| *** sdake has quit IRC | 00:00 | |
| *** sdake has joined #openstack-kolla | 00:02 | |
| *** sdake has quit IRC | 00:03 | |
| *** jistr has quit IRC | 00:03 | |
| *** sdake has joined #openstack-kolla | 00:03 | |
| *** jistr has joined #openstack-kolla | 00:07 | |
| *** williamcaban has quit IRC | 00:08 | |
| *** dave-mcc_ has joined #openstack-kolla | 00:08 | |
| *** dave-mccowan has quit IRC | 00:10 | |
| *** schwicht has joined #openstack-kolla | 00:11 | |
| *** dave-mcc_ has quit IRC | 00:13 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP memcached operator https://review.openstack.org/418073 | 00:16 |
|---|---|---|
| *** Pavo has joined #openstack-kolla | 00:23 | |
| *** hfu has joined #openstack-kolla | 00:24 | |
| *** sayantani01 has joined #openstack-kolla | 00:26 | |
| *** salv-orl_ has quit IRC | 00:27 | |
| *** sdake has quit IRC | 00:28 | |
| *** adrian_otto has joined #openstack-kolla | 00:30 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Mariadb service deployment https://review.openstack.org/412237 | 00:32 |
| *** dave-mccowan has joined #openstack-kolla | 00:40 | |
| *** liyifeng has joined #openstack-kolla | 00:50 | |
| *** dave-mccowan has quit IRC | 00:51 | |
| *** adrian_otto has quit IRC | 00:52 | |
| *** schwicht has quit IRC | 00:53 | |
| *** hfu has quit IRC | 00:56 | |
| *** zhubingbing has joined #openstack-kolla | 01:10 | |
| *** williamcaban has joined #openstack-kolla | 01:23 | |
| *** tovin07 has joined #openstack-kolla | 01:27 | |
| *** salv-orlando has joined #openstack-kolla | 01:28 | |
| *** tonanhngo has joined #openstack-kolla | 01:28 | |
| *** tonanhngo has quit IRC | 01:30 | |
| *** sacharya has joined #openstack-kolla | 01:31 | |
| *** hfu has joined #openstack-kolla | 01:31 | |
| *** goldyfruit has joined #openstack-kolla | 01:33 | |
| *** caowei has joined #openstack-kolla | 01:33 | |
| *** duonghq has joined #openstack-kolla | 01:35 | |
| duonghq | morning guys | 01:35 |
| *** salv-orlando has quit IRC | 01:35 | |
| duonghq | morning sbezverk portdirect | 01:39 |
| *** l4yerffeJ_ has joined #openstack-kolla | 01:53 | |
| *** l4yerffeJ has quit IRC | 01:53 | |
| openstackgerrit | Duong Ha-Quang proposed openstack/kolla-kubernetes: [WIP] Helm Ceph OSD https://review.openstack.org/418294 | 01:55 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Mariadb service deployment https://review.openstack.org/412237 | 01:57 |
| sbezverk | duonghq: good morning | 01:58 |
| openstackgerrit | Duong Ha-Quang proposed openstack/kolla-kubernetes: Helm Ceph OSD Bootstrap https://review.openstack.org/410696 | 01:59 |
| *** sayantani01 has quit IRC | 02:03 | |
| openstackgerrit | Duong Ha-Quang proposed openstack/kolla-kubernetes: Helm Ceph OSD Bootstrap https://review.openstack.org/410696 | 02:06 |
| *** sayantani01 has joined #openstack-kolla | 02:08 | |
| openstackgerrit | Duong Ha-Quang proposed openstack/kolla-kubernetes: [WIP] Helm Ceph OSD https://review.openstack.org/418294 | 02:20 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Mariadb service deployment https://review.openstack.org/412237 | 02:21 |
| *** stevemar has quit IRC | 02:21 | |
| *** stevemar has joined #openstack-kolla | 02:21 | |
| duonghq | sbezverk, | 02:22 |
| duonghq | around? | 02:22 |
| *** hfu has quit IRC | 02:28 | |
| *** sayantani01 has quit IRC | 02:29 | |
| *** salv-orlando has joined #openstack-kolla | 02:31 | |
| zhubingbing | hello guys | 02:34 |
| *** salv-orlando has quit IRC | 02:36 | |
| *** hfu has joined #openstack-kolla | 02:36 | |
| duonghq | sup zhubingbing | 02:37 |
| duonghq | kfox1111, around? | 02:37 |
| *** hfu has quit IRC | 02:38 | |
| openstackgerrit | Duong Ha-Quang proposed openstack/kolla-kubernetes: Helm Ceph Initial Mon https://review.openstack.org/410641 | 02:50 |
| openstackgerrit | Duong Ha-Quang proposed openstack/kolla-kubernetes: Helm Ceph Mon https://review.openstack.org/410660 | 02:55 |
| Jeffrey4l_ | sbezverk, kolla-k8s is tagged, check https://review.openstack.org/420380 | 02:56 |
| *** dave-mccowan has joined #openstack-kolla | 03:02 | |
| *** tonanhngo has joined #openstack-kolla | 03:04 | |
| *** tonanhngo has quit IRC | 03:05 | |
| *** dave-mccowan has quit IRC | 03:06 | |
| *** goldyfruit has quit IRC | 03:07 | |
| openstackgerrit | Duong Ha-Quang proposed openstack/kolla-kubernetes: Fix typo in keystone svc https://review.openstack.org/420479 | 03:09 |
| *** severion has joined #openstack-kolla | 03:10 | |
| *** jmontleon has quit IRC | 03:12 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Fix gate in iax-ord node when using libvirt 2.0 https://review.openstack.org/420480 | 03:13 |
| *** jmontleon has joined #openstack-kolla | 03:14 | |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible: Move ceph precheck into its own role https://review.openstack.org/416066 | 03:15 |
| *** Jeffrey4l_ has quit IRC | 03:21 | |
| *** Jeffrey4l has joined #openstack-kolla | 03:21 | |
| *** jmontleon has quit IRC | 03:21 | |
| *** jmontleon has joined #openstack-kolla | 03:27 | |
| openstackgerrit | Duong Ha-Quang proposed openstack/kolla-kubernetes: Helm Ceph Mon https://review.openstack.org/410660 | 03:30 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Bump epel-release to 7.9 https://review.openstack.org/420484 | 03:30 |
| *** dave-mccowan has joined #openstack-kolla | 03:32 | |
| *** williamcaban has quit IRC | 03:32 | |
| *** salv-orlando has joined #openstack-kolla | 03:32 | |
| *** salv-orlando has quit IRC | 03:36 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Install epel-release-latest rather than certain version https://review.openstack.org/420484 | 03:37 |
| *** williamcaban has joined #openstack-kolla | 03:38 | |
| *** tonanhngo has joined #openstack-kolla | 03:44 | |
| *** prameswar has joined #openstack-kolla | 03:44 | |
| *** tonanhngo has quit IRC | 03:45 | |
| *** dave-mccowan has quit IRC | 03:45 | |
| openstackgerrit | jianyi proposed openstack/kolla-ansible: Add zun ansible role https://review.openstack.org/416179 | 03:56 |
| *** williamcaban has quit IRC | 04:05 | |
| *** mdnadeem has joined #openstack-kolla | 04:05 | |
| *** liyifeng has quit IRC | 04:12 | |
| *** adrian_otto has joined #openstack-kolla | 04:23 | |
| *** adrian_otto has quit IRC | 04:24 | |
| *** williamcaban has joined #openstack-kolla | 04:24 | |
| *** adrian_otto has joined #openstack-kolla | 04:24 | |
| *** tonanhngo has joined #openstack-kolla | 04:30 | |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible: Move mariadb precheck into its own role https://review.openstack.org/415587 | 04:30 |
| *** tonanhngo has quit IRC | 04:31 | |
| *** adrian_otto has quit IRC | 04:33 | |
| *** salv-orlando has joined #openstack-kolla | 04:33 | |
| *** hfu has joined #openstack-kolla | 04:35 | |
| *** salv-orlando has quit IRC | 04:38 | |
| *** williamcaban has quit IRC | 04:39 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Install epel-release-latest rather than certain version https://review.openstack.org/420484 | 04:39 |
| *** adrian_otto has joined #openstack-kolla | 04:45 | |
| *** caowei has quit IRC | 04:52 | |
| *** sp__ has joined #openstack-kolla | 04:57 | |
| *** tonanhngo has joined #openstack-kolla | 05:04 | |
| *** caowei has joined #openstack-kolla | 05:05 | |
| *** tonanhngo has quit IRC | 05:05 | |
| *** unicell has joined #openstack-kolla | 05:11 | |
| *** adrian_otto has quit IRC | 05:11 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible: Try pull docker images from tarballs.o.o https://review.openstack.org/413720 | 05:14 |
| *** severion has quit IRC | 05:18 | |
| *** v1k0d3n has quit IRC | 05:18 | |
| *** sgordon has quit IRC | 05:31 | |
| *** salv-orlando has joined #openstack-kolla | 05:34 | |
| *** v1k0d3n has joined #openstack-kolla | 05:35 | |
| *** sacharya has quit IRC | 05:38 | |
| *** salv-orlando has quit IRC | 05:38 | |
| *** sacharya has joined #openstack-kolla | 05:39 | |
| *** sgordon has joined #openstack-kolla | 05:39 | |
| *** tonanhngo has joined #openstack-kolla | 05:44 | |
| *** tonanhngo has quit IRC | 05:46 | |
| *** v1k0d3n has quit IRC | 05:48 | |
| *** v1k0d3n has joined #openstack-kolla | 05:49 | |
| *** skramaja has joined #openstack-kolla | 05:56 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible: Try pull docker images from tarballs.o.o https://review.openstack.org/413720 | 06:01 |
| *** tonanhngo has joined #openstack-kolla | 06:04 | |
| *** tonanhngo has quit IRC | 06:05 | |
| *** adrian_otto has joined #openstack-kolla | 06:06 | |
| *** prameswar has quit IRC | 06:10 | |
| *** adrian_otto has quit IRC | 06:14 | |
| *** adrian_otto has joined #openstack-kolla | 06:16 | |
| *** adrian_otto has quit IRC | 06:16 | |
| *** adrian_otto has joined #openstack-kolla | 06:21 | |
| *** tonanhngo has joined #openstack-kolla | 06:22 | |
| *** unicell has quit IRC | 06:22 | |
| *** tonanhngo has quit IRC | 06:23 | |
| *** satyar has joined #openstack-kolla | 06:23 | |
| satyar | Hi jeffrey4l | 06:26 |
| Jeffrey4l | hey satyar | 06:27 |
| satyar | i see the problem at https://review.openstack.org/#/c/404944/22/ansible/library/rabbitmq_hostname.py@67 | 06:27 |
| satyar | i am using the if loop | 06:28 |
| satyar | and in the paste you used the for loop ;) | 06:28 |
| Jeffrey4l | oops. sorry. | 06:28 |
| satyar | so override is possible in your case | 06:28 |
| Jeffrey4l | got what u means. you are correct. | 06:28 |
| satyar | as if loops runs once so override is not possible | 06:28 |
| satyar | :) | 06:28 |
| Jeffrey4l | satyar, could u or will u change this purge python module to ansible module? | 06:29 |
| satyar | yes | 06:29 |
| Jeffrey4l | cool. | 06:29 |
| satyar | i will be fixing all other comments | 06:30 |
| Jeffrey4l | moreover, i think you can get the address by ansible facts rather than "ip addr" command. | 06:30 |
| satyar | using ip addr just to check if the ip belongs to the host | 06:31 |
| *** adrian_otto has quit IRC | 06:33 | |
| Jeffrey4l | ok | 06:33 |
| *** salv-orlando has joined #openstack-kolla | 06:35 | |
| *** salv-orlando has quit IRC | 06:39 | |
| *** tyrola has quit IRC | 06:47 | |
| *** sp__ has quit IRC | 06:48 | |
| *** tyrola has joined #openstack-kolla | 06:49 | |
| *** mnaser has quit IRC | 06:53 | |
| *** prameswar has joined #openstack-kolla | 07:02 | |
| *** mnaser has joined #openstack-kolla | 07:09 | |
| *** YuYangWang has joined #openstack-kolla | 07:14 | |
| *** nrado has joined #openstack-kolla | 07:16 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Removed unnecessary utf-8 encoding https://review.openstack.org/418837 | 07:21 |
| *** hogepodge_ has joined #openstack-kolla | 07:24 | |
| *** mdnadeem has quit IRC | 07:27 | |
| *** tonanhngo has joined #openstack-kolla | 07:27 | |
| *** tonanhngo has quit IRC | 07:29 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible: Pull kolla images from tarballs.o.o https://review.openstack.org/413720 | 07:32 |
| *** gfidente has joined #openstack-kolla | 07:33 | |
| *** gfidente has joined #openstack-kolla | 07:33 | |
| *** skramaja has quit IRC | 07:33 | |
| *** pcaruana has joined #openstack-kolla | 07:34 | |
| *** skramaja has joined #openstack-kolla | 07:35 | |
| *** salv-orlando has joined #openstack-kolla | 07:36 | |
| *** hogepodge_ has quit IRC | 07:37 | |
| *** salv-orlando has quit IRC | 07:40 | |
| *** mdnadeem has joined #openstack-kolla | 07:42 | |
| berendt | morning guys | 07:43 |
| duonghq | morning berendt | 07:45 |
| duonghq | berendt, sorry, but are you working on kolla-k8s? | 07:45 |
| berendt | duonghq: actually not, i think i will join in the future | 07:46 |
| duonghq | berendt, ok, thanks | 07:46 |
| openstackgerrit | chen.xing proposed openstack/kolla: [WIP] Add vitrage images https://review.openstack.org/417593 | 07:55 |
| openstackgerrit | Zeyu Zhu proposed openstack/kolla: Remove Ansible and Globals.yml in diag script https://review.openstack.org/420552 | 08:00 |
| *** zhurong has joined #openstack-kolla | 08:05 | |
| openstackgerrit | Duong Ha-Quang proposed openstack/kolla-kubernetes: Helm Ceph RGW https://review.openstack.org/418835 | 08:16 |
| openstackgerrit | zhubingbing proposed openstack/kolla-ansible: [WIP] Add fluentd role https://review.openstack.org/407392 | 08:20 |
| zhubingbing | sup berendt | 08:20 |
| *** tonanhngo has joined #openstack-kolla | 08:21 | |
| *** tonanhngo has quit IRC | 08:22 | |
| *** shardy has joined #openstack-kolla | 08:24 | |
| *** obre has joined #openstack-kolla | 08:32 | |
| *** salv-orlando has joined #openstack-kolla | 08:37 | |
| *** salv-orlando has quit IRC | 08:41 | |
| *** matrohon has joined #openstack-kolla | 08:42 | |
| *** egonzalez has joined #openstack-kolla | 08:44 | |
| *** tonanhngo has joined #openstack-kolla | 08:46 | |
| *** tonanhngo has quit IRC | 08:47 | |
| *** magnumbonum has joined #openstack-kolla | 08:50 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Install epel-release-latest rather than certain version https://review.openstack.org/420484 | 08:51 |
| *** sdake has joined #openstack-kolla | 08:51 | |
| sp_ | Jeffrey4l: please check https://review.openstack.org/#/c/419450/ | 08:51 |
| *** sdake has quit IRC | 08:51 | |
| *** sdake has joined #openstack-kolla | 08:52 | |
| egonzalez | sdake: congrats with the election :) | 08:53 |
| sdake | egonzalez its more like a service - like what a core reviewer does - but thanks :) | 08:53 |
| Jeffrey4l | sp_, i think you need add coverage into test-requirements.txt file | 08:55 |
| Jeffrey4l | then `tox -ecoverage` works in a new env. | 08:55 |
| sp_ | Jeffrey4l: BTB | 08:56 |
| duonghq | sup sdake | 08:59 |
| *** sdake_ has joined #openstack-kolla | 09:00 | |
| duonghq | sdake_, you just wakeup or have not slept yet? | 09:00 |
| sdake_ | fell asleep around 5pm | 09:01 |
| sdake_ | its 2am | 09:01 |
| sdake_ | groan | 09:01 |
| sdake_ | working really hard on a project friday/saturday | 09:01 |
| duonghq | ya | 09:01 |
| *** sdake has quit IRC | 09:02 | |
| sdake_ | https://twitter.com/steven_dake/status/820752300289339393 | 09:02 |
| sdake_ | the phto doesn't really do justice to the amount of work involved | 09:02 |
| sdake_ | there were 5 lights - one of htem weighed over 200 poudsn | 09:02 |
| *** prameswar has quit IRC | 09:02 | |
| sdake_ | took them out and replaced them with led ceiling lights | 09:03 |
| sdake_ | (2 instead) | 09:03 |
| duonghq | ah, cool | 09:03 |
| sdake_ | my wife speculates the previous owners were growing marijuana in the garage | 09:03 |
| sdake_ | the wiring in there was crazy | 09:03 |
| sdake_ | anyway i took all the crap down and put in something sane | 09:03 |
| duonghq | how much does the LED cost you? | 09:04 |
| sdake_ | 100 bucks per lamp - 40 bucks per motion sensor | 09:04 |
| portdirect | lol - maybe they just had a home lab bigger than yours sdake :) oh also moring kolla o/ | 09:04 |
| duonghq | oh, | 09:05 |
| duonghq | portdirect, nice to see you | 09:05 |
| sdake_ | https://twitter.com/steven_dake/status/820753240966840320 | 09:05 |
| sdake_ | sup portdirect | 09:05 |
| sdake_ | portdirect the wiring was on teh roof but only on 1 5 amp circuit | 09:05 |
| duonghq | sdake_, what is your LED power rate? | 09:05 |
| sdake_ | since its a flat roof | 09:06 |
| sdake_ | duonghq i am sure i went from about 1500 watts to about 80 :) | 09:06 |
| sdake_ | the big thing is i got rid of the cfcs | 09:06 |
| sdake_ | those things area pain in the ass | 09:06 |
| duonghq | sdake_, you mean 80W/LED? | 09:06 |
| sdake_ | something like that - not sure | 09:06 |
| sdake_ | its pretty bright | 09:06 |
| sdake_ | moment | 09:06 |
| duonghq | portdirect, do you do vagrant up recently? | 09:06 |
| portdirect | duonghq: yes | 09:07 |
| duonghq | sdake_, ya, LEDs make me surprise in 1st on | 09:07 |
| portdirect | set up a new cluster about 9 hours ago | 09:07 |
| sdake_ | 41 watts | 09:07 |
| duonghq | portdirect, so, something mess up with my network :(, it cannot do system upgrade | 09:07 |
