 04:21 |
| Jeffrey4l | no. i change nothing in the machine. | 04:21 |
| sdake | k | 04:26 |
| sdake | thx | 04:26 |
| *** lamt has quit IRC | 04:27 | |
| *** lamt has joined #openstack-kolla | 04:39 | |
| openstackgerrit | Steven Dake proposed openstack/kolla-kubernetes master: Fix placement API fails to deploy https://review.openstack.org/457131 | 04:49 |
| sdake | Jeffrey4l canyou ack that plz | 04:49 |
| Jeffrey4l | sdake, LGTM. +2ed | 04:50 |
| sdake | [sdake@beast-01 kolla-bringup(keystone_admin)]$ openstack server list | 04:53 |
| sdake | +--------------------------------------+-------+--------+-------------------+------------+ | 04:53 |
| sdake | | ID | Name | Status | Networks | Image Name | | 04:53 |
| sdake | +--------------------------------------+-------+--------+-------------------+------------+ | 04:53 |
| sdake | | 0eb6ab8a-0b84-4ad2-9fe2-a2c4b32bf51d | demo1 | ACTIVE | demo-net=10.0.0.4 | cirros | | 04:53 |
| sdake | +--------------------------------------+-------+--------+-------------------+------------+ | 04:53 |
| *** lamt has quit IRC | 04:54 | |
| openstackgerrit | Steven Dake proposed openstack/kolla-kubernetes master: Nova-control service chart dependencies https://review.openstack.org/451495 | 04:55 |
| sdake | Jeffrey4l that patch needs a merge | 04:56 |
| sdake | Jeffrey4l before we can tag | 04:56 |
| sdake | Jeffrey4l i'll work on getting the cr team to review it tomorrow | 04:56 |
| sdake | and we can tag around this same bat time, same bat channel | 04:56 |
| Jeffrey4l | +2ed for this patch. | 04:58 |
| Jeffrey4l | sdake, could u review https://review.openstack.org/457117 ? | 04:58 |
| Jeffrey4l | and https://review.openstack.org/457059 | 04:59 |
| Jeffrey4l | sdake, ^^ | 04:59 |
| sdake | Jeffrey4l i thought after al lthat you would give me something hard :) | 04:59 |
| Jeffrey4l | it is very simple ;) | 04:59 |
| sdake | boy kolla-ansible and kolla deliverables sure are grand in scope | 04:59 |
| *** lamt has joined #openstack-kolla | 05:00 | |
| Jeffrey4l | after these two patch, we can tag m1. | 05:00 |
| openstackgerrit | Merged openstack/kolla-ansible master: Add barbican_crypto_password key in passwords.yml file https://review.openstack.org/457117 | 05:00 |
| openstackgerrit | Steven Dake proposed openstack/kolla-kubernetes master: Nova-control service chart dependencies https://review.openstack.org/451495 | 05:04 |
| *** caowei has quit IRC | 05:06 | |
| *** zhubingbing has quit IRC | 05:07 | |
| *** zhubingbing_ has joined #openstack-kolla | 05:08 | |
| *** iceyao has quit IRC | 05:09 | |
| *** gkadam_ has joined #openstack-kolla | 05:10 | |
| *** tervancovan has joined #openstack-kolla | 05:10 | |
| *** bmace has quit IRC | 05:12 | |
| *** bmace has joined #openstack-kolla | 05:12 | |
| *** gkadam has quit IRC | 05:13 | |
| *** tervancovan has quit IRC | 05:13 | |
| openstackgerrit | Steven Dake proposed openstack/kolla-kubernetes master: Placement API now integrated into service chart https://review.openstack.org/457048 | 05:20 |
| *** gkadam_ is now known as gkadam | 05:21 | |
| sdake | ok mission accomplished- placement api functional | 05:21 |
| sdake | sbezverk was good with a +2 on this Jeffrey4l - although I had to udpate teh commit mesage 5 times because - tired :) | 05:23 |
| sdake | https://review.openstack.org/#/c/456980/14 | 05:23 |
| sdake | those are all the remaining patches I think for 0.6.0 | 05:23 |
| *** zhubingbing_ has quit IRC | 05:27 | |
| *** zhubingbing_ has joined #openstack-kolla | 05:33 | |
| *** iceyao has joined #openstack-kolla | 05:35 | |
| *** lamt has quit IRC | 05:37 | |
| *** unicell has quit IRC | 05:38 | |
| *** unicell has joined #openstack-kolla | 05:38 | |
| *** iceyao has quit IRC | 05:39 | |
| *** StephenWang1991 has quit IRC | 05:41 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible stable/ocata: Add barbican_crypto_password key in passwords.yml file https://review.openstack.org/457141 | 05:43 |
| *** caowei has joined #openstack-kolla | 05:47 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla stable/ocata: Install panko in ceilometer base container https://review.openstack.org/444681 | 05:48 |
| openstackgerrit | Merged openstack/kolla master: Remove legacy resource types creation https://review.openstack.org/457059 | 05:48 |
| *** StephenWang1991 has joined #openstack-kolla | 05:50 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible stable/ocata: Detect openstack_release variable in mariadb recovery playbooks https://review.openstack.org/457145 | 05:52 |
| *** ssurana has joined #openstack-kolla | 06:06 | |
| openstackgerrit | Chen proposed openstack/kolla master: Take a special key customization into account on Dockerfile.j2 https://review.openstack.org/457121 | 06:08 |
| *** StephenWang1991 has quit IRC | 06:08 | |
| *** MarginHu has joined #openstack-kolla | 06:09 | |
| *** gkadam_ has joined #openstack-kolla | 06:10 | |
| *** gkadam_ is now known as msg | 06:11 | |
| *** gkadam has quit IRC | 06:11 | |
| *** msg is now known as gkadam | 06:11 | |
| *** iceyao has joined #openstack-kolla | 06:14 | |
| *** jbadiapa has joined #openstack-kolla | 06:16 | |
| *** iceyao has quit IRC | 06:18 | |
| *** jayahn has quit IRC | 06:26 | |
| *** magicboiz has joined #openstack-kolla | 06:28 | |
| *** magicboiz has quit IRC | 06:33 | |
| openstackgerrit | jianyi proposed openstack/kolla-ansible master: replace 'node_config_directory' with 'node_custom_config' for code consistency https://review.openstack.org/457156 | 06:39 |
| *** StephenWang1991 has joined #openstack-kolla | 06:41 | |
| *** magicboiz has joined #openstack-kolla | 06:45 | |
| *** StephenWang1991 has quit IRC | 06:46 | |
| *** StephenWang1991 has joined #openstack-kolla | 06:48 | |
| *** magicboiz has quit IRC | 06:49 | |
| *** magicboiz has joined #openstack-kolla | 06:50 | |
| *** iceyao has joined #openstack-kolla | 06:51 | |
| *** magicboiz has quit IRC | 06:55 | |
| *** zhubingbing_ has quit IRC | 06:57 | |
| *** zhubingbing has joined #openstack-kolla | 06:59 | |
| openstackgerrit | Surya Prakash (spsurya) proposed openstack/kolla-kubernetes master: [WIP] Kubeadm-aio container for kolla-k8s dev and gates https://review.openstack.org/457042 | 07:01 |
| *** zhubingbing has quit IRC | 07:04 | |
| *** manheim has joined #openstack-kolla | 07:06 | |
| *** magicboiz has joined #openstack-kolla | 07:07 | |
| openstackgerrit | Duong Ha-Quang proposed openstack/kolla-kubernetes master: Heat service deployment https://review.openstack.org/437252 | 07:09 |
| *** manheim has quit IRC | 07:10 | |
| *** zhubingbing has joined #openstack-kolla | 07:11 | |
| *** zhubingbing has quit IRC | 07:11 | |
| *** tervancovan has joined #openstack-kolla | 07:15 | |
| tervancovan | "--use-kubernetes-version" was changed in 1.6.1 kubeadm | 07:17 |
| tervancovan | so now I can not "vagrant up" in halcyon-vagrant-kubernetes | 07:19 |
| tervancovan | https://github.com/att-comdev/halcyon-kubernetes/issues/58 | 07:22 |
| tervancovan | https://github.com/att-comdev/halcyon-vagrant-kubernetes/issues/66 | 07:27 |
| tervancovan | please check second one | 07:28 |
| *** zhubingbing_ has joined #openstack-kolla | 07:30 | |
| *** mgoddard has joined #openstack-kolla | 07:36 | |
| openstackgerrit | Mark Goddard proposed openstack/kolla-ansible master: Use ironic-inspector user for ironic inspector https://review.openstack.org/445587 | 07:41 |
| openstackgerrit | shaofeng cheng proposed openstack/kolla master: Add oslo-vmware module https://review.openstack.org/457163 | 07:42 |
| openstackgerrit | Chen proposed openstack/kolla master: Take a special key customization into account on Dockerfile.j2 https://review.openstack.org/457121 | 07:46 |
| openstackgerrit | Mark Goddard proposed openstack/kolla master: Use ironic-inspector user for ironic-inspector https://review.openstack.org/441605 | 07:48 |
| *** newmember has quit IRC | 07:50 | |
| *** dixiaoli has joined #openstack-kolla | 07:54 | |
| *** dixiaoli has quit IRC | 07:55 | |
| *** dixiaoli has joined #openstack-kolla | 07:56 | |
| openstackgerrit | Mark Goddard proposed openstack/kolla-ansible master: Use ironic-inspector user for ironic inspector https://review.openstack.org/445587 | 08:00 |
| *** ssurana has quit IRC | 08:05 | |
| *** iceyao_ has joined #openstack-kolla | 08:06 | |
| *** tervancovan has quit IRC | 08:07 | |
| *** StephenWang1991 has quit IRC | 08:07 | |
| *** StephenWang1991 has joined #openstack-kolla | 08:08 | |
| *** iceyao has quit IRC | 08:09 | |
| *** mgoddard has quit IRC | 08:13 | |
| *** yangyape_ has joined #openstack-kolla | 08:18 | |
| *** yangyapeng has quit IRC | 08:18 | |
| *** StephenWang1991 has quit IRC | 08:22 | |
| *** StephenWang1991 has joined #openstack-kolla | 08:23 | |
| *** manheim has joined #openstack-kolla | 08:24 | |
| openstackgerrit | Duong Ha-Quang proposed openstack/kolla-kubernetes master: Heat service deployment https://review.openstack.org/437252 | 08:36 |
| *** iceyao_ has quit IRC | 08:45 | |
| *** egonzalez has joined #openstack-kolla | 08:47 | |
| egonzalez | morning | 08:48 |
| *** lamt has joined #openstack-kolla | 08:48 | |
| *** iceyao has joined #openstack-kolla | 08:49 | |
| openstackgerrit | shaofeng cheng proposed openstack/kolla-ansible master: Add VMware DataStore support to glance https://review.openstack.org/452176 | 08:49 |
| *** iceyao has quit IRC | 08:52 | |
| openstackgerrit | shaofeng cheng proposed openstack/kolla-ansible master: Add VMware DataStore support to glance https://review.openstack.org/452176 | 08:52 |
