*** manheim has quit IRC | 00:01 | |
*** randomhack has joined #openstack-kolla | 00:05 | |
*** pradk has quit IRC | 00:09 | |
openstackgerrit | Merged openstack/kolla-ansible stable/ocata: Add makestep in chrony.conf https://review.openstack.org/472134 | 00:31 |
---|---|---|
*** randomhack has quit IRC | 00:39 | |
*** randomhack has joined #openstack-kolla | 00:40 | |
*** randomhack has quit IRC | 00:44 | |
*** randomhack has joined #openstack-kolla | 00:58 | |
*** salv-orlando has joined #openstack-kolla | 01:01 | |
*** tovin07_ has joined #openstack-kolla | 01:02 | |
*** cuongnv has joined #openstack-kolla | 01:02 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible master: Remove keystone-paste.ini file in kolla https://review.openstack.org/462535 | 01:03 |
*** salv-orlando has quit IRC | 01:05 | |
*** salv-orlando has joined #openstack-kolla | 01:13 | |
*** salv-orlando has quit IRC | 01:15 | |
*** salv-orlando has joined #openstack-kolla | 01:15 | |
*** sbezverk has joined #openstack-kolla | 01:17 | |
*** salv-orlando has quit IRC | 01:19 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible master: Use unified vars in task https://review.openstack.org/473166 | 01:20 |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible master: Flatten neutron_* variables https://review.openstack.org/473168 | 01:26 |
sbezverk | Jeffrey4l: ping | 01:27 |
Jeffrey4l | sup sbezverk | 01:27 |
sbezverk | Jeffrey4l are you on mac? | 01:27 |
Jeffrey4l | sbezverk, on. linux ( archlinux ) | 01:27 |
Jeffrey4l | on* | 01:27 |
Jeffrey4l | sbezverk, is kolla-k8s still using kolla-ansible's genconfig feature? | 01:28 |
sbezverk | Jeffrey4l thanks, I do not have "git review" on mac | 01:28 |
Jeffrey4l | sbezverk, found this https://www.mediawiki.org/wiki/Gerrit/git-review#Mac_OS_X :) | 01:29 |
sbezverk | Jeffrey4l : nope, inc0 build parallel version of configuration | 01:29 |
sbezverk | Jeffrey4l thanks for this link | 01:29 |
Jeffrey4l | sbezverk, could u show me the codes? | 01:29 |
*** ssurana has joined #openstack-kolla | 01:30 | |
sbezverk | Jeffrey4l : sure, https://github.com/openstack/kolla-kubernetes/tree/master/ansible | 01:30 |
Jeffrey4l | then kolla-ansible could remove all thing related to k8s? | 01:31 |
Jeffrey4l | like this " host_in_groups: "{{ True if orchestration_engine == 'KUBERNETES' else inventory_hostname in groups['neutron-dhcp-agent'] }}" | 01:31 |
Jeffrey4l | " | 01:31 |
Jeffrey4l | sbezverk, right? | 01:33 |
sbezverk | Jeffrey4l : yes, I think so | 01:34 |
Jeffrey4l | cool. thanks. | 01:34 |
Jeffrey4l | duonghq, around? | 01:36 |
duonghq | Jeffrey4l, I'm here | 01:36 |
Jeffrey4l | anything wanna to talk about https://review.openstack.org/#/c/452067/ duonghq | 01:37 |
Jeffrey4l | i saw you added it into next meeting agenda. | 01:37 |
duonghq | yes, just want to know if we bump the Ansbile version in P | 01:37 |
duonghq | if in this case, we should have plan to upgrade our kolla-ansible logic | 01:38 |
Jeffrey4l | should i -1 for my patch, in case it is merged before meeting. | 01:38 |
duonghq | it's depend on you :P | 01:38 |
Jeffrey4l | got, thanks. | 01:39 |
duonghq | np, thanks | 01:39 |
Jeffrey4l | btw, i think we have to fix all WARNING in current kolla-ansible once the bump ansible version patch is merged. saw lots of warning in CI. | 01:41 |
*** ssurana has left #openstack-kolla | 01:44 | |
*** zhubingbing has joined #openstack-kolla | 01:50 | |
*** ducttape_ has quit IRC | 01:50 | |
*** ducttape_ has joined #openstack-kolla | 01:51 | |
*** xinliang has quit IRC | 01:51 | |
*** ducttap__ has joined #openstack-kolla | 01:52 | |
*** ducttape_ has quit IRC | 01:55 | |
*** zhurong has joined #openstack-kolla | 01:56 | |
*** manheim has joined #openstack-kolla | 02:01 | |
*** xinliang has joined #openstack-kolla | 02:04 | |
*** xinliang has quit IRC | 02:04 | |
*** xinliang has joined #openstack-kolla | 02:04 | |
*** manheim has quit IRC | 02:06 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible master: Use unified vars in task https://review.openstack.org/473166 | 02:06 |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes master: Adding fallback to default keystone port https://review.openstack.org/473179 | 02:07 |
*** ducttap__ has quit IRC | 02:08 | |
*** ducttape_ has joined #openstack-kolla | 02:09 | |
*** ducttape_ has quit IRC | 02:13 | |
openstackgerrit | jimmygc proposed openstack/kolla master: Support VMware DVS https://review.openstack.org/459273 | 02:19 |
*** zhuzeyu has joined #openstack-kolla | 02:21 | |
*** itlinux_ has joined #openstack-kolla | 02:25 | |
*** itlinux has joined #openstack-kolla | 02:25 | |
*** ducttape_ has joined #openstack-kolla | 02:26 | |
*** kiennt26_ has quit IRC | 02:26 | |
openstackgerrit | jimmygc proposed openstack/kolla-ansible master: Add vmware DVS support to kolla-ansible https://review.openstack.org/459270 | 02:27 |
openstackgerrit | jimmygc proposed openstack/kolla master: Support VMware DVS https://review.openstack.org/459273 | 02:28 |
openstackgerrit | Yotaro Konishi proposed openstack/kolla-ansible master: Fix service_type of Octavia https://review.openstack.org/473183 | 02:29 |
*** zhubingbing has quit IRC | 02:34 | |
*** randomhack has quit IRC | 02:37 | |
*** ducttape_ has quit IRC | 02:37 | |
*** itlinux_ has quit IRC | 02:37 | |
*** itlinux has quit IRC | 02:37 | |
*** ducttape_ has joined #openstack-kolla | 02:38 | |
*** ducttape_ has quit IRC | 02:42 | |
*** yingjun_1 has joined #openstack-kolla | 02:44 | |
openstackgerrit | Yotaro Konishi proposed openstack/kolla-ansible master: Fix service_type of Octavia https://review.openstack.org/473185 | 02:44 |
*** jascott1 has joined #openstack-kolla | 02:45 | |
*** jascott1 has quit IRC | 02:50 | |
*** zhubingbing has joined #openstack-kolla | 03:02 | |
*** tonanhngo has joined #openstack-kolla | 03:05 | |
*** tonanhngo has quit IRC | 03:06 | |
*** salv-orlando has joined #openstack-kolla | 03:16 | |
*** salv-orlando has quit IRC | 03:20 | |
*** caowei has joined #openstack-kolla | 03:27 | |
*** randomha1k has joined #openstack-kolla | 03:33 | |
*** dave-mccowan has quit IRC | 03:56 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible stable/ocata: Add timeout parameter for precheck tasks https://review.openstack.org/473205 | 03:58 |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible stable/ocata: Rewrite haproxy precheck conditions to not use string interpolation https://review.openstack.org/473206 | 04:00 |
*** trozet has joined #openstack-kolla | 04:02 | |
*** zhubingbing has quit IRC | 04:05 | |
*** rwsu has joined #openstack-kolla | 04:06 | |
*** yingjun_1 has quit IRC | 04:23 | |
*** zhurong has quit IRC | 04:26 | |
*** trozet has quit IRC | 04:26 | |
*** eaguilar has quit IRC | 04:31 | |
*** zhurong has joined #openstack-kolla | 04:35 | |
openstackgerrit | Tuan Luong-Anh proposed openstack/kolla master: Remove unused function set_status https://review.openstack.org/473211 | 04:38 |
*** skramaja has joined #openstack-kolla | 04:38 | |
*** yingjun_1 has joined #openstack-kolla | 04:44 | |
*** spsurya has joined #openstack-kolla | 05:01 | |
spsurya | morning all | 05:01 |
spsurya | sup duonghq | 05:01 |
*** salv-orlando has joined #openstack-kolla | 05:05 | |
*** zhurong has quit IRC | 05:06 | |
*** zhurong has joined #openstack-kolla | 05:07 | |
*** rstarmer has joined #openstack-kolla | 05:12 | |
*** trozet has joined #openstack-kolla | 05:38 | |
*** omenv has joined #openstack-kolla | 05:39 | |
*** janki has joined #openstack-kolla | 05:56 | |
*** denaitre has quit IRC | 05:58 | |
*** jascott1 has joined #openstack-kolla | 06:05 | |
*** zhuzeyu has quit IRC | 06:09 | |
*** mrunge has quit IRC | 06:16 | |
*** mrunge has joined #openstack-kolla | 06:19 | |
*** salv-orlando has quit IRC | 06:19 | |
*** salv-orlando has joined #openstack-kolla | 06:20 | |
*** sbezverk has quit IRC | 06:21 | |
*** salv-orlando has quit IRC | 06:24 | |
*** sbezverk has joined #openstack-kolla | 06:32 | |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes master: Adding fallback to default keystone port https://review.openstack.org/473179 | 06:34 |
*** yingjun_1 has quit IRC | 06:34 | |
*** yingjun has joined #openstack-kolla | 06:43 | |
*** sbezverk has quit IRC | 06:49 | |
*** pcaruana has joined #openstack-kolla | 06:52 | |
openstackgerrit | Merged openstack/kolla-ansible master: Fix service_type of Octavia https://review.openstack.org/473185 | 06:53 |
*** skramaja is now known as skramaja_afk | 07:02 | |
*** shardy has joined #openstack-kolla | 07:03 | |
openstackgerrit | Merged openstack/kolla-ansible master: Nova oslo_messaging_notifications config error https://review.openstack.org/444504 | 07:07 |