| sdake_ | duonghq it has a strip of 100 leds in it | 09:07 |
| sdake_ | https://www.amazon.com/Lithonia-Lighting-SBL4-LP840-Wraparound/dp/B01359IGMG/ref=sr_1_2?s=hi&ie=UTF8&qid=1484557623&sr=1-2-spons&keywords=lithonia+lighting&psc=1 | 09:08 |
| duonghq | sdake_, ah, got it, | 09:08 |
| openstackgerrit | Surya Prakash Singh proposed openstack/kolla-ansible: Cover environment for kolla-ansible https://review.openstack.org/419450 | 09:08 |
| sdake_ | duonghq more accurate strip of 50 | 09:08 |
| duonghq | in Vietnam, t5 or t8 tube is much more common | 09:08 |
| sdake_ | 2 strips of 50 | 09:09 |
| portdirect | duonghq: do you mean on the hosts running vagrant or the vms? | 09:09 |
| *** matrohon has quit IRC | 09:09 | |
| sdake_ | duonghq ya - i got rid of the t8 tube lamps | 09:09 |
| duonghq | portdirect, the vms (in ansible playbook) | 09:09 |
| sdake_ | they may ahve been t12 | 09:09 |
| sdake_ | duonghq in the us, you can't buy the lights anymore | 09:09 |
| sdake_ | so they had to go | 09:09 |
| duonghq | sdake_, I mean LED in t-xx packaging | 09:10 |
| sdake_ | they make t8 leds, but they are still a pita to deal with | 09:10 |
| portdirect | duonghq: I've had that happen a few times with my machines and a reboot of the node running vagrant has always sorted it out (not the answer you want to hear I expect) | 09:10 |
| portdirect | duonghq: I've had my vms running centos timout while running yum before | 09:11 |
| duonghq | portdirect, I see somehow the no_proxy setting cannot get its way to vms when ansible is running (but not when I ssh to the vms) | 09:11 |
| sdake_ | these are 50k hours of lighting time -then replace them | 09:11 |
| sdake_ | 50k hours is a long long time | 09:11 |
| sdake_ | i'll probably be 60 before i have to deal with them again ;-) | 09:11 |
| duonghq | sdake_, sure, I'm waiting for the day one of my LED die | 09:11 |
| sdake_ | duonghq we have replaced all the household with leds over the last couple of years | 09:12 |
| sdake_ | we have a mismatch of led lights | 09:12 |
| duonghq | wonder how they die | 09:12 |
| portdirect | duonghq: could you rephrase that? Its morning for me too - though not as early as sdakes :) | 09:12 |
| sdake_ | we have one out back we leave on all the time andin front | 09:12 |
| sdake_ | they have been on ocntinuously since we moved in | 09:12 |
| sdake_ | they are sitll running | 09:12 |
| sdake_ | 5 years | 09:12 |
| portdirect | duonghq: though i need to warn you that I'm not that familar with the proxy setup as I've never needed to use it | 09:13 |
| duonghq | portdirect, I think it's the trouble I gotta | 09:13 |
| duonghq | portdirect, it's happy with normal repo (through proxy) | 09:13 |
| duonghq | but not with my local repo | 09:13 |
| duonghq | somehow the no_proxy settings do not go in when ansible playbook is triggered | 09:14 |
| sp_ | Jeffrey4l: working perfect now ${PYTHON:-python} -m subunit.run discover ${OS_TEST_PATH:-.} --load-list /tmp/tmpgRGevs | 09:15 |
| sp_ | Coverage.py warning: Module kolla-ansible was never imported. | 09:15 |
| sp_ | Coverage.py warning: No data was collected. | 09:15 |
| sp_ | Coverage.py warning: Module kolla-ansible was never imported. | 09:15 |
| sp_ | Coverage.py warning: No data was collected. | 09:15 |
| sp_ | Coverage.py warning: Module kolla-ansible was never imported. | 09:15 |
| sp_ | Coverage.py warning: No data was collected. | 09:15 |
| sp_ | Coverage.py warning: Module kolla-ansible was never imported. | 09:15 |
| sp_ | Coverage.py warning: No data was collected. | 09:15 |
| sp_ | Ran 51 tests in 0.179s | 09:15 |
| sp_ | PASSED (id=0) | 09:15 |
| sp_ | No data to report. | 09:15 |
| sp_ | ______________________________________________________________ summary _______________________________________________________________ | 09:15 |
| sp_ | cover: commands succeeded | 09:15 |
| sp_ | congratulations :) | 09:15 |
| sp_ | Jeffrey4l: thanks | 09:15 |
| sdake_ | sp_ note I noticed you submitted ( ithink it was you) a patch for kolla-kubernetes wihch broke the tagging there | 09:15 |
| Jeffrey4l | sp_, please use a paste service when copy long logs ;) | 09:15 |
| duonghq | sdake_, I see it's dimmable, is there any extra controller for it? | 09:15 |
| sdake_ | sp_ so i had to revert part of the patch | 09:15 |
| sdake_ | duonghq not sure - i haven't dimmed it | 09:16 |
| sdake_ | duonghq if there is, you probably have to buy one specially | 09:16 |
| duonghq | btw, this is 1st time I see Lithonia brand | 09:17 |
| portdirect | duonghq: looking at the playbook, I *think* I may see your issue (again I'm kind blind here) | 09:17 |
| duonghq | portdirect, something mess up with SSL, so I cannot use normal repo as usual, | 09:17 |
| duonghq | portdirect, I'm trying print out the env output | 09:17 |
| portdirect | you may need to so some work here: https://github.com/att-comdev/halcyon-kubernetes/blob/master/kube-deploy/roles/upgrade-os/tasks/centos.yml | 09:18 |
| portdirect | but if you could dump logs that would be great :) | 09:18 |
| sp_ | sdake_: sorry, but which patch you are talking about ? | 09:18 |
| sdake_ | sp_ moemnt | 09:18 |
| duonghq | portdirect, I patch it, it recognized my local repo, but cannot download metadata from it (I try curl manually) | 09:19 |
| sp_ | Jeffrey4l: Sorry I will do the same from next time. | 09:19 |
| duonghq | portdirect, sorry, brb in about 5mins | 09:19 |
| portdirect | duonghq: thats whats giving you the ssl issue? | 09:19 |
| portdirect | duonghq: np | 09:19 |
| sdake_ | sp_ i guess it was Tim Lane: d609177328465d50c63df210c76aaf679627c1f3 | 09:20 |
| sp_ | sdake_: Thanks for pointing out , I will take of care such thing from my side too. | 09:21 |
| sdake_ | sp_ cool thanks :) | 09:22 |
| duonghq | portdirect, our network is slow now, so repo cannot work normally, and proxy do something with SSl, so I cannot access some HTTPS website | 09:22 |
| duonghq | So I use my local repo (sync about 1 week ago) | 09:23 |
| portdirect | ok - can you curl that from a vm? | 09:23 |
| *** JeroenBo has joined #openstack-kolla | 09:23 | |
| portdirect | or is that giving error too? | 09:23 |
| duonghq | portdirect, I can curl repo metadata from VMs | 09:23 |
| *** JeroenBo has left #openstack-kolla | 09:24 | |
| portdirect | but yum upgrade fails? | 09:24 |
| duonghq | yup | 09:24 |
| portdirect | can you post log of that? | 09:24 |
| duonghq | [Errno 12] Timeout on http://10.164.180.104:2323/centos_7_epel/repodata/repomd.xml: (28, 'Operation too slow. Less than 1000 bytes/sec transferred the last 30 seconds') | 09:24 |
| sdake_ | duonghq are you able to ping 8.8.8.8 from vms? | 09:24 |
| duonghq | portdirect, http://paste.openstack.org/show/595007/ | 09:25 |
| duonghq | sdake_, I cannot ping 8.8.8.8 from any PC in our company network | 09:25 |
| sdake_ | duonghq how about can yo uping an addres syou can normally ping | 09:26 |
| sdake_ | duonghq the goal is to check the vm has network connecitvity to the external world | 09:26 |
| portdirect | ok can you post your .repo file? I;m a bit confused as you are talking about ssl issues, but that line you posed is http? | 09:26 |
| sdake_ | like pick your favorate website and ping that | 09:26 |
| sdake_ | duonghq that repo times out for me too | 09:26 |
| duonghq | sdake_, I curl it for check | 09:26 |
| sdake_ | duonghq is that na internal repo? | 09:27 |
| sdake_ | na/an | 09:27 |
| duonghq | portdirect, the ssl issues appears only with external repo, but not my local repo | 09:27 |
| duonghq | sdake_, yup | 09:27 |
| duonghq | so I posted env output too | 09:27 |
| sdake_ | duonghq ok - got it- the problem is your not running ssl on your internal repo | 09:27 |
| sdake_ | duonghq and I suspect not mirring it correclty | 09:28 |
| *** caowei has quit IRC | 09:28 | |
| *** JeroenBo has joined #openstack-kolla | 09:28 | |
| duonghq | sdake_, I used it fine for building my docker image | 09:28 |
| sdake_ | but ya - try curling the repomod.xml file | 09:28 |
| duonghq | curling the repomd.xml is ok | 09:28 |
| portdirect | can you so that with --verbose and dump log | 09:28 |
| portdirect | (and command) | 09:28 |
| sdake_ | sp_ - your tryign to use proxy server with kolla-kubernetes? | 09:29 |
| *** JeroenBo has left #openstack-kolla | 09:29 | |
| duonghq | sdake_, portdirect hmm, yum update failed too | 09:29 |
| duonghq | just curl work | 09:30 |
| sdake_ | duonghq I think what your after is to customize the repo halcyon uses | 09:30 |
| sp_ | sdake_: sorry I didn't get you. | 09:30 |
| duonghq | sdake_, ya, I'm adding local repo to the ansible playbook | 09:30 |
| sdake_ | sp_ coverage was broken in kolla-kubernetes recently to fix the 0.4.0 tagging | 09:30 |
| duonghq | portdirect, how can I run just ansible part? | 09:31 |
| sdake_ | duonghq cool that sounds like a godo feature | 09:31 |
| sp_ | sdake_: I will check. | 09:31 |
| portdirect | duonghq: just like any other playbook - though you will need to set up the inventory manually | 09:32 |
| portdirect | (and get ssh keys from each host) | 09:32 |
| duonghq | portdirect, so the vagrant pass these hosts to ansible playbook through CLI? | 09:33 |
| portdirect | but if yum update is not working then thats your problem - and you/we should get that fixed as otheriwse you are gonna keep on hitting probelms moving forwarf | 09:33 |
| duonghq | portdirect, sure | 09:34 |
| duonghq | I'm not sure why yum doesn't recognize no_proxy settings | 09:35 |
| portdirect | if you could post the output of curling the metadata adding --verbose to comannd that would help | 09:35 |
| duonghq | curl just fine | 09:35 |
| portdirect | and also the contents of the .repo file you are adding | 09:35 |
| duonghq | portdirect, http://paste.openstack.org/show/595008/ | 09:36 |
| portdirect | that does not have --verbose in the command | 09:36 |
| duonghq | portdirect, http://paste.openstack.org/show/595009/ | 09:37 |
| duonghq | the repo | 09:37 |
| *** salv-orlando has joined #openstack-kolla | 09:37 | |
| *** zhurong has quit IRC | 09:37 | |
| duonghq | portdirect, http://paste.openstack.org/show/595010/ | 09:38 |
| duonghq | the curl with verbose | 09:38 |
| portdirect | cool - could you get the url of a rpm and try to curl/wget that direct | 09:39 |
| portdirect | (wondering if mtu issue) | 09:39 |
| openstackgerrit | Surya Prakash Singh proposed openstack/kolla: py3 py2 implementation for kolla/image module https://review.openstack.org/414813 | 09:39 |
| duonghq | yum update cannot download even repomd.xml | 09:39 |
| *** sgordon has quit IRC | 09:41 | |
| *** sdake_ has quit IRC | 09:41 | |
| portdirect | also I'm assuming that http://10.164.180.104:2323/centos_7_epel/ is autualy a mirror of centos-base, updates, extras, and plus as well? | 09:41 |
| duonghq | ya | 09:42 |
| duonghq | sorry for naming | 09:42 |
| portdirect | np :) | 09:42 |
| *** salv-orlando has quit IRC | 09:42 | |
| duonghq | portdirect, base centosplus elasticsearch-2.x epel extras kubernetes noarch others rabbitmq_rabbitmq-server repodata updates | 09:42 |
| *** athomas has joined #openstack-kolla | 09:43 | |
| openstackgerrit | Surya Prakash Singh proposed openstack/kolla: Py3 and py2 implementation for kolla/image module https://review.openstack.org/414813 | 09:43 |
| portdirect | you can remove all the duplications in that file in that case | 09:43 |
| *** saneax-_-|AFK is now known as saneax | 09:43 | |
| portdirect | but should work fine as it | 09:44 |
| zhubingbing | sdake | 09:44 |
| zhubingbing | ping sdake | 09:44 |
| duonghq | portdirect, sure, | 09:44 |
| portdirect | I'm sorry though duonghq I dont think I'm actually gonna be much help here, your environment is sufficiently custom that its very hard to work out whats going on :( | 09:45 |
| duonghq | portdirect, anyway, thank you | 09:45 |
| duonghq | portdirect, got the network issues (again) when 3 submarine cable systems go out my country is broken, and the proxy is somehow messup with something | 09:47 |
| portdirect | :( | 09:48 |
| duonghq | portdirect, which part of kolla-k8s are you working on? | 09:49 |
| *** sgordon has joined #openstack-kolla | 09:49 | |
| portdirect | at the moment I've not been able to contribute as much as I'd like - but plan to get back in as soon as possible - workin on the consumption of ceph by k8s pod | 09:50 |
| portdirect | so the work you are doing on ceph deployment is really appreciated :) | 09:50 |
| duonghq | I'm stuck with ceph-osd, so I want to try something with operator or service, and comeback to osd microservice later, just to break my mindloop (if any) | 09:51 |
| portdirect | roger :) | 09:51 |
| duonghq | guess that I should ask somebody else :P | 09:51 |
| portdirect | you may want to reach out to kfox, as he did the original work | 09:52 |
| duonghq | portdirect, thank you | 09:52 |
| portdirect | and i think is the only one who really undersands it | 09:52 |
| portdirect | :) also a big :P to you too :) | 09:52 |
| duonghq | nice | 09:52 |
| duonghq | btw, can you review this small ps: https://review.openstack.org/#/c/420479/ | 09:53 |
| duonghq | and which is Kevin's tz? | 09:53 |
| portdirect | PST i think? but somewher in the states for sure :) | 09:53 |
| duonghq | ack | 09:54 |
| openstackgerrit | Merged openstack/kolla-kubernetes: Fix typo in keystone svc https://review.openstack.org/420479 | 09:58 |
| duonghq | portdirect, I figured out how to use local repo | 10:06 |
| duonghq | just add proxy=_none_ into .repo file | 10:06 |
| *** hfu has quit IRC | 10:08 | |
| *** Serlex has joined #openstack-kolla | 10:08 | |
| *** zhurong has joined #openstack-kolla | 10:08 | |
| *** prameswar has joined #openstack-kolla | 10:30 | |
| *** satyar has quit IRC | 10:36 | |
| *** salv-orlando has joined #openstack-kolla | 10:38 | |
| *** salv-orlando has quit IRC | 10:43 | |
| openstackgerrit | Nenad Radojevic proposed openstack/kolla-ansible: TLS loadbalancer creation fails due to unread neutron_lbaas.conf https://review.openstack.org/420634 | 10:46 |
| *** pbourke has quit IRC | 10:47 | |
| *** pbourke has joined #openstack-kolla | 10:48 | |
| *** zhubingbing has quit IRC | 10:48 | |
| *** rmart04 has joined #openstack-kolla | 10:50 | |
| *** williamcaban has joined #openstack-kolla | 10:52 | |
| *** sayantani01 has joined #openstack-kolla | 10:56 | |
| *** sayantani01 has quit IRC | 11:00 | |
| openstackgerrit | Surya Prakash Singh proposed openstack/kolla-ansible: Cover environment for kolla-ansible https://review.openstack.org/419450 | 11:02 |
| *** tonanhngo has joined #openstack-kolla | 11:06 | |
| *** msimonin has joined #openstack-kolla | 11:06 | |
| *** tonanhngo has quit IRC | 11:08 | |
| *** JoseMello has joined #openstack-kolla | 11:13 | |
| *** williamcaban has quit IRC | 11:14 | |
| *** egonzalez has quit IRC | 11:19 | |
| *** egonzalez has joined #openstack-kolla | 11:19 | |
| *** markus-k has quit IRC | 11:32 | |
| *** markus-k has joined #openstack-kolla | 11:33 | |
| *** mliima has joined #openstack-kolla | 11:39 | |
| *** salv-orlando has joined #openstack-kolla | 11:39 | |
| mliima | morning guys | 11:40 |
| egonzalez | morning mliima | 11:40 |
| openstackgerrit | Zeyu Zhu proposed openstack/kolla: Modify the description in README.rst https://review.openstack.org/420665 | 11:41 |
| *** ccesario has joined #openstack-kolla | 11:43 | |
| *** salv-orlando has quit IRC | 11:43 | |
| *** ccesario has quit IRC | 11:58 | |
| *** nrado has quit IRC | 11:59 | |
| *** zhubingbing has joined #openstack-kolla | 11:59 | |
| mliima | hi zhubingbing, morning | 12:01 |
| *** saneax is now known as saneax-_-|AFK | 12:01 | |
| mliima | can you review https://review.openstack.org/#/c/419976/ ? | 12:01 |
| zhubingbing | ok | 12:01 |
| egonzalez | zhubingbing: i'm testing fluentd patch, is a must having elasticsearch enabled? | 12:02 |
| zhubingbing | done | 12:02 |
| *** ccesario has joined #openstack-kolla | 12:02 | |