| *** iceyao has joined #openstack-kolla | 08:54 | |
| *** zhurong has quit IRC | 08:54 | |
| openstackgerrit | jeck proposed openstack/kolla-ansible master: Removing swift references from fstab https://review.openstack.org/456805 | 08:56 |
| duonghq | morning egonzalez | 08:57 |
| *** lamt has quit IRC | 08:59 | |
| *** mgoddard has joined #openstack-kolla | 09:02 | |
| *** StephenW_ has joined #openstack-kolla | 09:04 | |
| *** iceyao has quit IRC | 09:07 | |
| *** StephenWang1991 has quit IRC | 09:08 | |
| *** iceyao has joined #openstack-kolla | 09:11 | |
| openstackgerrit | Merged openstack/kolla-ansible stable/ocata: Detect openstack_release variable in mariadb recovery playbooks https://review.openstack.org/457145 | 09:14 |
| openstackgerrit | Merged openstack/kolla-ansible stable/ocata: Add barbican_crypto_password key in passwords.yml file https://review.openstack.org/457141 | 09:15 |
| spsurya | egonzalez: morning | 09:20 |
| *** mgoddard has quit IRC | 09:31 | |
| *** guoguo has quit IRC | 09:32 | |
| openstackgerrit | shaofeng cheng proposed openstack/kolla-ansible master: Add VMware DataStore support to glance https://review.openstack.org/452176 | 09:33 |
| openstackgerrit | Surya Prakash (spsurya) proposed openstack/kolla-kubernetes master: [WIP] Kubeadm-aio container for kolla-k8s dev and gates https://review.openstack.org/457042 | 09:41 |
| *** StephenWang1991 has joined #openstack-kolla | 09:41 | |
| *** StephenW_ has quit IRC | 09:44 | |
| *** dixiaoli has quit IRC | 10:00 | |
| *** tovin07_ has quit IRC | 10:05 | |
| *** MarginHu has quit IRC | 10:06 | |
| openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible master: Update simple_cell_setup to manual creation https://review.openstack.org/456920 | 10:11 |
| *** cuongnv has quit IRC | 10:12 | |
| openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible master: Update simple_cell_setup to manual creation https://review.openstack.org/456920 | 10:12 |
| *** mgoddard has joined #openstack-kolla | 10:16 | |
| *** dixiaoli has joined #openstack-kolla | 10:19 | |
| *** dixiaoli has quit IRC | 10:19 | |
| *** dixiaoli has joined #openstack-kolla | 10:20 | |
| *** dixiaoli has quit IRC | 10:20 | |
| *** dixiaoli has joined #openstack-kolla | 10:20 | |
| openstackgerrit | Duong Ha-Quang proposed openstack/kolla-kubernetes master: Heat service deployment https://review.openstack.org/437252 | 10:21 |
| *** manheim has quit IRC | 10:22 | |
| *** manheim has joined #openstack-kolla | 10:23 | |
| *** dixiaoli has quit IRC | 10:30 | |
| *** zhurong has joined #openstack-kolla | 10:52 | |
| *** Pavo has joined #openstack-kolla | 10:52 | |
| portdirect | mornin' 🐨's o/ | 10:53 |
| *** Pavo has quit IRC | 10:55 | |
| *** StephenWang1991 has quit IRC | 10:57 | |
| *** mgoddard has quit IRC | 10:57 | |
| *** mgoddard has joined #openstack-kolla | 11:00 | |
| *** caowei has quit IRC | 11:02 | |
| *** zhurong has quit IRC | 11:05 | |
| *** manheim has quit IRC | 11:07 | |
| openstackgerrit | Martin André proposed openstack/kolla master: Bypass docker client memoization in TasksTest https://review.openstack.org/457214 | 11:07 |
| *** manheim has joined #openstack-kolla | 11:08 | |
| openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible master: Support osprofile usage https://review.openstack.org/455628 | 11:10 |
| *** dave-mccowan has joined #openstack-kolla | 11:10 | |
| *** zhurong has joined #openstack-kolla | 11:12 | |
| openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible master: Support osprofile usage https://review.openstack.org/455628 | 11:12 |
| *** kolla-slack has quit IRC | 11:16 | |
| *** kolla-slack has joined #openstack-kolla | 11:19 | |
| sdake | morning | 11:19 |
| sdake | sup portdirect | 11:19 |
| sdake | sup duonghq | 11:20 |
| sdake | sup spsurya | 11:20 |
| sdake | sup egonzalez | 11:20 |
| sdake | sup mandre :) | 11:20 |
| egonzalez | sup sdake | 11:20 |
| spsurya | morning sdake | 11:20 |
| portdirect | Word up sdake, egonzalez and spsurya | 11:21 |
| *** Pavo has joined #openstack-kolla | 11:28 | |
| *** Pavo has quit IRC | 11:30 | |
| *** rwallner has joined #openstack-kolla | 11:30 | |
| *** rwallner has quit IRC | 11:30 | |
| *** rwallner has joined #openstack-kolla | 11:31 | |
| openstackgerrit | Merged openstack/kolla-ansible master: Fix ceph role set target_max_bytes for cache pools https://review.openstack.org/452444 | 11:37 |
| *** jbadiapa has quit IRC | 11:39 | |
| sdake | zhubingbing_ ping | 11:47 |
| zhubingbing_ | hear | 11:50 |
| *** jmontleo has quit IRC | 11:50 | |
| *** StephenWang1991 has joined #openstack-kolla | 11:51 | |
| *** jmontleon has joined #openstack-kolla | 11:52 | |
| sdake | zhubingbing_ did you start libvirt on my machine | 11:53 |
| zhubingbing_ | yeah | 11:54 |
| zhubingbing_ | but the libvirtd don't running | 11:55 |
| zhubingbing_ | i give up | 11:55 |
| zhubingbing_ | ;) | 11:55 |
| *** jbadiapa has joined #openstack-kolla | 11:55 | |
| zhubingbing_ | https://docs.openstack.org/developer/kolla-kubernetes/development-environment.html | 11:55 |
| *** mgoddard has quit IRC | 11:56 | |
| *** zhuzeyu has quit IRC | 11:57 | |
| *** egonzalez has quit IRC | 11:58 | |
| sdake | zhubingbing_ you see that big red box at the top | 12:06 |
| sdake | zhubingbing_ halcyon is broken | 12:06 |
| openstackgerrit | Bertrand Lallau proposed openstack/kolla-ansible master: Grafana: automatically set InfuxDB datasource https://review.openstack.org/457094 | 12:06 |
| zhubingbing_ | sdake i see | 12:07 |
| sdake | zhubingbing_ spsurya is working on a new dev env that will be in our repo and maintained | 12:07 |
| zhubingbing_ | sdake so, i give up to deploy halcyon | 12:07 |
| sdake | right | 12:07 |
| sdake | use the deployment guide | 12:07 |
| sdake | it does work | 12:07 |
| openstackgerrit | Bertrand Lallau proposed openstack/kolla-ansible master: Cleanup /run/netns host directory volume mount https://review.openstack.org/454179 | 12:07 |
| zhubingbing_ | but. the box have your enivirment | 12:08 |
| sdake | i am switcing to using vms | 12:08 |
| sdake | as soon as I can sort out how to do that | 12:08 |
| zhubingbing_ | I am afraid that will affect your environment | 12:08 |
| sdake | I dont really need baremetal | 12:08 |
| sdake | starting libvirt interacts with me using baremetal :) | 12:08 |
| sdake | so lets sort that problem out | 12:08 |
| zhubingbing_ | hmm yeah | 12:08 |
| sdake | I think those docs are pretty close - its the git clone halycon part that is busted | 12:08 |
| openstackgerrit | Bertrand Lallau proposed openstack/kolla-ansible master: Cleanup /lib/modules volume mounts in containers https://review.openstack.org/454214 | 12:09 |
| sdake | we have had nothing but trouble with halycon - for months on end | 12:09 |
| sdake | we gave it a go - too complicated | 12:09 |
| zhubingbing_ | ok | 12:10 |
| portdirect | zhubingbing_: if you just need a quick k8s env to get up and running this may be worth a look? https://github.com/openstack/openstack-helm/tree/master/tools/kubeadm-aio#kubeadm-aio-container | 12:12 |
| sdake | zhubingbing_ its not vagrant that is the problem - ithink people can live with that | 12:12 |
| sdake | portdirect he is using my gear, so that wont work sadly | 12:12 |
| kolla-slack | <portdirect> np - though it should work on pretty much anything | 12:13 |
| kolla-slack | <portdirect> I'm running on both real hardware and VM's across Ubuntu, CentOS and Fedora | 12:13 |
| kolla-slack | <portdirect> but appreciate you probably have some specific requirements | 12:14 |
| *** rwellum has joined #openstack-kolla | 12:15 | |
| sdake | portdirect fedora? :) | 12:16 |
| sdake | portdirect how do you launch the vm in which to run kubeadm-aio-container? | 12:16 |
| kolla-slack | <portdirect> always :slightly_smiling_face: I'm running OSH via that container on my laptop | 12:17 |
| kolla-slack | <portdirect> wither vagrant, but mostly just run on baremetal nodes | 12:17 |
| sdake | cool | 12:17 |
| sdake | i ran fedora for about 9 years | 12:17 |
| kolla-slack | <portdirect> I've been using AWS quite a bit as well recently :S | 12:17 |
| sdake | although centos is more my speed now | 12:17 |
| zhubingbing_ | hmm | 12:18 |
| rwellum | I've used both bare-metal and centos vm sdake for the AIO. For centos vm note that it's better to use minimal image and don't enable nesting, as that messes up docker. | 12:18 |
| rwellum | Morning btw. | 12:18 |
| zhubingbing_ | i will test it https://github.com/openstack/openstack-helm/tree/master/tools/kubeadm-aio#kubeadm-aio-container | 12:19 |
| zhubingbing_ | thanks sdake | 12:19 |
| portdirect | zhubingbing_: just make sure kubelet is not running on the host itself and you should be good to go | 12:20 |
| sdake | morning rwellum | 12:20 |
| *** zhurong has quit IRC | 12:24 | |
| openstackgerrit | Pradeep Kilambi proposed openstack/kolla master: Add ceilometer agent ipmi docker file https://review.openstack.org/454281 | 12:25 |
| sbezverk | portdirect: hey man, if you have a second could you check this and provide your thoughts https://review.openstack.org/456906 | 12:33 |
| *** lamt has joined #openstack-kolla | 12:36 | |
| *** lamt has quit IRC | 12:36 | |
| *** iceyao has quit IRC | 12:37 | |
| sbezverk | portdirect: specifically container lifecycle hook.. | 12:38 |
| *** iceyao has joined #openstack-kolla | 12:44 | |
| sdake | sbezverk quick Q for ya - entrypoint - see: https://review.openstack.org/#/c/451495/5/helm/service/nova-compute/values.yaml | 12:49 |
| sdake | sbezverk this results in compute never coming up on 2.0.0 images | 12:49 |
