*** adrian_otto has quit IRC | 00:04 | |
*** jrist_ has joined #openstack-kolla | 00:09 | |
*** jrist has quit IRC | 00:10 | |
*** Pavo has joined #openstack-kolla | 00:15 | |
*** yingjun has joined #openstack-kolla | 00:17 | |
*** absubram has quit IRC | 00:20 | |
*** inc0 has quit IRC | 00:22 | |
openstackgerrit | Marcus Williams proposed openstack/kolla-ansible: [WIP] Add OpenDaylight to kolla-ansible https://review.openstack.org/416367 | 00:22 |
---|---|---|
*** adrian_otto has joined #openstack-kolla | 00:23 | |
*** sayanta__ has quit IRC | 00:25 | |
*** sayantan_ has joined #openstack-kolla | 00:26 | |
*** jrobinson is now known as jrobinson-afk | 00:27 | |
*** sayantan_ has quit IRC | 00:30 | |
*** sayantan_ has joined #openstack-kolla | 00:31 | |
*** sayanta__ has joined #openstack-kolla | 00:33 | |
*** sayantan_ has quit IRC | 00:33 | |
*** Pavo has quit IRC | 00:36 | |
*** sayanta__ has quit IRC | 00:37 | |
*** jrobinson-afk is now known as jrobinson | 00:42 | |
*** Marcellin__ has quit IRC | 00:49 | |
*** tovin07 has joined #openstack-kolla | 00:51 | |
*** jmccarthy has quit IRC | 00:53 | |
openstackgerrit | zhongshengping proposed openstack/kolla: Add support for installing binary packages of watcher https://review.openstack.org/416426 | 00:53 |
*** sayantan_ has joined #openstack-kolla | 00:54 | |
*** jmccarthy has joined #openstack-kolla | 00:54 | |
*** Pavo has joined #openstack-kolla | 00:58 | |
*** kong_ is now known as kong | 01:00 | |
*** awiddersheim has quit IRC | 01:09 | |
*** sdake has quit IRC | 01:21 | |
*** zhubingbing has joined #openstack-kolla | 01:25 | |
openstackgerrit | Marcus Williams proposed openstack/kolla-ansible: [WIP] Add OpenDaylight role https://review.openstack.org/416367 | 01:26 |
openstackgerrit | Marcus Williams proposed openstack/kolla-ansible: [WIP] Split Openvswitch into own role https://review.openstack.org/416793 | 01:26 |
*** Pavo has quit IRC | 01:30 | |
*** awiddersheim has joined #openstack-kolla | 01:31 | |
*** Pavo has joined #openstack-kolla | 01:31 | |
sbezverk | any kolla-ansible cores online? | 01:43 |
*** adrian_otto has quit IRC | 01:50 | |
*** dave-mccowan has joined #openstack-kolla | 02:02 | |
*** celo_r has joined #openstack-kolla | 02:02 | |
*** l4yerffeJ__ has joined #openstack-kolla | 02:10 | |
openstackgerrit | caoyuan proposed openstack/kolla: Remove kolla-genpwd from setup.cfg https://review.openstack.org/416532 | 02:11 |
*** l4yerffeJ_ has quit IRC | 02:13 | |
*** unicell has quit IRC | 02:20 | |
*** unicell has joined #openstack-kolla | 02:23 | |
*** Pavo has quit IRC | 02:25 | |
*** Pavo has joined #openstack-kolla | 02:25 | |
*** Pavo has joined #openstack-kolla | 02:26 | |
*** prameswar has joined #openstack-kolla | 02:27 | |
*** jtriley has quit IRC | 02:27 | |
*** dave-mccowan has quit IRC | 02:33 | |
*** duonghq has joined #openstack-kolla | 02:36 | |
*** goldyfruit has joined #openstack-kolla | 02:39 | |
*** caowei has joined #openstack-kolla | 02:41 | |
*** unicell has quit IRC | 02:42 | |
*** Pavo has quit IRC | 02:51 | |
openstackgerrit | zhongshengping proposed openstack/kolla: Indent all binary part https://review.openstack.org/416819 | 02:55 |
*** dave-mccowan has joined #openstack-kolla | 02:55 | |
*** jrobinson is now known as jrobinson-afk | 03:01 | |
zhubingbing | hello | 03:01 |
zhubingbing | guys | 03:01 |
*** harlowja has quit IRC | 03:05 | |
*** tonyb has quit IRC | 03:05 | |
*** apuimedo has quit IRC | 03:05 | |
*** jascott1 has quit IRC | 03:05 | |
*** otavio_ has quit IRC | 03:05 | |
*** timss has quit IRC | 03:05 | |
*** srwilker- has quit IRC | 03:05 | |
*** jgriffith has quit IRC | 03:05 | |
*** cloudnull has quit IRC | 03:05 | |
*** cloudnull has joined #openstack-kolla | 03:05 | |
*** tonyb has joined #openstack-kolla | 03:05 | |
*** srwilkers_ has joined #openstack-kolla | 03:05 | |
*** otavio has joined #openstack-kolla | 03:05 | |
*** otavio has quit IRC | 03:05 | |
*** otavio has joined #openstack-kolla | 03:05 | |
*** timss has joined #openstack-kolla | 03:05 | |
*** jgriffith has joined #openstack-kolla | 03:06 | |
*** apuimedo has joined #openstack-kolla | 03:06 | |
openstackgerrit | Marcus Williams proposed openstack/kolla-ansible: [WIP] Add OpenDaylight role https://review.openstack.org/416367 | 03:11 |
Jeffrey4l | sean-k-mooney, still around? | 03:11 |
Jeffrey4l | sup sbezverk | 03:12 |
*** mgiles has quit IRC | 03:14 | |
*** jascott1 has joined #openstack-kolla | 03:14 | |
*** sayanta__ has joined #openstack-kolla | 03:14 | |
*** sayantan_ has quit IRC | 03:18 | |
*** zhurong has joined #openstack-kolla | 03:21 | |
*** mdnadeem has joined #openstack-kolla | 03:23 | |
sp_ | Jeffrey4l: ping.. | 03:25 |