| mliima | great zhubingbing | 12:02 |
| mliima | thank you | 12:02 |
| zhubingbing | is a must having elasticsearch enabled? yes | 12:02 |
| zhubingbing | it same as to heka | 12:03 |
| egonzalez | never enabled elasticsearch with heka for an allinone ;) | 12:03 |
| egonzalez | guess, that enable_elasticsearch must be set to yes by default | 12:04 |
| openstackgerrit | Merged openstack/kolla-ansible: Move prechecks into its own role https://review.openstack.org/419976 | 12:04 |
| zhubingbing | because I fluentd the output to es above | 12:05 |
| egonzalez | hrm, in my second attemp wasn't a problem not having elasticsearch, but logs are full of errors | 12:06 |
| zhubingbing | errors？ | 12:07 |
| zhubingbing | fluentd logs error? | 12:07 |
| egonzalez | error_class=Fluent::ElasticsearchOutput::ConnectionFailure error="Can not reach Elasticsearch cluster | 12:08 |
| zhubingbing | 2017-01-14 14:33:37 +0000 [info]: fluentd worker is now running | 12:09 |
| zhubingbing | 2017-01-14 14:33:53 +0000 [info]: Connection opened to Elasticsearch cluster => {:host=>"192.168.23.254", :port=>9200, :scheme=>"http"} | 12:09 |
| zhubingbing | ;) | 12:09 |
| egonzalez | if we make this configurable with enable_elasticseach|bool , will fluentd works in allinone? https://review.openstack.org/#/c/407392/36/ansible/roles/common/templates/td-agent.conf.j2 | 12:11 |
| egonzalez | part of elastic config | 12:11 |
| *** rhallisey has joined #openstack-kolla | 12:12 | |
| openstackgerrit | zhubingbing proposed openstack/kolla-ansible: [WIP] Add fluentd role https://review.openstack.org/407392 | 12:12 |
| zhubingbing | fluentd works in allinone ; yes | 12:12 |
| egonzalez | zhubingbing: yeah, is working but without elasticsearch fluentd logs are full of those errors (makes sense) | 12:14 |
| egonzalez | just wondering if can avoid sending to elastic when is not enabled | 12:14 |
| zhubingbing | ;) understand | 12:14 |
| zhubingbing | i add it now | 12:14 |
| egonzalez | i can test it before if you want | 12:16 |
| zhubingbing | good | 12:17 |
| *** caowei has joined #openstack-kolla | 12:18 | |
| egonzalez | if i remove that section, fluentd seems still working | 12:20 |
| zhubingbing | yes | 12:22 |
| zhubingbing | done | 12:23 |
| openstackgerrit | zhubingbing proposed openstack/kolla-ansible: [WIP] Add fluentd role https://review.openstack.org/407392 | 12:23 |
| zhubingbing | now fluentd default output to es inside, so if not enable es it have cant | 12:24 |
| zhubingbing | contact es cluster error | 12:24 |
| zhubingbing | egonzalez https://review.openstack.org/#/c/407392/38 | 12:26 |
| egonzalez | yup, i'll test again | 12:26 |
| *** liyifeng has joined #openstack-kolla | 12:26 | |
| *** krtaylor has quit IRC | 12:29 | |
| *** williamcaban has joined #openstack-kolla | 12:30 | |
| *** williamcaban has quit IRC | 12:34 | |
| *** salv-orlando has joined #openstack-kolla | 12:40 | |
| openstackgerrit | zhubingbing proposed openstack/kolla-ansible: [WIP] Add fluentd role https://review.openstack.org/407392 | 12:41 |
| *** salv-orlando has quit IRC | 12:44 | |
| des_consolado | hey I've posted a question about neutron_external_interface https://ask.openstack.org/en/question/101562/confused-about-neutron_external_interface/ | 12:45 |
| *** sayantani01 has joined #openstack-kolla | 12:45 | |
| *** salv-orlando has joined #openstack-kolla | 12:46 | |
| *** sayantani01 has quit IRC | 12:49 | |
| *** williamcaban has joined #openstack-kolla | 12:51 | |
| *** williamcaban has quit IRC | 12:55 | |
| pbourke | des_consolado: have you tried setting ens2f1 as your neutorn_external_interface? | 12:59 |
| pbourke | des_consolado: and ens2f0 as your api_interface | 12:59 |
| sbezverk | good morning | 13:02 |
| *** zhurong has quit IRC | 13:03 | |
| *** duonghq_ has joined #openstack-kolla | 13:04 | |
| *** duonghq has quit IRC | 13:04 | |
| *** duonghq_ is now known as duonghq | 13:04 | |
| *** caowei has quit IRC | 13:04 | |
| des_consolado | pbourke: but then what about the external vip interface? | 13:04 |
| duonghq | evening guys | 13:04 |
| *** duonghq_ has joined #openstack-kolla | 13:04 | |
| des_consolado | that will also be the same interface as the neutron_external_interface? | 13:05 |
| *** williamcaban has joined #openstack-kolla | 13:05 | |
| sp_ | Jeffrey4l: it would be great. now if you review this patch set https://review.openstack.org/#/c/419450/ | 13:06 |
| sp_ | Jeffrey4l: after this kolla-ansible must be able to run coverage | 13:07 |
| *** matrohon has joined #openstack-kolla | 13:07 | |
| sp_ | Jeffrey4l: one more patch set on BP | 13:07 |
| sp_ | Jeffrey4l: ignore last message | 13:09 |
| *** goldyfruit has joined #openstack-kolla | 13:10 | |
| *** goldyfruit has quit IRC | 13:10 | |
| openstackgerrit | Mauricio Lima proposed openstack/kolla: install murano dashboard package in horizon https://review.openstack.org/395957 | 13:11 |
| *** masuberu has joined #openstack-kolla | 13:15 | |
| *** l4yerffeJ_ has quit IRC | 13:19 | |
| *** l4yerffeJ_ has joined #openstack-kolla | 13:19 | |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible: Move swift precheck into its own role https://review.openstack.org/418762 | 13:22 |
| openstackgerrit | zhubingbing proposed openstack/kolla-ansible: [WIP] Add fluentd role https://review.openstack.org/407392 | 13:23 |
| openstackgerrit | Merged openstack/kolla: Fix gate in iax-ord node when using libvirt 2.0 https://review.openstack.org/420340 | 13:23 |
| *** schwicht has joined #openstack-kolla | 13:24 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible: Pull kolla images from tarballs.o.o https://review.openstack.org/413720 | 13:26 |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible: Move iscsi precheck into its own role https://review.openstack.org/417725 | 13:28 |
| *** krtaylor has joined #openstack-kolla | 13:30 | |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible: Move mongodb precheck into its own role https://review.openstack.org/416060 | 13:31 |
| egonzalez | des_consolado: not, external_vip_address is the IP for external API connections in HAproxy, neutron_external_interface is to provide external connectivity to instances | 13:32 |
| egonzalez | both should be different | 13:32 |
| *** nrado has joined #openstack-kolla | 13:33 | |
| des_consolado | egonzalez: but how does neutron_external_interface get its IP configuration? At the moment I have a public IP address configured manually on ens2f0 so to me it would make sense to make that the interface for neutron to provide external connectivity to instances, but in the comments it says that it shouldn't have an IP address? | 13:35 |
| egonzalez | des_consolado: does not need IP address because neutron/openvswitch is going to manage networking there, neutron_external_interface is the interface added in br-ex | 13:38 |
| openstackgerrit | Mauricio Lima proposed openstack/kolla-ansible: Add karbor ansible role https://review.openstack.org/402229 | 13:38 |
| *** sdake has joined #openstack-kolla | 13:40 | |
| des_consolado | Oh I see, I will have to check the openvswitch configuration to see how it configures the IP addressing for this then, hopefully all will become clear and I will answer the question on the question board, ty | 13:40 |
| des_consolado | I just destroyed and am rebuilding all images for 3.0.2 right now so might be a while | 13:41 |
| *** mgoddard has joined #openstack-kolla | 13:41 | |
| sdake | morning peeps | 13:43 |
| *** rhallisey has quit IRC | 13:43 | |
| openstackgerrit | Mauricio Lima proposed openstack/kolla-ansible: Add karbor ansible role https://review.openstack.org/402229 | 13:43 |
| *** salv-orlando has quit IRC | 13:46 | |
| *** mdnadeem has quit IRC | 13:46 | |
| *** schwicht has quit IRC | 13:48 | |
| *** sacharya has quit IRC | 13:52 | |
| *** fguillot has joined #openstack-kolla | 13:55 | |
| *** goldyfruit has joined #openstack-kolla | 13:57 | |
| openstackgerrit | zhubingbing proposed openstack/kolla-ansible: [WIP] Add fluentd role https://review.openstack.org/407392 | 13:57 |
| *** l4yerffeJ_ has quit IRC | 13:58 | |
| *** saneax-_-|AFK is now known as saneax | 13:59 | |
| *** l4yerffeJ_ has joined #openstack-kolla | 13:59 | |
| *** l4yerffeJ_ has quit IRC | 14:00 | |
| *** l4yerffeJ_ has joined #openstack-kolla | 14:00 | |
| *** schwicht has joined #openstack-kolla | 14:01 | |
| sp_ | duonghq: ping.. | 14:04 |
| sbezverk | sdake: morning, do you have tiem to review this: https://review.openstack.org/#/c/412237/ | 14:06 |
| *** tonanhngo has joined #openstack-kolla | 14:08 | |
| zhubingbing | sbezverk | 14:09 |
| *** tonanhngo has quit IRC | 14:10 | |
| zhubingbing | do you have Kolla-K8S presentation documents ？ | 14:10 |
| duonghq | sp_, pong | 14:11 |
| duonghq | morning sdake | 14:11 |
| *** mattmceuen has joined #openstack-kolla | 14:11 | |
| duonghq | kfox1111, around? | 14:12 |
| sp_ | duonghq: I need your help to complete this BP for ocata. | 14:13 |
| sp_ | duonghq: will provide some info to proceed this BP | 14:13 |
| *** dave-mccowan has joined #openstack-kolla | 14:14 | |
| duonghq | sp_, sure, | 14:14 |
| duonghq | what can I do for you? | 14:14 |
| sp_ | how to proceed this ? | 14:15 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: install ansible from pypi in bifrost https://review.openstack.org/420732 | 14:15 |
| duonghq | moment, I'm reviewing your ps | 14:15 |
| sp_ | what exactly I require to do ? | 14:15 |
| duonghq | 1st, increase test coverage | 14:16 |
| duonghq | we have not measured test coverage yet | 14:17 |
| duonghq | the py35 test still green while it should not due to low coverage | 14:17 |
| *** jheroux has joined #openstack-kolla | 14:17 | |
| sp_ | duonghq: you mean i need to add some more unit test cases. right ? | 14:18 |
| duonghq | moment, I'm trying to recall in OpenStack, do we have any tool for measure test coverage | 14:18 |
| openstackgerrit | Merged openstack/kolla: Modify the description in README.rst https://review.openstack.org/420665 | 14:19 |
| sdake | sbezverk i see y ou got reno rolling :) | 14:19 |
| sdake | sbezverk thanks for that | 14:19 |
| sdake | sbezverk the patch looks good although I didn't pllus 2 because of whitespace issue | 14:19 |
| sp_ | duonghq: we add coverage tool in tox for coverage. | 14:20 |
| duonghq | sp_, https://coverage.readthedocs.io/en/coverage-4.3.1/# -> that it | 14:20 |
| sdake | i'm not sure if that bothers kubernetes - but it bothers git to have trailing whitespace (whith the cli tools) | 14:20 |
| *** mkoderer has joined #openstack-kolla | 14:20 | |
| sdake | with the git cli tools i mean | 14:20 |
| duonghq | sp_, you mean we already had it, or we are adding it? | 14:21 |
| duonghq | we have this tool | 14:21 |
| sp_ | duonghq: I have added it in kolla and kolla-ansible deliverable. | 14:22 |
| duonghq | nice | 14:22 |
| sp_ | duonghq: need to add it to kolla-kubernetes too | 14:22 |
| duonghq | what is coverage percent now? | 14:22 |
| sdake | duonghq - coverage is a rockin tool | 14:22 |
| *** Satyap has joined #openstack-kolla | 14:22 | |
| sdake | sp_ there is also a gate job pioneered by someone which can do A-B comparisons on coverage | 14:22 |
| sdake | and giveou a delta | 14:22 |
| duonghq | I think you should increase percent for python code | 14:23 |
| sdake | duonghq clearly | 14:23 |
| sdake | duonghq 85% is a good target | 14:23 |
| duonghq | sdake, guy from magnum project already done this stuff, through shell script | 14:23 |
| sdake | duonghq maybe multipel proejcts have this kind of gate job now | 14:23 |
| duonghq | sdake, sure | 14:23 |
| sdake | sp_ note there is coverage tool in kolla-kubernetes, but i think it is busted | 14:24 |
| *** bachp has quit IRC | 14:24 | |
| sdake | using cover (probably what you used) | 14:24 |
| sp_ | duonghq: sdake then i have to add more test cases | 14:25 |
| sdake | sp_ nobody requires adding test cases - but if you want to do that separately that rocks :) | 14:25 |
| sdake | sp_ people work on what they want in this project or what interests them | 14:25 |
| sp_ | sdake: yes. its is busted. I am still investigating this issue | 14:25 |
| Satyap | i am using kolla to deploy my openstack services. external interface for network node and compute node is different. so I added below line multinode file to differentiate the external interface for compute node | 14:25 |
| zhubingbing | sup sdake | 14:26 |
| zhubingbing | do you have Kolla-K8S presentation documents ？ | 14:26 |
| sdake | sup zhubingbing | 14:26 |
| sdake | holdiay today | 14:26 |
| sdake | so i'll be out most of the day | 14:26 |
| sdake | zhubingbing i don't have any at all no | 14:26 |
| zhubingbing | i will test kolla-k8s | 14:26 |
| sdake | zhubingbing sbezverk probqably has the presentation material form summit | 14:26 |
| Satyap | in [compute] section of multinode file i added compute01 neutron_external_interface=eno1.808 | 14:26 |
| zhubingbing | wow | 14:26 |
| Satyap | but my change is not reflecting in compute node | 14:26 |
| sdake | zhubingbing although there is no presentation documentation that is up to date | 14:26 |
| Satyap | doeas anybody has any idea about this | 14:26 |
| sdake | zhubingbing if you wnat to see whats implemented check out the release notes | 14:27 |
| sdake | zhubingbing those are up to date | 14:27 |
| zhubingbing | ok | 14:27 |
| sdake | zhubingbing the dev env setup guide will get you started | 14:27 |
| sdake | zhubingbing after that, the docs are sparse | 14:27 |
| zhubingbing | I mainly want to understand what the next kolla-k8s degree | 14:27 |
| sdake | zhubingbing or incorrect | 14:27 |
| sdake | zhubingbing the question didn't parse on my end - could you rephrase | 14:28 |
| sdake | i don't know what you mean by the word degree above | 14:28 |
| zhubingbing | ；） | 14:28 |
| zhubingbing | Kolla-K8s completed the project | 14:28 |
| zhubingbing | ；） | 14:28 |
| zhubingbing | sdake happy holiday | 14:29 |
| zhubingbing | I want to play kolla-k8s | 14:30 |
| zhubingbing | ；） | 14:30 |
| sdake | zhubingbing cool we can get you there pretty fast i think | 14:30 |
| sdake | read release notes, read development environment | 14:30 |
| zhubingbing | emm | 14:30 |
| sdake | get to that step | 14:30 |
| zhubingbing | ok | 14:30 |
| sdake | then we will sort out whats next | 14:30 |
| sdake | you can use a single box | 14:30 |
| sdake | with 16gb of ram | 14:30 |
| *** sayantani01 has joined #openstack-kolla | 14:30 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Mariadb service deployment https://review.openstack.org/412237 | 14:31 |
| sdake | the development environment imo is the best way to setup an eval environment too | 14:31 |
| sdake | when you make a commitment to dpeloying kolla-kubernetes, we haven't sovled how to pave bremetal or configure it and get kubernetes on it like kolla-ansible has done | 14:31 |
| zhubingbing | understand | 14:32 |
| sdake | s othat commitment comes with figuring that part out on your own during deployment | 14:32 |
| sdake | whereas the dev env is virtualized | 14:32 |
| sdake | and designed to standaridaze a deploy of kubernetes on virtual machines | 14:32 |
| openstackgerrit | Merged openstack/kolla-ansible: TLS loadbalancer creation fails due to unread neutron_lbaas.conf https://review.openstack.org/420634 | 14:32 |
| *** srwilkers has joined #openstack-kolla | 14:34 | |
| openstackgerrit | Merged openstack/kolla-ansible: Cover environment for kolla-ansible https://review.openstack.org/419450 | 14:34 |
| sdake | morning srwilkers | 14:34 |
| srwilkers | hey sdake | 14:34 |
| *** sayantani01 has quit IRC | 14:35 | |
| openstackgerrit | Merged openstack/kolla-ansible: Add missed octavia_ca_password variable in passwords.yml file https://review.openstack.org/419774 | 14:35 |
| srwilkers | have a pleasant weekend? | 14:35 |
| sdake | srwilkers meh - hung up lighting:https://twitter.com/steven_dake/status/820752300289339393 | 14:36 |