| sdake | sbezverk does entrypoint have any conditional logic at all? the docs are very poor | 12:50 |
| *** manheim has quit IRC | 12:50 | |
| sdake | sbezverk this is blocking the release of 0.6.0 - (which happens today) | 12:50 |
| *** manheim has joined #openstack-kolla | 12:50 | |
| sdake | sbezverk portdirect also need a review on https://review.openstack.org/#/c/456980/ - also blocking release | 12:51 |
| kolla-slack | <portdirect> Roger sdake / sbezverk, will look in 10 :) | 12:52 |
| *** StephenWang1991 has quit IRC | 12:58 | |
| *** srwilkers has joined #openstack-kolla | 12:59 | |
| srwilkers | morning | 12:59 |
| sdake | sup srwilkers - where ya been :) | 13:00 |
| sbezverk | srwilkers: morning | 13:00 |
| srwilkers | sdake, a little all over the place :) | 13:00 |
| srwilkers | have a good holiday weekend? | 13:00 |
| sdake | srwilkers survived another day - so guess so :) | 13:01 |
| *** zhubingbing_ has quit IRC | 13:01 | |
| *** mbruzek has joined #openstack-kolla | 13:01 | |
| *** jtriley has joined #openstack-kolla | 13:01 | |
| srwilkers | sdake, awesome dude. | 13:02 |
| sdake | sbezverk re https://review.openstack.org/#/c/456980/ it was green in patchset #14 | 13:04 |
| sdake | looks a little flakey | 13:04 |
| *** mbruzek has quit IRC | 13:05 | |
| sbezverk | sdake: maybe it is worth to look into it, no? not every ps should be just delete ;) | 13:05 |
| sdake | sbezverk ya - so we need to merge basically one patch before tagging | 13:05 |
| sdake | as master as is is broken | 13:05 |
| sdake | and that patch is : https://review.openstack.org/#/c/451495/ | 13:05 |
| *** mbruzek has joined #openstack-kolla | 13:06 | |
| *** mbruzek has joined #openstack-kolla | 13:06 | |
| sdake | sbezverk I need your help on this patch please - i can't do it alone - i understand what needs to be done | 13:06 |
| sdake | and why | 13:06 |
| sdake | just not how to do it | 13:06 |
| sdake | we are tagging today | 13:06 |
| sdake | so i'd appreciate any help you can provide on this patch | 13:07 |
| sdake | https://review.openstack.org/#/c/451495/5/helm/service/nova-compute/values.yaml - line 12 and 21 need to be conditional of some sort | 13:07 |
| sbezverk | sdake: this ps https://review.openstack.org/#/c/451495/ cannot be merged as it is failing on area which you touched.. today is really bad day for me in terms of time as we had lots of issue.. so the earliest I can spare is tomorrow.. | 13:08 |
| sdake | sbezverk can you just answer one question - how do you have conditionals in entrypoint | 13:08 |
| sdake | it is failing because placement is not enabled (2.0 gates) - yet compute must run after control for 4.0 | 13:09 |
| sbezverk | sdake: you list dependency in values.yaml fr the presense of service/socket etc | 13:10 |
| sbezverk | entrypoint will not start the service until these conditions are met | 13:10 |
| sbezverk | sdake: I think you are confusing the purpose of entrypoint and this specific case. entrypoint checks for KUBERNETES level dependency | 13:12 |
| sbezverk | here you have openstack versions issue | 13:12 |
| sdake | right | 13:12 |
| portdirect | :( | 13:13 |
| sdake | i am not confusing the purpose of entrypoint | 13:13 |
| sdake | i think entrypoint doesn't completely do the job | 13:13 |
| sdake | i am curious if it can | 13:13 |
| sbezverk | sdake: entrypoint is excellent tool when it is used properly | 13:14 |
| portdirect | sdake: agree with you there - inc0 probably has best grasp of the ability to extend entrypoint | 13:14 |
| portdirect | but for what it says on the tin, it does that no problem | 13:14 |
| sbezverk | sdake: this issue shold be handled not with entrypoint but with helm conditions and logic in the script | 13:14 |
| sdake | sbezverk cool - logic in which script? | 13:16 |
| sdake | we already have a helm condition | 13:16 |
| *** eanylin_ has quit IRC | 13:17 | |
| sbezverk | sdake: gate script, what else ?!? | 13:17 |
| sdake | sbezverk operators are not going to run the gate scripts dude ;) | 13:20 |
| portdirect | sbezverk: re https://review.openstack.org/#/c/456906 looks good | 13:21 |
| portdirect | though I've not tested it yet | 13:22 |
| portdirect | think we could drop some privs, and the cleanup could perhaps move to iproute2? | 13:22 |
| *** lamt has joined #openstack-kolla | 13:23 | |
| portdirect | do you have a copy of the systemd unit for keepalived? | 13:23 |
| * portdirect should not be lazy and look himself, but only on 2nd coffee of the day :) | 13:24 | |
| *** MarginHu has joined #openstack-kolla | 13:24 | |
| sbezverk | portdirect: sure sure it is still WIP, I tested all except removing ip address on the container delete | 13:25 |
| sbezverk | portdirect: thought you might have tried it before | 13:29 |
| portdirect | sbezverk: (un)fortunately I've never had to use it in a k8s context, but I'll try and have a play later today | 13:30 |
| *** al498u has joined #openstack-kolla | 13:31 | |
| openstackgerrit | Steven Dake proposed openstack/kolla-kubernetes master: Nova-control service chart dependencies https://review.openstack.org/451495 | 13:33 |
| *** al498u_ has joined #openstack-kolla | 13:34 | |
| sdake | Jeffrey4l just tag the release as is | 13:37 |
| sdake | Jeffrey4l its gtg | 13:37 |
| *** al498u has quit IRC | 13:37 | |
| mnaser | duonghq https://review.openstack.org/#/c/449952/ -- can you please give me more info about this? | 13:39 |
| mnaser | you mention a confirmation and left a -1.. its just a backport for a bug in vpnaas neutron | 13:39 |
| *** eanylin has joined #openstack-kolla | 13:41 | |
| *** lamt has quit IRC | 13:43 | |
| Jeffrey4l | sdake, cool. prepare the tag | 13:44 |
| *** dave-mccowan has quit IRC | 13:46 | |
| *** dave-mccowan has joined #openstack-kolla | 13:46 | |
| *** magicboiz has quit IRC | 13:47 | |
| *** jayahn has joined #openstack-kolla | 13:49 | |
| *** magicboiz has joined #openstack-kolla | 13:50 | |
| *** schwicht has joined #openstack-kolla | 13:52 | |
| *** magicboiz has quit IRC | 13:54 | |
| *** magicboiz has joined #openstack-kolla | 13:57 | |
| *** jtriley_ has joined #openstack-kolla | 14:03 | |
| Jeffrey4l | sdake, check https://review.openstack.org/457253 | 14:03 |
| *** jtriley has quit IRC | 14:05 | |
| *** Pavo has joined #openstack-kolla | 14:06 | |
| *** haplo37 has quit IRC | 14:14 | |
| *** goldyfruit has joined #openstack-kolla | 14:15 | |
| *** zhubingbing has joined #openstack-kolla | 14:16 | |
| *** goldyfruit has quit IRC | 14:17 | |
| *** goldyfruit has joined #openstack-kolla | 14:17 | |
| zhubingbing | hi guys | 14:19 |
| spsurya | zhubingbing: hi | 14:21 |
| *** StephenWang1991 has joined #openstack-kolla | 14:22 | |
| *** haplo37 has joined #openstack-kolla | 14:23 | |
| *** rwallner has quit IRC | 14:23 | |
| SamYaple | /win/win 27 | 14:24 |
| *** rwallner has joined #openstack-kolla | 14:24 | |
| *** rwallner has quit IRC | 14:26 | |
| *** StephenWang1991 has quit IRC | 14:26 | |
| *** rwallner has joined #openstack-kolla | 14:26 | |
| zhubingbing | hispsurya | 14:26 |
| *** inc0 has joined #openstack-kolla | 14:31 | |
| zhubingbing | sup egonzalez | 14:31 |
| inc0 | good morning | 14:32 |
| *** rwallner has quit IRC | 14:32 | |
| *** rwallner has joined #openstack-kolla | 14:32 | |
| openstackgerrit | Merged openstack/kolla-ansible master: Fix panko mongo idempotency https://review.openstack.org/455624 | 14:33 |
| sdake | if spock were here and I were there what would he do? | 14:34 |
| sdake | "Hed let you die" | 14:34 |
| sdake | morning inc0 | 14:34 |
| magicboiz | Hi everyone | 14:35 |
| magicboiz | I need a clue from the fluentd masters :) | 14:35 |
| magicboiz | I did upgrade from newton to ocata (heka -> fluentd), but kibana/elasticsearch aren't receiving anything? How could I debug this? | 14:36 |
| magicboiz | thanks :) | 14:36 |
| * inc0 not a fluentd master | 14:36 | |
| inc0 | but let's start from beginnging | 14:36 |
| inc0 | /var/lib/docker/volumes/kolla_logs is ok | 14:37 |
| inc0 | ? | 14:37 |
| *** lamt has joined #openstack-kolla | 14:37 | |
| inc0 | I mean sevices logs into logs? | 14:37 |
| inc0 | into files I mean | 14:37 |
| inc0 | (didn't have caffeine today yet, sorry) | 14:38 |
| magicboiz | jajaja | 14:38 |
| magicboiz | In which container do you suggest to start looking at? | 14:38 |
| inc0 | well, all of services will log there | 14:39 |
| inc0 | and this is host path | 14:39 |
| inc0 | inside container this will be in /var/log | 14:39 |
| magicboiz | ok, for example, ls -l /var/lib/docker/volumes/kolla_logs/_data/nova/ seems ok, with recent logs there | 14:40 |
| magicboiz | nova, trove, etc etc | 14:40 |
| inc0 | then check health of fluentd container | 14:41 |
| inc0 | should be running on every node | 14:41 |
| magicboiz | ok, are we looking for something like this??? error_class="Fluent::ElasticsearchOutput::ConnectionFailure" error="Can not reach Elasticsearch cluster | 14:42 |
| magicboiz | :) | 14:42 |
| magicboiz | thanks master | 14:42 |
| magicboiz | docker logs fluentd shows this error.... | 14:42 |
| inc0 | well then | 14:43 |
| inc0 | is elasticsearch fine? | 14:44 |
| inc0 | because it surely looks like issue;) | 14:44 |
| magicboiz | yeah, I think it's my fault. I changed elasticsearch nodes but didn't update haproxy with new config.... | 14:45 |
| magicboiz | thx inc0 | 14:45 |
| inc0 | np | 14:45 |
| inc0 | glad I could help | 14:45 |
| *** blallau has joined #openstack-kolla | 14:46 | |
| inc0 | ha, it's not even 8am and I already made my first good deed of a day! | 14:46 |