openstackgerrit | Merged openstack/kolla-ansible master: Neutron oslo_messaging_notifications config error https://review.openstack.org/444506 | 07:07 |
*** magicboiz has joined #openstack-kolla | 07:09 | |
*** salv-orlando has joined #openstack-kolla | 07:11 | |
*** magicboiz has quit IRC | 07:14 | |
*** magicboiz has joined #openstack-kolla | 07:14 | |
*** omenv has quit IRC | 07:17 | |
*** manheim has joined #openstack-kolla | 07:17 | |
openstackgerrit | jimmygc proposed openstack/kolla-ansible master: Add vmware DVS support to kolla-ansible https://review.openstack.org/459270 | 07:21 |
*** lpetrut has joined #openstack-kolla | 07:23 | |
openstackgerrit | jimmygc proposed openstack/kolla-ansible master: Add vSphere support to nova_compute https://review.openstack.org/449529 | 07:24 |
*** karimb has joined #openstack-kolla | 07:30 | |
*** trozet has quit IRC | 07:30 | |
openstackgerrit | jimmygc proposed openstack/kolla-ansible master: Add VMware NSXV support to neutron https://review.openstack.org/450092 | 07:30 |
openstackgerrit | zhubingbing proposed openstack/kolla-ansible master: Change Solume keystone v2 to v3 https://review.openstack.org/472931 | 07:33 |
*** zhubingbing_ has joined #openstack-kolla | 07:33 | |
zhubingbing_ | ping Bertrand Lallau | 07:33 |
openstackgerrit | Juan J. MartÃnez proposed openstack/kolla-ansible master: Update Vagrant dev environment https://review.openstack.org/470280 | 07:33 |
reidrac | morning; +w on https://review.openstack.org/#/c/470270/ anyone? :) | 07:34 |
*** unicell1 has joined #openstack-kolla | 07:38 | |
*** ducttape_ has joined #openstack-kolla | 07:38 | |
*** unicell has quit IRC | 07:39 | |
*** mrunge_ has joined #openstack-kolla | 07:39 | |
*** mrunge has quit IRC | 07:40 | |
*** ducttape_ has quit IRC | 07:43 | |
*** numans_ has joined #openstack-kolla | 07:45 | |
*** salv-orlando has quit IRC | 07:47 | |
*** numans has quit IRC | 07:48 | |
*** salv-orlando has joined #openstack-kolla | 07:48 | |
*** egonzalez has joined #openstack-kolla | 08:02 | |
*** daidv has quit IRC | 08:03 | |
*** sbezverk has joined #openstack-kolla | 08:04 | |
*** karimb has quit IRC | 08:12 | |
*** blallau has joined #openstack-kolla | 08:13 | |
*** jaosorior has joined #openstack-kolla | 08:16 | |
*** daidv has joined #openstack-kolla | 08:19 | |
*** rstarmer has quit IRC | 08:27 | |
*** mgoddard has joined #openstack-kolla | 08:28 | |
*** trozet has joined #openstack-kolla | 08:37 | |
manheim | hi | 08:43 |
*** pbourke has quit IRC | 08:44 | |
*** pbourke has joined #openstack-kolla | 08:45 | |
manheim | i have an issue deploying the keystone chart for kolla-k8s | 08:47 |
manheim | Running command: '/usr/sbin/httpd -DFOREGROUND' | 08:47 |
manheim | /usr/local/bin/kolla_start: line 24: /usr/sbin/httpd: No such file or directory | 08:47 |
manheim | they other keystone pods are running | 08:48 |
manheim | this is keystone-1902127961-2gldb | 08:48 |
sbezverk | manheim kubectl get pod -n kolla | grep keystone | 08:50 |
manheim | https://paste.ubuntu.com/24840040/ | 08:51 |
manheim | ubuntu 16.04 | 08:51 |
manheim | also, at the forst deploy time none of the keystone pods worked, they started working after a reboot, same for mariadb | 08:52 |
*** karimb has joined #openstack-kolla | 08:52 | |
*** athomas has joined #openstack-kolla | 08:52 | |
sbezverk | manheim how did you start it? iit does not look right at all. you have keystone bootstrap things running | 08:52 |
manheim | i followed the docs | 08:53 |
manheim | helm install --debug kolla-kubernetes/helm/service/keystone --namespace kolla --name keystone --values ./cloud.yaml | 08:53 |
sbezverk | manheim keystone-manage-db-vt7q4 means it tries to prepare keystone db | 08:54 |
manheim | ok, so it should go away after it finishes, right? | 08:54 |
sbezverk | it should not happen, it should be running for a couple of minutes and then go away | 08:54 |
manheim | aha | 08:54 |
manheim | well, it throws the httpd command not existing | 08:55 |
manheim | could it be the docker image? | 08:55 |
manheim | it's the first time working with kolla-k8s so I'm kinda in the dark | 08:56 |
sbezverk | manheim kubectl get pods -n kolla | 08:56 |
manheim | https://paste.ubuntu.com/24840055/ | 08:56 |
sbezverk | manheim : helm delete keystone --purge | 08:58 |
sbezverk | helm install helm/service/keystone-cleanup. -- you need to add here your cloud | 08:59 |
sbezverk | and then reinstall keystone | 09:00 |
openstackgerrit | Merged openstack/kolla-ansible master: Fix sed replacement command in Ubuntu's install_docker https://review.openstack.org/470270 | 09:00 |
reidrac | thanks sdake! \o/ | 09:01 |
manheim | will do | 09:01 |
manheim | thanks | 09:01 |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible master: Remove port-security in gates https://review.openstack.org/473321 | 09:03 |
openstackgerrit | Merged openstack/kolla-kubernetes master: Allow overriding images for ironic pods with multiple containers https://review.openstack.org/469483 | 09:04 |
sdake | sbezverk did you see https://review.openstack.org/#/c/472349/1 | 09:06 |
manheim | lol, even keystone delete enters crashloop state | 09:07 |
sbezverk | sdake: not yet, looking at it.. | 09:08 |
*** zhubingbing_ has quit IRC | 09:08 | |
*** zhubingbing_ has joined #openstack-kolla | 09:09 | |
*** rmart04 has joined #openstack-kolla | 09:17 | |
manheim | hmm, I get connection refused from the DB | 09:18 |
*** jascott1 has quit IRC | 09:19 | |
*** jascott1 has joined #openstack-kolla | 09:19 | |
sbezverk | manheim : probably your mariadb does not function properly | 09:20 |
sbezverk | try to test it by using mysql tool | 09:20 |
openstackgerrit | Merged openstack/kolla master: Add "Vitrage" to "Openstack services" in README.rst https://review.openstack.org/463228 | 09:21 |
*** karimb has quit IRC | 09:21 | |
*** karimb has joined #openstack-kolla | 09:21 | |
*** jascott1 has quit IRC | 09:24 | |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla master: Fix kolla-toolbox build https://review.openstack.org/473332 | 09:26 |
sdake | nothing like waking up to $1300 tube blown | 09:29 |
* sdake growls | 09:29 | |
openstackgerrit | Merged openstack/kolla-ansible master: Add the ceph keyring files related task condition https://review.openstack.org/458417 | 09:32 |
openstackgerrit | Merged openstack/kolla-ansible master: Remove symlink from fluentd global conf https://review.openstack.org/458791 | 09:32 |
*** trozet has quit IRC | 09:35 | |
*** wagiel has joined #openstack-kolla | 09:36 | |
openstackgerrit | jimmygc proposed openstack/kolla-ansible master: Make glance filesystem_store_datadir configurable. https://review.openstack.org/452094 | 09:36 |
*** skramaja_afk is now known as skramaja | 09:37 | |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla master: Implement blazar Docker images https://review.openstack.org/454745 | 09:38 |
openstackgerrit | jimmygc proposed openstack/kolla-ansible master: Make glance filesystem_store_datadir configurable. https://review.openstack.org/452094 | 09:41 |
openstackgerrit | jimmygc proposed openstack/kolla-ansible master: Make glance filesystem_store_datadir configurable. https://review.openstack.org/452094 | 09:42 |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible master: Change zun rewrite rule number in fluentd https://review.openstack.org/473341 | 09:44 |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible master: Implement blazar ansible role https://review.openstack.org/454752 | 09:46 |
*** tvignaud has quit IRC | 09:46 | |
openstackgerrit | jimmygc proposed openstack/kolla-ansible master: Add VMware NSXV support to neutron https://review.openstack.org/450092 | 09:51 |
openstackgerrit | Merged openstack/kolla-ansible master: Add default_boot_option configuration https://review.openstack.org/472112 | 09:55 |
openstackgerrit | Merged openstack/kolla-ansible master: Remove deprecated options designate https://review.openstack.org/450715 | 09:57 |
*** magicboiz has quit IRC | 09:58 | |
*** david-lyle has quit IRC | 10:02 | |
*** tvignaud has joined #openstack-kolla | 10:04 | |
*** daidv has quit IRC | 10:12 | |
*** zhurong has quit IRC | 10:12 | |
*** magicboiz has joined #openstack-kolla | 10:14 | |
*** duonghq has quit IRC | 10:16 | |
*** magicboiz has quit IRC | 10:18 | |
*** magicboiz has joined #openstack-kolla | 10:18 | |
*** wagiel has quit IRC | 10:19 | |
*** tovin07_ has quit IRC | 10:20 | |
*** david-lyle has joined #openstack-kolla | 10:24 | |
*** wagiel has joined #openstack-kolla | 10:25 | |
egonzalez | can I have reviews https://review.openstack.org/#/c/473332/ , kolla-build is broken in master and stable branchs due missing make package for rpm distros | 10:26 |