| sdake | srwilkers and took some down | 14:36 |
| sdake | today is a work holiday and it has stopped raining so getting lights off the house today :) | 14:37 |
| sdake | (christmsa lights) | 14:37 |
| duonghq | sdake, it has rained for some days? | 14:37 |
| sdake | duonghq abouta week | 14:37 |
| sp_ | duonghq: will review this why gate is failing here ? https://review.openstack.org/#/c/414813/7 | 14:37 |
| sdake | duonghq little nervous to get on the roof since it may be soft | 14:38 |
| duonghq | sdake, what does it make from? | 14:38 |
| sdake | duonghq wood i think | 14:38 |
| sdake | duonghq then coated | 14:38 |
| sdake | duonghq i get on it occasicoanlly | 14:38 |
| sdake | there is one part where i wont go | 14:39 |
| sdake | i ran an aircondition power line when i hooked up my nest (needed 7 wires) through my flat roof | 14:39 |
| sdake | tha twas a huge pitia - because a flat roof cannot be drilled | 14:39 |
| srwilkers | yeah, that sounds miserable | 14:39 |
| duonghq | oh, so stand on it after 1 week of raining is not a good idea | 14:39 |
| sdake | hooking up the wires to the air conditioning unit was frightening | 14:39 |
| sdake | "going to blow this 7k air conditoiner" | 14:40 |
| srwilkers | haha | 14:40 |
| sdake | i actually dont know how much it cost, but that is how much it would be to replace | 14:40 |
| sdake | its a really nice unit - 5 ton 15 seer | 14:40 |
| sdake | and top of the line model | 14:41 |
| sdake | brand i mean | 14:41 |
| sdake | i was surprised to see it on the house when we bougth it | 14:41 |
| srwilkers | thats awesome | 14:41 |
| sdake | it was the best feature of the house imo | 14:41 |
| sdake | however they had a janky install of it before | 14:41 |
| duonghq | sp_, I'm looking why bifrost-base failed | 14:41 |
| sdake | and i fixed that without an air conditioner degree :) | 14:41 |
| sp_ | duonghq: thanks | 14:42 |
| sdake | srwilkers now i pay people to maintain it | 14:42 |
| sdake | srwilkers however i couldn't get anyone to do he actua lwork of hooking it up properly | 14:42 |
| sdake | srwilkers nobody would touch it -too much liably i guess | 14:42 |
| *** v1k0d3n has quit IRC | 14:43 | |
| srwilkers | sdake, yeah, i can see that happening | 14:43 |
| sdake | running the wire i get - why they wouldn't wnat t o do that | 14:43 |
| sdake | the hooking it up i actually never asked anyone | 14:43 |
| sdake | i probably shoudl have | 14:43 |
| *** srwilkers is now known as srwilkers_coffee | 14:43 | |
| sdake | i just did it - and it worked after reading an 18 page phd thesis on air conditioners | 14:43 |
| srwilkers_coffee | lol | 14:44 |
| sdake | and air controls | 14:44 |
| sdake | it was fantastic phd work | 14:44 |
| srwilkers_coffee | the engineer way | 14:44 |
| sdake | it may hav ebeen a masters paper | 14:44 |
| sdake | i dont recall | 14:44 |
| sdake | i wish i had a copy of it | 14:44 |
| sdake | i can't find it on the internet anymore | 14:44 |
| srwilkers_coffee | #trustmeimanengineer | 14:44 |
| sdake | i searched for 8 hourso n google | 14:44 |
| sdake | finally i typed something in and google ralphed ou tthe right thing | 14:44 |
| *** lrensing has joined #openstack-kolla | 14:45 | |
| sdake | oddly my dad afterwards replaced a control on his oven later which was way more compelx | 14:45 |
| sdake | and he is like 'if you can hook up an airconditioner i can fix my oven" :) | 14:45 |
| duonghq | srwilkers_coffee, can you give me some advice on ceph mon? | 14:45 |
| *** ccesario has quit IRC | 14:45 | |
| sdake | although he didn't hook up his ac later :) - he got his brother to help him :) | 14:45 |
| sdake | anyway thats the hardest around the house thing i've done ;) | 14:46 |
| sdake | and had copious help from my wife | 14:46 |
| *** salv-orlando has joined #openstack-kolla | 14:47 | |
| duonghq | sdake, have not seen any a/c in the USA yet :( | 14:49 |
| sdake | duonghq carrier is he brand i am pretty sure | 14:49 |
| sdake | its made specifically for the arizona market | 14:49 |
| duonghq | In Vietnam the hardest part of a/c maintenance is recharge the gas, | 14:50 |
| sdake | duonghq oh ya - need a pro for that | 14:50 |
| sdake | duonghq if you have a leak in your ac, you ahve a problem tho :) | 14:50 |
| *** masuberu has quit IRC | 14:50 | |
| duonghq | ya, but the wiring is very simple, just plug and play | 14:50 |
| sdake | wiring isn't plug and play here | 14:51 |
| sdake | wirenuts, 6 wires in use, gotta hook em up correctly | 14:51 |
| sdake | on both ends | 14:51 |
| sdake | or bad things hapepn | 14:51 |
| *** salv-orlando has quit IRC | 14:51 | |
| sdake | one of them is a common run - so it can blow up if you hook it up improperly | 14:51 |
| duonghq | seem dangerous? | 14:52 |
| sdake | i am not really an electrician but learning the basics :) | 14:52 |
| sdake | duonghq like i siad, I was frieghtened hooking it up the whole thing was going to expode | 14:52 |
| duonghq | explode literally? | 14:52 |
| sdake | burn out the circuitboard yes | 14:52 |
| sdake | blow up capacitors (which are 1fard or so in size0 | 14:53 |
| sdake | i think it could have been bad ;) | 14:53 |
| sdake | the ac had a 4x4 foot circuitboard | 14:53 |
| sdake | maybe 3x4 | 14:53 |
| sdake | it was big | 14:53 |
| duonghq | ya | 14:54 |
| sdake | what they need is little connectors on the thing | 14:54 |
| sdake | to make it plug and play | 14:54 |
| sdake | oh well | 14:54 |
| *** bachp has joined #openstack-kolla | 14:54 | |
| sdake | atleast i used the correct coloring scheme pros use | 14:54 |
| sdake | unlike the last install which was totally random | 14:54 |
| sdake | and used a hacked device to make 1 wire carry two signals | 14:54 |
| sdake | the ac guy told me if you use those your ac motors will blow | 14:54 |
| duonghq | what does user manual say? | 14:54 |
| sdake | user manual | 14:55 |
| sdake | LOL | 14:55 |
| sdake | no user manual | 14:55 |
| sdake | a schematic | 14:55 |
| sdake | is what I had to work with | 14:55 |
| duonghq | pretty awesome | 14:55 |
| sdake | lucky i kind of paid attention in university on the elecitrcal enginerring part of my degree | 14:55 |
| sdake | although i didn't earn the best grades in those courses, mostly in computing and math i was an a student - elwehre not so much ;) | 14:56 |
| sdake | duonghq in the us, NAU (where I went to school) had a CSE degree which is half computer enginerring have hardware engineering | 14:56 |
| sdake | i'd have preferred just a computer science degree, but that wasn't taught out of the engineering school | 14:57 |
| sdake | i was born to be an engineer I think :) | 14:57 |
| duonghq | sdake, I heard from my friend, and I'm always wonder how students can handle both discipline together | 14:57 |
| sdake | i never knew it until i was 25 or so | 14:57 |
| sdake | duonghq people typicallyl excelled in one adn did poorly in the other | 14:57 |
| sdake | in my university | 14:57 |
| sdake | i graded papers for computer science | 14:58 |
| sdake | you wouldn't believe the mistakes people made that were juniors/seniors | 14:58 |
| sdake | teh prfoessors i graded for told me i needed to back off on the grading harslhy | 14:58 |
| sdake | they said I was making them look bad ;) | 14:58 |
| sdake | i guess i was harsh - although i got graded harsly in other courses | 14:59 |
| duonghq | ya, so technically you are a scientist? | 14:59 |
| sdake | 5 pages of math for 1 problem - get one minus sign wrong but 20 different types of porblem solving | 14:59 |
| sdake | no i'm an engineer - CSE degree | 14:59 |
| sdake | and earn a 33% on the homework (this was diffeq) | 15:00 |
| sdake | complain to professor response was "If you get a minus sign wrong in engineering, you kill people" | 15:00 |
| Pavo | morning gents | 15:00 |
| duonghq | morning Pavo | 15:00 |
| sdake | he had it out for engineers | 15:00 |
| sdake | held us to a higher level of requirements | 15:01 |
| sdake | we didn't get along ;) | 15:01 |
| Pavo | so far so good on 3.0.2 | 15:01 |
| duonghq | ya | 15:01 |
| Pavo | using kolla-ansible upgrade feature | 15:01 |
| duonghq | anybody see Kevin around? | 15:01 |
| sdake | duonghq today is a us holdiay - probably for where he works too | 15:01 |
| duonghq | sdake, sorry but what is working holiday? | 15:02 |
| sdake | some people call it civil rights day or martin luther king day | 15:02 |
| Pavo | in US we have holidays that we get off of work for | 15:02 |
| sdake | i dont know what the correct term is | 15:02 |
| *** liyifeng has quit IRC | 15:03 | |
| sdake | cisco has it off too so i am about to jet :) | 15:03 |
| duonghq | ya | 15:03 |
| duonghq | sdake, I awake from 2AM till now? | 15:04 |
| duonghq | *you | 15:04 |
| sdake | nah at 3am went back to bed i guess | 15:04 |
| sdake | so i got like 15 hrs of sleep | 15:04 |
| sbezverk | sdake: addressed your comments, do you mind to ack before you go for your holiday? | 15:04 |
| sdake | kind of odd for me | 15:04 |
| sdake | sbezverk you got it | 15:04 |
| duonghq | I'm always have trouble with pronoun even in my primary language :( | 15:04 |
| duonghq | 15hrs of sleep is quite much then usual you're | 15:05 |
| sdake | sbezverk you have one more whitespace in the release note | 15:05 |
| sdake | sbezverk i guess i didn't see that earlier | 15:05 |
| duonghq | sbezverk, you're working on service layer of kolla-k8s, aren't you? | 15:05 |
| sdake | sbezverk it was off the right side of my gerrit screen | 15:06 |
| sdake | duonghq anyone can work on it - sbezverk has a model that canbe stamped out | 15:06 |
| sbezverk | duonghq: yes | 15:06 |
| *** srwilkers_coffee is now known as srwilkers | 15:06 | |
| srwilkers | morning sbezverk o/ | 15:06 |
| duonghq | sbezverk, I'm trying to get in the layer | 15:07 |
| sdake | duonghq if you wnt to work on that part - could use help there | 15:07 |
| duonghq | sdake, just want to get in other part and comeback to ceph microservice some days later, maybe I can find out how to make it work fine | 15:07 |
| sdake | duonghq we just released 0.4.0 yesterday | 15:07 |
| sdake | duonghq this tracker needs refinement (some items need pushign to 0.6.0) | 15:08 |
| sdake | https://launchpad.net/kolla-kubernetes/+milestone/0.5.0 | 15:08 |
| sdake | https://blueprints.launchpad.net/kolla-kubernetes/+spec/helm-services | 15:08 |
| sdake | duonghq sign up for work there | 15:08 |
| sbezverk | sdake: release notes do not count | 15:08 |
| sbezverk | ;-) | 15:08 |
| duonghq | sdake, I'm reading it | 15:08 |
| sdake | sbezverk the problem is with git cli tools - it makes using them onerous for peole | 15:09 |
| sdake | sbezverk for example git frmat-patch and git apply dont work if there are whitespace in the ptch | 15:09 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Mariadb service deployment https://review.openstack.org/412237 | 15:09 |
| duonghq | sbezverk, srwilkers which part should I get in? | 15:09 |
| sbezverk | sdake: done | 15:09 |
| sdake | sbezverk yup i already reviewed it :) | 15:09 |
| sdake | duonghq srwilkers portdirect kfox1111 i think what makes sens eis i'll unassign eveyrone for the service blueprint | 15:10 |
| sdake | except sbezverk - since he has already claimed some work | 15:10 |
| duonghq | ya | 15:10 |
| sdake | duonghq oh maybe i already did that :) | 15:11 |
| sdake | srwilkers you still up for doing cinder service level package? | 15:11 |
| sbezverk | duonghq: you can start servicizing one the completed in microservices service | 15:11 |
| srwilkers | sdake, yessir | 15:11 |
| sdake | srwilkers cool i'll leave that one alone then | 15:12 |
| *** l4yerffeJ__ has joined #openstack-kolla | 15:12 | |
| *** goldyfruit has quit IRC | 15:12 | |
| sdake | someone else may have edited the list - i think i adde dmyself for heat | 15:12 |
| sdake | and don't see myself there now | 15:12 |
| *** l4yerffeJ_ has quit IRC | 15:12 | |
| duonghq | sbezverk, okay, I need help from you guys, but which part should be done sooner? | 15:12 |
| sdake | duonghq keystone + mariadb + glance are critical first | 15:12 |
| sdake | duonghq keystone is the most important one | 15:13 |
| sdake | and it apperas unassigned unless sbezverk is already working on it | 15:13 |
| duonghq | roger | 15:13 |
| *** l4yerffeJ__ has quit IRC | 15:14 | |
| sdake | then next most importnat batch are nova and neutron | 15:14 |
| sdake | oh and rabbitmq is in the first batch too | 15:14 |
| sbezverk | duonghq: I was going to do keystone, but if you promise to do it fast then go ahead ;) | 15:14 |
| *** l4yerffeJ__ has joined #openstack-kolla | 15:14 | |
| sdake | duonghq basically we want to build up the compute kit from the bottom to top | 15:14 |
| duonghq | sbezverk, I'll take something just lower priority then keystone | 15:14 |
| duonghq | sdake, saw it | 15:15 |
| sdake | duonghq memcache woudl be in teh first batch too i think | 15:15 |
| sdake | duonghq but lower priority | 15:15 |
| sdake | duonghq i think our team is big enough to do all of them in parlalel | 15:15 |
| duonghq | sdake, so I think I can get it done, | 15:15 |
| duonghq | thank you | 15:16 |
| sdake | duonghq ya for inspiration take a look at the above patch from sbezverk | 15:16 |
| duonghq | sure | 15:16 |
| sdake | duonghq https://blueprints.launchpad.net/kolla-kubernetes/+spec/helm-services | 15:17 |
| sdake | duonghq that has been updated | 15:17 |
| duonghq | sdake, thank you | 15:18 |
| duonghq | I'll start working on it tomorrow (in my tz) | 15:18 |
| sdake | duonghq sweet - keep a bookmark of sbezverk 's work | 15:18 |
| sdake | duonghq or its in the blueprint as well | 15:19 |
| duonghq | hope that when I get to my dev machine, the halcyon runs again, | 15:19 |
| sdake | duonghq actually reading hte blueprint reviews start to finish would help you get started :) | 15:19 |
| duonghq | last weekend I turn it of and network messed up with these things | 15:19 |
| sdake | duonghq having trouble with halcyon? | 15:19 |
| duonghq | *off | 15:19 |
| sdake | duonghq so you had to reboot or something? | 15:20 |
| duonghq | before I leave, I get the local repo work (at least as the logs told me) | 15:20 |
| sdake | duonghq oh right - you have an added complexity of the local repo rthing iirc | 15:21 |
| duonghq | sdake, in my company, machine is put in working room, so the a/c is turn off night and weekend, so I should not leave the machine run through whole weekend | 15:21 |
| sdake | duonghq one way to test this stuff in the gate | 15:21 |
| sdake | duonghq understood | 15:21 |
| sdake | duonghq is in the gate that is | 15:22 |
| *** saneax is now known as saneax-_-|AFK | 15:22 | |
| duonghq | sdake, sure, just want one local test machine | 15:22 |
| sdake | duonghq right of course | 15:22 |
| duonghq | last time I tried helm install microservices/.... | 15:22 |
| duonghq | it told it cannot find set function | 15:23 |
| duonghq | the helm stuff | 15:23 |
| sdake | duonghq - you had old versoin of helm i think | 15:23 |
| sdake | duonghq helm 2.1.3 is required | 15:23 |
| duonghq | currently, which kubernetes version are you using? (beside helm) | 15:23 |
| sdake | 1.51 | 15:23 |
| sdake | 1.5.1 | 15:23 |
| duonghq | I used halcyon from Pete | 15:24 |
| sdake | the set function is a fetuer of helm not kubernetes | 15:24 |
| sdake | from pete's fork? | 15:24 |
| sdake | or from the dev guide upstream? | 15:24 |
| sdake | portdirect 's fork is not maintianed iirc - conical source is att-dev repo for our dev env | 15:24 |
| duonghq | from attdev | 15:24 |
| duonghq | att-comdev | 15:24 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: install ansible from pypi in bifrost https://review.openstack.org/420732 | 15:25 |
| *** satyar has joined #openstack-kolla | 15:25 | |
| sdake | duonghq hrm - that needs looking into | 15:25 |
| duonghq | tomorrow I will check helm version, today I can get only 2/4 vms up, hope that local repo'll do the trick | 15:26 |