*** goldyfruit has quit IRC | 03:26 | |
cliles | is there an easy way to run kolla-ansible deploy to just deploy 1 service? I tried pruning the inventory file down to only contain the group I want to redeploy, but seems then I miss all the prechecks. | 03:26 |
sp_ | Jeffrey4l: should I abandon this patch ? please suggest https://review.openstack.org/#/c/390179/8 | 03:29 |
sp_ | Jeffrey4l: I have put much effort on this. | 03:31 |
*** l4yerffeJ__ has quit IRC | 03:31 | |
*** jrobinson-afk is now known as jrobinson- | 03:31 | |
*** jrobinson- is now known as jrobinson | 03:31 | |
*** l4yerffeJ__ has joined #openstack-kolla | 03:31 | |
*** mdnadeem has quit IRC | 03:41 | |
*** zhurong has quit IRC | 03:41 | |
sp_ | duonghq: ping.. | 03:43 |
*** srwilkers has quit IRC | 03:45 | |
duonghq | sp_, pong | 03:45 |
*** Pavo has joined #openstack-kolla | 03:47 | |
*** Pavo has quit IRC | 03:47 | |
sp_ | duonghq: for this https://blueprints.launchpad.net/kolla/+spec/support-python-35 what approach should i take ? | 03:48 |
sp_ | duonghq: your help would be highly appreciable | 03:48 |
sp_ | i have submitted i patch too regarding this. | 03:49 |
Jeffrey4l | sp_, i think so. static gid/uid have include you patch, which can not be removed. it is meaningless to merge your patch and implement again in that patch. | 03:50 |
Jeffrey4l | but thanks for your effort ;) | 03:50 |
sp_ | Jeffrey4l: can i be Co-Author to that at least ? :( | 03:50 |
duonghq | sp_, atm, the only I can think is trying to run our script w/ python 3.5, if is there any error isn't addressed in unit test, update the test and make it green, | 03:51 |
sp_ | duonghq: Yes, I was doing the same. Thanks for the suggestion. | 03:53 |
duonghq | sp_, we lack many test case | 03:53 |
sp_ | duonghq: Yes, need to add those | 03:54 |
*** dave-mccowan has quit IRC | 03:54 | |
*** MasterOfBugs has quit IRC | 03:55 | |
*** tyrola has quit IRC | 04:04 | |
*** tyrola has joined #openstack-kolla | 04:04 | |
*** Jeffrey4l has quit IRC | 04:06 | |
*** Jeffrey4l has joined #openstack-kolla | 04:07 | |
*** yingjun has quit IRC | 04:12 | |
*** haplo37 has quit IRC | 04:12 | |
*** tyrola has quit IRC | 04:14 | |
*** tyrola has joined #openstack-kolla | 04:14 | |
*** haplo37 has joined #openstack-kolla | 04:21 | |
*** yingjun has joined #openstack-kolla | 04:22 | |
*** dave-mccowan has joined #openstack-kolla | 04:23 | |
openstackgerrit | Marcus Williams proposed openstack/kolla-ansible: [WIP] Split Openvswitch into own role https://review.openstack.org/416793 | 04:29 |
*** l4yerffeJ__ has quit IRC | 04:32 | |
*** l4yerffeJ__ has joined #openstack-kolla | 04:33 | |
*** l4yerffeJ__ has quit IRC | 04:33 | |
*** mdnadeem has joined #openstack-kolla | 04:34 | |
*** l4yerffeJ has joined #openstack-kolla | 04:36 | |
openstackgerrit | Marcus Williams proposed openstack/kolla-ansible: [WIP] Add OpenDaylight role https://review.openstack.org/416367 | 04:42 |
portdirect | Jeffrey4l (or any other core) can someone look at this: https://review.openstack.org/#/c/416754/ | 04:48 |
portdirect | kolla-k8s gate was broken by https://review.openstack.org/#/c/412225/ and the above should resolve it | 04:49 |
sbezverk | Jeffrey4l: please ack this PS https://review.openstack.org/#/c/412225 | 05:02 |
Jeffrey4l | i -1 for that. should use action != genconfig | 05:03 |
Jeffrey4l | sbezverk, portdirect ^^ | 05:03 |
sbezverk | ok let me pull it change it and then re-submit | 05:04 |
portdirect | cheers sbezverk | 05:05 |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-ansible: Limit check containers to reconfigure action https://review.openstack.org/416754 | 05:07 |
sbezverk | Jeffrey4l: please check it is ok now | 05:07 |
portdirect | sbezverk: https://review.openstack.org/#/c/416754/2/ansible/roles/memcached/tasks/config.yml@32 | 05:09 |
Jeffrey4l | roger. | 05:10 |
Jeffrey4l | sbezverk, fix the memcached, too. | 05:10 |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-ansible: Limit check containers to reconfigure action https://review.openstack.org/416754 | 05:11 |
sbezverk | done | 05:11 |
sbezverk | Jeffrey4l: there were some other PS of similar nature pushed yesterday, do we need to check those as well? | 05:12 |
Jeffrey4l | sbezverk, yep. that's all my patch. i have -1 for them and will fix them. | 05:12 |
sbezverk | for example this one: Merge "Optimize reconfigure action for glance" | 05:13 |
sbezverk | Jeffrey4l: thanks | 05:13 |
Jeffrey4l | sbezverk, - action != "geconfig" . | 05:13 |
Jeffrey4l | not reconfigure | 05:13 |
Jeffrey4l | https://review.openstack.org/#/c/416754/3/ansible/roles/memcached/tasks/config.yml | 05:13 |