egonzalez | *kolla-toolbox | 10:26 |
*** daidv has joined #openstack-kolla | 10:26 | |
reidrac | egonzalez: just fixed this in my SUSE environment; PyNaCl now requires make for libsodium | 10:30 |
reidrac | I can confirm that adding GNU make to the installed packages fixes the problem | 10:30 |
egonzalez | reidrac, yep | 10:30 |
reidrac | ah, oh; that's the actual text of the commit hah | 10:31 |
egonzalez | reidrac, the change add make to rpm distros, .deb distros already have make as part of build-essential package | 10:33 |
reidrac | yes, exacly what happens with SUSE | 10:35 |
egonzalez | reidrac, are you working with suse in kolla? | 10:36 |
reidrac | this sounds a bit like the issue with setuptools; perhaps pip installed dependencies would be best pinned to a specific version | 10:36 |
reidrac | egonzalez: yes, I'm woring on it; got the dev environment with vagrant + keystone and dependencies up an running; for now | 10:37 |
egonzalez | reidrac, cool | 10:37 |
reidrac | all running openSUSE 42.2 but fluentd, because td-agent doesn't support SUSE | 10:37 |
reidrac | but that container can be run with CentOS and is fine; until I get to fix that | 10:38 |
*** MarginHu has joined #openstack-kolla | 10:38 | |
*** sbhojpur has joined #openstack-kolla | 10:39 | |
*** zhurong has joined #openstack-kolla | 10:39 | |
openstackgerrit | jimmygc proposed openstack/kolla-ansible master: Fix fwaas options https://review.openstack.org/463882 | 10:40 |
reidrac | when this is merged: https://review.openstack.org/#/c/470280/ ; I'll start pushing some of the SUSE bits, at least the dev environment that is self contained | 10:40 |
*** Yumeng has joined #openstack-kolla | 10:41 | |
sbhojpur | Need help with Kolla Magnum installation | 10:43 |
sbhojpur | Somehow my setup is stuck in cluster creation till it times out | 10:44 |
sbhojpur | Heat engine logs says 'creating HeatWaitCondition "master_wait_condition" Stack "test_ashish-kmdzvrbp5byu-kube_masters-dbjvn7oxpd7q-0-aagbote25v4t" [747d1952-bd43-4dab-b06d-e915e903d9d7] ' | 10:46 |
*** caowei has quit IRC | 10:49 | |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla master: Change ubuntu ODL repo to ppa's https://review.openstack.org/473359 | 10:51 |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla master: Change ubuntu ODL repo to ppa's https://review.openstack.org/473359 | 10:52 |
*** yingjun has quit IRC | 10:54 | |
egonzalez | reidrac, there you go :) | 10:56 |
reidrac | egonzalez: cheers! :) | 10:57 |
openstackgerrit | Bertrand Lallau proposed openstack/kolla-ansible master: Fix "Enable influxdb datasource" random error https://review.openstack.org/472977 | 10:57 |
openstackgerrit | Merged openstack/kolla-ansible master: Update Vagrant dev environment https://review.openstack.org/470280 | 10:59 |
*** wagiel has quit IRC | 11:00 | |
egonzalez | reidrac, btw, dont need to wait for dependant changes to merge, if adds Depends-On: <CHANGE_ID> in commit message, dependant change will be used in gates | 11:00 |
reidrac | egonzalez: OK, thanks for the tip | 11:00 |
*** sbezverk has quit IRC | 11:05 | |
*** cuongnv has quit IRC | 11:08 | |
*** sbezverk has joined #openstack-kolla | 11:16 | |
*** jascott1 has joined #openstack-kolla | 11:20 | |
*** jascott1 has quit IRC | 11:24 | |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes master: Do not merge. Review test https://review.openstack.org/473378 | 11:25 |
manheim | how can I delete keystone-cleanup? | 11:31 |
*** pradk has joined #openstack-kolla | 11:31 | |
manheim | does not work w helm delete | 11:33 |
*** randomha1k has quit IRC | 11:36 | |
*** ducttape_ has joined #openstack-kolla | 11:39 | |
sbezverk | manheim you should be able to delete any helm deployment with helm delete --purge | 11:40 |
sbezverk | helm ls | 11:40 |
manheim | if i do helm list I don;t see it | 11:41 |
manheim | note: it's in crashloop state | 11:41 |
sbezverk | manheim it should not matter | 11:42 |
sbezverk | manheim kubectl get job -n kolla | grep keystone | 11:42 |
manheim | Error: Unable to lock release keystone-delete-db: release not found | 11:43 |
*** ducttape_ has quit IRC | 11:43 | |
sbezverk | manheim kubectl get job -n kolla | grep keystone | awk '{print "kubectl delete job -n kolla "$1}' | sh -l | 11:44 |
sbezverk | that should clean up all the jobs | 11:44 |
sbezverk | you also must make sure mariadb is working | 11:45 |
sbezverk | and if the name resolution working too, if keystone cannot locate mariadb service it will be very upset and will start crash loop | 11:46 |
manheim | resolution should woek | 11:56 |
manheim | work | 11:57 |
*** ducttape_ has joined #openstack-kolla | 11:59 | |
manheim | if I reinstall the DB pod, the main container from mariadb-0 does not start | 12:00 |
manheim | Error from server (BadRequest): container "main" in pod "mariadb-0" is waiting to start: PodInitializing | 12:01 |
egonzalez | anyone using kuryr with kolla? | 12:02 |
egonzalez | as far I can see, only kuryr api is implemented, no backend at all(libnetwork or k8s) | 12:03 |
sbezverk | manheim that is the clear indication that there is problem with your setup | 12:03 |
manheim | hmm, I see, makes sense | 12:04 |
manheim | tho i followed the docs | 12:04 |
sbezverk | are you using aio? | 12:04 |
manheim | yes | 12:06 |
manheim | in a ubuntu VM | 12:06 |
manheim | thing is after a reboot the pods start correctly | 12:07 |
*** ducttape_ has quit IRC | 12:11 | |
*** ducttape_ has joined #openstack-kolla | 12:11 | |
openstackgerrit | Merged openstack/kolla-ansible master: Change zun rewrite rule number in fluentd https://review.openstack.org/473341 | 12:12 |
*** ducttape_ has quit IRC | 12:16 | |
*** zhurong has quit IRC | 12:16 | |
*** mrunge_ is now known as mrunge | 12:17 | |
openstackgerrit | Bertrand Lallau proposed openstack/kolla-ansible stable/ocata: Fix Telegraf container restarting issue https://review.openstack.org/464039 | 12:18 |
openstackgerrit | zhubingbing proposed openstack/kolla-ansible master: Change Solum keystone v2 to v3 https://review.openstack.org/472931 | 12:19 |
*** sbhojpur has quit IRC | 12:22 | |
*** pomac_ is now known as pomac | 12:22 | |
*** hrw has quit IRC | 12:24 | |
*** karimb has quit IRC | 12:26 | |
*** schwicht has joined #openstack-kolla | 12:28 | |
*** hrw has joined #openstack-kolla | 12:29 | |
openstackgerrit | Andy Smith proposed openstack/kolla-ansible master: Add support for hybrid messaging backends https://review.openstack.org/468966 | 12:30 |
sdake | manheim if you delete the pod related to mariadb you ahve to remove all backoding data the pod created | 12:32 |
*** janki has quit IRC | 12:33 | |
*** schwicht has quit IRC | 12:34 | |
*** numans has joined #openstack-kolla | 12:36 | |
*** numans_ has quit IRC | 12:39 | |
*** dave-mccowan has joined #openstack-kolla | 12:39 | |
*** omenv has joined #openstack-kolla | 12:40 | |
manheim | and that is done with helm delete --purge, right? | 12:41 |
sdake | no | 12:41 |
*** omenv has quit IRC | 12:43 | |
*** omenv has joined #openstack-kolla | 12:43 | |
bmaltais | Hi, trying to deploy to a new node this morning and I get "The requested image does not exist: registry.ssclab.com:5000/kolla/centos-binary-fluentd:5.0.0" | 12:48 |
bmaltais | I did a git pull this morning and notices a lot of changes | 12:49 |
bmaltais | Was something introduced that would break the deploy based on a non existand fluentd 5.0.0 pre-built container? | 12:49 |
*** schwicht has joined #openstack-kolla | 12:49 | |
*** karimb has joined #openstack-kolla | 12:50 | |
*** schwicht has quit IRC | 12:52 | |
*** ducttape_ has joined #openstack-kolla | 12:52 | |
bmaltais | Setting "openstack_release: "4.0.0"" has fixed the issue | 12:54 |
egonzalez | bmaltais, kolla 5.0.0 (master) was tagged on friday | 12:54 |
bmaltais | egonzalez: Are pre-build containers for it available? Is this why it fail? | 12:55 |
bmaltais | egonzalez: Or do I have to run an upgrade on my setup? | 12:55 |
egonzalez | bmaltais, nope, containers are master images | 12:56 |
egonzalez | but if you not set openstack_release at globals it will try to use 5.0.0 due pbr versioning | 12:56 |
bmaltais | egonzalez: I never built my containers... so forcing 4.0.0 is probably safer for now? | 12:56 |
egonzalez | bmaltais, yep | 12:57 |
*** ducttape_ has quit IRC | 12:57 | |
bmaltais | egonzalez: OK. I am in the process of converting docker to btrfs... I have some strange issues withlvm I think | 12:57 |
MarginHu | I have used btrfs for 2 weeks as docker backend , no issue happened | 13:00 |
*** rwallner has joined #openstack-kolla | 13:00 | |
manheim | how can I remove all the data mariadb pod created? | 13:01 |
sbezverk | manheim normally volumes are created | 13:01 |