| srwilkers | duonghq, poke me tomorrow and ill help you work through getting it working | 15:27 |
| duonghq | srwilkers, thanks in advanced | 15:27 |
| duonghq | sorry but what is your tz | 15:27 |
| srwilkers | CST in US | 15:28 |
| srwilkers | however, im usually awake for all but 4 hours of the day, so anytime other than 12am-4am US CST im typically awake | 15:28 |
| duonghq | how do you guys can sleep as little time? | 15:29 |
| duonghq | I can only withstand in this condition for 1 week (maximum) | 15:30 |
| duonghq | *month | 15:31 |
| srwilkers | duonghq, no idea. ive only slept 4 hours a night for as long as i can remember | 15:31 |
| sdake | srwilkers - thats not a healthy sleep schedule :) | 15:32 |
| sdake | srwilkers if it works for you tho - it works :0 | 15:32 |
| sdake | :) i mean | 15:32 |
| srwilkers | sdake, :) | 15:32 |
| sdake | i get about 6-7 hours a day | 15:33 |
| sdake | whic his plenty for me | 15:33 |
| sdake | sometimes i get 12-15 like as happend last night | 15:33 |
| *** goldyfruit has joined #openstack-kolla | 15:33 | |
| *** mliima has quit IRC | 15:33 | |
| sdake | when i really do alot of physical work | 15:33 |
| sdake | or after a week of travel ;) | 15:33 |
| sdake | nothign like hitting the bed after getting off the airplane | 15:34 |
| duonghq | I haven't travel outside my country yet :( | 15:34 |
| Serlex | I will state the obvious, sleep longer guys for healthy mind and body | 15:34 |
| srwilkers | sdake, no kidding. i crashed for about 20 hours after getting home from barcelona | 15:34 |
| *** l4yerffeJ__ has quit IRC | 15:35 | |
| srwilkers | sdake, that was after sleeping almost the entire way from barcelona to DC. i took 3 dramamine to help calm my nerves. i guess i was sleeping so heavily, the flight attendant put a napkin in front of my mouth to make sure i was still breathing because she couldnt wake me up for dinner | 15:35 |
| *** l4yerffeJ__ has joined #openstack-kolla | 15:36 | |
| duonghq | few days ago, I did annual health check and doctor told me that I should sleep more and lower duration I look to PC monitor, so I think I should sleep now, bye guys | 15:36 |
| duonghq | my eyes need some recover | 15:36 |
| *** l4yerffeJ__ has quit IRC | 15:37 | |
| *** l4yerffeJ__ has joined #openstack-kolla | 15:37 | |
| srwilkers | duonghq, gnight \o | 15:38 |
| sdake | duonghq sounds like a plan :) | 15:38 |
| sdake | duonghq dr's know best ;) | 15:38 |
| duonghq | sdake, yup, cya | 15:39 |
| duonghq | srwilkers, cya | 15:39 |
| sdake | Serlex i wish sometimes i could sleep longer then 7 hrs a day - however - i wake up - and then i'm awake | 15:39 |
| *** duonghq has quit IRC | 15:39 | |
| sdake | Serlex fortunately idont have insomnia | 15:39 |
| *** srwilkers is now known as srwilkers_brb | 15:40 | |
| sdake | the blue lighiting in al lthe led displays is bad for sleep | 15:40 |
| sdake | they make glass coatings for glasses that filter out the blue light | 15:40 |
| sdake | i'm getting them on my next glasses | 15:40 |
| sdake | whih should be this or next week :) | 15:40 |
| *** tonanhngo has joined #openstack-kolla | 15:41 | |
| *** tonanhngo has quit IRC | 15:42 | |
| sdake | anyway i am off- later :) | 15:42 |
| sdake | sbezverk never seen this before: http://logs.openstack.org/37/412237/39/check/gate-kolla-kubernetes-deploy-centos-binary-2-ceph-multi-nv/a7ed001/console.html#_2017-01-16_15_35_14_981456 | 15:48 |
| *** salv-orlando has joined #openstack-kolla | 15:48 | |
| sdake | sbezverk during reviews i am collecting a list of common bugs we run into in teh gated | 15:48 |
| openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible: Fix cleanup-containers script https://review.openstack.org/420775 | 15:48 |
| sdake | sbezverk this is runnign on osic cloud so it should be flawlesss - the osic cloud provider is our most reliable provider | 15:48 |
| sdake | sbezverk i ran a recheck on your review, however, have you seen this problem before? | 15:49 |
| *** dave-mccowan has quit IRC | 15:50 | |
| *** salv-orlando has quit IRC | 15:52 | |
| *** celo_r has joined #openstack-kolla | 15:54 | |
| sbezverk | sdake: we started etherpad with gate issues sometime ago. I do not have link handy, kfox1111 must have it.. | 15:55 |
| sdake | i'm filing bugs for them | 15:55 |
| sdake | rather then etherpad - so they are tracked and fixed | 15:55 |
| sdake | gate problems are highest priority if we can fix em - since they slow down dev | 15:55 |
| sdake | sbezverk i'll ping kfox1111 on tuesday and see if he has the etherpad handy | 15:56 |
| sdake | if so i'll make bugs out of em all | 15:56 |
| *** apuimedo is now known as apuimedo|away | 15:57 | |
| *** zhubingbing has quit IRC | 16:00 | |
| *** sayantani01 has joined #openstack-kolla | 16:02 | |
| *** dave-mccowan has joined #openstack-kolla | 16:05 | |
| *** dave-mcc_ has joined #openstack-kolla | 16:08 | |
| *** dave-mccowan has quit IRC | 16:10 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP Keystone service https://review.openstack.org/420791 | 16:13 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP Keystone service https://review.openstack.org/420791 | 16:15 |
| sbezverk | srwilkers: ping | 16:17 |
| *** mliima has joined #openstack-kolla | 16:18 | |
| mliima | sweet core reviews, I need some eyes on https://review.openstack.org/#/c/417094/ | 16:21 |
| *** srwilkers_brb is now known as srwilkers | 16:22 | |
| sdake | sbezverk - check this out: https://bugs.launchpad.net/kolla-kubernetes/+bug/1656870 | 16:23 |
| openstack | Launchpad bug 1656870 in kolla-kubernetes "Zuul reports standard output/error still open after child exited" [High,Confirmed] - Assigned to Steven Dake (sdake) | 16:23 |
| sdake | sbezverk i am curious how often sinc the 20th people run recheck? | 16:23 |
| sdake | i didn't see a whole lot of rechecks prior to teh 20th, however, the code was just getting started | 16:23 |
| sdake | and haven't actively reviewed sinigifcantly | 16:23 |
| *** tonanhngo has joined #openstack-kolla | 16:28 | |
| *** salv-orlando has joined #openstack-kolla | 16:30 | |
| *** adrian_otto has joined #openstack-kolla | 16:32 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP Keystone service https://review.openstack.org/420791 | 16:32 |
| kfox1111 | morning. | 16:33 |
| openstackgerrit | Mauricio Lima proposed openstack/kolla: install murano dashboard package in horizon https://review.openstack.org/395957 | 16:33 |
| *** sdake has quit IRC | 16:33 | |
| *** tonanhngo has quit IRC | 16:34 | |
| sbezverk | kfox1111: good morning | 16:34 |
| kfox1111 | hey. :) | 16:34 |
| sbezverk | sdake: we already have gate job for microservices | 16:34 |
| sbezverk | kfox1111: mariadb service is ready for your review | 16:35 |
| kfox1111 | sdake: I bet that error is helm still running in the background? | 16:35 |
| openstackgerrit | Mauricio Lima proposed openstack/kolla: Fix kibana dockerfile format https://review.openstack.org/420387 | 16:35 |
| kfox1111 | sbezverk: k. I'll have a look. | 16:35 |
| sbezverk | kfox1111: https://review.openstack.org/#/c/412237/39 | 16:35 |
| srwilkers | morning kfox1111 o/ | 16:35 |
| sbezverk | kfox1111: somebody reconfigured gate, do you know by any chance who and why? | 16:35 |
| kfox1111 | sbezverk: me. | 16:36 |
| kfox1111 | adding hooks for v3 and trunk builds. | 16:37 |
| sbezverk | kfox1111: I see | 16:37 |
| kfox1111 | since both of those patches are in flight already but had no place to land without breaking other tests. | 16:37 |
| sbezverk | kfox1111: cool I was just curious | 16:38 |
| sbezverk | it seems working fine.. | 16:38 |
| kfox1111 | yeah. submitted it late friday. didn't expect it quite this early. | 16:38 |
| kfox1111 | its doing nothhing at the moment. an additional param is tacked on the setup_gate.sh script now, of 2, 3, or t | 16:39 |
| kfox1111 | nothing's acting on it yet, so its always acting like 2. | 16:39 |
| kfox1111 | I'll submit a patch this monring to setup some infrastructure and disable it running on those other modes until implemented. | 16:40 |
| kfox1111 | sbezverk: should the pv/pvc/svc be part of the mariadb servie package too? | 16:43 |
| sbezverk | kfox1111: for sure | 16:43 |
| kfox1111 | shoudl we wait then on the mariadb service package? | 16:44 |
| sbezverk | what do you mean? PS I sent you is mariadb service package | 16:45 |
| kfox1111 | looks good as is, with one deps question. the rest could go either way. | 16:45 |
| sbezverk | check the requirements.yaml | 16:45 |
| *** nrado has quit IRC | 16:45 | |
| sbezverk | kfox1111: what about deps? | 16:46 |
| kfox1111 | yeah. I see it there. but I think the deps should probably be in all_values? | 16:46 |
| kfox1111 | that way non service package users can enable them if they want to. | 16:46 |
| sbezverk | kfox1111: disagree | 16:46 |
| sbezverk | otherwise all.yaml will grow like crazy | 16:46 |
| sbezverk | very difficult to manage | 16:46 |
| sbezverk | this is very specific scenario, deployment with entry point | 16:47 |
| sbezverk | no need to hang these setting for ALL cases | 16:47 |
| kfox1111 | With the api stabalizing, I'm almost thinking we're about ready to just split them back out to flat values.yaml in each microservice package? | 16:47 |
| kfox1111 | I'd still really like to see a common way of maintaining shared values for kolla-common, but | 16:48 |
| kfox1111 | at this point, not sure what will be more maintenence. ps conflicts in all_values, or maintaining the common vars in the microservice packages. | 16:48 |
| *** JeroenBo has joined #openstack-kolla | 16:48 | |
| *** JeroenBo has left #openstack-kolla | 16:48 | |
| sbezverk | I think keeping common in all.yaml and more specific in separate files makes sense to me | 16:48 |
| kfox1111 | that could work too. | 16:49 |
| sbezverk | resolving conflicts in all.yaml is very easy | 16:49 |
| sbezverk | we do not change there much, they are relatively static | 16:49 |
| kfox1111 | git workflow wouldn't work so nice if we just added the common stuff to values.yaml and had the base stuff in values.yaml too. | 16:49 |
| kfox1111 | we could build values.yaml from a package_values.yaml in each microservice dir, and all_alu[C[C[C[C[C[C[C[C[C[C[4~?es.yaml | 16:50 |
| kfox1111 | all_values.yaml | 16:50 |
| kfox1111 | then, once helm supports some kind of shared values.yaml, we can switch over to using that? | 16:52 |
| sbezverk | kfox1111: I do not see at least any advatanges of increasing complexity. I would go with all.yaml and each service with its own values.yaml allowing to deploy this specific service with entrypoint | 16:52 |
| kfox1111 | here's the problem with that. | 16:53 |
| sbezverk | for other scenarios like operator, we will not need these at all | 16:53 |
| kfox1111 | every time you prebuild, it will copy all_values into each microservices values.yaml. | 16:53 |
| kfox1111 | then, when you do a git diff, you will see all the changes to values.yaml. | 16:53 |
| kfox1111 | it will be really hard to seperate what tweaks to values.yaml devs do, vs what the build system does. | 16:54 |
| kfox1111 | well, I think we're talking about two different things maybe too. | 16:54 |
| sbezverk | kfox1111: I am lost.. | 16:54 |
| sbezverk | I was talking about service level values.yaml | 16:55 |
| kfox1111 | yeah, I was talking about microservice level values. | 16:55 |
| kfox1111 | talking about greatly shrinking the stuff stored in all_values.yaml for microservices. | 16:55 |
| kfox1111 | then putting deps in the microservices won't be so bad. | 16:55 |
| *** jrich523 has joined #openstack-kolla | 16:56 | |
| sbezverk | kfox1111: why would we need to keep dependency if they are used only for 1 signle scenario? | 16:56 |
| kfox1111 | I could see folks wanting to manually deploy microservices, but still support entrypoint? | 16:56 |
| sbezverk | I think it will add more confusion than help | 16:56 |
| kfox1111 | if its disabled by default, it doesn't do anything. | 16:57 |
| kfox1111 | if enabled, the user doens't have to define their own deps unless they are different then the defaults already there? | 16:57 |
| *** nrado has joined #openstack-kolla | 16:58 | |
| sbezverk | kfox1111: well you are not consistent, in some cases you want clear separation as in the case of repo, what if service repo will be separate, it would make sense to keep vlaues.yaml there.. | 16:58 |
| sbezverk | microservice do not have need and visibility to entrypoint dependencies | 16:59 |
| sbezverk | as they are kind of atomic things | 16:59 |
| kfox1111 | heh. disagreement on consistency I guess. | 16:59 |
| kfox1111 | ok, what if.... | 16:59 |
| kfox1111 | tripleo decided to deploy with microservices, but wanted to use entrypoint? | 17:00 |
| *** rmart04_ has joined #openstack-kolla | 17:00 | |
| kfox1111 | I'm for having a seperate repo for deployment/orchestration level things where people can't agree on the right path forward. | 17:00 |
| kfox1111 | putting shared code in the microservices, if it can be assured is disabled, I think is ok though? | 17:01 |
| sbezverk | kfox1111: that would require if new microservice is added to update dependency file, which is not part of microservice development but service development | 17:02 |
| *** markmcclain has quit IRC | 17:02 | |
| *** rmart04_ has quit IRC | 17:02 | |
| kfox1111 | the dependency handling code is already in each container. part of the init container. all it needs is defintions of dependencies, and a flag enabling it. | 17:02 |
| *** rmart04 has quit IRC | 17:02 | |
| kfox1111 | so we already went 90% of the way there in the microservices. | 17:03 |
| kfox1111 | were disagreeing on adding the remaining 9%. | 17:03 |
| kfox1111 | I guess it really doesn't matter too much at this point. we'll have to remove dependencies from all_values from whats there already, if thats what we decide to do. | 17:03 |
| kfox1111 | if someone else wants to share the deps, we can always move them back into the microservice packages. | 17:04 |
| sbezverk | kfox1111: it is not about 9% I do not mind to do something 90% it is just does not seem to be logic to add dependecies at microservice level | 17:04 |
| *** egonzalez has quit IRC | 17:04 | |
| sbezverk | kfox1111: I think we need to have broader discussion then.. | 17:04 |
| kfox1111 | I can see where your coming from. | 17:05 |
| sbezverk | kfox1111: here is my problem I was showing kolla-kube to some folks and they were overwhelmed by HUGE amount of setting in all.yaml | 17:06 |
| *** dave-mcc_ has quit IRC | 17:06 | |
| kfox1111 | that I agree with. | 17:06 |
| sbezverk | when you want to increase it even a bit but wihtout huge advantage | 17:06 |
| kfox1111 | but, that doesn't mean we should "fix" that problem by avoiding putting things where they potentially belong. | 17:07 |
| kfox1111 | the other way would be to tackle the real problem, that file being too big. | 17:07 |
| sbezverk | I cannot agree with it at least wihtout expression | 17:07 |
| kfox1111 | so, that file contains two different sets of info. | 17:07 |
| kfox1111 | vals that are package specific, and vals that are shared. | 17:07 |
| sbezverk | kfox1111: the problem here is that I do not see they belong there at all. I think they are perfectly fine in service package, exectly where they are used. | 17:08 |
| kfox1111 | can we split it such that, there is a vals file per package with the package specific stuff in it, and one remaining shared file? | 17:08 |
| kfox1111 | sbezverk: I understand your arguement, and I'm coming around to that logic. but you raise a seperate issue, that of all_values being too big. | 17:08 |
| kfox1111 | that I do think we need to handle, reguardless if deps are in it or not. | 17:09 |
| sbezverk | kfox1111: reducing all.yaml for sure if we can, we should do it | 17:09 |
| sbezverk | no disagreement here | 17:09 |
| kfox1111 | k. | 17:09 |