| *** jayahn has quit IRC | 14:50 | |
| SamYaple | inc0: in the real world its almost 11am | 14:50 |
| inc0 | well I live in "Dream of the 90s" | 14:50 |
| sdake | inc0 how about reviewing: https://review.openstack.org/#/c/451495/6 | 14:50 |
| inc0 | sdake, I already made my good deed today... | 14:50 |
| inc0 | :P | 14:51 |
| *** blallau has quit IRC | 14:51 | |
| sdake | inc0 and https://review.openstack.org/#/c/456980/ | 14:51 |
| inc0 | yeah I'm getting to reviews in a moment | 14:51 |
| inc0 | need caffeine | 14:51 |
| sdake | inc0 and https://review.openstack.org/#/c/457131/ | 14:52 |
| *** MarginHu has quit IRC | 15:00 | |
| sdake | zhubingbing please remove your -1 from : https://review.openstack.org/#/c/452857/ | 15:01 |
| zhubingbing | em | 15:01 |
| zhubingbing | ok | 15:01 |
| zhubingbing | done sdake | 15:02 |
| sdake | zhubingbing thanks | 15:02 |
| *** iceyao has quit IRC | 15:03 | |
| sdake | for the love of god, can I get a review on : https://review.openstack.org/#/c/457048/2/doc/source/deployment-guide.rst | 15:04 |
| sdake | and a second one | 15:04 |
| sdake | thanks! | 15:04 |
| zhubingbing | roger | 15:05 |
| *** iceyao has joined #openstack-kolla | 15:06 | |
| *** huzhengchuan has joined #openstack-kolla | 15:13 | |
| *** Reepicheep has joined #openstack-kolla | 15:15 | |
| *** mgoddard has joined #openstack-kolla | 15:17 | |
| *** manheim has quit IRC | 15:17 | |
| *** huzhengchuan has quit IRC | 15:21 | |
| openstackgerrit | Javier Castillo proposed openstack/kolla-ansible master: Fix incorrect fluend destination address https://review.openstack.org/457262 | 15:23 |
| *** Pavo has quit IRC | 15:27 | |
| *** athomas has joined #openstack-kolla | 15:28 | |
| *** huzhengchuan has joined #openstack-kolla | 15:33 | |
| sdake | inc0 sbezverk kfox1111 can you please review the reviews i pointed out above | 15:37 |
| sbezverk | sdake: done | 15:37 |
| sdake | thanks | 15:37 |
| sbezverk | well at least 2 | 15:37 |
| *** unicell has quit IRC | 15:37 | |
| openstackgerrit | Javier Castillo proposed openstack/kolla-ansible master: Fix incorrect fluend destination address https://review.openstack.org/457262 | 15:38 |
| *** unicell has joined #openstack-kolla | 15:38 | |
| openstackgerrit | Merged openstack/kolla stable/ocata: Install panko in ceilometer base container https://review.openstack.org/444681 | 15:40 |
| *** srwilkers has quit IRC | 15:44 | |
| openstackgerrit | Steven Dake proposed openstack/kolla-kubernetes master: Fix placement API fails to deploy https://review.openstack.org/457131 | 15:44 |
| sdake | sbezverk thanks ;) | 15:45 |
| *** srwilkers has joined #openstack-kolla | 15:45 | |
| *** gkadam has quit IRC | 15:46 | |
| *** rwallner has quit IRC | 15:53 | |
| *** rwallner has joined #openstack-kolla | 15:54 | |
| *** iceyao has quit IRC | 15:59 | |
| openstackgerrit | Steven Dake proposed openstack/kolla-kubernetes master: Placement API now integrated into service chart https://review.openstack.org/457048 | 15:59 |
| kfox1111 | sdake: ok, that comment of yours about simplified devtool environment thing made me grumpy... | 15:59 |
| sdake | which comment? | 16:00 |
| kfox1111 | we never, ever have had the dev env integrated with the gate properly. | 16:00 |
| sdake | right - i think we do | 16:00 |
| kfox1111 | we keep reinventing the gate, over and over in dev envs, from scratch. | 16:00 |
| sdake | maybe I mistated it | 16:00 |
| *** iceyao has joined #openstack-kolla | 16:00 | |
| kfox1111 | and then wonder why it bitrots. :/ | 16:00 |
| kfox1111 | https://review.openstack.org/#/c/457075/ | 16:00 |
| sdake | we dont need to reinvent the gate | 16:00 |
| sdake | we need to demystify the development experience | 16:00 |
| sdake | atm its seriously mystified | 16:00 |
| sdake | it took me 2+ months of constant banging to learn how things work | 16:01 |
| sdake | this is part of that vision :) | 16:01 |
| kfox1111 | I agree we need to demistify it. but a lot of it is trying to redo work over and over thats been solved for a long time. :/ | 16:01 |
| kfox1111 | we just need to make it more generic. :/ | 16:01 |
| kfox1111 | like: https://review.openstack.org/#/c/457042/6/devtools/kubeadm-aio-launcher.sh | 16:01 |
| kfox1111 | how many times are we going to write that script? | 16:02 |
| kfox1111 | I've seen it at least a dozen times now. | 16:02 |
| kfox1111 | lets focus as a team, on writing a single version, that works for everyone, thats tested. | 16:03 |
| kfox1111 | am I wrong here? | 16:05 |
| sdake | kfox1111 i understand your pov | 16:06 |
| sdake | you want a gated set of deployment scripts that exists only in one place | 16:07 |
| kfox1111 | yeah. and to stop having churn for folks coming and going rewriting from scratch the same code over and over. :/ | 16:07 |
| kfox1111 | we need a solid set of scripts and we're spending a lot of time rewriting rather then enhancing. | 16:08 |
| kfox1111 | the all in one script for example, hardocdes all the stuff all over again, amking it only appliciable to a small subset of folks. | 16:08 |
| kfox1111 | setup_kubernetes.sh is pretty generic. it just needs some way of making the networks configurable. like some config flags or env variables. | 16:09 |
| *** newmember has joined #openstack-kolla | 16:10 | |
| kfox1111 | POD_NETWORK=10.x.x/16 or something. | 16:10 |
| kfox1111 | https://review.openstack.org/#/c/457075/ feels like, yet another fork of the gate stuff. lets figure out how to share it intead of forking. we've proven to ourselves we can't maintain forks very well. | 16:12 |
| sdake | my position is the gate is specialized and very complex to understand - it is not obvious to people how it works - whereas 457075 is totally obvious | 16:14 |
| kfox1111 | I'm ok completely refactoring the gate if need be to chunk it up in a way thats more sharable. that lets us test most of it, and lesson our burdon of code maintenenc,e while stil growing folks konwlege of how the gate works. | 16:14 |
| sdake | so there are the two POVs, lets sort out how t o converge | 16:14 |
| kfox1111 | sdake: when 457075 gets the error handling it needs to function properly, it will end up looking basically like the gate. | 16:15 |
| sbezverk | kfox1111: morning, can you check keepalived soltuion? | 16:15 |
| kfox1111 | sbezverk: k. sec. | 16:15 |
| sdake | kfox1111 for development tools, I want humans to understand error handling | 16:15 |
| sdake | kfox1111 i understand your pov that humans are not scalable | 16:15 |
| kfox1111 | sdake: if you don't have some of the orchestration bits in there, like the wait for pods, its not going to work very well at all. | 16:16 |
| sdake | kfox1111 that workflow works for me perfectly | 16:16 |
| sdake | kfox1111 without waiting | 16:16 |
| sdake | and the waits - I am removing | 16:16 |
| sdake | it does not however record errors or check for errors or any of that | 16:17 |
| sdake | that is what the gate does differently | 16:17 |
| *** mgoddard has quit IRC | 16:18 | |
| kfox1111 | I still think 90+ of the gate code can be shared between the two use cases in a way that everyone should be happy with it. | 16:18 |
| sdake | that totally works for me | 16:18 |
| kfox1111 | 90+% | 16:18 |
| sdake | if the gate calls out to these operations | 16:19 |
| sdake | rather then having them built directly into the gate | 16:19 |
| kfox1111 | yes, pwe ull the bits of the gate that are generic, out to their own callable scripts and tweak the gate to call them. | 16:19 |
| sdake | sounds good - whos going to do the work :) | 16:19 |
| kfox1111 | thats why stuff is called "setup_kubernetes.sh", | 16:19 |
| kfox1111 | not "gatespecifick8ssetup" | 16:20 |
| kfox1111 | :/ | 16:20 |
| kfox1111 | I tried. and got largely ignored. :/ | 16:20 |
| sdake | i find much of the gate shell very difficult to flllow | 16:20 |
| kfox1111 | then lets fix it. | 16:20 |
| sdake | ok - well - like I said devenv take two | 16:20 |
| sdake | we have more people interested in solving this problem now | 16:20 |
| sdake | the fact that you were ignored was a statement of the small community size - community that wants to contribute is larger now | 16:20 |
| sbezverk | sdake: I do not know but the idea of try to script everything, based current activities, looks like it consumes more time than anybody would spend bringing up the environment. Instead of doing it once and start actual developming, you start yet another "let's script everything" activity. | 16:20 |
| kfox1111 | +1 for take 2. but the premice of the commit message is really salt in a wound. | 16:21 |
| sdake | kfox1111 sorry i didn't mean it to come off that way | 16:21 |
| sdake | kfox1111 i didn't perceive any wounds | 16:21 |
| sdake | except hte fat that hte devenv has woudned this project for 2+ months | 16:21 |
| sdake | tha has nothing to do with the gate | 16:21 |
| kfox1111 | yeah. its ok. its just there. no offence taken or anything. | 16:22 |
| kfox1111 | its a lot of hard faught/earned code that is getting ignored and feels like its getting poo'poo'ed. | 16:22 |
| kfox1111 | and then we try and move forward and we break the dev env's constantly. | 16:22 |
| sdake | kfox1111 I am pretty sure without the gate, I couldn't have ever understood how to deploy kolla-kubernetes or write a deployment document | 16:23 |
| sdake | kfox1111 i think one thing that will help it from being ignored is to place it in a separate directory from tools/ | 16:23 |
| kfox1111 | yeah. I really wanted to get it documented better, the whole minikube thing. but getting a repeatable testing framework took priroity. :/ | 16:23 |