sbezverk | at /var/lib/kolla/volumes | 13:02 |
manheim | sudo rm -rf /var/lib/kolla/volumes/* | 13:02 |
manheim | got it | 13:02 |
sbezverk | yep | 13:02 |
*** randomhack has joined #openstack-kolla | 13:06 | |
*** ducttape_ has joined #openstack-kolla | 13:08 | |
*** ducttap__ has joined #openstack-kolla | 13:10 | |
*** schwicht has joined #openstack-kolla | 13:11 | |
*** rwallner has quit IRC | 13:12 | |
*** rwallner has joined #openstack-kolla | 13:12 | |
*** ducttape_ has quit IRC | 13:12 | |
reidrac | I'm having a weird issue deploying fluentd; ansible_eth0 variable not defined replacing the template for 01.syslog.conf; this is after api_interface_address generalisation | 13:14 |
*** eanylin has quit IRC | 13:15 | |
*** ducttap__ has quit IRC | 13:16 | |
reidrac | is like is not reading all.yml | 13:16 |
*** skramaja has quit IRC | 13:21 | |
openstackgerrit | Mark Goddard proposed openstack/kolla-ansible master: Fix bootstrap-servers with invalid api_interface https://review.openstack.org/434940 | 13:25 |
*** ducttape_ has joined #openstack-kolla | 13:25 | |
*** krtaylor has joined #openstack-kolla | 13:26 | |
*** lucasxu has joined #openstack-kolla | 13:29 | |
*** ducttape_ has quit IRC | 13:31 | |
*** ducttape_ has joined #openstack-kolla | 13:33 | |
*** tumble has joined #openstack-kolla | 13:35 | |
*** randomhack has quit IRC | 13:35 | |
*** eanylin has joined #openstack-kolla | 13:36 | |
*** emccormick has joined #openstack-kolla | 13:42 | |
*** karimb has quit IRC | 13:48 | |
*** ducttape_ has quit IRC | 13:49 | |
blallau | @reidrac this is due to a Facts loading issue | 13:54 |
blallau | it should be fix by https://review.openstack.org/#/c/472660/ | 13:55 |
*** ducttape_ has joined #openstack-kolla | 13:57 | |
*** ducttape_ has quit IRC | 13:58 | |
*** ducttape_ has joined #openstack-kolla | 13:59 | |
sdake | sbezverk yup the macs rock | 14:02 |
sdake | sbezverk compared to windows for dev atleast - mac blows pc away | 14:02 |
*** karimb has joined #openstack-kolla | 14:02 | |
*** schwicht has quit IRC | 14:04 | |
*** schwicht has joined #openstack-kolla | 14:05 | |
*** ducttape_ has quit IRC | 14:05 | |
reidrac | blallau: thanks for the pointer | 14:16 |
reidrac | blallau: yes! that was it, thanks! | 14:18 |
blallau | @reidac you are welcome ;) | 14:18 |
*** ducttape_ has joined #openstack-kolla | 14:21 | |
*** bmace has quit IRC | 14:25 | |
*** jtriley has joined #openstack-kolla | 14:25 | |
*** bmace has joined #openstack-kolla | 14:25 | |
hrw | sdake: mac blows pc or is it rather macos blows mswindows? | 14:26 |
sdake | something blows tahts for sure | 14:27 |
sdake | like my 1300$ elrog 845 tube in my tube amp this mroning | 14:27 |
* sdake FURIOUS | 14:27 | |
*** ducttape_ has quit IRC | 14:28 | |
*** jamesbenson has joined #openstack-kolla | 14:34 | |
*** sgrasley has joined #openstack-kolla | 14:37 | |
*** sgrasley has left #openstack-kolla | 14:37 | |
*** jtriley has quit IRC | 14:38 | |
*** jtriley has joined #openstack-kolla | 14:38 | |
*** eaguilar has joined #openstack-kolla | 14:39 | |
openstackgerrit | Christian Berendt proposed openstack/kolla-ansible master: Use templates for keystone domain specific configurations https://review.openstack.org/473483 | 14:42 |
*** MarginHu has quit IRC | 14:43 | |
*** ducttape_ has joined #openstack-kolla | 14:45 | |
openstackgerrit | Mathias Ewald proposed openstack/kolla-ansible master: Fix Horizon Static Files Path for Ubuntu https://review.openstack.org/473485 | 14:46 |
*** mewald has joined #openstack-kolla | 14:46 | |
*** jtriley has quit IRC | 14:46 | |
*** jtriley has joined #openstack-kolla | 14:50 | |
*** srwilkers has joined #openstack-kolla | 14:50 | |
reidrac | question: how much disk needs all-in-one deployment? | 14:52 |
reidrac | I wild guess would be OK; say... around 40GB? | 14:52 |
reidrac | "official" openSUSE image is 20GB and I run out disk deploying keystone and deps | 14:53 |
*** jtriley has quit IRC | 14:57 | |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla master: Add openvswitch and bridge-utils in kuryr https://review.openstack.org/473490 | 14:58 |
*** jtriley has joined #openstack-kolla | 14:58 | |
*** hongbin has joined #openstack-kolla | 14:59 | |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla master: Add openvswitch and bridge-utils in kuryr https://review.openstack.org/473490 | 15:01 |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla master: Add openvswitch and bridge-utils in kuryr https://review.openstack.org/473490 | 15:01 |
*** ducttape_ has quit IRC | 15:03 | |
openstackgerrit | Wayne Chan proposed openstack/kolla master: Optimize the link address https://review.openstack.org/473493 | 15:04 |
*** ducttape_ has joined #openstack-kolla | 15:04 | |
*** jtriley has quit IRC | 15:05 | |
*** dklyle has joined #openstack-kolla | 15:07 | |
openstackgerrit | Wayne Chan proposed openstack/kolla master: Use https links for documentation https://review.openstack.org/473493 | 15:07 |
*** jtriley has joined #openstack-kolla | 15:07 | |
*** omenv has quit IRC | 15:07 | |
*** omenv has joined #openstack-kolla | 15:07 | |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible master: WIP: fix kuryr https://review.openstack.org/473495 | 15:08 |
*** fooliouno has joined #openstack-kolla | 15:08 | |
*** david-lyle has quit IRC | 15:10 | |
openstackgerrit | Andy Smith proposed openstack/kolla-ansible master: Add qdrouterd role for messaging infrastructure component https://review.openstack.org/469462 | 15:16 |
egonzalez | please, review. kolla-toolbox fail to build in all rpm distros for stable and master branchs https://review.openstack.org/#/c/473332/ | 15:16 |
*** blallau has quit IRC | 15:17 | |
*** jtriley has quit IRC | 15:20 | |
*** lpetrut has quit IRC | 15:21 | |
*** srwilkers has quit IRC | 15:21 | |
*** srwilkers has joined #openstack-kolla | 15:25 | |
*** rmart04 has quit IRC | 15:29 | |
sdake | egonzalez its still fialing many gates? | 15:30 |
egonzalez | sdake, failed in ubuntu gates which are unrelated to the change. make package is a dependency of build-essential | 15:30 |
*** pcaruana has quit IRC | 15:31 | |
*** eaguilar has quit IRC | 15:36 | |
*** hieulq_ has joined #openstack-kolla | 15:37 | |
inc0 | good morning | 15:37 |
*** vhosakot has joined #openstack-kolla | 15:38 | |
*** eaguilar has joined #openstack-kolla | 15:38 | |
*** goldyfruit has joined #openstack-kolla | 15:40 | |
*** manheim has quit IRC | 15:40 | |
kfox1111 | morning. | 15:41 |
sbezverk | good morning | 15:42 |
*** salv-orlando has quit IRC | 15:46 | |
*** salv-orlando has joined #openstack-kolla | 15:47 | |
openstackgerrit | Paul Bourke (pbourke) proposed openstack/kolla-ansible master: Update cinder sanity check to use built in shade https://review.openstack.org/437920 | 15:47 |
*** jtriley has joined #openstack-kolla | 15:47 | |
*** iniazi has quit IRC | 15:48 | |
*** krtaylor has quit IRC | 15:48 | |
*** dklyle has quit IRC | 15:49 | |
*** itlinux has joined #openstack-kolla | 15:49 | |
*** itlinux_ has joined #openstack-kolla | 15:49 | |
*** david-lyle has joined #openstack-kolla | 15:49 | |
spsurya | inc0: morning | 15:51 |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible master: Implement blazar ansible role https://review.openstack.org/454752 | 15:53 |
*** rhallisey has quit IRC | 16:00 | |
*** unicell1 has quit IRC | 16:05 | |
*** egonzalez has quit IRC | 16:06 | |
*** krtaylor has joined #openstack-kolla | 16:07 | |
*** itlinux_ has quit IRC | 16:07 | |
*** itlinux has quit IRC | 16:07 | |
inc0 | kfox1111: http://logs.openstack.org/15/466815/10/experimental/gate-kolla-ansible-dsvm-deploy-ceph-centos-source-centos-7-2-node-nv/6d7e75e/logs/docker_logs/bootstrap_osd_1 have you seen anythin like this fron your ceph in gates thing? | 16:07 |
*** itlinux has joined #openstack-kolla | 16:08 | |
*** itlinux_ has joined #openstack-kolla | 16:08 | |
kfox1111 | inc0: looking.... | 16:10 |
inc0 | thanks | 16:10 |
kfox1111 | hmm... probably not relevent, but it looks odd to see only quorum members. you trying to launch a third there? | 16:11 |
kfox1111 | have not seen that problem before.... | 16:11 |
inc0 | it's gonna be hard on 2 node gate | 16:11 |
*** daidv_ has joined #openstack-kolla | 16:11 | |
kfox1111 | maybe zap the partition table before running? | 16:11 |
kfox1111 | yeah. but it won't hbe happy with an even number. | 16:11 |
kfox1111 | a single node ceph will be ok for testing. | 16:11 |
inc0 | meh | 16:12 |
kfox1111 | there's nothing really different between a single node and multi as far as openstack's concerned. | 16:12 |
inc0 | well, I can fix that | 16:12 |