sbezverk | damn :-( one sec | 05:14 |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-ansible: Limit check containers to reconfigure action https://review.openstack.org/416754 | 05:14 |
Jeffrey4l | +2ed | 05:15 |
sbezverk | portdirect: I am off for now, please keep an eye on this and if you have a chance re-check Kevin's PS | 05:16 |
sbezverk | Jeffrey4l: thank you | 05:16 |
sbezverk | talk to you later gents.. | 05:16 |
portdirect | Thanks Jeffrey4l :) | 05:16 |
portdirect | sbezverk: will do - though I just work up (5am here - so gonna try and get some more sleep :) ) | 05:17 |
*** skramaja|afk is now known as skramaja | 05:17 | |
portdirect | soon as its though I'll get kevings ps rechecked if I'm awake - catch you later man | 05:18 |
*** celo_r has quit IRC | 05:21 | |
openstackgerrit | Wei Cao proposed openstack/kolla-ansible: Add freezer ansible role https://review.openstack.org/402224 | 05:22 |
*** tyrola has quit IRC | 05:27 | |
*** tyrola has joined #openstack-kolla | 05:28 | |
openstackgerrit | Wei Cao proposed openstack/kolla-ansible: Add .project and .pydevproject to .gitignore https://review.openstack.org/416837 | 05:30 |
*** Jeffrey4l_ has joined #openstack-kolla | 05:31 | |
*** l4yerffeJ_ has joined #openstack-kolla | 05:31 | |
*** l4yerffeJ has quit IRC | 05:34 | |
*** Jeffrey4l has quit IRC | 05:34 | |
openstackgerrit | Marcus Williams proposed openstack/kolla-ansible: [WIP] Split Openvswitch into own role https://review.openstack.org/416793 | 05:39 |
*** salv-orlando has joined #openstack-kolla | 05:41 | |
*** mdnadeem has quit IRC | 05:41 | |
openstackgerrit | Marcus Williams proposed openstack/kolla-ansible: [WIP] Add OpenDaylight role https://review.openstack.org/416367 | 05:41 |
*** unicell has joined #openstack-kolla | 05:41 | |
*** mdnadeem has joined #openstack-kolla | 05:45 | |
*** sayanta__ has quit IRC | 05:45 | |
*** skramaja has quit IRC | 06:05 | |
*** skramaja has joined #openstack-kolla | 06:05 | |
openstackgerrit | Wei Cao proposed openstack/kolla-ansible: Use openstack_service_workers variable in horizon.conf.j2 https://review.openstack.org/416847 | 06:09 |
*** qiliang27 has joined #openstack-kolla | 06:12 | |
*** unicell has quit IRC | 06:14 | |
*** unicell has joined #openstack-kolla | 06:14 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible: Optimize reconfiguration for keystone https://review.openstack.org/408933 | 06:15 |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible: Optimize reconfigure action for nova https://review.openstack.org/407416 | 06:17 |
openstackgerrit | jianyi proposed openstack/kolla-ansible: Add zun ansible role https://review.openstack.org/416179 | 06:18 |
openstackgerrit | jianyi proposed openstack/kolla-ansible: Add zun ansible role https://review.openstack.org/416179 | 06:19 |
*** abc__ has joined #openstack-kolla | 06:22 | |
abc__ | exit | 06:25 |
*** unicell1 has joined #openstack-kolla | 06:27 | |
*** unicell has quit IRC | 06:29 | |
openstackgerrit | Cao Xuan Hoang proposed openstack/kolla: Fix a typo in releasenotes https://review.openstack.org/416853 | 06:31 |
*** bjolo has joined #openstack-kolla | 06:38 | |
openstackgerrit | Wei Cao proposed openstack/kolla-ansible: Add freezer ansible role https://review.openstack.org/402224 | 06:40 |
bjolo | morning | 06:44 |
*** sdake has joined #openstack-kolla | 06:51 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible: Add enable_horizon_* parameter for dashboard https://review.openstack.org/413484 | 06:55 |
*** satyar has joined #openstack-kolla | 07:05 | |
*** pcaruana has joined #openstack-kolla | 07:05 | |
satyar | Hi jeffrey4l | 07:06 |
satyar | the boot from vol issue is solved | 07:07 |
satyar | and got another issue which is because of cinder | 07:07 |
bjolo | Jeffrey4l_, nice patch ^^ | 07:08 |
Jeffrey4l_ | bjolo, thanks ;) | 07:09 |
Jeffrey4l_ | satyar, how it is solved? and what is happening now? | 07:09 |
satyar | the rbd_secret | 07:09 |
satyar | the rbd_secret_uuid need to be in the cinder.conf | 07:10 |
satyar | added that to conf file solved the issye | 07:10 |
satyar | issue | 07:10 |
Jeffrey4l_ | satyar, rbd_secret_uuid is exit in cinder. | 07:10 |
bjolo | satyar, are you doing external ceph? | 07:10 |
bjolo | or kolla ceph? | 07:11 |
satyar | yeah the problen is we are looking at ceph_enable =yes to pass the conf files | 07:12 |
satyar | {% elif enable_ceph | bool and cinder_backend_ceph | bool %} | 07:12 |
satyar | for external ceph also | 07:12 |
*** salv-orl_ has joined #openstack-kolla | 07:13 | |
*** salv-orl_ has quit IRC | 07:13 | |
satyar | so i was using the config folder to configure the external ceph | 07:13 |