| kfox1111 | so, lets just keep size of all_values out of the discussion on where deps should go. cause I think thats a seperate thing. | 17:10 |
| sbezverk | kfox1111: I think the way how values.yaml is used now is more "helm native" | 17:10 |
| sbezverk | as we provided where they needed and they without any black magic | 17:10 |
| kfox1111 | I don't think helm has a concept of native when multiple shared packages are involved. :/ | 17:10 |
| kfox1111 | I'd kind of like to see a _values.yaml or something added to each microservice package, with that packages specific values added. | 17:11 |
| kfox1111 | then _values.yaml and all_values.yaml are merged to build that packages values.yaml. | 17:11 |
| kfox1111 | that woudl drastically shrink the all_values.yaml file. | 17:12 |
| kfox1111 | and keep package specific values close to the package. | 17:12 |
| kfox1111 | what do you think of that idea? | 17:12 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP Keystone service https://review.openstack.org/420791 | 17:13 |
| kfox1111 | I think the all_values.yaml file would shrink by something like 90%. | 17:13 |
| sbezverk | kfox1111: 1 question, where do you think people deploying service package would go to change service package dependencies? | 17:14 |
| kfox1111 | their own values.yaml if they wanted to. | 17:15 |
| kfox1111 | they can override anything they want there. | 17:15 |
| sbezverk | kfox1111: if they know how to do it. but most likely they will not know, so for example they would go to service package for example/default values | 17:16 |
| sbezverk | and it make sense. I would not start looking for service level setting in a microservice package | 17:16 |
| kfox1111 | users should never edit a kolla provided values.yaml in either the service or microservice packages. | 17:17 |
| *** matrohon has quit IRC | 17:17 | |
| kfox1111 | thats what their own values.yaml file is for. | 17:17 |
| *** gfidente is now known as gfidente|afk | 17:17 | |
| sbezverk | kfox1111: but they need to base it on something | 17:17 |
| kfox1111 | yeah. they need some way to discover the values they will want to set. | 17:17 |
| sbezverk | not everybody like you who can write it from scratch ;-) | 17:17 |
| kfox1111 | agreed. I can't do it from scratch either. | 17:18 |
| kfox1111 | but should we be encuraging them to look in values.yaml in the various packages, | 17:18 |
| sbezverk | I think keeping values.yaml with deployable settings will help users enourmously | 17:18 |
| kfox1111 | or give them a tool to help them find the values? | 17:18 |
| kfox1111 | either docs, or a program or something? | 17:18 |
| kfox1111 | the other option is a reference user values.yaml. | 17:19 |
| kfox1111 | provide all possible values with them all commented or something. | 17:19 |
| kfox1111 | then users can easily uncomment out what they want to use and fill in values. | 17:19 |
| *** markmcclain has joined #openstack-kolla | 17:19 | |
| kfox1111 | other projects do that with their config files. | 17:19 |
| sbezverk | kfox1111: I think it is already done with openstack, they provide cinder.conf with basica deployable setting with cinder package, not cinder volume ot sccheduler or api | 17:20 |
| sbezverk | we should do the same | 17:20 |
| kfox1111 | yeah. | 17:20 |
| *** sayantani01 has quit IRC | 17:20 | |
| sbezverk | otherwise you break a stereotape | 17:20 |
| kfox1111 | I think they have some infra in oslo to looka at all possible values and build the example | 17:20 |
| sbezverk | which will make people confused and frastrated | 17:20 |
| kfox1111 | rather then try and maintain an example over time statically. | 17:20 |
| *** bmace has quit IRC | 17:20 | |
| kfox1111 | but we can't reuse that infra, as its ini based. | 17:21 |
| *** bmace has joined #openstack-kolla | 17:21 | |
| sbezverk | kfox1111: I was just using it as am example | 17:21 |
| kfox1111 | so, maybe we do need to write a little python script that looks at the packages values file and builds an example-values.yaml. | 17:21 |
| sbezverk | why I think packaging values.yaml with service package makes sense | 17:21 |
| kfox1111 | I agree service packages need values. just not sure what exactly goes in it. | 17:22 |
| kfox1111 | trying to balance usablility and maintainability. | 17:22 |
| kfox1111 | so, for now, I guess I'm ok putting deps in the service packages. | 17:25 |
| *** mdnadeem has joined #openstack-kolla | 17:25 | |
| sbezverk | by looking at glance or mariadb's values.yaml I think they are very self explanatory and self sufficient :-) | 17:25 |
| kfox1111 | yeah, but if they want to find any value but a dep, then they have to end up looking at the microservices values.yaml anyway. | 17:26 |
| kfox1111 | sharingn the deps may be premature optomaization though. they may be different for other orchestration. | 17:26 |
| kfox1111 | so, I'll go along with putting them in servcies values.yaml. | 17:26 |
| kfox1111 | we only have one dep listed in trunk at the moment. | 17:27 |
| kfox1111 | should be easy to move that when we do the rabbit service package. | 17:27 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Common infra for enabling 3.x and trunk builds. https://review.openstack.org/420872 | 17:30 |
| sbezverk | kfox1111: glance was merged yesterday | 17:32 |
| sbezverk | kfox1111: sorry I did not get what you suggest to mve wit hrabbit ps? | 17:33 |
| kfox1111 | theres a dep in all_values for rabbit. it should be moved to the rabbit service package when it gets started. | 17:33 |
| *** Serlex has quit IRC | 17:33 | |
| kfox1111 | ah. thats why the neutron package merge conflicted. ok. thanks for the heads up. | 17:33 |
| sbezverk | kfox1111: got it | 17:34 |
| kfox1111 | hmm... | 17:34 |
| *** msimonin has quit IRC | 17:34 | |
| kfox1111 | for consistency, can you lowercase the JOBS/SERVICE ? | 17:34 |
| sbezverk | sure | 17:36 |
| kfox1111 | thx. | 17:36 |
| openstackgerrit | Eduardo Gonzalez proposed openstack/kolla: TLS loadbalancer creation fails due to unread neutron_lbaas.conf https://review.openstack.org/420880 | 17:37 |
| kfox1111 | so, mariadb doesn't have the pv/pvc/svc. but glance does. | 17:37 |
| *** tonanhngo has joined #openstack-kolla | 17:37 | |
| kfox1111 | can you please add the pv/pvc/svc to the mariadb one? Then I think its good to go? | 17:37 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Mariadb service deployment https://review.openstack.org/412237 | 17:38 |
| sbezverk | kfox1111: there is no dependecy for these three | 17:39 |
| sbezverk | they are not needed | 17:39 |
| sbezverk | in values.yaml | 17:39 |
| kfox1111 | hmm... | 17:39 |
| kfox1111 | actually, I think I see a problem with this approach and mariadb? | 17:39 |
| sbezverk | kfox1111: I have proof, it is working :-) | 17:40 |
| kfox1111 | yeah, I'm sure it is. | 17:40 |
| kfox1111 | but, what if element_name != mariadb? | 17:40 |
| kfox1111 | ie, you have more then one mariadb. | 17:40 |
| kfox1111 | the name of the job/service should change... | 17:40 |
| kfox1111 | in that case, the deps probably need to be generated based on code. | 17:41 |
| kfox1111 | probably only ever will affect rabbit/mariadb/memcached. | 17:41 |
| *** tonanhngo has quit IRC | 17:42 | |
| sbezverk | kfox1111: that would require a separate tool, which will generate values.yaml file for a user requiring this special case | 17:42 |
| sbezverk | all they need is to run it, provide desired name and then put values.yaml in the right folder | 17:43 |
| kfox1111 | or put the logic in the mariadb template, as it has access to .Values.element_name to build the deps. | 17:43 |
| *** athomas has quit IRC | 17:44 | |
| sbezverk | kfox1111: would it not be cleaner approach wit hexternal tool? template looks fairly complex even now, addinf this logic will make it a rocket sience level ;-) | 17:44 |
| kfox1111 | not sure. | 17:45 |
| kfox1111 | external tool caries with it a different kidn of complexity. | 17:45 |
| kfox1111 | helm install kolla/mariadb --set element_name=foo #and it working sems logical. | 17:46 |
| kfox1111 | having it work the first time, and failing on: | 17:46 |
| kfox1111 | helm install kolla/mariadb --set element_name=bar #and it then fails to launch properly is unexpected. | 17:46 |
| srwilkers | i agree kfox1111. if we're allowing this functionality out of the box, it should be handled internally, not with an external tool | 17:47 |
| kfox1111 | well...hmm... it probably does fail on the first one, as the name is different in that case too. | 17:47 |
| *** adrian_otto1 has joined #openstack-kolla | 17:47 | |
| *** inc0 has joined #openstack-kolla | 17:47 | |
| inc0 | good morning | 17:48 |
| kfox1111 | inc0: good morning. | 17:48 |
| kfox1111 | inc0: don't know if you saw it, but we had an idea for the helm repo thing. | 17:48 |
| kfox1111 | inc0: keep seperate repo's for each major version. | 17:48 |
| kfox1111 | but provide a metapackage in kubernetes/charts that when isntalled, add the other repo's. | 17:49 |
| kfox1111 | so you just helm install stable/openstack-newton | 17:49 |
| sbezverk | kfox1111: well I cannot say I completely understand how it could be implemented, maybe srwilkers can develop it since he agrees ;) | 17:49 |
| kfox1111 | and then a helm search shows the kolla newton stuff. | 17:49 |
| sbezverk | inc0: good morning | 17:49 |
| kfox1111 | sbezverk: I think we just need a macro, that if defined, gets added to the dep list. | 17:50 |
| kfox1111 | kind of like some of the other {{ if .foo }} {{ include .foo }} {{ end }} things we have. | 17:50 |
| *** adrian_otto has quit IRC | 17:51 | |
| kfox1111 | then we put in a dynamic dep macro in the mariadb main template that has the service: - .Values.element_name logic in it? | 17:51 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP Keystone service https://review.openstack.org/420791 | 17:51 |
| sbezverk | kfox1111: sure, let me merge my service, then you can add your macro and even tested at the gate, how about that? | 17:54 |
| kfox1111 | k. | 17:54 |
| kfox1111 | it just means the deps may need to get removed from that service.values and moved back to the microservice package. | 17:54 |
| sbezverk | just for 3 flakes mariadb/rabbitmq and memcached | 17:55 |
| kfox1111 | yeah. | 17:56 |
| inc0 | kfox1111, hmm, so really what you'll be doing is something like installing repo via apt | 17:56 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP Keystone service https://review.openstack.org/420791 | 17:56 |
| inc0 | this is quite good imho | 17:56 |
| kfox1111 | inc0: yeah. thats where the idea came from. :) | 17:56 |
| inc0 | do we need to do something in helm to make it work? | 17:56 |
| inc0 | kfox1111, would you mind summarizing this idea in https://github.com/kubernetes/charts/issues/395 ? | 17:58 |
| *** mgoddard has quit IRC | 17:58 | |
| inc0 | I like it | 17:58 |
| kfox1111 | inc0: they are working on adding a feature that they said, once implemented would make it really simple to do. | 17:58 |
| openstackgerrit | Merged openstack/kolla-ansible: Move mongodb precheck into its own role https://review.openstack.org/416060 | 17:59 |
| inc0 | sounds good, I'm glad we get this out of the way | 17:59 |
| kfox1111 | so we can still maintain our own versioned repo's, and have visibility in kubernetes/charts. :) | 18:01 |
| sbezverk | kfox1111: since you know how it should be, part of your macro ps would be moving deps to whatever you want. moving it now does not seems right as I do not grasp the idea.. | 18:01 |
| sbezverk | at the same time moving deps would be a part of testing.. | 18:01 |
| kfox1111 | sbezverk: I'm ok with that plan. | 18:02 |
| inc0 | yeah, maintaining main repo in gerrit is hands down a must | 18:02 |
| openstackgerrit | Merged openstack/kolla-ansible: Move swift precheck into its own role https://review.openstack.org/418762 | 18:02 |
| inc0 | cool, so kfox1111 tell me, with our current repo structure, we will achieve helm search compatibility right? | 18:03 |
| inc0 | I mean all helmy features will work | 18:03 |
| kfox1111 | initially, helm search would show kolla major versions. | 18:03 |
| kfox1111 | then once one is installed, | 18:03 |
| kfox1111 | then helm search would show the rest. | 18:03 |
| inc0 | that's cool | 18:04 |
| inc0 | let's focus on getting this done too please | 18:04 |
| kfox1111 | its mostly in helms court. | 18:04 |
| kfox1111 | we just continue along as normal. | 18:04 |
| inc0 | well, focusing may mean sticking with them, asking around and reviewing their code | 18:05 |
| kfox1111 | yeah. | 18:05 |
| inc0 | tell me, our current repo structure won't stand in the way? I mean with our microservice arch we'll be compatible with all the helm features? | 18:06 |
| inc0 | and to get rid of our build script, we'll need to figure out "templated" requirements? anything else? | 18:06 |
| sbezverk | inc0: are you refering to helm_prebuild script? | 18:07 |
| kfox1111 | they donn't really care how we build our packages. but if we want to be more "helm native", for whatever that means, yeah, they need some more features. | 18:07 |
| inc0 | I think I do..thing that we do to template stuff | 18:07 |
| sbezverk | inc0: the main thing it does is packaging, which is not in helm | 18:08 |
| sbezverk | even att folks use make tool for packaging | 18:08 |
| sbezverk | same as sap | 18:08 |
| kfox1111 | yeah. helm doesn't have any multipackage building features. | 18:09 |
| inc0 | sbezverk, yeah, well, but I understand that it's not ideal | 18:09 |
| kfox1111 | heh. | 18:09 |
| kfox1111 | what is "ideal" ? :) | 18:09 |
| sbezverk | inc0: :-) | 18:09 |
| kfox1111 | "computer, do my job for me"... ideal.... :) | 18:09 |
| inc0 | kfox1111, ideal would be helm install openstack/newton -> helm install kolla/nova | 18:10 |
| sbezverk | kfox1111: if helm does not provide tool, they cannot keep us accountable for not using what does not exist ;) | 18:10 |
| inc0 | with config file | 18:10 |
| kfox1111 | sbezverk: +1 | 18:10 |
| kfox1111 | inc0: yeah, but the ydo that from packages usually, not source charts. | 18:10 |
| inc0 | sbezverk, kfox1111 but it's open source which means we can help them grow features we need | 18:10 |
| *** tonanhngo has joined #openstack-kolla | 18:10 | |
| kfox1111 | inc0: its opensource, so devs can do that, if they care. | 18:10 |
| srwilkers | ;) | 18:11 |
| kfox1111 | I really prefer to deploy from packages, not source. | 18:11 |
| inc0 | or we can do this, or sobmebody | 18:11 |
| inc0 | what I keep saying is, nothing will get done unless someone frame the problem correctly | 18:11 |
| kfox1111 | so prebuild doesn't bother me at al. | 18:11 |
| *** tonanhngo has quit IRC | 18:11 | |
| inc0 | but if we'd build packages on the fly, that's same thing | 18:12 |
| inc0 | I mean, at the end of the day it's just yaml for k8s | 18:12 |
| kfox1111 | no, its a repo of prebuild packages. | 18:12 |
| kfox1111 | users don't nessisarily need to build from source. | 18:12 |
| kfox1111 | they just helm install kolla/foo | 18:12 |
| kfox1111 | rather the checkou from git, do some building, then helm install kolla/foo | 18:13 |
| kfox1111 | same with kubernetes/charts. | 18:13 |
| kfox1111 | users don't chckout kubernetes/charts. they use prebuilt ones. | 18:13 |
| inc0 | so it's like helm install openstack/newton, kolla-k8s build --conf /my/conf, helm install kolla/nova? | 18:13 |
| kfox1111 | just helm search/helm install. | 18:13 |
| kfox1111 | helm install stable/openstack-newton | 18:14 |
| kfox1111 | helm install kolla/nova --values ~/mycloud.yaml | 18:14 |
| kfox1111 | (ultimately... still not sure how config gen wil lwork though.) | 18:14 |
| *** portdirect is now known as intlabs | 18:14 | |
| inc0 | when will build happen? on helm install stable/openstack-newton? | 18:14 |
| openstackgerrit | Mauricio Lima proposed openstack/kolla-ansible: Move some prechecks into its own role https://review.openstack.org/420899 | 18:15 |
| kfox1111 | when we release. | 18:15 |