sbezverk | kfox1111: in your comment, you do not like me removing port from the workflow common config or falling back to 5000? | 16:12 |
inc0 | error doesn't seem to be related tho | 16:12 |
kfox1111 | sbezverk: oh. thought I was looking at all_values.yaml. which it should be in there? | 16:13 |
kfox1111 | yeah. removing it from common_workflow_config.sh is good. | 16:13 |
inc0 | ok, off to ceph debug I go | 16:13 |
kfox1111 | defaulting it I think shouldn't be needed as the values.yaml is the defaults. | 16:14 |
sbezverk | kfox1111: unfortunately it does not default | 16:14 |
kfox1111 | inc0: I don't think they ever test 2 node ceph as its not a supported config. | 16:14 |
kfox1111 | sbezverk: is the search path not picking it up then? | 16:14 |
inc0 | I'll do 1 node | 16:14 |
sbezverk | kfox1111: right, by some reason it does not find it.. | 16:15 |
*** manheim has joined #openstack-kolla | 16:16 | |
kfox1111 | sbezverk: oh. I see.... | 16:18 |
kfox1111 | sbezverk: search for 5000 on all_values.yaml | 16:18 |
inc0 | I think this is sgdisk rather than ceph | 16:18 |
kfox1111 | its set on each package explicitly. | 16:18 |
kfox1111 | inc0: quite possible. | 16:18 |
inc0 | https://github.com/openstack/kolla/blob/master/docker/ceph/ceph-osd/extend_start.sh#L20-L28 | 16:18 |
kfox1111 | inc0: hmm.... the journal dev is not configurable there... | 16:19 |
kfox1111 | maybe its using too much space for your volumes? | 16:19 |
inc0 | you can configure journal differently | 16:19 |
*** rwallner has quit IRC | 16:20 | |
*** manheim has quit IRC | 16:20 | |
*** karimb has quit IRC | 16:21 | |
inc0 | kfox1111: plan to watch E3? | 16:21 |
kfox1111 | not sure. | 16:21 |
inc0 | new star wars looks interesting | 16:22 |
*** pbourke has quit IRC | 16:24 | |
*** jascott1 has joined #openstack-kolla | 16:35 | |
*** jascott1 has quit IRC | 16:36 | |
*** eaguilar has quit IRC | 16:37 | |
*** jascott1 has joined #openstack-kolla | 16:48 | |
*** rwallner has joined #openstack-kolla | 16:50 | |
vhosakot | egonzalez: good work in https://review.openstack.org/#/c/473332/, I've reviewed it. I saw the same issue related to pynacl when building. | 16:51 |
*** mewald has quit IRC | 16:53 | |
sdake | its a good thing my tubes have a 1 year warranty | 16:53 |
sdake | i am less ragiing now | 16:53 |
sdake | the vendor grilled me "any power problems" "any fuse problems" "what type of wiring do you have" "blah blah blah" | 16:54 |
sdake | tubes failing after 200 hours of use (for a 10k hr rated tube) is completely unacceptable | 16:54 |
*** hieulq_ has quit IRC | 16:55 | |
*** iniazi has joined #openstack-kolla | 16:55 | |
*** mgoddard has quit IRC | 16:57 | |
*** rhallisey has joined #openstack-kolla | 16:59 | |
*** salv-orl_ has joined #openstack-kolla | 17:02 | |
vhosakot | inc0: does stable/oacat also have multinode gates now? --> https://review.openstack.org/#/c/471401/ | 17:04 |
*** salv-orlando has quit IRC | 17:04 | |
inc0 | vhosakot: yes, but you need to run check experimental for that | 17:05 |
inc0 | https://review.openstack.org/#/c/471428/ until this merges | 17:06 |
*** rhallisey_ has joined #openstack-kolla | 17:10 | |
*** rhallisey has quit IRC | 17:10 | |
vhosakot | inc0: cool, thought so.. I want to get started with upgrade gates... will sync with duonghq later today. | 17:10 |
inc0 | this needs to merge;) https://review.openstack.org/#/c/471428/ | 17:11 |
inc0 | for upgrade gates too | 17:11 |
inc0 | it includes new gate - upgrade | 17:11 |
inc0 | just centos source for now | 17:11 |
vhosakot | inc0: infra pre-work.. got it.. ncie work | 17:11 |
*** mgoddard has joined #openstack-kolla | 17:11 | |
vhosakot | sdake: sorry to hear about your tubes.. why do you use those vacuum tubes for? | 17:12 |
sdake | listening to toons | 17:12 |
openstackgerrit | Marcus Williams proposed openstack/kolla-ansible master: Add OpenDaylight role https://review.openstack.org/416367 | 17:13 |
vhosakot | sdake: I see | 17:13 |
*** jtriley has quit IRC | 17:15 | |
*** unicell has joined #openstack-kolla | 17:15 | |
*** jtriley has joined #openstack-kolla | 17:22 | |
*** lucasxu has quit IRC | 17:24 | |
*** egonzalez has joined #openstack-kolla | 17:25 | |
*** sambetts is now known as sambetts|afk | 17:27 | |
*** itlinux_ has quit IRC | 17:33 | |
*** itlinux has quit IRC | 17:33 | |
*** itlinux has joined #openstack-kolla | 17:38 | |
*** itlinux_ has joined #openstack-kolla | 17:38 | |
inc0 | hey guys, I need sencond +2 on https://review.openstack.org/#/c/471428/ <- needed for gate | 17:40 |
*** jaosorior is now known as jaosorior_away | 17:41 | |
*** eaguilar has joined #openstack-kolla | 17:42 | |
inc0 | btw kfox1111 I figured out this GPU passthrough thing | 17:46 |
kfox1111 | yeah? sweet. :) | 17:47 |
inc0 | it turns out 1. intel skylake put my amd gpu and nvidia gpu in same iommu group | 17:47 |
inc0 | so I bought PCI extention riser and switched pci ports | 17:47 |
inc0 | for nvidia | 17:47 |
kfox1111 | wow. all 3 types of gpu's in one box? :) | 17:47 |
*** kolla-slack2 has quit IRC | 17:47 | |
*** kolla-slack has joined #openstack-kolla | 17:47 | |
inc0 | that split iommu groups but intell still puts it's audio driver on nvidia audio device | 17:47 |
kfox1111 | ah. | 17:47 |
inc0 | so I need to unbind it before I start vm | 17:48 |
inc0 | yeah I have all 3 types of them;) | 17:48 |
inc0 | next stop, make them work together on same tensorflow computation | 17:48 |
inc0 | that's a strech tho:P | 17:48 |
kfox1111 | helm install tensorflow. :) | 17:48 |
kfox1111 | https://github.com/kubernetes/charts/tree/master/incubator/tensorflow-inception | 17:49 |
inc0 | right now I'm still using vagrant because I actually need pci passthrough | 17:49 |
inc0 | ..because of all 3 gpus in my box | 17:49 |
*** daidv_ has quit IRC | 17:49 | |
inc0 | I need that to make a vm with only one type;) | 17:49 |
kfox1111 | hmm.... ok. | 17:50 |
inc0 | installing tensorflow is easy | 17:50 |
inc0 | making it work with gpu is harder | 17:50 |
kfox1111 | yeah. I could believe that. | 17:50 |
inc0 | it was fun ride, but got it working and yesterday made my first gpu based computatuon | 17:51 |
inc0 | I was damn happy;) | 17:51 |
kfox1111 | yeah. the gpu passthrough stuff can be really tricky. :/ | 17:52 |
kfox1111 | figured with all the virutlaization push over the years, they would have cracked that problem a lot sooner. | 17:52 |
inc0 | well problem is quite limited to few use cases | 17:53 |
kfox1111 | there's a whole indestry around virtual desktops. | 17:54 |
*** eaguilar_ has joined #openstack-kolla | 18:01 | |
inc0 | kfox1111: well fwiw virtualbox was super easy to pci passthrough | 18:02 |
inc0 | reason I dropped virtualbox is because nvidia drivers hates pci passthrough | 18:02 |
*** eaguilar has quit IRC | 18:04 | |
*** tonanhngo has joined #openstack-kolla | 18:05 | |
*** tonanhngo has quit IRC | 18:05 | |
*** tonanhngo has joined #openstack-kolla | 18:06 | |
*** lucasxu has joined #openstack-kolla | 18:08 | |
*** Manheim has joined #openstack-kolla | 18:09 | |
*** rhallisey has joined #openstack-kolla | 18:10 | |
*** manheim_ has joined #openstack-kolla | 18:12 | |
*** rhallisey_ has quit IRC | 18:13 | |
openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla-ansible master: Enable multi-scenerio gates https://review.openstack.org/466815 | 18:19 |
openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla-ansible master: Enable deploying ceph on loopback devices https://review.openstack.org/472771 | 18:19 |
*** cinerama` is now known as cinerama | 18:24 | |
*** manheim_ has quit IRC | 18:32 | |
*** erlon has joined #openstack-kolla | 18:38 | |
*** mgoddard has quit IRC | 18:40 | |
*** eaguilar_ has quit IRC | 18:43 | |
*** bjolo has quit IRC | 18:52 | |
*** bjolo has joined #openstack-kolla | 18:53 | |
Manheim | inc0, did you have a change to take a look on my latest patchset? | 18:53 |
inc0 | https://review.openstack.org/#/c/455684/ gates are red | 18:53 |
inc0 | ;) | 18:53 |
inc0 | let me do recheck | 18:54 |
Manheim | ok | 18:55 |
*** vhosakot has quit IRC | 19:04 | |
*** itlinux_ has quit IRC | 19:09 | |
*** itlinux has quit IRC | 19:09 | |
*** ducttape_ has quit IRC | 19:21 | |
*** itlinux_ has joined #openstack-kolla | 19:21 | |
*** ducttape_ has joined #openstack-kolla | 19:21 | |
*** itlinux has joined #openstack-kolla | 19:21 | |
*** itlinux has quit IRC | 19:22 | |
*** itlinux_ has quit IRC | 19:22 | |
*** mgoddard has joined #openstack-kolla | 19:29 | |
Manheim | they failed again :( | 19:30 |
*** eaguilar has joined #openstack-kolla | 19:33 | |
*** mgoddard has quit IRC | 19:35 | |
*** itlinux has joined #openstack-kolla | 19:37 | |
*** itlinux_ has joined #openstack-kolla | 19:37 | |
kfox1111 | inc0: yeah, its not very hard to do the passthrough in kvm either. but its making it actually work, thats the ugly. ;) | 19:39 |