*** jianyi has joined #openstack-kolla | 07:14 | |
satyar | planing to edit the template file for cinder.conf.j2 to get the configs populated | 07:14 |
satyar | what you say jeffrey4l | 07:14 |
Jeffrey4l_ | so you are using extenal ceph? | 07:15 |
satyar | yes | 07:16 |
*** salv-orlando has quit IRC | 07:16 | |
satyar | raising a bug to fix the external ceph piece for cinder.conf | 07:16 |
Jeffrey4l_ | hrm. i haven't test extenal ceph. ;( | 07:17 |
satyar | hhmmm | 07:17 |
satyar | we just need to separate out the {% elif enable_ceph | bool and cinder_backend_ceph | bool %} | 07:17 |
Jeffrey4l_ | to what? could u push a fix? | 07:18 |
satyar | cinder_backend_ceph will be under another elif | 07:18 |
satyar | yes | 07:18 |
satyar | raising the bug then will push the fix | 07:18 |
Jeffrey4l_ | cool. | 07:18 |
satyar | :) | 07:18 |
bjolo | i have also noticed some issues with external ceph | 07:19 |
bjolo | have not opened any bugs for it yet since i just found out before the holidays | 07:20 |
bjolo | the rbd_secret file does not get copied into the containers if not using the exact same name. and they are also copied to the wrong path in the containers | 07:22 |
satyar | jeffrey4l added https://bugs.launchpad.net/kolla/+bug/1654157 | 07:23 |
openstack | Launchpad bug 1654157 in kolla "cinder.conf is not getting populated with the rbd values if ceph_enabled is no but cinder_backend_ceph =yes" [Undecided,New] | 07:23 |
satyar | is it possible to triage it? | 07:23 |
*** jrobinson has quit IRC | 07:24 | |
satyar | bojlo can you please raise a bug | 07:25 |
satyar | i will try to look into it | 07:25 |
bjolo | yes | 07:25 |
satyar | Thanks :) | 07:25 |
bjolo | let me dig up the details, my brain is still booting from the hollidays :) | 07:25 |
bjolo | doing an upgrade atm. off the top of the head, anyone got a hunch why it hangs on this: Sighup nova-api _raw_params=docker exec -t nova_api kill -1 1 | 07:27 |
*** sdake_ has joined #openstack-kolla | 07:28 | |
*** sdake has quit IRC | 07:30 | |
bjolo | satyar, looking at your bug 1654157 | 07:31 |
openstack | bug 1654157 in kolla "cinder.conf is not getting populated with the rbd values if ceph_enabled is no but cinder_backend_ceph =yes" [Undecided,New] https://launchpad.net/bugs/1654157 - Assigned to Satya Sanjibani Routray (satroutr) | 07:31 |
bjolo | the docs for external ceph says that you should create your own cinder-volume.conf. | 07:32 |
bjolo | with the ceph information that is needed | 07:33 |
bjolo | http://docs.openstack.org/developer/kolla/external-ceph-guide.html | 07:33 |
bjolo | can you please elaborate on what you mean in your bug? | 07:33 |
*** zhurong has joined #openstack-kolla | 07:37 | |
satyar | yes | 07:50 |
*** openstackgerrit has quit IRC | 07:50 | |
satyar | bjolo this bug will help in adding the ceph piece when external ceph is enabled also | 07:51 |
satyar | as you can see in the document the doc is not mentioning anything rbd_secret_uuid | 07:52 |
satyar | inside cinder.conf | 07:52 |
bjolo | yes, i kinda understand that, but how should kolla know what information to put in there. in our setup for example we dont use the standard names for the pools for ex. | 07:52 |
satyar | anyways we can provide the rbd_secret_uuid inthe globals.conf | 07:53 |
satyar | which is read by the cinder.conf.j2 template | 07:53 |
*** zhugaoxiao has quit IRC | 07:54 | |
bjolo | afk meeting. to be continueed :) | 07:54 |
*** zhugaoxiao has joined #openstack-kolla | 07:54 | |
satyar | sure | 07:54 |
*** matrohon has joined #openstack-kolla | 07:55 | |
sdake_ | hey folks | 08:09 |
sdake_ | see channel is busy as usual :) | 08:09 |
*** athomas has joined #openstack-kolla | 08:27 | |
*** openstackgerrit has joined #openstack-kolla | 08:29 | |
openstackgerrit | Marcus Williams proposed openstack/kolla-ansible: [WIP] Add OpenDaylight role https://review.openstack.org/416367 | 08:29 |
sdake_ | Jeffrey4l_ did mgiles get unblocked? | 08:32 |
openstackgerrit | Merged openstack/kolla-ansible: Limit check containers to reconfigure action https://review.openstack.org/416754 | 08:32 |
*** openstackgerrit has quit IRC | 08:33 | |
*** prameswar has quit IRC | 08:33 | |
*** shardy has joined #openstack-kolla | 08:35 | |
*** openstackgerrit has joined #openstack-kolla | 08:37 | |
openstackgerrit | zhubingbing proposed openstack/kolla: Change freezer user to root https://review.openstack.org/416889 | 08:37 |
zhubingbing | hi sdake | 08:39 |
*** egonzalez90 has joined #openstack-kolla | 08:39 | |
egonzalez90 | morning o/ | 08:40 |
egonzalez90 | zhubingbing: around? | 08:40 |
zhubingbing | yes | 08:40 |
zhubingbing | hi | 08:40 |
sdake_ | sup egonzalez90 | 08:41 |