| sdake | kfox1111 i hear ya - thats why I'm assisting here | 16:24 |
| kfox1111 | yeah. | 16:24 |
| kfox1111 | so, what shoudl it be called if not tools? | 16:24 |
| kfox1111 | I agree, devenv should be rooted somewhere easily found. but the shared bits between the gate/devenv should be tools or somehwere else? | 16:25 |
| sdake | tools might be right, and then gating code should go to /gate? | 16:25 |
| kfox1111 | works for me. | 16:25 |
| sdake | kfox1111 can I get a review on https://review.openstack.org/#/c/457131/ plz | 16:26 |
| sdake | kfox1111 just so you are aware, my frustration expressed with that commit message had little to do with the gate and has everything to do with the developer experience | 16:26 |
| kfox1111 | done. +2. | 16:27 |
| sdake | thx | 16:27 |
| kfox1111 | yeah. I get the frustration. :/ | 16:27 |
| sdake | kfox1111 https://review.openstack.org/#/c/456980/ | 16:28 |
| sdake | kfox1111 iI got a couple more too :) | 16:28 |
| sdake | kfox1111 https://review.openstack.org/#/c/457048/ | 16:28 |
| sdake | sbezverk - are you good with the discussion we just had - I'm not sure I can tackle it alone before summit - this would be a great opportunity to get rwallner and spsurya and zhubingbing up to speed on how gating works | 16:29 |
| kfox1111 | sdake: https://review.openstack.org/#/c/456980/14 <- could this be the start of the sharing we just talked about? | 16:29 |
| sdake | rwellum that is ) | 16:29 |
| kfox1111 | the gate has a few forks too. it would be nice to clean them all up and share that kind of thing. | 16:29 |
| sdake | kfox1111 i agree we need to clean that forking up | 16:30 |
| sdake | kfox1111 atm, master is broken - which is higher priorit yto me | 16:30 |
| openstackgerrit | Merged openstack/kolla-kubernetes master: Fix placement API fails to deploy https://review.openstack.org/457131 | 16:30 |
| kfox1111 | we can kill 2 birds with one stone. fix the gate forking and the devenv forking together. | 16:30 |
| *** Pavo has joined #openstack-kolla | 16:30 | |
| kfox1111 | sdake: k. | 16:30 |
| portdirect | o/ | 16:31 |
| sdake | https://review.openstack.org/#/c/452853/ | 16:31 |
| sdake | sup portdirect | 16:31 |
| *** iceyao has quit IRC | 16:31 | |
| *** jascott1 has joined #openstack-kolla | 16:31 | |
| *** unicell has quit IRC | 16:32 | |
| portdirect | just dropping my 2c on minikube, its awesome for many things, but I found it a bit hard recently | 16:33 |
| portdirect | its been tuned so far, that things like neutron are often pretty borked :( | 16:33 |
| kfox1111 | portdirect: yeah? whats the kind of issues that have been coming up? | 16:33 |
| portdirect | kfox1111: they *love* removing the ipset (and other) kernel modules | 16:34 |
| sdake | kfox1111 re https://review.openstack.org/#/c/451495/ - I have no idea what enabled: false means | 16:34 |
| kfox1111 | ah. | 16:34 |
| sdake | kfox1111 I can tell you it does work with a manual deployment of 4.0.0 | 16:34 |
| sdake | it may not be necessary at all or ignored - I can't find any docs that explain what the enabled: false thing does for entrypoint | 16:34 |
| kfox1111 | sdake: yeah, I think its actually dead code and just needs to be removed. | 16:34 |
| sdake | and reading the code turned up no answers | 16:35 |
| sdake | kfox1111 let me test it out -and i'll submit a new review | 16:35 |
| kfox1111 | portdirect: is ipset required in neutron now or is it still optional? | 16:35 |
| kfox1111 | sdake: its just a helm value.... | 16:35 |
| sdake | kfox1111 portdirect i really want whatever we gate with to develop with | 16:35 |
| sdake | kfox1111 so i am no treally pro-minikube | 16:36 |
| kfox1111 | an unused helm value I think. | 16:36 |
| sdake | kfox1111 k - let me remove and retry locally with 4.0.0 images | 16:36 |
| kfox1111 | sdake: I want to gat eas much as possible. I like minikube primarily for the reason I was able to get a physist to do it on his mac so easily. if our dev env is that light weight, we will have something going for us. | 16:36 |
| sdake | kfox1111 4 commands ;) | 16:37 |
| sdake | kfox1111 pretty lightweight - assuming you dont have to cusotmize for your personal network environment | 16:37 |
| kfox1111 | sdake: and the 4 commands work on a mac or a windows box? | 16:37 |
| sdake | linux only | 16:37 |
| portdirect | kfox1111: I think so ? but +2 to minikube if you can get it working | 16:37 |
| *** SamYaple is now known as puppet-master | 16:37 | |
| kfox1111 | thats not as easy. | 16:37 |
| *** puppet-master is now known as SamYaple | 16:38 | |
| sdake | there is always vagrant for the macwindows use case | 16:38 |
| sdake | most of our devs use linux or have linux :) | 16:38 |
| sdake | so lets focus on the 80% case;) | 16:38 |
| kfox1111 | yeah, but then there's instructions for that stuff too. | 16:38 |
| kfox1111 | so, not 4 commands. | 16:38 |
| sdake | i hear ya | 16:38 |
| kfox1111 | just trying to look at the big picture. :) | 16:38 |
| kfox1111 | user has a dev laptop. | 16:39 |
| sdake | what i'd like to do is stop wasting contributors time with stuff that is busted | 16:39 |
| kfox1111 | it happens to be whatever it is. odds are windows, but maybe mac. | 16:39 |
| kfox1111 | sdake: +1. | 16:39 |
| sdake | and help them ramp up their knowledge (like how does this thing work?) | 16:39 |
| kfox1111 | so, I'm afraid to say, the 80% is probably a windows box. | 16:40 |
| sdake | i didn't get a very good handle on kolla-kubernetes unitl the last coule of weeks | 16:40 |
| kfox1111 | yeah. the spinup time is very long right now. :/ | 16:40 |
| sdake | kfox1111 out of all the years developing kolla-ansible, only one dude has ever wanted a windows vm :) | 16:40 |
| sdake | kfox1111 and only one dude was ever a windows devleoper | 16:40 |
| openstackgerrit | Merged openstack/kolla-kubernetes master: Fix a bug related to genconfig and iscsi https://review.openstack.org/456980 | 16:40 |
| kfox1111 | sdake: not saying windwos vm's, windows dev laptop? | 16:41 |
| sdake | ya - virtualbox is a solution there - although everyone i know with a windows laptop has lab access to a linux machine | 16:41 |
| sdake | i run mac, but develop on linxu | 16:41 |
| sdake | mac doesn't have enough ram | 16:41 |
| sdake | (laptops) | 16:41 |
| kfox1111 | labs are hard to work on if your doing it in your spair time though. | 16:41 |
| sdake | and my poor laptop only has 256gb disk :( | 16:41 |
| kfox1111 | if it works on what you already got, then its more likely to spawn contributors. | 16:42 |
| portdirect | sdake - I'm running fine on 16gb 2011 Macbook pro? | 16:42 |
| sdake | portdirect mac 13" has 8gb (what I hvae) | 16:42 |
| kfox1111 | I think I did the full testign workflow in 6gb? | 16:42 |
| sdake | portdirect everyone has different gear | 16:42 |
| sdake | kfox1111 ya its possible | 16:42 |
| kfox1111 | that included 2 vm's. | 16:42 |
| sdake | we got poeple ready to go today with access to linux machines | 16:42 |
| sdake | lets get them ramped up | 16:42 |
| portdirect | why? we should standardise on my old laptop? that would be much easier for me | 16:42 |
| sdake | if windows devs come along ,we can sole that proble then :) | 16:43 |
| kfox1111 | I've got 16gb on mine too, and gave minikube 8g and things were happy. I think shaving it off to 6 gigs may be possible never tried it though. | 16:43 |
| kfox1111 | 4g was for sure too small though. | 16:43 |
| kfox1111 | sdake: its kind of related though, cause if you can assume a vm, the instructions can crap it up, and its not a problem. | 16:44 |
| kfox1111 | if its their main box though, we have to be much more careful. | 16:44 |
| kfox1111 | sbezverk: so, the helm ps. | 16:47 |
| kfox1111 | you asked for a review, but the ip setting seems to be done outside of keepalived? | 16:48 |
| kfox1111 | are you asking for help debuggin it, or think its ready? | 16:48 |
| *** harlowja has quit IRC | 16:49 | |
| openstackgerrit | Steven Dake proposed openstack/kolla-kubernetes master: Nova-control service chart dependencies https://review.openstack.org/451495 | 16:49 |
| *** zhubingbing has quit IRC | 16:50 | |
| sbezverk | kfox1111: it is all working | 16:52 |
| kfox1111 | sbezverk: is this line really needed: sudo ifconfig br-ex 172.18.0.254/24 ? | 16:52 |
| sbezverk | kfox1111: I would like your opinion on the solution | 16:52 |
| *** harlowja has joined #openstack-kolla | 16:52 | |
| kfox1111 | I could maybe see the ifcofnig br-ex up being needed. | 16:52 |
| sbezverk | kfox1111: think how we use keepalived normally | 16:53 |
| sbezverk | we have an interfaec with ip and subnet | 16:53 |
| sbezverk | then keepalived adds a virtual interface | 16:53 |
| sbezverk | sorry virtual IP | 16:53 |
| kfox1111 | yeah. but the machine will have two ips then. 172.18.0.254 and 172.18.0.1 | 16:53 |
| sbezverk | what I found that virtual IP has subnet mask of /32 | 16:53 |
| kfox1111 | yeah. | 16:54 |
| sbezverk | no .254 it "kind of physical addredd" | 16:54 |
| sbezverk | .1 is VIP | 16:54 |
| kfox1111 | 254 is a valid ip. | 16:54 |
| sbezverk | we need .254 with subnet mask | 16:54 |
| *** mgoddard has joined #openstack-kolla | 16:54 | |
| sbezverk | to make sure we have connectivity wehn you ping from 1 vm to another | 16:54 |
| kfox1111 | I'm wondering if we can just make the vip /24 instead of /32. that would fix it without the extra ip. | 16:54 |
| sbezverk | kfox1111: nope | 16:55 |
| sbezverk | VIP ios always /32 | 16:55 |
| sbezverk | it make sense you do not want your subnet starting to float around ;) | 16:55 |