| inc0 | ahh so *we* do the building? | 18:15 |
| kfox1111 | yes. | 18:15 |
| inc0 | not users? | 18:15 |
| inc0 | then I'm totally cool with this | 18:15 |
| kfox1111 | users can, but most of the time, they don't need to. | 18:15 |
| inc0 | that was my misunderstanding | 18:15 |
| kfox1111 | yeah. | 18:15 |
| inc0 | sorry guys, I'm still wrapping my head around this all | 18:15 |
| kfox1111 | sokay. | 18:16 |
| *** mdnadeem has quit IRC | 18:16 | |
| *** intlabs has left #openstack-kolla | 18:16 | |
| inc0 | so, back to the requirements templating, I thought we do building to for example turn off neutron-lbaas | 18:16 |
| *** Satyap has quit IRC | 18:16 | |
| inc0 | how to we determine which neutron packages gets installed today? | 18:16 |
| *** portdirect has joined #openstack-kolla | 18:17 | |
| kfox1111 | thats containers? | 18:18 |
| kfox1111 | this is k8s objects. | 18:19 |
| inc0 | what I'm askin is, in values, we provide "enable_lbaas" or something right? | 18:19 |
| kfox1111 | if there was a need for lbaas specific k8s object code, we would. | 18:20 |
| kfox1111 | not sure we need that though. | 18:20 |
| kfox1111 | or if its a pure container tweak. | 18:20 |
| inc0 | how do user define which openstack services helm install will install? | 18:20 |
| inc0 | like, helm install kolla/neutron will install lbaas or not? | 18:20 |
| kfox1111 | hmm.... depends on how config happens I guess. and if lbaas needs its own objects. | 18:21 |
| kfox1111 | (and lbaas is going away... ) | 18:21 |
| inc0 | there are more neutron subservices I can use as example;) | 18:21 |
| inc0 | I know, octavia | 18:21 |
| inc0 | nova network is going away too btw | 18:22 |
| kfox1111 | octavia is a seperate thing, so probably would get its own service package? | 18:22 |
| *** srwilkers is now known as srwilkers_away | 18:22 | |
| kfox1111 | yeah. I'm not implementing nova-network. ;) | 18:22 |
| inc0 | yeah, I'd say so | 18:22 |
| *** shardy has quit IRC | 18:24 | |
| *** satyar has quit IRC | 18:31 | |
| sbezverk | inc0: if you have time please: https://review.openstack.org/#/c/412237/ | 18:31 |
| *** g3ek has quit IRC | 18:33 | |
| *** tonanhngo has joined #openstack-kolla | 18:33 | |
| *** mgoddard has joined #openstack-kolla | 18:34 | |
| *** salv-orlando has quit IRC | 18:35 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Neutron service package https://review.openstack.org/412239 | 18:36 |
| *** salv-orlando has joined #openstack-kolla | 18:36 | |
| inc0 | guys did we code in galera or not? | 18:36 |
| inc0 | we still run 1 node mariadb? | 18:36 |
| kfox1111 | https://review.openstack.org/#/c/420872/ is ready. | 18:37 |
| kfox1111 | no galera yet. | 18:37 |
| *** prameswar has quit IRC | 18:37 | |
| *** tonanhngo has quit IRC | 18:38 | |
| openstackgerrit | Merged openstack/kolla-kubernetes: Mariadb service deployment https://review.openstack.org/412237 | 18:41 |
| *** g3ek has joined #openstack-kolla | 18:42 | |
| *** dave-mccowan has joined #openstack-kolla | 18:44 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: Switch gate to 3.0.2 https://review.openstack.org/420009 | 18:44 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/kolla-kubernetes: Updated from global requirements https://review.openstack.org/419343 | 18:46 |
| inc0 | kfox1111, when I finish repo download, can we switch to master | 18:49 |
| *** tonanhngo has joined #openstack-kolla | 18:49 | |
| inc0 | ? | 18:49 |
| kfox1111 | inc0: no/yes/sortof. | 18:49 |
| *** tonanhngo has quit IRC | 18:49 | |
| kfox1111 | I just got in multiversion jobs. | 18:49 |
| inc0 | that was helmful | 18:49 |
| kfox1111 | so we have hooks for 2/3/trunk. | 18:49 |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla-kubernetes: WIP - local registry in gates https://review.openstack.org/418550 | 18:52 |
| *** sacharya has joined #openstack-kolla | 18:54 | |
| dmsimard | Noob question -- is btrfs the fastest docker storage driver ? I see that's what is set up in the Kolla gate. | 18:54 |
| inc0 | dmsimard, well, btrfs and overlay are quite head-to-head | 18:55 |
| inc0 | question is what distro and stuff like that | 18:55 |
| dmsimard | yeah I think overlay needs >=3.18 | 18:55 |
| inc0 | aufs is not horrible speed too (it's different horrible) | 18:55 |
| inc0 | really devicemapper is only really slow one | 18:55 |
| kfox1111 | inc0: should rebase on to https://review.openstack.org/#/c/420872/ | 18:56 |
| portdirect | overlay2 is probably fastest (certainly more memory efficient), but as inc0 says distro support is varied | 18:56 |
| dmsimard | yeah the default is devicemapper it looks like | 18:56 |
| kfox1111 | and hook in kind of like: https://review.openstack.org/#/c/420009/ | 18:56 |
| inc0 | also overlay2 is young | 18:56 |
| dmsimard | So I should use something like btrfs on CentOS instead ? I'm not deploying, just interested in building stuff faster. | 18:56 |
| kfox1111 | if centos, you get devicemapper or btrfs. | 18:56 |
| kfox1111 | btrfs seems more reliable. | 18:56 |
| inc0 | I've heard of stranger things happening | 18:56 |
| inc0 | so I'd roll with btrfs | 18:56 |
| dmsimard | Thanks, I'll try that. | 18:57 |
| inc0 | if you had default (devicemapper), you'll see quite a difference once you move to btrfs | 18:57 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: test rebooting in the gate https://review.openstack.org/413193 | 18:57 |
| portdirect | inc0: rememebr that device mapper comes in multiple forms | 18:57 |
| inc0 | yeah, but default is with loopbacks right? | 18:58 |
| portdirect | the default loopback is terrible thinlvm not so bad | 18:58 |
| inc0 | I mean docker creates loopback for each layer | 18:58 |
| inc0 | and that shit is slow | 18:58 |
| kfox1111 | portdirect: review please? https://review.openstack.org/#/c/420872/ | 18:58 |
| kfox1111 | inc0: slow's not my concert. its the fact it ate itself 3 times on me. :/ | 18:59 |
| kfox1111 | concern | 18:59 |
| *** Mr_Broken has joined #openstack-kolla | 18:59 | |
| *** tonanhngo has joined #openstack-kolla | 18:59 | |
| dmsimard | eating yourself is quite an inconvenience :) | 18:59 |
| *** tonanhngo has quit IRC | 18:59 | |
| kfox1111 | yeah. you really don't want your storage layer to start having problems. :/ | 19:00 |
| portdirect | inc0: it does not create a loopback for each layer, but a two loopback devices that it then runs lvm inside - but i still world 100% not recomend it :) | 19:00 |
| inc0 | my bad then, well, btrfs is what I'd run | 19:02 |
| inc0 | bottom line | 19:02 |
| kfox1111 | +1 | 19:02 |
| portdirect | +1 :) | 19:02 |
| kfox1111 | at least on centos. | 19:02 |
| inc0 | on buntu too | 19:03 |
| kfox1111 | till at least 7.4. :) | 19:03 |
| inc0 | again, I've heard some strange shit happening with overlay | 19:03 |
| kfox1111 | ah. ok. | 19:03 |
| inc0 | but that might be resolved now | 19:03 |
| inc0 | still, I don't trust it yet | 19:03 |
| portdirect | inc0: that is with kernels <4.2 i think | 19:03 |
| kfox1111 | it would be nice if zuul showed you the ps name.... | 19:03 |
| inc0 | quite possibly, it's my superstitious nature speaking | 19:04 |
| portdirect | inc0: used to be whiteout issues with the way docker did it, that was fixed with overlay2 | 19:04 |
| portdirect | inc0: i used to have a docker file to demonstrate it somewhere - you would delete a file and then two layers later it would be back :) | 19:05 |
| kfox1111 | fun. | 19:05 |
| openstackgerrit | Merged openstack/kolla-kubernetes: Common infra for enabling 3.x and trunk builds. https://review.openstack.org/420872 | 19:05 |
| inc0 | I've seen stuff happening to your files | 19:05 |
| inc0 | like names changed spontaneusly to $*!(#*!@(# | 19:05 |
| *** srwilkers_away is now known as srwilkers | 19:06 | |
| inc0 | bottom line, I'm mistrustruf towards it, but it shows promise | 19:06 |
| portdirect | wow thats new to me :/ | 19:06 |
| inc0 | well, that was over year ago | 19:06 |
| inc0 | so well | 19:07 |
| inc0 | stuff changed since probably | 19:07 |
| inc0 | and it was not overlay2, it was overlay | 19:07 |
| jrich523 | is there a CLI command i can run to see if a network is a provider network or not? | 19:07 |
| inc0 | neutron net-show I think shows type of network | 19:07 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP Keystone service https://review.openstack.org/420791 | 19:07 |
| jrich523 | cool thanks | 19:08 |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla-kubernetes: WIP - local registry in gates https://review.openstack.org/418550 | 19:13 |
| *** matrohon has joined #openstack-kolla | 19:14 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/kolla-kubernetes: Updated from global requirements https://review.openstack.org/419343 | 19:16 |
| kfox1111 | if your an admin net show will give you details. | 19:16 |
| Pavo | can someone tese ddi.hopto.org:8080 please | 19:17 |
| jrich523 | hmm well i get details, and, i would expect either provider:network_type or router:external to tell me its a provider | 19:17 |
| jrich523 | but looking at a neutron (right term?) network vs a provider and both of those are the same values | 19:17 |
| jrich523 | vlan/False | 19:17 |
| inc0 | mattermost cames up PAv | 19:17 |
| inc0 | Pavo, | 19:17 |
| Pavo | ty | 19:17 |
| Pavo | setup a Mattermost server | 19:18 |
| inc0 | Pavo, after this few months, still liking Kolla?:) | 19:18 |
| Pavo | absolutely | 19:18 |
| Pavo | WAY better than packstack | 19:18 |
| Pavo | more customizable | 19:18 |
| inc0 | well, that's not hard frankly | 19:18 |
| Pavo | but still docs think need to more user friendly than operator frendly | 19:19 |
| *** tonanhngo has joined #openstack-kolla | 19:19 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: Support 3.0.2 in the gate https://review.openstack.org/420009 | 19:20 |
| *** tonanhngo has quit IRC | 19:20 | |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla-kubernetes: WIP - local registry in gates https://review.openstack.org/418550 | 19:23 |
| Pavo | inc0 try ddi.hopto.org:8080/openstack-kolla | 19:25 |
| *** mgoddard has quit IRC | 19:28 | |
| *** tonanhngo has joined #openstack-kolla | 19:29 | |
| *** tonanhngo has quit IRC | 19:30 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Neutron service package https://review.openstack.org/412239 | 19:35 |
| inc0 | Pavo, redirects to login, but that seems right | 19:35 |
| *** msimonin has joined #openstack-kolla | 19:37 | |
| *** fguillot has quit IRC | 19:38 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: Support 3.0.2 in the gate https://review.openstack.org/420009 | 19:39 |
| *** sdake has joined #openstack-kolla | 19:57 | |
| *** nrado has left #openstack-kolla | 20:11 | |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla-kubernetes: WIP - local registry in gates https://review.openstack.org/418550 | 20:13 |
| sdake | sup peeps | 20:17 |
| sdake | hey kfox1111 | 20:17 |
| kfox1111 | hey | 20:17 |
| sdake | sup inc0 | 20:17 |
| sdake | inc0 did you call? | 20:17 |
| *** salv-orlando has quit IRC | 20:23 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: test rebooting in the gate https://review.openstack.org/413193 | 20:26 |
| sdake | sbezverk can any of these gates be made from experimental to nonvoting? | 20:27 |
| sdake | or kfox1111 | 20:28 |
| sdake | it would speed up the gating process if they were | 20:28 |
| sdake | kfox1111 sbezverk the idea of experimental gates is you keep them experiemetnal for 1-2 weeks then make them real | 20:28 |
| sdake | if we are check experiemtanl on every commit - might as weell promote the parts that matter | 20:28 |
| kfox1111 | sdake: I think prdetty much everything that is non experimental now can be protomoted to voting. | 20:29 |
| kfox1111 | the iscsi stuff can be promoted to nonvoting | 20:29 |
| sdake | ok - one word of of caution there is once they are voting, if the gate job blocks there is no "force it through" method | 20:29 |
| *** inc0 has quit IRC | 20:29 | |
| kfox1111 | the rest are still experimental. | 20:30 |
| sdake | kfox1111 roger | 20:30 |
| kfox1111 | yeah. we have been treating the non voting ones largely as voting already. | 20:30 |
| sdake | kfox1111 yup | 20:30 |
| sdake | kfox1111 kolla-ansible was in this sme spot however we didn't move to voting gates because of the general flakiness of the gate | 20:30 |
| sdake | and the fact we gated on master | 20:31 |
| sdake | (of openstack) | 20:31 |
| kfox1111 | its pretty stable now. if we could cache deps, would probably fix most of the random everyonce in a while's too. | 20:31 |
| sdake | fair enough i'll make it so | 20:31 |
| *** rhallisey has joined #openstack-kolla | 20:32 | |
| kfox1111 | so, gate-kolla-kubernetes-deploy[centos-binary-2-ceph-multi, centos-binary-2-external-ovs, centos-binary-2-ceph and ubuntu-binary-2-ceph] go to voting. | 20:32 |
| kfox1111 | and gate-kollla-kubernetes-deploy[centos, ubuntu]-2-iscsi go to check with non voting. | 20:33 |
| kfox1111 | thanks. :) | 20:33 |
| sdake | kfox1111 i've been filing bugs in launchpad for flakey gate scenarios | 20:33 |
| sdake | and pla nto fix them | 20:33 |
| sdake | kfox1111 sbezverk said you had an ehterpad for these issues | 20:33 |
| kfox1111 | we've been doing it a bit differently. | 20:33 |
| sdake | can you link it? | 20:33 |
| kfox1111 | https://etherpad.openstack.org/p/kolla-kubernetes-gate-failure-tracking | 20:33 |
| kfox1111 | keeping patterns of crashed gates there. | 20:33 |
| sdake | thanks - would like it in bug tracker if possible - nobody can find that etherpad :) | 20:33 |
| *** dave-mccowan has quit IRC | 20:34 | |
| kfox1111 | once we get enough of them, usually a trend starts showing, and we can start addressing it. | 20:34 |
| sdake | I told him I would move that etherpad to bugs | 20:34 |
| kfox1111 | its hard to track the issues in bugs. :/ | 20:34 |
| sdake | i disagree - we can use a gate tag | 20:34 |
| kfox1111 | cause you don't know where a given failure run belongs. | 20:34 |
| *** msimonin1 has joined #openstack-kolla | 20:34 | |
| sdake | let me look at your etherpad moment :) | 20:34 |
| kfox1111 | 90% of the gate issues became prety obvious when enough of them hit the etherpad. | 20:35 |
| kfox1111 | then I fixed it, then deleted the entries. | 20:35 |
| *** msimonin has quit IRC | 20:35 | |
| sdake | ya we already know rax is flakey | 20:35 |
| sdake | i'd be surprised ifyou ever have a gate from rax that works :) | 20:35 |
| kfox1111 | so far, surprisingly there hasn't been too many rax only issues to show. | 20:35 |
| sdake | i mean a gate jobs run | 20:35 |
| sdake | that is because rax is rarely scheduled | 20:35 |
| kfox1111 | I've seen quite a few rax jobs actually. | 20:36 |
| sdake | that were successful? | 20:36 |
| kfox1111 | yup. | 20:36 |
| sdake | that is good data | 20:36 |
| sdake | i guess i can file a blueprint instead with a link to the etherpad | 20:36 |
| kfox1111 | I'd say I've seen more osic failures, but I think thats due to them getting scheduled more. | 20:36 |
| kfox1111 | yeah. that would work. | 20:36 |
| sdake | u guys are both pink ;) | 20:37 |
| kfox1111 | random color selection. | 20:37 |
| sdake | yup | 20:37 |
| kfox1111 | somehow I changed to purple too. | 20:37 |
| kfox1111 | I never changed it. | 20:38 |
| sdake | just kidding around :) | 20:38 |
| kfox1111 | no worries. :) | 20:38 |
| kfox1111 | bleh. the big tent has been such a desaster. :/ | 20:38 |
| *** adrian_otto1 has quit IRC | 20:40 | |
| sdake | kfox1111 what is binary-2 and binary-3 - number of nodes? | 20:46 |
| kfox1111 | 2.x, 3.x, and trunk. | 20:47 |
| sdake | multi is 2 nodes then? | 20:48 |
| kfox1111 | yeah. | 20:48 |
| sdake | cool | 20:48 |
| kfox1111 | multinode. | 20:48 |
| sdake | waht is the master designiation - just nothing? | 20:48 |
| kfox1111 | t | 20:48 |
| kfox1111 | trunk. | 20:48 |
| sdake | 2.x = of images and ansible genconfig? | 20:49 |
| sdake | or just of images | 20:49 |