sdake_ | sup zhubingbing | 08:41 |
sdake_ | how u peeps been | 08:41 |
*** saneax-_-|AFK is now known as saneax | 08:46 | |
*** saneax is now known as saneax-_-|AFK | 08:47 | |
egonzalez90 | hey sdake_ , how was christmas hollidays? | 08:49 |
sdake_ | egonzalez90 too short | 08:49 |
sdake_ | egonzalez90 how about you :) | 08:49 |
*** saneax-_-|AFK is now known as saneax | 08:50 | |
*** sdake_ is now known as sdake | 08:54 | |
satyar | Hi sdake | 09:15 |
satyar | How are you | 09:15 |
egonzalez90 | sdake or any involved in kolla-k8s, take a look at this question. Free adversitement ;) | 09:19 |
egonzalez90 | https://ask.openstack.org/en/question/101159/containers-and-openstack-in-ha/ | 09:19 |
sdake | still alive satyar | 09:19 |
sdake | actually heading to bed | 09:19 |
satyar | ohh ok :) | 09:19 |
sdake | egonzalez90 your busy with ask.openstack.org | 09:20 |
satyar | Good Night ;) | 09:20 |
sdake | \o/ :) | 09:20 |
egonzalez90 | good night | 09:20 |
zhubingbing | sdake good night, | 09:21 |
*** imcsk8 has quit IRC | 09:27 | |
*** openstackgerrit has quit IRC | 09:33 | |
*** jianyi has quit IRC | 09:35 | |
*** Serlex has joined #openstack-kolla | 09:35 | |
*** jascott1 has quit IRC | 09:35 | |
*** zhurong has quit IRC | 09:40 | |
*** sdake_ has joined #openstack-kolla | 09:42 | |
*** openstackgerrit has joined #openstack-kolla | 09:42 | |
openstackgerrit | zhongshengping proposed openstack/kolla: Add support for installing binary packages of watcher https://review.openstack.org/416426 | 09:42 |
*** saneax is now known as saneax-_-|AFK | 09:43 | |
*** sdake has quit IRC | 09:44 | |
*** saneax-_-|AFK is now known as saneax | 09:47 | |
Jeffrey4l_ | sdake_, yep. he unblocked. | 09:50 |
*** rmart04 has joined #openstack-kolla | 09:52 | |
*** rmart04 has quit IRC | 09:52 | |
*** msimonin1 has quit IRC | 09:58 | |
*** msimonin has joined #openstack-kolla | 09:59 | |
openstackgerrit | Marcus Williams proposed openstack/kolla-ansible: [WIP] Add OpenDaylight role https://review.openstack.org/416367 | 10:00 |
*** rmart04 has joined #openstack-kolla | 10:01 | |
*** msimonin has quit IRC | 10:03 | |
*** yingjun has quit IRC | 10:05 | |
*** sdake_ has quit IRC | 10:06 | |
openstackgerrit | Merged openstack/kolla-ansible: Fix python3 compatibility in kolla_docker https://review.openstack.org/402399 | 10:07 |
*** hieulq has quit IRC | 10:13 | |
*** openstackstatus has quit IRC | 10:13 | |
*** openstack has joined #openstack-kolla | 10:17 | |
*** satyar has quit IRC | 10:19 | |
*** prameswar has joined #openstack-kolla | 10:19 | |
*** bando has joined #openstack-kolla | 10:25 | |
*** msimonin has joined #openstack-kolla | 10:28 | |
openstackgerrit | Merged openstack/kolla: Fix designate container https://review.openstack.org/413186 | 10:28 |
*** rmart04 has quit IRC | 10:31 | |
openstackgerrit | Marcus Williams proposed openstack/kolla: [WIP] Add OpenDaylight Container to kolla https://review.openstack.org/416369 | 10:31 |
*** rmart04 has joined #openstack-kolla | 10:33 | |
*** bando has quit IRC | 10:35 | |
*** msimonin has left #openstack-kolla | 10:37 | |
openstackgerrit | Marcus Williams proposed openstack/kolla: [WIP] Add OpenDaylight Container to kolla https://review.openstack.org/416369 | 10:37 |
*** zhubingbing has quit IRC | 10:40 | |
berendt | morning guys | 10:40 |
egonzalez90 | morning berendt | 10:40 |
egonzalez90 | contrats for your travel support :) | 10:41 |
berendt | egonzalez90: thanks. i just received my tickets. | 10:41 |
*** pbourke has quit IRC | 10:41 | |
berendt | are you accepted, too? | 10:41 |
egonzalez90 | don't know, dont have an answer yet | 10:42 |
*** pbourke has joined #openstack-kolla | 10:42 | |
berendt | egonzalez90: i received the confirmation before christmas | 10:42 |
egonzalez90 | had an email saying that continue checking and mid-january they answer | 10:43 |
berendt | egonzalez90: ah ok maybe they have a 2nd round | 10:44 |
egonzalez90 | berendt: they are evaluating travel rates to try accomodate as much people they can | 10:45 |
openstackgerrit | Marcus Williams proposed openstack/kolla-ansible: [WIP] Split Openvswitch into own role https://review.openstack.org/416793 | 10:49 |
*** salv-orlando has joined #openstack-kolla | 10:52 | |
*** salv-orlando has quit IRC | 10:57 | |
*** salv-orlando has joined #openstack-kolla | 10:57 | |
mgkwill | egonzalez90: Thanks for all your reviews of opendaylight_support patches. Time to get some sleep for me. | 10:58 |
egonzalez90 | mgkwill: np, good night | 10:59 |
egonzalez90 | Jeffrey4l_: do we have local repositories in gates or still using centos? | 11:00 |
Jeffrey4l_ | egonzalez90, openstack infra team provides from repo. | 11:01 |