| sbezverk | anyway in all other possibilities something gets broken, this one give 100% success | 16:56 |
| sbezverk | kfox1111: btw that set external job was doing exactly the same :) | 16:57 |
| kfox1111 | sbezverk: this looks promicing: | 16:57 |
| kfox1111 | http://paste.openstack.org/show/sbGnUzE6iKLtay6O9enO/ | 16:57 |
| kfox1111 | oh... | 16:58 |
| kfox1111 | no, I think you can just add a /24 on to the end of the vip: | 16:58 |
| kfox1111 | http://paste.openstack.org/show/jsoUYGsMxN2RM28dqiDM/ | 16:59 |
| *** krtaylor has joined #openstack-kolla | 16:59 | |
| *** jascott1 has quit IRC | 17:00 | |
| sbezverk | kfox1111: do you have a keepalived doc link? I was looking for options and could not find it.. | 17:00 |
| kfox1111 | sbezverk: http://manpages.ubuntu.com/manpages/precise/man5/keepalived.conf.5.html | 17:01 |
| sbezverk | kfox1111: hm figures, should have checked ubuntu docs | 17:03 |
| sbezverk | kfox1111: ok let me try it then.. | 17:03 |
| openstackgerrit | Merged openstack/kolla-kubernetes master: Placement API now integrated into service chart https://review.openstack.org/457048 | 17:04 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes master: WIP Converting keepalived into helm charts https://review.openstack.org/456906 | 17:07 |
| sbezverk | kfox1111: what about deleting ip in lifecycle? it should solve the issue you brought up the other day.. | 17:07 |
| kfox1111 | I think its probably a good enough solution for now. but should kill off the keepalived right after just to make sure it doesn't end up putting it back. | 17:08 |
| sdake | quick q | 17:09 |
| sbezverk | kfox1111: that might trigger container crash, not sure how kubernetes will react on it.. | 17:09 |
| sdake | i ahte to ask this now | 17:09 |
| sdake | but i will anyway | 17:09 |
| sdake | has anyone considered nginx | 17:09 |
| kfox1111 | sbezverk: should be ok. prestop can stop the pod and it will be fine. cause it enters 'terminating' state right before prestop. | 17:09 |
| kfox1111 | and if it exists during termination, it is happy. | 17:09 |
| sdake | i have heard it solves this use case (keepalived+haproxy) | 17:10 |
| kfox1111 | sdake: yeah, I use nginx. but your question is, for what? | 17:10 |
| sdake | although have no evidence to prove it :) | 17:10 |
| sdake | for vip serving | 17:10 |
| kfox1111 | oh. no, it wouldn't solve this use case. | 17:10 |
| sdake | could you expand the use case :) | 17:10 |
| kfox1111 | nginx doesn't provide vip keepalive. just serving on a vip. | 17:10 |
| sdake | so nginx is like haproxy | 17:10 |
| sdake | but not like keepalived? | 17:10 |
| kfox1111 | k8s needs a vip for svc external_vip running somewhere. | 17:11 |
| sdake | got it | 17:11 |
| sdake | that part | 17:11 |
| kfox1111 | yeah. nginx is half way between apache and haproxy. | 17:11 |
| sdake | so keepalived is the only VIP solution available - and nginx doesn't solve that the multinode VIP use case? | 17:11 |
| portdirect | kube proxy? | 17:12 |
| kfox1111 | its mostly just keepalived or pacemaker. | 17:12 |
| sdake | right | 17:13 |
| kfox1111 | and keepalived's much much simpler. :) | 17:13 |
| sdake | agreed | 17:13 |
| sbezverk | kfox1111: I built multinode with external ha proxy, works really well | 17:13 |
| kfox1111 | sbezverk: yeah. that works. | 17:13 |
| sdake | kfox1111 internal wouldn't work tho? | 17:13 |
| kfox1111 | sbezverk: but then your external haproxy is your ha single point of failure unless you make that ha, and then your back to keepalived/pacemaker. | 17:13 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/kolla master: Updated from global requirements https://review.openstack.org/455928 | 17:14 |
| sbezverk | kfox1111: then you need a pair of loadbalancers, that is who neworking gear manufacturers make money ;) | 17:14 |
| kfox1111 | sdake: with keepalived running in k8s, its fully ha. | 17:14 |
| kfox1111 | sbezverk: yeah. a lot of money. :) | 17:14 |
| kfox1111 | sbezverk: but keepalived running in k8s solves it too. | 17:15 |
| sbezverk | kfox1111: you want redundancy, please double your infra mister customer ;) | 17:15 |
| kfox1111 | so long as the vip runs on a node that's got kube-proxy running, your good. | 17:15 |
| kfox1111 | :) | 17:15 |
| *** unicell has joined #openstack-kolla | 17:15 | |
| *** unicell has quit IRC | 17:17 | |
| sdake | so to summarize, haproxy or nginx could work - but would not be redundant - hence the only viable software solution is keepalived - and the only viable hardware solution is a pair of redundant load balancers? | 17:17 |
| kfox1111 | sdake: traffic comes in http -> vip (managed by keepalived) -> iptables (managed by kube-proxy) -> randomly selected openstack-api pod. | 17:18 |
| *** unicell has joined #openstack-kolla | 17:18 | |
| sdake | i notice kargo deploys nginx - which makes me wonder why they do so - trying to understand if that provides the whole solution - it sounds like from this exchange it lacks redundancy | 17:19 |
| kfox1111 | haproxy/nginx are l7 loadbalancing. its l4 traffic all the way to the openstack pod. | 17:19 |
| kfox1111 | theres another way to do it with k8s. called ingress. | 17:19 |
| kfox1111 | the most common ingress implementation is nginx based. | 17:20 |
| kfox1111 | which looks like: | 17:20 |
| kfox1111 | http -> vip (managed by keepalived) -> iptables (managed by kube-proxy -> nginx-ingress -> openstack-api pod | 17:20 |
| kfox1111 | for ha configs. | 17:21 |
| kfox1111 | or, | 17:21 |
| kfox1111 | http -> vips (managed by keepalived) -> nginx-ingress net=host -> openstack-api pod | 17:21 |
| kfox1111 | ingress is still kind of baking. though much more stable then it was. I'm starting to use it in a few places. | 17:21 |
| *** tonanhngo has joined #openstack-kolla | 17:22 | |
| *** StephenWang1991 has joined #openstack-kolla | 17:25 | |
| *** mgoddard has quit IRC | 17:26 | |
| *** tonanhngo has quit IRC | 17:26 | |
| *** tonanhngo has joined #openstack-kolla | 17:27 | |
| *** StephenWang1991 has quit IRC | 17:29 | |
| openstackgerrit | Merged openstack/kolla-kubernetes master: Remove support for py34 https://review.openstack.org/440895 | 17:30 |
| *** unicell has quit IRC | 17:32 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes master: Test Ceph OSD Chart https://review.openstack.org/457302 | 17:34 |
| sbezverk | kfox1111: it looks like even if I use /24 VIP without actual ip on the interface, connectivity does not work. the ping test between two vms getting stuck forever.. | 17:35 |
| kfox1111 | hmm... | 17:35 |
| kfox1111 | i'll have a look. | 17:36 |
| *** rwallner has quit IRC | 17:37 | |
| *** zhugaoxiao has quit IRC | 17:38 | |
| *** zhugaoxiao has joined #openstack-kolla | 17:38 | |
| *** rwallner has joined #openstack-kolla | 17:39 | |
| openstackgerrit | Justin Scott proposed openstack/kolla-kubernetes master: WIP Playbook implements UserGuide https://review.openstack.org/457317 | 17:41 |
| *** jascott1 has joined #openstack-kolla | 17:41 | |
| kfox1111 | which one did you change? I see sudo ifconfig br-ex 172.18.0.254/24 in all of them? | 17:41 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes master: WIP Converting keepalived into helm charts https://review.openstack.org/456906 | 17:41 |
| kfox1111 | oh. its commented in the iscsi one... | 17:42 |
| kfox1111 | yeah. that. :) | 17:42 |
| *** lamt has quit IRC | 17:43 | |
| *** lamt has joined #openstack-kolla | 17:44 | |
| *** schwicht has quit IRC | 17:44 | |
| *** srwilkers has quit IRC | 17:45 | |
| *** athomas has quit IRC | 18:02 | |
| *** unicell has joined #openstack-kolla | 18:02 | |
| *** magicboiz has quit IRC | 18:02 | |
| *** unicell has quit IRC | 18:03 | |
| *** unicell has joined #openstack-kolla | 18:04 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes master: Converting keepalived into helm charts https://review.openstack.org/456906 | 18:05 |
| *** srwilkers has joined #openstack-kolla | 18:06 | |
| sbezverk | kfox1111: it worked :), glad you can always add something to improve.. https://review.openstack.org/456906 | 18:06 |
| kfox1111 | cool. thanks. :) | 18:07 |
| openstackgerrit | Javier Castillo proposed openstack/kolla-ansible master: Fix incorrect fluend destination address https://review.openstack.org/457262 | 18:07 |
| kfox1111 | so, maybe we should make it a variable and default it to 32 still. | 18:07 |
| kfox1111 | I'd be curious if the ifconfig up is still needed too? maye by getting the subnet in there fixes the issue entirely? | 18:09 |
| *** krtaylor has quit IRC | 18:09 | |
| kfox1111 | hmm.. openvswitch-set-external-ip does do an ip link set {{ neutron_bridge_name }} up; | 18:10 |
| kfox1111 | maybe we should add that to the keepalived chart? | 18:10 |
| *** srwilkers has quit IRC | 18:11 | |
| sbezverk | kfox1111: this step is ver specific to the gate | 18:13 |
| sbezverk | normally you would have your interface up and with ip address | 18:13 |
| sbezverk | kfox1111: the only reason we do it is because openvswitch does not bring up its bridges by default | 18:14 |
| *** newmember has quit IRC | 18:14 | |
| *** newmember has joined #openstack-kolla | 18:15 | |
| *** MasterOfBugs has joined #openstack-kolla | 18:15 | |
| kfox1111 | sbezverk: yeah, but we can make it conditional. helm values. | 18:16 |
| sbezverk | kfox1111: what is the value? nobody would use this config outside of the gate? | 18:16 |
| kfox1111 | folks doing devenvs will run into the same issue. | 18:16 |
| *** david-lyle has joined #openstack-kolla | 18:16 | |