| kfox1111 | just images. | 20:49 |
| kfox1111 | kolla trunk for genconfig. | 20:49 |
| sdake | cool thought so - just wanted to valicate | 20:49 |
| kfox1111 | yeah. 2.x didn't have neough genconfig bits to do kolla-kubernetes. | 20:49 |
| sdake | yup i knwo | 20:49 |
| sdake | know | 20:49 |
| kfox1111 | so we're using kolla-ansible trunk for all the jobs there. | 20:50 |
| sdake | got it | 20:50 |
| kfox1111 | might be nice to switch to 3.x release at somepoint for 2.x/3.x though. | 20:50 |
| sdake | i thik we need another set of gate jobs | 20:50 |
| sdake | as we add services, sbezverk is removing the microservice gate jobs | 20:50 |
| sdake | ideally we can test at each layer - especially since the code is written | 20:51 |
| kfox1111 | no. | 20:53 |
| kfox1111 | we forked the gate code. | 20:53 |
| sdake | just o be clear, you want master gate to be voting? | 20:53 |
| kfox1111 | microservices have a job, services have a job. | 20:53 |
| kfox1111 | so that test code isn't getting lost. | 20:53 |
| sdake | what is the job delta name? | 20:54 |
| kfox1111 | for now, 2 is voting as its proven stable. | 20:54 |
| kfox1111 | 3 will become voting as it becomes stable. | 20:54 |
| kfox1111 | trunk will be nonvoting as it can break randomly but still good to catch issues early with it. | 20:54 |
| kfox1111 | job delta name? | 20:54 |
| sdake | how do the names distinguish between the services and microservices jobs in the gate? | 20:55 |
| sdake | i understand the code has been copied | 20:55 |
| kfox1111 | helm-entrypoint is the helm+kubernetes-entrypoint service packages. | 20:55 |
| sdake | cool thanks | 20:55 |
| kfox1111 | helm-operator is for helm microservices deployed with operators | 20:56 |
| sdake | the reboot job is which? | 20:57 |
| kfox1111 | all in one install that hard reboots after deployed. | 20:57 |
| kfox1111 | work in progress. | 20:57 |
| kfox1111 | but will check that it can stand back up after power failure. | 20:57 |
| sdake | pretty sure you can't hard reboot in the gate | 20:57 |
| sdake | zuul loses contact with the node under management | 20:57 |
| kfox1111 | sdake: you can. :) | 20:57 |
| sdake | cool that must be a new feature ;) | 20:58 |
| kfox1111 | you just cant reboot the primary. so I do everything on a secondary node and watch it. :) | 20:58 |
| kfox1111 | https://review.openstack.org/#/c/413193/27/tests/bin/gate_reboot_master.sh | 20:58 |
| sdake | thats a good workaround! | 20:58 |
| kfox1111 | yup. :) | 20:58 |
| sdake | kolla-ansible needs some of this gate-foo :) | 20:58 |
| kfox1111 | would be nice. | 20:59 |
| kfox1111 | time is always a constraint though. :/ | 20:59 |
| sdake | right | 20:59 |
| sdake | long term i mean | 20:59 |
| sdake | kolla-ansible deliverable has struggled making more advanced gate jobs | 20:59 |
| kfox1111 | yeah. I think part of the problem is it is almost more an operator task then a dev task. | 21:00 |
| sdake | sort of - although we have to cross both boundaries when developing gate jobs | 21:01 |
| sdake | gate jobs are very difficult to write :) | 21:01 |
| sdake | learning the mechanics of the gate itself is daunting | 21:01 |
| kfox1111 | ops are more familior with the tricks/weirdness that can show up across different machines. | 21:01 |
| kfox1111 | and there are quite the variaty across the openstack gates. | 21:01 |
| sdake | every provider is slightly different | 21:02 |
| kfox1111 | yeah. and distro too. | 21:02 |
| sbezverk | kfox1111: hey what is up with the gate? why I see skipped ? | 21:02 |
| sdake | why have 1 snowflake when you can have 9 :) | 21:02 |
| kfox1111 | what works for ubuntu wont work for centos... | 21:02 |
| kfox1111 | sbezverk: which thing skipped? | 21:02 |
| kfox1111 | sdake: heh. yah... story of my life. :/ | 21:02 |
| sbezverk | kfox1111: it is in the email notification | 21:03 |
| sdake | each infra also uses sometimes not kvm - but other virt infrastructure in the host kernel | 21:03 |
| kfox1111 | sbezverk: email? did you send me one? | 21:04 |
| sbezverk | kfox1111: forwarded to you | 21:04 |
| sdake | sbezverk the best place to ask those questions in #openstack-infra | 21:04 |
| sbezverk | kfox1111: I get email from jenkins | 21:04 |
| sdake | sbezverk looking at your gate jobs, they all look green to me | 21:04 |
| kfox1111 | ah. | 21:04 |
| sbezverk | sdake: not when kfox1111 is playing with gate jobs | 21:04 |
| kfox1111 | sbezverk: yeah, I get email too. but doing 4 patches in parallel right now, so way to much to keep track of. :/ | 21:04 |
| kfox1111 | sbezverk: never seen that before. no idea why skipped. | 21:05 |
| kfox1111 | sbezverk: did you submit a newer ps maybe and it skiped running the older? | 21:05 |
| sbezverk | sdake: you see, no need yo bother infra folks :-) it is internal stuff | 21:05 |
| sdake | sbezverk I have seen skipped before - not sure what the root cause was | 21:05 |
| kfox1111 | I'm submitting new gate jobs, but not affecting the current stuff. | 21:05 |
| kfox1111 | they are still in review. | 21:05 |
| kfox1111 | so that skipped thing is unknown. might ask infra about it. | 21:06 |
| sdake | mifhgt be a hesinenbug | 21:06 |
| sdake | typically openstack ci is 5-10% hesienbugs | 21:06 |
| sdake | because project gate code is wrong or the providers are wrong or something just goes wrong | 21:07 |
| sdake | hence the recheck feature :) | 21:07 |
| sdake | what we dont want is 90% failure rate on a particular gate job that is nonvoting | 21:07 |
| sdake | kfox1111 i htink promoting those gate jobs while your reviews are outstanding may cause a conflict in the redviews | 21:08 |
| sdake | kfox1111 so rather then forcey ou to rebase, ill hold off until your done adding jobs :) | 21:08 |
| *** salv-orlando has joined #openstack-kolla | 21:09 | |
| kfox1111 | sdake: no, there's a shell script interface. | 21:09 |
| kfox1111 | the gate jobs call the shell script with some static args. | 21:10 |
| kfox1111 | the new jobs are just looking at the args. | 21:10 |
| kfox1111 | so we're good. | 21:10 |
| sdake | yes but you have to add the jobs to gerrit | 21:10 |
| kfox1111 | no. | 21:10 |
| kfox1111 | they are already in the gate. | 21:10 |
| kfox1111 | just stubbed out in the code to always emediately pass. | 21:10 |
| sdake | you said you have 4 gate jobs outstanding | 21:11 |
| kfox1111 | the work in progress is to just make them actually do something useful. :) | 21:11 |
| sdake | ah god it | 21:11 |
| sdake | hot it | 21:11 |
| sdake | got it | 21:11 |
| sdake | groan | 21:11 |
| kfox1111 | https://github.com/openstack/kolla-kubernetes/blob/master/tools/setup_gate.sh#L6 | 21:11 |
| kfox1111 | :) | 21:11 |
| *** nihilifer has quit IRC | 21:14 | |
| *** JoseMello has quit IRC | 21:15 | |
| *** nihilifer has joined #openstack-kolla | 21:15 | |
| *** rhallisey has quit IRC | 21:19 | |
| sdake | does the external ovs sgate actually work? | 21:20 |
| sdake | i mean test anything :) | 21:20 |
| kfox1111 | yup. | 21:20 |
| kfox1111 | I'm going to need that functionality for my clouds. | 21:20 |
| sdake | got it | 21:20 |
| sdake | just wondering if it was finished implementeing yet | 21:20 |
| kfox1111 | yup. | 21:20 |
| kfox1111 | should be promoted to voting. | 21:20 |
| sbezverk | iscsi were stable since we moved to 3.0.2 | 21:22 |
| sbezverk | could be also used for voting | 21:22 |
| kfox1111 | sbezverk: i asked for it to be promoted to check/nv. | 21:22 |
| kfox1111 | so it will always run. | 21:22 |
| sbezverk | cool, thanks | 21:22 |
| kfox1111 | its been stable for only a few days, so maybe in a few weeks we promote to voting if it stays that way. | 21:22 |
| *** tonanhngo has joined #openstack-kolla | 21:23 | |
| *** sdake_ has joined #openstack-kolla | 21:23 | |
| kfox1111 | I'm pretty confident it will stay stable. | 21:23 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: test rebooting in the gate https://review.openstack.org/413193 | 21:24 |
| *** sdake has quit IRC | 21:24 | |
| *** sayantani01 has joined #openstack-kolla | 21:27 | |
| *** sdake_ has quit IRC | 21:27 | |
| *** sdake has joined #openstack-kolla | 21:28 | |
| *** tonanhngo has joined #openstack-kolla | 21:28 | |
| *** nihilifer has quit IRC | 21:29 | |
| *** nihilifer has joined #openstack-kolla | 21:29 | |
| *** dave-mccowan has joined #openstack-kolla | 21:32 | |
| *** adrian_otto has joined #openstack-kolla | 21:33 | |
| *** l4yerffeJ has joined #openstack-kolla | 21:34 | |
| *** l4yerffeJ__ has quit IRC | 21:34 | |
| *** Jeffrey4l has quit IRC | 21:34 | |
| *** Jeffrey4l has joined #openstack-kolla | 21:35 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: Support 3.0.2 in the gate https://review.openstack.org/420009 | 21:40 |
| *** inc0 has joined #openstack-kolla | 21:41 | |
| *** sdake has quit IRC | 21:41 | |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla-kubernetes: WIP - local registry in gates https://review.openstack.org/418550 | 21:43 |
| *** absubram_ has joined #openstack-kolla | 21:46 | |
| *** salv-orlando has quit IRC | 21:47 | |
| *** v1k0d3n has joined #openstack-kolla | 21:50 | |
| *** dave-mccowan has quit IRC | 21:54 | |
| kfox1111 | inc0: did you test that the insecure flag is really actually needed for localhost repo's? | 21:54 |
| *** sdake has joined #openstack-kolla | 21:57 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP memcached operator https://review.openstack.org/418073 | 21:57 |
| *** inc0 has quit IRC | 21:59 | |
| *** absubram_ has quit IRC | 22:00 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Keystone service deployment with kubernetes entrypoint https://review.openstack.org/420791 | 22:03 |
| *** lrensing has quit IRC | 22:04 | |
| *** gfidente|afk has quit IRC | 22:07 | |
| *** sdake_ has joined #openstack-kolla | 22:12 | |
| *** salv-orlando has joined #openstack-kolla | 22:12 | |
| *** sdake has quit IRC | 22:14 | |
| *** jheroux has quit IRC | 22:22 | |
| *** salv-orl_ has joined #openstack-kolla | 22:24 | |
| *** salv-orlando has quit IRC | 22:26 | |
| *** matrohon has quit IRC | 22:35 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: test rebooting in the gate https://review.openstack.org/413193 | 22:42 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Keystone service deployment with kubernetes entrypoint https://review.openstack.org/420791 | 22:43 |
| *** msimonin1 has quit IRC | 22:57 | |
| *** msimonin has joined #openstack-kolla | 22:58 | |
| *** inc0 has joined #openstack-kolla | 23:02 | |
| inc0 | kfox1111, sorry went for a tea, so without this flag it threw TLS error | 23:03 |
| *** schwicht_at_work has joined #openstack-kolla | 23:03 | |
| kfox1111 | would it work just http:? | 23:04 |
| kfox1111 | no real reason to encrypt traffic that wont leave the node? | 23:04 |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla-kubernetes: WIP - local registry in gates https://review.openstack.org/418550 | 23:04 |
| inc0 | well let me put it this way, so far I couldn't get it work anyhow | 23:05 |
| inc0 | when it works with flag, I'll try removing it | 23:05 |
| *** schwicht has quit IRC | 23:05 | |
| inc0 | and see if it keeps working | 23:05 |
| inc0 | keep working | 23:05 |
| inc0 | omg my english is metalocalypse | 23:06 |
| *** krtaylor has quit IRC | 23:06 | |
| *** schwicht has joined #openstack-kolla | 23:08 | |
| *** v1k0d3n has quit IRC | 23:08 | |
| inc0 | kfox1111, that looks like fun http://logs.openstack.org/50/418550/10/check/gate-kolla-kubernetes-deploy-ubuntu-binary-2-ceph-nv/28ad501/console.html#_2017-01-16_20_21_24_106773 | 23:08 |
| kfox1111 | fair enough. :) | 23:09 |
| *** schwicht_at_work has quit IRC | 23:10 | |
| kfox1111 | just throught it might be easier to not have to mess with the docker unit files. | 23:10 |
| inc0 | we do it anyway to setup overlay2 | 23:11 |
| kfox1111 | it wasn't required before. | 23:11 |
| inc0 | ok | 23:12 |
| inc0 | for some reason I forgot I added this code | 23:12 |
| inc0 | yeah, when it gets up, I'll check if it's needed | 23:13 |
| *** srwilkers has quit IRC | 23:13 | |
| inc0 | I think I copied it from kolla registry | 23:13 |
| inc0 | well, yeah, and overlay2 will just break on centos | 23:15 |
| inc0 | so meh | 23:15 |
| inc0 | well it does seem to work even on centos | 23:18 |
| kfox1111 | really? | 23:21 |
| kfox1111 | did overlay2 make it into 7.3? | 23:21 |
| inc0 | kfox1111, http://logs.openstack.org/50/418550/12/check/gate-kolla-kubernetes-deploy-centos-binary-2-ceph-nv/d6d8a02/console.html#_2017-01-16_23_14_05_534328 | 23:22 |
| sbezverk | kfox1111: ready for review when you have time https://review.openstack.org/#/c/420791/ | 23:23 |
| kfox1111 | inc0: was that line before or after the docker restart? | 23:25 |
| inc0 | after | 23:25 |
| inc0 | you can see tee and restarts above | 23:25 |
| kfox1111 | ah. cool. | 23:26 |
| kfox1111 | that seemed to be missing from the centos 7.3 change logs. :/ | 23:26 |
| kfox1111 | I may have to play with that more. :) | 23:26 |
| inc0 | let me know would actually be cool to do overlay2 in gates | 23:27 |
| kfox1111 | sbezverk: is keystone-public-service not already port_external port 5000 in the microservice? | 23:27 |
| kfox1111 | inc0: if its stable/faster, I'm ok with it. | 23:27 |
| inc0 | we talked about it today;) portdirect says it's fine | 23:28 |
| kfox1111 | cool. | 23:28 |
| portdirect | That's with Lt kernel, I didn't know they backported it either. | 23:30 |
| portdirect | But if they have and it works then :) | 23:30 |
| sbezverk | kfox1111: by some reason it is set to false in all.yaml | 23:30 |
| sbezverk | kfox1111: you override it in cepg_workflow | 23:31 |
| sbezverk | ceph_workflow I meant | 23:31 |
| *** sayantani01 has quit IRC | 23:33 | |
| kfox1111 | it may have just been an oversite. | 23:34 |
| kfox1111 | can you think of any reason you wouldn't want the public api to not be exposed? | 23:34 |
| kfox1111 | (or commonly not I mean?) | 23:34 |
| portdirect | inc0: commuting ATM so can't look at that PS properly but are you sure its running with overlay2? | 23:34 |
| inc0 | portdirect, yeah, pretty sure, will check more tho | 23:35 |
| portdirect | I think it will stick with what it was originally started with unless they updated that too | 23:35 |
| sbezverk | kfox1111: not really | 23:35 |
| kfox1111 | sbezverk: maybe we should just tweak the public microservices port_external to be true then? | 23:37 |
| inc0 | hmf it seems my dev vm lived up it's usability | 23:37 |
| *** sayantani01 has joined #openstack-kolla | 23:37 | |
| kfox1111 | seems like a sane default everyone would probably want. | 23:37 |
| inc0 | ls /home/ubuntu -> permission denied | 23:37 |
| inc0 | as ubuntu | 23:37 |
| inc0 | :( | 23:37 |
| sbezverk | kfox1111: sure I can do that | 23:40 |
| sbezverk | about your second comment, we need that $1 since we pass it to the function | 23:40 |
| *** sayantani01 has quit IRC | 23:41 | |
| kfox1111 | where? I didn't see it used anywhere. | 23:42 |
| kfox1111 | maybe I missed it? | 23:42 |
| sbezverk | in if for ceph | 23:42 |
| sbezverk | if [ "x$1" == "xceph-multi" ]; then | 23:43 |
| kfox1111 | oh. I'm just blind... | 23:44 |
| kfox1111 | ok. nm. just ignore me. :) | 23:45 |
| sbezverk | so I changed other things you asked | 23:46 |
| kfox1111 | k. cool. thanks. | 23:46 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Keystone service deployment with kubernetes entrypoint https://review.openstack.org/420791 | 23:46 |
| inc0 | yup, my dev vm is oficially broken | 23:49 |
| inc0 | how fun | 23:49 |
| *** salv-orl_ has quit IRC | 23:50 | |
| inc0 | 6 months it served me | 23:50 |
| inc0 | good night sweet prince | 23:50 |
| sdake_ | kfox1111 i find it highly unlikely that overlay2 works well in centos 7.3 :) | 23:57 |
| *** mattmceuen has quit IRC | 23:57 | |
| sdake_ | kfox1111 overlyafs was a mess in centos7.2 | 23:58 |
| *** adrian_otto has quit IRC | 23:58 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!