Jeffrey4l_ | but kolla do not use now ;( | 11:01 |
Jeffrey4l_ | egonzalez90, check this https://review.openstack.org/365721 | 11:01 |
egonzalez90 | I'm having errors of repoxml not present, just waas wondering if is local issue or general as gates are not failing now | 11:02 |
*** salv-orlando has quit IRC | 11:02 | |
*** eaguilar has quit IRC | 11:03 | |
Jeffrey4l_ | you can check the repoxml url. | 11:04 |
Jeffrey4l_ | we are not using infra mirror repo now. | 11:04 |
*** salv-orlando has joined #openstack-kolla | 11:04 | |
egonzalez90 | Is trying to download http://mirror.centos.org/centos/7/os/x86_64/repodata/c6411f1cc8a000ed2b651b49134631d279abba1ec1f78e5dcca79a52d8c1eada-primary.sqlite.bz2 | 11:06 |
egonzalez90 | but is not present | 11:06 |
egonzalez90 | probably they changed the sqlite package ID recently and is not in sync with repoxml | 11:09 |
egonzalez90 | yup, repoxml points to the missing sqlite, but the repo have other one | 11:09 |
mliima | morning guys | 11:12 |
*** yingjun has joined #openstack-kolla | 11:13 | |
*** salv-orlando has quit IRC | 11:19 | |
*** salv-orlando has joined #openstack-kolla | 11:19 | |
*** Bico_Fino has joined #openstack-kolla | 11:20 | |
*** yingjun has quit IRC | 11:24 | |
*** salv-orlando has quit IRC | 11:24 | |
*** jascott1 has joined #openstack-kolla | 11:36 | |
*** rmart04 has quit IRC | 11:40 | |
*** rmart04 has joined #openstack-kolla | 11:41 | |
*** jascott1 has quit IRC | 11:41 | |
*** rmart04 has quit IRC | 11:41 | |
*** rmart04 has joined #openstack-kolla | 11:42 | |
*** timss has quit IRC | 11:44 | |
*** timss has joined #openstack-kolla | 11:44 | |
*** matrohon has quit IRC | 11:55 | |
*** williamcaban has quit IRC | 11:55 | |
*** britthouser has quit IRC | 11:55 | |
*** esmiurium has quit IRC | 11:55 | |
*** ajafo has quit IRC | 11:55 | |
*** nihilifer has quit IRC | 11:55 | |
*** devananda has quit IRC | 11:55 | |
*** QuentinM has quit IRC | 11:55 | |
*** zeus has quit IRC | 11:55 | |
*** zeus- has joined #openstack-kolla | 11:55 | |
*** britthouser has joined #openstack-kolla | 11:56 | |
*** matrohon has joined #openstack-kolla | 11:56 | |
*** zeus- is now known as zeus | 11:56 | |
*** devananda has joined #openstack-kolla | 11:56 | |
*** nihilifer has joined #openstack-kolla | 11:56 | |
*** esmiurium has joined #openstack-kolla | 11:56 | |
*** zeus is now known as Guest74290 | 11:56 | |
*** williamcaban has joined #openstack-kolla | 11:56 | |
*** serverascode has quit IRC | 11:59 | |
*** serverascode has joined #openstack-kolla | 12:00 | |
*** masuberu has joined #openstack-kolla | 12:05 | |
*** JoseMello has joined #openstack-kolla | 12:05 | |
masuberu | hi | 12:05 |
masuberu | would like to ask, is kolla production ready? | 12:05 |
*** Bico_Fino_ has joined #openstack-kolla | 12:10 | |
*** saneax is now known as saneax-_-|AFK | 12:11 | |
*** mdnadeem has quit IRC | 12:12 | |
egonzalez90 | masuberu: yeah | 12:12 |
masuberu | good | 12:12 |
masuberu | how is the HA provided? do we need something like kubernetes, mesos+marathon or ocker swarm in place to do health-check? | 12:13 |
*** Bico_Fino has quit IRC | 12:13 | |
*** Bico_Fino_ is now known as Bico_Fino | 12:13 | |
*** haplo37 has quit IRC | 12:14 | |
*** g3ek has quit IRC | 12:14 | |
*** zhubingbing has joined #openstack-kolla | 12:14 | |
*** Serlex has quit IRC | 12:17 | |
egonzalez90 | masuberu: actually kolla provides docker images, and we have two deliverables kolla-ansible and kolla-kubernetes | 12:18 |
egonzalez90 | kolla-ansible is production ready, kolla-kubernetes is under development at good pace | 12:19 |
masuberu | egonzalez90, so how is the HA provided? is galera and pacemaker running on containers aswell? | 12:19 |
openstackgerrit | Merged openstack/kolla: Add a 'fallback mode' to disable udev in find_disks https://review.openstack.org/384496 | 12:20 |
egonzalez90 | in kolla-ansible: SQL with galera, rabbitmq with mirror queues, and API services with HAproxy/keepalived | 12:20 |
masuberu | ok, what is the roadmap for kolla-kubernetes? | 12:21 |
*** Serlex has joined #openstack-kolla | 12:23 | |
*** g3ek has joined #openstack-kolla | 12:24 | |
*** haplo37 has joined #openstack-kolla | 12:24 | |
egonzalez90 | masuberu: not really sure, prefer not to answer something wrong. | 12:26 |
*** mdnadeem has joined #openstack-kolla | 12:26 | |
egonzalez90 | masuberu: people working on k8s are mostly in US timezone | 12:28 |
masuberu | egonzalez90, thanks would like to know if they have any date in mind | 12:41 |
masuberu | because I am really looking forward for this | 12:41 |