| sbezverk | kfox1111: and the gate script hints them what needs to be done | 18:17 |
| *** manheim has joined #openstack-kolla | 18:17 | |
| sbezverk | also it becomes kind openvswitch specific | 18:17 |
| kfox1111 | if its conditional though, is it a problem? | 18:18 |
| sbezverk | example I plan to switch to networking-vpp soon | 18:18 |
| kfox1111 | maybe we move the ifconfig up to the openvswitch module? | 18:18 |
| sbezverk | kfox1111: can you guarantee brining up br-ex as an interface does not impact anything else? | 18:19 |
| kfox1111 | we already have add-br and add-port in openvswitch-vswitchd. maybe we just make it go in there? | 18:19 |
| kfox1111 | $localVals.interface_up or something? | 18:19 |
| sbezverk | kfox1111: maybe as a part of separate ps.. it has nothing to do with keepalived chart.. | 18:20 |
| kfox1111 | well, tangentially, cause it doesn't work without the sudo ifconfig. | 18:20 |
| kfox1111 | which feels a bit dirty to me. | 18:20 |
| kfox1111 | 1. start up containerized openvswithch. 2. poke stuff at it, 3, run other container on it. | 18:21 |
| sbezverk | kfox1111: well, we have lots of shortcuts, example playing with dns | 18:21 |
| kfox1111 | ok. lets merge it, then I'll do a ps to clean it up. | 18:21 |
| sbezverk | why do not you do it in the chart ;) ?? | 18:21 |
| kfox1111 | yeah, thats what I'm saying. do it in the chart. :) | 18:22 |
| *** manheim has quit IRC | 18:22 | |
| sbezverk | kfox1111: would you really like as an operator the idea that somehitng playing with your interfaces in chart? (not taking about gate) | 18:23 |
| sbezverk | I am not real operator and I do not like that idea | 18:24 |
| kfox1111 | on the fence on that one. we've got a lot of instances of that already. | 18:24 |
| kfox1111 | its wrapped up in a single setup_bridge=false/true value today. | 18:24 |
| sbezverk | kfox1111: but they are all gate specific. I think we are fairly clean for "normal" deployment scenario | 18:24 |
| kfox1111 | I think the use case is more general then you think? | 18:25 |
| sbezverk | kfox1111: never heard anybody using br-ex as a real interface.. curious what would neutron folks say about it | 18:25 |
| kfox1111 | I do on one cloud. | 18:26 |
| sbezverk | kfox1111: ok, if you want to push additional ps to do it, ok, but I do not think it is right thing and especially in keepalived chart | 18:27 |
| sbezverk | if it goes in then it should be openvswitch-vswitchd chart | 18:27 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes master: Add support for external interface bringup https://review.openstack.org/457326 | 18:29 |
| kfox1111 | if that looks good, can you rebase the keepalived one on it and drop the ifconfig up's? they should then be unnessisary. Then I think we're good to merge? | 18:30 |
| *** lpetrut has joined #openstack-kolla | 18:34 | |
| sbezverk | kfox1111: sorry I do not have time for this lots of emergencies I need to address. Please merge keepalive as is and then you can apply your on top.. | 18:36 |
| kfox1111 | in the same boat here. :/ | 18:36 |
| kfox1111 | k... this gets us at least farther away from kolla-kube... | 18:37 |
| *** srwilkers has joined #openstack-kolla | 18:46 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes master: WIP: Test out flannel helm package. https://review.openstack.org/456964 | 18:46 |
| *** krtaylor has joined #openstack-kolla | 18:48 | |
| *** lamt has quit IRC | 18:50 | |
| *** lamt has joined #openstack-kolla | 18:52 | |
| *** lpetrut has quit IRC | 18:57 | |
| *** rwallner has quit IRC | 19:01 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes master: WIP: Test out flannel helm package. https://review.openstack.org/456964 | 19:06 |
| kfox1111 | sbezverk: ping | 19:14 |
| kfox1111 | I forgot, the hardcoded /24 is still in the keepalived. we should probably change that to a subnet value before merging. do you have time to do that, or should I add it in quick? | 19:14 |
| *** mgoddard has joined #openstack-kolla | 19:15 | |
| *** schwicht has joined #openstack-kolla | 19:26 | |
| *** lamt has quit IRC | 19:29 | |
| *** pbourke has quit IRC | 19:35 | |
| *** pbourke has joined #openstack-kolla | 19:36 | |
| *** Pavo has quit IRC | 19:43 | |
| *** lamt has joined #openstack-kolla | 19:56 | |
| *** gfhellma has joined #openstack-kolla | 19:58 | |
| *** mgoddard has quit IRC | 19:59 | |
| *** jtriley_ has quit IRC | 20:05 | |
| *** lamt has quit IRC | 20:06 | |
| *** david-lyle has quit IRC | 20:08 | |
| sbezverk | kfox1111: vip comes from kolla config as it is used in keepalived config map | 20:13 |
| kfox1111 | sbezverk: yeah, but the /24 is hardcoded now. | 20:13 |
| sbezverk | keepalived config map template is still rendered by kolla kube | 20:13 |
| kfox1111 | we should make it a /{{ $subnet }} or something | 20:14 |
| kfox1111 | oh. | 20:14 |
| sbezverk | I mean I can add jinja2 var | 20:14 |
| kfox1111 | yeah. that would work. | 20:14 |
| sbezverk | but I would rather address it properly once all configmaps are moved to helm | 20:14 |
| kfox1111 | keepalived's not really useful though until it can be set to 32/24 appritely for the situation though. | 20:15 |
| sbezverk | kfox1111: if you feet too strngly it has to be done now with jinja I will do it | 20:15 |
| sbezverk | later today | 20:15 |
| kfox1111 | yeah, I think its probably a good idea to make it settable. I know others will need that to be able to use it at there sites. | 20:16 |
| kfox1111 | https://review.openstack.org/#/c/457326/ shoudl be ready for review. | 20:17 |
| kfox1111 | maybe I shoudl rename it to ext_interface_up though? | 20:18 |
| sbezverk | kfox1111: ok I will do then two vars one in jinja2 for now and one currently not used in helm, once configmap migration is completed, I will switch on using helm var.. | 20:18 |
| kfox1111 | ok. sounds good. | 20:18 |
| kfox1111 | thanks. | 20:18 |
| sbezverk | kfox1111: yep, you should rename it, it sounds ver confusing | 20:19 |
| kfox1111 | k. | 20:19 |
| kfox1111 | ext_bridge_interface_up ? | 20:19 |
| kfox1111 | or ext_bridge_up ? | 20:19 |
| kfox1111 | the former is slightly more accurite. the latter is probably good enough, and shorter. | 20:21 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes master: Add support for external interface bringup https://review.openstack.org/457326 | 20:22 |
| *** esharao has joined #openstack-kolla | 20:25 | |
| *** esharao has quit IRC | 20:26 | |
| *** esharao has joined #openstack-kolla | 20:26 | |
| *** StephenWang1991 has joined #openstack-kolla | 20:27 | |
| *** japestinho has quit IRC | 20:30 | |
| *** StephenWang1991 has quit IRC | 20:31 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes master: Converting keepalived into helm charts https://review.openstack.org/456906 | 20:39 |
| sbezverk | kfox1111: yep, looks good | 20:40 |
| sbezverk | kfox1111: do we want to have iscsi storage backend in 1.0? | 20:41 |
| *** manheim has joined #openstack-kolla | 20:52 | |
| kfox1111 | yeah. some folks will want it, and its been solid for a long time. no reason to remove it. | 20:55 |
| *** manheim has quit IRC | 20:57 | |
| sbezverk | kfox1111:I meant add iscsi storage backend to kube | 20:58 |
| kfox1111 | oh. | 20:59 |
| kfox1111 | is there such a beast? | 20:59 |
| kfox1111 | you talking a pvc auto provisioner? | 20:59 |
| sbezverk | right now we have ceph and file, I just built 5 node kube cluster using iscsi pv/pvc | 20:59 |
| sbezverk | and storage box shares volumes via iscsi interface | 21:00 |
| kfox1111 | did you manually create the pv's in the backend storage? tgtd based? | 21:00 |
| sbezverk | kfox1111: manually, right now autocreation is not supported for iscsi afaik | 21:00 |
| kfox1111 | right. cuase I think to actually do so, they would have to reimplment all of cinder. | 21:01 |
| sbezverk | kfox1111: not sure these are two separate things | 21:01 |
| sbezverk | the storage I mentioned does not have to be used by openstack | 21:02 |
| kfox1111 | cinder is an api to allow users to request backend volumes to be previsioned in a vendor neutral way via storage provisioner plugins. | 21:02 |
| sbezverk | basically any container can claim iscsi pvc/pv | 21:02 |
| kfox1111 | right. | 21:02 |
| sbezverk | kfox1111: I see what you mean | 21:02 |
| kfox1111 | if the admin preprevisions the pv's, then no need for something like cinder. | 21:02 |
| kfox1111 | but for autoprovisioning, its basically just what cinder does. | 21:03 |
| sbezverk | make "cinder" as a part of kube cluster for autoprovisioning | 21:03 |
| kfox1111 | yeah. | 21:03 |
| kfox1111 | but hit a circular dep right around, how do you store cinder's state. :/ | 21:03 |
| *** inc0 has quit IRC | 21:03 | |
| kfox1111 | if cinder had an etcd db backend... :) | 21:03 |
| sbezverk | kfox1111: well, even canal has and uses its own etcd, why cinder cannot do the same ;) | 21:04 |
| kfox1111 | indeed. :) | 21:05 |
| kfox1111 | I've been using etcd'less canal for a little while though. seems to work ok. :) | 21:05 |
| kfox1111 | same with flannel. | 21:05 |
| sbezverk | kfox1111: what do you use then? main etcd? | 21:05 |
| kfox1111 | it ends up using k8s api as etcd, passing through the data. so it doesn't need one of its own. | 21:05 |
| *** matrohon has joined #openstack-kolla | 21:05 | |
| kfox1111 | see: https://github.com/projectcalico/canal/blob/master/k8s-install/canal-etcdless.yaml | 21:06 |
| *** krtaylor has quit IRC | 21:08 | |
| *** inc0 has joined #openstack-kolla | 21:11 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes master: WIP: Test out flannel helm package. https://review.openstack.org/456964 | 21:11 |