dave-mccowan | kfox1111 i'm looking at the haproxy support for openstack services in kolla-kubernetes. i see it is there for nova/cinder/glance, but not horizon and keystone. do you know if is there any reason why? | 19:46 |
kfox1111 | yeah. | 19:46 |
kfox1111 | horizon and keystone both embed nicely in apache in kolla <= 4. | 19:47 |
kfox1111 | while the rest are more often done python stand alone. | 19:47 |
kfox1111 | its a goal of pike I think to support all of them in apache. | 19:47 |
*** salv-orl_ has quit IRC | 19:55 | |
*** salv-orlando has joined #openstack-kolla | 19:55 | |
*** salv-orlando has quit IRC | 19:59 | |
*** shardy has quit IRC | 20:07 | |
inc0 | kfox1111: I made progress with ceph thing, so, ever seen this? http://logs.openstack.org/15/466815/11/experimental/gate-kolla-ansible-dsvm-deploy-ceph-centos-source-centos-7-2-node-nv/92345a1/logs/kolla/ceph/ceph-osd.0.log | 20:12 |
sbezverk | kfox1111: when do we plan to use parent grouping? the reason I am asking is if it is soon, then no reason to merge this ps, but if it is in "near" future, maybe it is worth to merge it., thought? | 20:16 |
*** nea1 has joined #openstack-kolla | 20:17 | |
nea1 | FAILED! => {"failed": true, "msg": "The conditional check ''{{ hostvars[item['item']]['ansible_' + hostvars[item['item']]['api_interface']]['ipv4']['address'] }}' not in '{{ item.stdout }}'' failed. The error was: Invalid conditional detected: EOL while scanning string literal (<unknown>, line 1)\n\nThe error appears to have been in | 20:17 |
nea1 | '/home/openstack/.venv/share/kolla-ansible/ansible/roles/rabbitmq/tasks/precheck.yml': line 69, column 3, but may\nbe elsewhere in the file depending on the exact syntax problem.\n\nThe offending line appears to be:\n\n\n- fail: msg=\"Hostname has to resolve to IP address of api_interface\"\n ^ here\n"} | 20:17 |
nea1 | anyone with an idea what could be wrong? | 20:18 |
inc0 | nea1: which task? | 20:18 |
inc0 | ahh I see | 20:18 |
nea1 | i did a debug output of both variables, and they seam fine | 20:18 |
inc0 | Hostname has to resolve to IP address of api_interface | 20:18 |
nea1 | yep, it does | 20:18 |
*** rwallner has quit IRC | 20:19 | |
inc0 | all controllers has to resolve each other | 20:19 |
nea1 | debug of item.stdout ""msg": "172.31.15.1 STREAM rz15n1\n172.31.15.1 DGRAM \n172.31.15.1 RAW "" | 20:19 |
inc0 | using api_interface address | 20:19 |
nea1 | debug of hostvars[item['item']]['ansible_' + hostvars[item['item']]['api_interface']]['ipv4']['address'] "msg": "172.31.15.1" | 20:19 |
inc0 | all of control nodes can resolve all other using api_interface? | 20:20 |
inc0 | hostname -> api interface address | 20:20 |
inc0 | all control nodes to all other | 20:21 |
nea1 | its only one node | 20:21 |
*** MarginHu has joined #openstack-kolla | 20:21 | |
*** MarginHu has quit IRC | 20:21 | |
inc0 | so it's aio? | 20:21 |
*** MarginHu has joined #openstack-kolla | 20:21 | |
inc0 | all in one | 20:21 |
nea1 | yep, but i'm using a modified hosts file | 20:22 |
nea1 | as i'm not runing it as root | 20:22 |
inc0 | modified how? mind pasting it? | 20:22 |
nea1 | not runing kolla-ansible as root | 20:22 |
nea1 | ansible_user=root and with the ip | 20:23 |
nea1 | [network] | 20:23 |
nea1 | 172.31.15.1 ansible_user=root | 20:23 |
inc0 | /etc/hosts | 20:23 |
inc0 | plz | 20:23 |
nea1 | # BEGIN ANSIBLE GENERATED HOSTS | 20:24 |
nea1 | 172.31.15.1 rz15n1 | 20:24 |
nea1 | # END ANSIBLE GENERATED HOSTS | 20:24 |
inc0 | aaand rz15n1 doesn't resolve to anything else? | 20:25 |
nea1 | in the lines above only localhost and some ipv6 hosts like ip6-allhosts | 20:25 |
inc0 | as in, ping rz15n1 will show 172...? | 20:25 |
nea1 | yep | 20:25 |
nea1 | thats what the test above does | 20:25 |
inc0 | and there is only one node in [control] ? | 20:25 |
nea1 | and the test is working but the fail: when: part is failing | 20:26 |
inc0 | yeah, and fails if there is something wrong with hostnames | 20:26 |
nea1 | all the same | 20:26 |
nea1 | well then the test should fail not the "The conditional check [..." | 20:26 |
inc0 | also, you're deploying on node you run ansible from? | 20:26 |
nea1 | yep, but ansible is installed in a venv under the user openstack, not as root | 20:27 |
inc0 | that's fine | 20:27 |
inc0 | plz add debug with {{ hostvars[item['item']]['ansible_' + hostvars[item['item']]['api_interface']]['ipv4']['address'] }}' | 20:28 |
nea1 | 22:19:52 nea1 | debug of hostvars[item['item']]['ansible_' + hostvars[item['item']]['api_interface']]['ipv4']['address'] "msg": "172.31.15.1" | 20:29 |
inc0 | hmm, which ansible version? | 20:29 |
*** emccormick has quit IRC | 20:30 | |
nea1 | ansible==2.3.1.0 | 20:30 |
inc0 | try rollinb back to 2.2.* | 20:30 |
inc0 | if that helps, I guess we have bug | 20:30 |
nea1 | yep, that fixed it | 20:31 |
*** manheim_ has joined #openstack-kolla | 20:32 | |
*** jtriley has quit IRC | 20:33 | |
*** salv-orlando has joined #openstack-kolla | 20:36 | |
*** manheim_ has quit IRC | 20:37 | |
egonzalez | backport was merged for that issue, will be available in the next ocata release | 20:38 |
inc0 | right | 20:39 |
inc0 | speaking of which | 20:39 |
inc0 | Jeffrey4l: when you wake up :) ^ | 20:39 |
egonzalez | inc0 can you +w, will need backport befora tagging https://review.openstack.org/#/c/473332/ | 20:39 |
inc0 | done egonzalez | 20:40 |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla stable/ocata: Fix kolla-toolbox build https://review.openstack.org/473580 | 20:44 |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla stable/ocata: Fix kolla-toolbox build https://review.openstack.org/473580 | 20:45 |
*** emccormick has joined #openstack-kolla | 20:47 | |
*** eaguilar has quit IRC | 20:51 | |
jascott1 | kfox1111 ping | 20:53 |
*** iniazi has quit IRC | 21:00 | |
*** eaguilar has joined #openstack-kolla | 21:00 | |
*** harlowja has quit IRC | 21:02 | |
openstackgerrit | Bradley Jones proposed openstack/kolla-kubernetes master: (WIP) Deploy playbook using microservices https://review.openstack.org/473588 | 21:02 |
*** rwsu has quit IRC | 21:03 | |
*** rhallisey has quit IRC | 21:07 | |
*** itlinux_ has quit IRC | 21:08 | |
*** itlinux has quit IRC | 21:08 | |
*** iniazi has joined #openstack-kolla | 21:08 | |
*** lucasxu has quit IRC | 21:12 | |
*** eaguilar has quit IRC | 21:18 | |
sdake | inc0 kfox1111 https://review.openstack.org/#/c/473588/1 | 21:22 |
inc0 | sdake: soooo....that needs a lot of work | 21:22 |
sdake | agreed | 21:22 |
inc0 | biggest issue we have (and reason we need entrypoint) is that we need to wait till things finish | 21:23 |
sdake | that is why gerrit dudes invented the comment button | 21:23 |
inc0 | and that's biggest and most important thing to build here | 21:23 |
inc0 | we need design discussion | 21:23 |
sdake | tell brad where the sync points on | 21:23 |
inc0 | not just review | 21:23 |
sdake | on/are | 21:24 |
inc0 | bradjones, sdake can we have webex/zoom/hangouts session about that? | 21:24 |
sdake | inc0 bradjones is in uk timezone | 21:24 |
sdake | inc0 i htink morning time shoudl work | 21:24 |
inc0 | wed before meeting? | 21:25 |
inc0 | or tomorrow around meeting time? | 21:25 |
inc0 | 16UTC that is | 21:25 |
sdake | might as well schedule it i guess | 21:25 |
sdake | if you could leave comments in the meantime - would appreciate it | 21:25 |
inc0 | sdake: would you do me a favor and schedule it? since you have your cisco webex thingy | 21:25 |
sdake | yup | 21:26 |
inc0 | thank you good sir | 21:26 |
inc0 | there are few points to consider | 21:26 |
sdake | can ou leave comments in the meantime | 21:26 |
sdake | bradjones doesn't understand why entrypoint is there today (the synchronization points) | 21:26 |
sdake | bradjones comments will help him understand that | 21:26 |
inc0 | 1. how to put it into kolla-toolbox (or rather sub container) | 21:26 |
inc0 | 2. Create wait for tasks in every required spot | 21:27 |
inc0 | I'd focus on mariadb+keystone role first as it has to be first and it is quite complex in terms of orchiestration | 21:28 |
inc0 | we could drop few microservices too, things like registering users | 21:28 |
inc0 | since we'd have kolla-toolbox, we might as well use it | 21:29 |
sdake | we need the users registered | 21:29 |
inc0 | yes, with ansible | 21:29 |
inc0 | we have all the plumbing | 21:29 |
sdake | microcharts read cloud.yaml | 21:29 |
sdake | our plumbing does not | 21:29 |
inc0 | we'll run orch within container | 21:30 |
inc0 | we can modify tasks, but general idea is still sound | 21:30 |
sdake | right | 21:30 |
inc0 | depends how we want to do it | 21:30 |