*** msimonin has joined #openstack-kolla | 12:46 | |
masuberu | egonzalez90, another question, do you require all kolla containers to run on same host? | 13:01 |
masuberu | if not, how do you manage the service discovery or the networking between the hosts? | 13:01 |
masuberu | *networing across containers on different hosts? | 13:01 |
*** shardy has quit IRC | 13:02 | |
egonzalez90 | masuberu: depends of the service, nova libvirt have to be in the same as nova-compute as example | 13:02 |
egonzalez90 | we use --net= host for networking, share the IP of the host | 13:02 |
*** Serlex has left #openstack-kolla | 13:04 | |
masuberu | so you bind the containers and the hosts nics | 13:05 |
egonzalez90 | yes in ansible, k8s use different networksetup IIRC | 13:06 |
hawi | hi. has anyone used browbeat with kolla installation? | 13:19 |
hawi | just searching some way to benchmark setup | 13:21 |
portdirect | masuberu: the roadmap for kolla kubernetes is pretty flexible at the moment | 13:27 |
portdirect | masuberu: we are currently changing from using a jinja2 templating to using helm to generate k8s manifests | 13:27 |
masuberu | portdirect, I see | 13:28 |
masuberu | portdirect, thanks for this | 13:28 |
masuberu | however I saw a video, that the people from tectonic already got it working | 13:28 |
masuberu | but I believe that is a paid solution | 13:28 |
portdirect | masuberu: and are about 50% of the way through that work, all individual components are moving | 13:28 |
portdirect | whoop moving should have been working | 13:29 |
portdirect | and we are now starting to het service leve components made | 13:29 |
portdirect | masuberu: Theres also my project, thats up and running https://github.com/portdirect/harbor | 13:30 |
portdirect | masuberu: and also https://github.com/att-comdev/openstack-helm | 13:30 |
portdirect | as well as https://github.com/sapcc/openstack-helm | 13:31 |
portdirect | they all serve slightly different use cases atm (thou I hope to retire harbor, once att-comdev's openstack-helm is a bit further along) | 13:32 |
*** jheroux has joined #openstack-kolla | 13:33 | |
masuberu | ok, I will need to spend time looking into this | 13:33 |
masuberu | I basically would like to have an openstack HA mode running | 13:34 |
portdirect | what are you planning on using openstack/k8s for? | 13:34 |
portdirect | (all of the above do that :) ) | 13:34 |
masuberu | to create an environment so our users can prototype | 13:34 |
masuberu | at the moment out cluster runs HPC which is very strict | 13:34 |
portdirect | HPC? | 13:35 |
masuberu | we have users who has special needs and we would like to give something they can customise | 13:35 |
masuberu | High Performance Computing | 13:35 |
portdirect | ok - thats a pretty broad spectrum :) | 13:36 |
masuberu | yes | 13:36 |
portdirect | CFD? | 13:36 |
egonzalez90 | customise in which way? | 13:36 |
masuberu | we are a medical research intitute | 13:36 |
masuberu | we have users know who needs to run tools like tensorflow | 13:37 |
masuberu | apache spark | 13:37 |
portdirect | ^^ egonzalez90's question is very pertinent, do you mena cusom hardware or just software running ontop of openstack? | 13:37 |
masuberu | and we can't provide those type of solutions with our current cluster | 13:37 |
egonzalez90 | so, need a dynamic and scalable infra? | 13:38 |
masuberu | egonzalez90, correct | 13:38 |
masuberu | we don't know yet if we are going to use openstack for production | 13:38 |
portdirect | ok - so right now probably either harbor, or att-comdev's is going to be closest to what you need | 13:39 |
*** rhallisey has joined #openstack-kolla | 13:39 | |
masuberu | but definetly we need openstack so out users can have kind of an isolated environment where they can experiement | 13:39 |
portdirect | (I would not recommend using harbor for more than kicking the tyres atm) | 13:39 |
masuberu | I see | 13:39 |
egonzalez90 | masuberu: there is a scientific working group, they may help you in HPC stuff https://wiki.openstack.org/wiki/Scientific_working_group | 13:40 |
portdirect | but moving forward kolla-k8s should be a viable solution as well, though i think that may be about 2 months away | 13:40 |
masuberu | 2 months is pretty quick | 13:40 |
sbezverk | Jeffrey4l: ping | 13:40 |
sbezverk | good morning | 13:40 |
portdirect | morning sbezverk ! :) | 13:41 |
masuberu | I would also like to do some presentation on the openstack summit (if I can deliver something interesting to the community) | 13:41 |
masuberu | they are coming to Sydney this year | 13:41 |
*** eaguilar has joined #openstack-kolla | 13:41 | |
masuberu | but that is a different story | 13:41 |
portdirect | though giving back to the community is awesome :) | 13:41 |