| sbezverk | kfox1111: I see, make sense, since new version of etcd is powerfull enough, no reason to run multiple instances | 21:12 |
| kfox1111 | yeah. less to manage, secure, etc. | 21:13 |
| kfox1111 | if I can get the above ps working, then we can get the feature added to helm, and we can build apon that flannel chart to do canal too. | 21:14 |
| sbezverk | kfox1111: confused a bit, to deploy helm chart we need kubernetes cluster up, but for it to be up we need networking.. am I missing something? | 21:19 |
| kfox1111 | I've gotten tiller to work as a static pod on one of my systems. | 21:23 |
| kfox1111 | hostNetwork: True | 21:23 |
| kfox1111 | I'm working with jascott1 to get that to be a supported option to 'helm init'. | 21:24 |
| kfox1111 | in that case you can kubeadm init; helm init --net-host, and helm install canal --set .... | 21:24 |
| kfox1111 | rather then sed'ing yaml to configure it. :) | 21:24 |
| *** srwilkers has quit IRC | 21:26 | |
| *** inc0 has quit IRC | 21:27 | |
| sbezverk | kfox1111: hm, interesting.. | 21:30 |
| kfox1111 | should be able to upgrade your sdn easily then too. :) | 21:31 |
| kfox1111 | I've gotten kube-dashboard working as a helm chart. | 21:31 |
| kfox1111 | and would like to get kube-proxy/kube-dns too. | 21:31 |
| sbezverk | kfox1111: man, soon Bill Gates will send you an invoice for all these requests you feed to helm/msft people ;) | 21:34 |
| kfox1111 | he should thank me for sending people his way. ;) | 21:34 |
| sbezverk | kfox1111: :) dream on | 21:35 |
| kfox1111 | was worth a shot. ;) | 21:35 |
| kfox1111 | nice.. that got tiller to work... ok. now to fix flannel. | 21:37 |
| *** matrohon has quit IRC | 21:38 | |
| *** inc0 has joined #openstack-kolla | 21:38 | |
| *** inc0 has quit IRC | 21:39 | |
| *** Pavo has joined #openstack-kolla | 21:40 | |
| *** Pavo has quit IRC | 21:40 | |
| *** dave-mcc_ has joined #openstack-kolla | 21:43 | |
| *** dave-mccowan has quit IRC | 21:44 | |
| *** manheim has joined #openstack-kolla | 21:48 | |
| *** dave-mccowan has joined #openstack-kolla | 21:48 | |
| *** dave-mcc_ has quit IRC | 21:50 | |
| *** lamt has joined #openstack-kolla | 21:50 | |
| *** manheim has quit IRC | 21:52 | |
| *** Pavo has joined #openstack-kolla | 21:57 | |
| *** manheim has joined #openstack-kolla | 21:59 | |
| *** Pavo has quit IRC | 22:01 | |
| *** jascott1 has quit IRC | 22:01 | |
| *** inc0 has joined #openstack-kolla | 22:01 | |
| *** jascott1 has joined #openstack-kolla | 22:02 | |
| *** inc0 has quit IRC | 22:02 | |
| *** jascott1 has quit IRC | 22:04 | |
| *** jascott1 has joined #openstack-kolla | 22:04 | |
| *** esharao has quit IRC | 22:06 | |
| *** jascott1 has quit IRC | 22:06 | |
| *** jascott1 has joined #openstack-kolla | 22:06 | |
| *** jascott1 has quit IRC | 22:10 | |
| *** rwellum has quit IRC | 22:11 | |
| *** mbruzek has quit IRC | 22:25 | |
| *** jascott1 has joined #openstack-kolla | 22:27 | |
| *** krtaylor has joined #openstack-kolla | 22:35 | |
| *** lamt has quit IRC | 22:37 | |
| sdake | kfox1111 sbezverk can you guys review https://review.openstack.org/#/c/451495/ plz | 22:38 |
| *** lamt has joined #openstack-kolla | 22:39 | |
| *** MarginHu has joined #openstack-kolla | 22:41 | |
| *** MarginHu has quit IRC | 22:41 | |
| *** MarginHu has joined #openstack-kolla | 22:42 | |
| *** lamt has quit IRC | 22:43 | |
| sbezverk | sdake: ideally to ack this PS we would need to have compute kit with 4.0.0 and 2.0.0 running | 22:44 |
| sdake | sbezverk it runs locally | 22:45 |
| sdake | if your asking me to implement the gate, its on my todo | 22:45 |
| sdake | computekit runs with 2.0.0 | 22:45 |
| sbezverk | kfox1111: any chance you can complete 4.0.0 gate images? | 22:45 |
| kfox1111 | gate jobs? | 22:48 |
| sbezverk | kfox1111: I think I saw you PS and 4.0.0 compute kit was working | 22:48 |
| sbezverk | or I am missing something? | 22:48 |
| kfox1111 | cinder is still broken. :/ | 22:48 |
| sdake | i'd rather not block that patch on the gate as it works locally and is tested | 22:49 |
| openstackgerrit | Marcus Williams proposed openstack/kolla master: Add OpenDaylight Container to kolla https://review.openstack.org/416369 | 22:49 |
| sdake | i am happy to take on kevin's gate patch when I have time | 22:49 |
| sdake | we need 4.0.0 gates operational prior to summit | 22:49 |
| sbezverk | kfox1111: ah right ceph thing :( | 22:49 |
| kfox1111 | https://review.openstack.org/#/c/451495/7/helm/service/nova-control/requirements.yaml ? | 22:50 |
| kfox1111 | why reorder? | 22:50 |
| sbezverk | kfox1111: :) had the same question in my mind | 22:50 |
| sdake | kfox1111 that is how the aptch was submitted when i otok it over | 22:50 |
| sbezverk | sdake: what did you change? | 22:50 |
| sdake | i dont particularly think it matters however can test without a reorder | 22:50 |
| sdake | i took out cell0 | 22:51 |
| sdake | which needs to be a separate service chart | 22:51 |
| *** sayantan_ has joined #openstack-kolla | 22:51 | |
| kfox1111 | cell0 could be conditional. | 22:51 |
| sdake | sbezverk to be more precise, here is my lsit of changes: https://review.openstack.org/#/c/451495/4..7 | 22:52 |
| sdake | kfox1111 i think it can't be because of the requirement for entrypoint | 22:52 |
| openstackgerrit | Marcus Williams proposed openstack/kolla master: Add OpenDaylight Container to kolla https://review.openstack.org/416369 | 22:53 |
| sdake | kfox1111 cell0 needs to register after computekit launches | 22:53 |
| sbezverk | sdake: it has to be conditional | 22:53 |
| sbezverk | as placement and cell are not supported with 2.0.0 images | 22:53 |
| sdake | sbezverk agreed it needs to be conditional | 22:54 |
| sbezverk | these two things only applicable to 4.0.0 images | 22:54 |
| sdake | sbezverk do you recall this review: | 22:54 |
| *** jrobinson has joined #openstack-kolla | 22:54 | |
| kfox1111 | sdake: I think there's a better way. | 22:54 |
| kfox1111 | we have a job to register compute nodes with the cells. | 22:55 |
| sdake | better way for which | 22:55 |
| kfox1111 | we just need to ensure that runs after cell0 and computes. | 22:55 |
| sdake | cells need to run after comput eis running | 22:55 |
| kfox1111 | no. | 22:55 |
| kfox1111 | only in order to skip rerunning the compute registration job. | 22:55 |
| kfox1111 | we've got one of those now, so we could change it to just launch the cell api in paralel, then add all the nodes at the hend. | 22:56 |
| kfox1111 | end | 22:56 |
| kfox1111 | nova-api-cell-discover-host-job | 22:56 |
| kfox1111 | so long as that runs to completion after cell's are ready and computes are ready, and before nova commands, we're good. | 22:57 |
| sdake | i'm all for not having a cell service chart | 22:58 |
| sbezverk | sdake: we cannot not have cell chart it is mandatory | 22:58 |
| sdake | ok - before we get into the cell discussion | 22:59 |
| sdake | which I want to have | 22:59 |
| kfox1111 | sbezverk: I think we're talking about making the cell stuff in the nova-controller chart, rather then a sepreate one. | 22:59 |
| sdake | can both of you please ack: https://review.openstack.org/#/c/451495/ | 22:59 |
| sdake | placement api is orthognoal to cells api | 23:00 |
| sdake | or cell db creation | 23:01 |
| sdake | or whtaever its called | 23:01 |
| sdake | either ack it or -1 it or -2 it, it is ready for merge imo, if you want further work done, pleas e-1 it so the work can be done | 23:03 |
| sdake | i wouldn't be asking if i didn't need it done now | 23:03 |
| sdake | or don't ack it | 23:04 |
| sdake | i'm heading out | 23:04 |
| sdake | when that merges, then i'll be happy to have the cells discussion | 23:05 |
| sdake | bye | 23:05 |
| kfox1111 | sdake: the revew looks ok, other then the weird rearengment of nova-placement. | 23:09 |
| kfox1111 | would it be easy to drop that out of the ps? | 23:10 |
| *** sbezverk has quit IRC | 23:10 | |
| *** sbezverk has joined #openstack-kolla | 23:10 | |
| sdake | kfox1111 -1 the review and i'll change it htanks | 23:12 |
| *** manheim has quit IRC | 23:12 | |
| kfox1111 | k. | 23:16 |
| *** Margin has joined #openstack-kolla | 23:16 | |
| kfox1111 | sdake: done. | 23:18 |
| kfox1111 | sdake: https://review.openstack.org/#/c/457326/ please. | 23:18 |
| *** MarginHu has quit IRC | 23:20 | |
| *** huzhengchuan has quit IRC | 23:21 | |
| *** jtriley has joined #openstack-kolla | 23:22 | |
| *** Margin has quit IRC | 23:25 | |
| *** jascott1 has quit IRC | 23:25 | |
| *** jtriley has quit IRC | 23:27 | |
| *** StephenWang1991 has joined #openstack-kolla | 23:29 | |
| *** jtriley has joined #openstack-kolla | 23:31 | |
| *** StephenWang1991 has quit IRC | 23:34 | |
| sdake | my wife complaining that my cardbox box i have for a rack needs to go | 23:36 |
| * sdake hands wife a screwdriver to build the rack :) | 23:36 | |
| openstackgerrit | Merged openstack/kolla-kubernetes master: Add support for external interface bringup https://review.openstack.org/457326 | 23:41 |
| openstackgerrit | Steven Dake proposed openstack/kolla-kubernetes master: Nova-control service chart dependencies https://review.openstack.org/451495 | 23:48 |
| openstackgerrit | Steven Dake proposed openstack/kolla-kubernetes master: A start at a simplified devtool environment https://review.openstack.org/457075 | 23:48 |
| sbezverk | kfox1111: ping | 23:51 |
| *** gfhellma has quit IRC | 23:51 | |
| *** yangyape_ has quit IRC | 23:57 | |
| spsurya | morning all | 23:58 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!