inc0 | that's why we need design discussion | 21:30 |
sdake | see how smple that is - add in some sync points - wrap in a container - wrap container in a helm chart - bam HELM NATIVE helm install openstack | 21:32 |
inc0 | devil's in the details tho | 21:32 |
inc0 | I like idea of ansible driving k8s, I said that before | 21:33 |
inc0 | and I was proponent for ansible-based operators, which this kinda is | 21:33 |
inc0 | or however you call it | 21:33 |
sdake | k8s-controller is the nerd-term | 21:33 |
sdake | an operator combines several controllers into one | 21:34 |
inc0 | yeah, controller | 21:34 |
openstackgerrit | Merged openstack/kolla master: Fix kolla-toolbox build https://review.openstack.org/473332 | 21:38 |
sdake | kfox1111 if your alive - please review https://review.openstack.org/#/c/473588/1 | 21:39 |
*** jamesbenson has quit IRC | 21:39 | |
*** itlinux_ has joined #openstack-kolla | 21:44 | |
*** itlinux has joined #openstack-kolla | 21:45 | |
sbezverk | sdake: do you have PoC of this solution somewhere? | 21:46 |
hawi | i do have question regarding designate: is it safe to deploy it to existing environment? i mean, it needs to restart some other containers as well? | 21:46 |
sdake | sbezverk no - i asked brad to write the code about a week ago | 21:46 |
sdake | that is what has written | 21:47 |
sdake | he says it works for him | 21:47 |
sbezverk | sdake: I would think before that massive coding, there should be PoC or it might be a waste of time | 21:47 |
sdake | although as is, it looks racy | 21:47 |
sdake | there is no massive coding needed | 21:47 |
sbezverk | sdake: I have hard time to see it working unless some specialized kubernetes aware library is used. | 21:48 |
sdake | the one thing that patch is missing is sync points | 21:48 |
sbezverk | I mean ansible specialized library, just curling will not cut it.. | 21:48 |
sdake | you mean for the sync points? | 21:49 |
sdake | kubectl can tell you when a job is finished, no? | 21:49 |
openstackgerrit | Dan Ardelean proposed openstack/kolla-ansible master: Add Hyper-V role https://review.openstack.org/455684 | 21:49 |
Manheim | think i know where the issue was | 21:50 |
Manheim | i'll see if the passes this time | 21:50 |
sbezverk | sdake: of course not | 21:50 |
egonzalez | hawi what you mean by restart other containers? | 21:50 |
sbezverk | you just start it | 21:50 |
sdake | and check for its exit code | 21:50 |
sdake | in a loop | 21:51 |
sbezverk | and then you get control back right away | 21:51 |
sdake | you use kubectl to get the exit code of the job in a loop - when its done its done | 21:51 |
sbezverk | job might be hanging or crashing | 21:51 |
sbezverk | etc | 21:51 |
egonzalez | hawi yeah, will need to restart neutron containers | 21:51 |
sdake | if a job crashes, the deployment is cratered | 21:52 |
sdake | if a job hangs, we can set timeouts on the loop | 21:52 |
egonzalez | hawi designate will enable port_security extension in neutron, saw issues enabling it in existing networks | 21:52 |
hawi | egonzalez: i see, thanks. will try to find some safe time for enabling it. | 21:52 |
sdake | in either case, the deployment is cratered - as with the entyrpoint solution | 21:53 |
egonzalez | hawi what I saw is that existing networks not get dhcp, just a--disable-dhcp and--enable-dhcp on each subnet fixes the issue | 21:53 |
*** srwilkers has quit IRC | 21:54 | |
hawi | i do have already live setup, had some issues finding proper solution for dns and outside recursers pointing to designate, it's almost the last chance to apply it :( | 21:54 |
sbezverk | sdake: you are starting to build infra which is already exist in entrypoint, why reinvent the wheel? | 21:54 |
sdake | entrypoint is a dead upstream | 21:54 |
sdake | because of its governance structure, even if i wanted ot, as a cisco employee I cannot contribute to it | 21:55 |
sbezverk | it does what it supposed to do and does it well | 21:55 |
sdake | just like I can't contribute to ansible itself | 21:55 |
sdake | does it well is debateable | 21:55 |
sdake | we cannot base our solution on a dead upstream | 21:55 |
sbezverk | if you are not happy with macos thing, you do not throu yourself in re-writing it right? | 21:55 |
sdake | coreos has abandoned stackanetes | 21:56 |
sbezverk | sdake: other than baseless comments you cannot show any problem with entrypoint.. I am sorry but at least me I have not seen any issue after services piece got fixed. | 21:56 |
sdake | wit hthis model, we can do upgrades easily | 21:56 |
sdake | how is "DEAD UPSTREAM" baseless? | 21:56 |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla stable/ocata: Fix kolla deploy gates https://review.openstack.org/473605 | 21:57 |
sbezverk | sdake: where does it day that? | 21:57 |
sbezverk | sorry I meant say that | 21:57 |
sbezverk | sdake: I heard it so far only from you | 21:58 |
sdake | inc0 is entrypoint upstream dead? | 21:58 |
inc0 | sdake: dead upstream can be fixed by simple fork | 21:58 |
sdake | inc0 i cannot contribute ot that because of the attorneys..... | 21:58 |
*** jamesbenson has joined #openstack-kolla | 21:58 | |
inc0 | ok, but let's assume we get code entrypoint like that will not be dead | 21:59 |
sdake | right - so rewrite entrypoint | 21:59 |
sdake | or altenately, ust use this | 21:59 |
inc0 | yeah | 21:59 |
inc0 | both are hard tbh | 21:59 |
sdake | writing sync points in this ansible code is trivially easy | 21:59 |
inc0 | don't underestimate this | 22:00 |
inc0 | especially if we get to a world of dying ansible containers that needs to be restarted | 22:00 |
sdake | we dont have to write a genreal purpose tool | 22:00 |
sdake | (which entyrpoint is) | 22:00 |
inc0 | it's good idea, but by no means easy | 22:00 |
sdake | we have to write speicfic sycn points for specific things | 22:00 |
sdake | which is easy | 22:00 |
inc0 | but we already wrote sync points for entrypoint | 22:00 |
inc0 | we just need entrypoitn that is alive | 22:01 |
sbezverk | sdake: if it is that easy as you say how come you have not showed that in all its beauty ;) | 22:01 |
sdake | the sync points dont manage upgrades | 22:01 |
sdake | sbezverk its patchset #1 | 22:01 |
sdake | written in 2 hours | 22:01 |
sdake | i'm not really sure what you expect - miracles? :) | 22:01 |
inc0 | upgrades can be orchiestrated by ansible | 22:01 |
inc0 | we can do all of that | 22:01 |
sbezverk | sdake: but it is just lines without any logic in them ;) | 22:02 |
inc0 | sdake: right, hard part are still to come | 22:02 |
inc0 | right now it's equivalent of shellscript | 22:02 |
sbezverk | inc0: in shell we have waiting subroutines | 22:02 |
sdake | granted using the mysql module is a little ocmpicated in ansible | 22:02 |
sbezverk | here there are none so far.. | 22:02 |
inc0 | in ansible we can have them too | 22:02 |
sdake | granted writing a loop in ansible is hard waiting on a kubectl output | 22:02 |
inc0 | sbezverk: so idea in general is good | 22:03 |
*** jamesbenson has quit IRC | 22:03 | |
inc0 | I like this, I have really good experiences with ansible driving k8s | 22:03 |
sdake | those are the two things we ahve t odo | 22:03 |
*** salv-orlando has quit IRC | 22:03 | |
sdake | not write a genreal purpose orchestration replacement for the dead upstream entyrpoint | 22:03 |
inc0 | sdake: no it's not | 22:03 |
inc0 | we can actually write ansible module that will wait for k8s output | 22:03 |
*** salv-orlando has joined #openstack-kolla | 22:04 | |
sdake | that could be done - seems like overkill -but wfm :) | 22:04 |
sbezverk | inc0: maybe but when it comes for complex logic, I am not sure if ansible will be better than entrypoint calling k8s api functions to get status of k8s objects | 22:04 |
inc0 | sbezverk: it's easy | 22:04 |
inc0 | hard part is what's not in k8s | 22:05 |
sbezverk | inc0: listening to you guys, everything is easy ;) until it gets hard :) | 22:05 |
sdake | sbezverk w hae a team of people that are ansile experts | 22:05 |
*** jamesbenson has joined #openstack-kolla | 22:05 | |
sdake | everything is easy when you can ask an expert :) | 22:05 |
inc0 | sbezverk: it's not bad, entrypount is really really simple thing | 22:06 |
*** harlowja has joined #openstack-kolla | 22:06 | |
inc0 | ansible won't be any more difficult | 22:06 |
sbezverk | inc0: I see it as advatage | 22:06 |
sdake | sbezverk try upgrading with entyrpoint. | 22:06 |
inc0 | sbezverk: limitations of entrypoint are issye tho | 22:06 |
inc0 | sdake is right about upgrades | 22:06 |
inc0 | they're non-trivial | 22:06 |
inc0 | possible, but non-trivial | 22:07 |
sbezverk | inc0: that part I have not seen it yet.. | 22:07 |
sdake | this really isn't about deployment - any money could write that - this is about upgrades | 22:07 |
*** schwicht has quit IRC | 22:07 | |
*** salv-orlando has quit IRC | 22:08 | |
sdake | and with helm containing a container containint hte ansible, helm native baby- all the way | 22:08 |
inc0 | sdake: how about we focus with portdirect and srwilkers on writing brave new entrypoint | 22:08 |
sdake | this microchart ansible code is 2 weeks out max | 22:09 |
inc0 | and make ansible upgrade mechanism | 22:09 |
sdake | new entrypoint - who knows - several months | 22:09 |
inc0 | sdake: uhh...that's unlikely. | 22:09 |
sdake | with no real advantage | 22:09 |
*** dave-mccowan has quit IRC | 22:09 | |
inc0 | let me rephase it - ansible *will not be easy* | 22:09 |
*** jamesbenson has quit IRC | 22:09 | |
inc0 | it looks easy on paper, but kolla-ansible took us a year to make | 22:10 |
sdake | easier than an entrypoint that is upgradable | 22:10 |
sdake | upgrades mut happen in a certain order | 22:10 |
sdake | entrypoint has no ordering | 22:10 |
inc0 | ofc it has | 22:10 |
*** spsurya has quit IRC | 22:10 | |
inc0 | dependencies determine order | 22:10 |
inc0 | ordering is one and only reason entrypoint exists | 22:10 |
sdake | entrypoint is an orchestration tool | 22:11 |
sdake | which is more powerful, entrypoint or ansible? | 22:11 |
inc0 | orchiestration is glorified word for dependnecy management | 22:11 |
*** Manheim has quit IRC | 22:11 | |
kolla-slack | <srwilkers> Oh hello | 22:13 |
sdake | give it 2 weeks | 22:13 |
sdake | will be done working and gating | 22:13 |
*** rwsu has joined #openstack-kolla | 22:13 | |
sdake | what more could you ask for? | 22:13 |
inc0 | sdake: right, I'll be happy to see that;) | 22:13 |
inc0 | srwilkers, we need to move spaceship at full speed ahead | 22:14 |
sbezverk | sdake: as a generic tool ansible is more powerfull, but we do not need all its generic beauty, we need lean and mean dependency thing | 22:14 |
sdake | sbezverk this stuff needs to be upgradeable | 22:14 |
sdake | entyrpoitn will never be lean and mean with ugprades in mind | 22:14 |
sbezverk | swilkers: if you have something to share in term of development, please let me know.. | 22:15 |
*** fooliouno has quit IRC | 22:15 | |
SamYaple | jgriffith: youre all good. im going to add you to the core group for loci | 22:16 |
kolla-slack | <srwilkers> Yeah. Driving currently so hard to respond. Will be in front of a keyboard in 30 | 22:16 |
*** manheim has joined #openstack-kolla | 22:19 | |
sdake | we have atleast 3 workflows to handle - reconfigure, ugprade, deploy | 22:19 |
sdake | entrypoint could probably be made to do these things | 22:19 |
sdake | ansible can do them faster | 22:20 |
sdake | i have tech support duty for my parents, gotta jet - review the patch - be constructive - keep an open mind | 22:21 |
sdake | sbezverk also keep in mind ansible -> container -> helm -> HELM NATIVE | 22:22 |
sbezverk | sdake: a bit strange sequence, in your case ansible drives helm and then helm launches releases. | 22:23 |
sdake | sbezverk nah - ansible goes into a container - container goes into helm - helm native | 22:23 |
sdake | i htink you misunderstood me | 22:23 |
portdirect | o/ | 22:23 |
inc0 | sbezverk: helm creates job based of kolla-toolbox, kolla-toolbox runs ansible play that builds rest of resrouces | 22:23 |
inc0 | hey portdirect | 22:23 |
* portdirect reading scrollback :) | 22:23 | |
inc0 | we're having orchiestration discussion and 2 competing ideas | 22:24 |
portdirect | i pick the 3rd one | 22:24 |
* portdirect argues point bravely | 22:24 | |
sbezverk | portdirect: why am I not surprised ;) | 22:24 |
portdirect | what are they? | 22:24 |
inc0 | portdirect: you're in luck, it exists | 22:24 |
* portdirect regrets joke again | 22:25 | |
inc0 | 1. entrypoint rewritten in spaceship | 22:25 |
inc0 | 2. ansible in container driving rest of helm | 22:25 |
sdake | sadly 2 only works with microcharts inc0 | 22:25 |
inc0 | 3. https://github.com/helm/rudder-appcontroller this thing | 22:25 |
sdake | which is good, because we aren't shipping spaceship we are shipping kolla-kubernetes ;) | 22:25 |
inc0 | yeah, dependencies within same chart still have to be addressed | 22:26 |
portdirect | s/spaceship/spaceport :P | 22:26 |
inc0 | well | 22:26 |
inc0 | yeah | 22:27 |
inc0 | anyway | 22:27 |
inc0 | competing ideas | 22:27 |
portdirect | ok - so I'll need some more context, but (and I'm obviously biased) but my gut says 1, and 3 combined have most utility? | 22:27 |
manheim | heh, it passed | 22:27 |
*** emccormick has quit IRC | 22:27 | |
inc0 | 3 isn't something I'd put my money on | 22:28 |
inc0 | but that's my biases | 22:28 |
sbezverk | inc0: that app controller is what Mirantis was pushing? | 22:28 |
inc0 | ansible is good in stuff like quorum building, ugprades | 22:28 |
inc0 | these won't be trivial without external (or semi external) orch tool | 22:28 |
inc0 | sbezverk: yuo | 22:28 |
inc0 | yup | 22:28 |
sbezverk | inc0: I see, curious if anybody played with that thing.. | 22:29 |
sdake | path of least resistance is to implement orchestration in an orchestration engine with a rockin upstream - which is ansible | 22:29 |
inc0 | sdake: again, it won't be as easy as you think, but feel free to prove me wrong | 22:30 |
inc0 | to me 1 and 2 are equally appealing | 22:30 |
*** egonzalez has quit IRC | 22:30 | |
portdirect | whats missing from k8s entrypoint thats needed here | 22:31 |
inc0 | portdirect: stuff like making mariadb quorum | 22:31 |
portdirect | (I have a list of things i want - but not sure if they are the same pain points.) | 22:31 |
inc0 | would be much easier | 22:31 |
portdirect | ah - so having the ability to work as a sidecar doing things like leader election? | 22:32 |
inc0 | yeah | 22:32 |
inc0 | upgrades too | 22:32 |
*** goldyfruit has quit IRC | 22:32 | |
inc0 | that kind of fun | 22:32 |
inc0 | also ansible has a lot of utility built into it already | 22:33 |
inc0 | waiting for stuff and such | 22:33 |
portdirect | ++ | 22:33 |
inc0 | I think some hybrid solution might actually be best | 22:33 |
portdirect | (oh -- on the waiting, but ++ on the utility :) | 22:33 |
portdirect | ^^ yup, I think theres not gonna be a single tool that does (r perhaps should) fit the job here | 22:33 |
portdirect | I'm a big fan of a lot of very simple tiny tools over a multi-tool | 22:34 |
portdirect | and ansible def has a space there | 22:34 |
portdirect | hows the ansible integration with k8s these days? | 22:35 |
inc0 | not horrible | 22:35 |
inc0 | but definetly could be better | 22:35 |
inc0 | ansible is really good templating engine | 22:35 |
inc0 | I think my favorite solution would be: use dependency-manager (let's stop using word entrypoint plz) | 22:39 |
inc0 | for 90% of stuff when it's easy | 22:40 |
inc0 | and use ansible sidecar for 10% that is hard | 22:40 |
kolla-slack | <srwilkers> Yes. Entry points been used generally in this discussion but it really comes down to dependency checking that makes sense | 22:40 |
inc0 | or ansible-driven jobs rather than sidecar | 22:40 |
inc0 | it's glorified sleep until deps are met | 22:40 |
inc0 | that's all it is | 22:40 |
kolla-slack | <srwilkers> We need to distinguish between dependency checking and orchestration here | 22:40 |
*** n0isyn0i1e has joined #openstack-kolla | 22:42 | |
inc0 | hey, I need to step out to do some grocery shopping to avoid anger of hungry and tired wife, can we continue discussion in like 30? | 22:42 |
inc0 | or tomorrow | 22:42 |
inc0 | bbiaf | 22:43 |
kolla-slack | <srwilkers> Hey me too. I'm chatting while grocery shopping now | 22:43 |
*** n0isyn0ise has quit IRC | 22:45 | |
*** rhallisey has joined #openstack-kolla | 22:49 | |
*** itlinux has quit IRC | 22:50 | |
*** itlinux_ has quit IRC | 22:50 | |
inc0 | back | 23:10 |
*** manheim has quit IRC | 23:10 | |
jascott1 | in black? | 23:12 |
inc0 | I was trying to reply with another rhyme, but only thing I could come up with was crack, and that wasn't greatest word to reply | 23:13 |
inc0 | I'd be horrible rapper | 23:13 |
jascott1 | most of them are | 23:13 |
jascott1 | :D | 23:13 |
inc0 | yeah, I guess, I'll keep it as backup career | 23:14 |
inc0 | guys need critical +2s on these two changes https://review.openstack.org/#/c/473580/ https://review.openstack.org/#/c/473605/1 | 23:16 |
inc0 | in this order | 23:16 |
*** schwicht has joined #openstack-kolla | 23:18 | |
*** ducttape_ has quit IRC | 23:31 | |
Daviey | 6.1006.80 | 23:37 |
*** schwicht has quit IRC | 23:43 | |
*** eaguilar has joined #openstack-kolla | 23:43 | |
*** schwicht has joined #openstack-kolla | 23:45 | |
*** schwicht has quit IRC | 23:47 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!