sbezverk | portdirect: it looks like last night fix is not really a fix :( | 13:42 |
portdirect | :( | 13:42 |
portdirect | masuberu: the Scientific working group are a great place to start to find out about openstack for you use case - and once you have and idea of the services you need, there are people here who can help you get them up and running | 13:43 |
masuberu | yeah, well we really appreciate people like you guys is what you do | 13:43 |
sbezverk | portdirect: let me quickly nail to b2 version of kolla-ansible and you can ack it | 13:43 |
sbezverk | so we restore our gate | 13:43 |
portdirect | sbezverk: roger | 13:43 |
*** tovin07_ has joined #openstack-kolla | 13:44 | |
masuberu | yeah, well at the moment we just want to run VMs | 13:44 |
masuberu | something very simple | 13:44 |
masuberu | but with HA | 13:44 |
portdirect | masuberu: if you find an openstack on k8s thats not HA, then I'll be impressed :) | 13:45 |
masuberu | I will have a look to openstack-helm | 13:45 |
masuberu | I think k8s will be very impressive, I just saw he video from the people from tectonic and I was really impressed | 13:46 |
masuberu | plus I think kubernetes needs to move quick as docker swarm is pushing hard | 13:47 |
*** tovin07 has quit IRC | 13:47 | |
*** tovin07_ is now known as tovin07 | 13:47 | |
*** tovin07_ has joined #openstack-kolla | 13:48 | |
*** zhubingbing has quit IRC | 13:50 | |
*** goldyfruit has joined #openstack-kolla | 13:52 | |
*** eaguilar has quit IRC | 13:52 | |
*** tovin07 has quit IRC | 13:53 | |
*** nihilifer has quit IRC | 13:54 | |
*** nihilifer has joined #openstack-kolla | 13:54 | |
*** mgiles has joined #openstack-kolla | 13:54 | |
*** salv-orlando has joined #openstack-kolla | 13:56 | |
*** sp__ has joined #openstack-kolla | 13:57 | |
*** fguillot has joined #openstack-kolla | 13:59 | |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Nailing ansible version at the kube gate https://review.openstack.org/416981 | 14:01 |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Remove footer block and variable in *-base images https://review.openstack.org/415880 | 14:05 |
*** zhubingbing has joined #openstack-kolla | 14:07 | |
sbezverk | portdirect: have you tried to build kolla images behind proxy ? is there anything special needs to be done? | 14:07 |
portdirect | sbezverk: no mgiles has though i think | 14:08 |
masuberu | what do you think about stackanetes? | 14:09 |
mgiles | sbezverk Yes, I have. It's covered in the docs: http://docs.openstack.org/developer/kolla/image-building.html#building-behind-a-proxy | 14:10 |
*** srwilkers has joined #openstack-kolla | 14:10 | |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Nailing ansible version at the kube gate https://review.openstack.org/416981 | 14:10 |
sbezverk | mgiles: thank you | 14:11 |
portdirect | masuberu: that its a nice PoC, was relased shortly after my harbor work. Most of the stackinetes developers are now contributing to att-comdev's openstack-helm | 14:11 |
masuberu | so in your opinion what would be the difference between openstack-helm and stackanetes? | 14:12 |
portdirect | stackanetes is based round kpm vs helm. | 14:13 |
v1k0d3n | morning all | 14:13 |
portdirect | hey v1k0d3n o/ | 14:14 |
v1k0d3n | yo/ | 14:14 |
*** pc_m has quit IRC | 14:15 | |
*** pc_m has joined #openstack-kolla | 14:15 | |
*** l4yerffeJ_ has quit IRC | 14:15 | |
masuberu | portdirect, thanks, I need to read more about kpm and helm, all this is new to me | 14:15 |
v1k0d3n | masuberu: lots to read up on. things get opinionated quickly. :) | 14:16 |
v1k0d3n | be very open-minded and aware of trends | 14:16 |
v1k0d3n | i saw you were asking about stackanetes/openstack-helm, etc. | 14:17 |
v1k0d3n | there are in fact two openstack-helms now. so in total like 4 or 5 openstack on kubernetes projects going on. | 14:17 |
v1k0d3n | we are one of those os-helm projects. | 14:17 |
*** Guest74290 is now known as zeus` | 14:18 | |
openstackgerrit | Merged openstack/kolla-ansible: Allow cinder-volume to be configured to use HNAS iscsi https://review.openstack.org/409883 | 14:18 |
*** zeus` is now known as zeus | 14:18 | |
*** zeus has quit IRC | 14:18 | |
*** zeus has joined #openstack-kolla | 14:18 | |
masuberu | v1k0d3n, I see, yes a lot to read. Opensource community is quite impressive | 14:20 |
masuberu | I follow trends just because I assume it will be easier to find help | 14:21 |
*** sp__ has quit IRC | 14:21 | |
*** Jeffrey4l_ has quit IRC | 14:21 | |
masuberu | to be honest | 14:22 |
v1k0d3n | masuberu: it is easier to find help. takes a while to fully take it all in. i'm not sure there's a person in this channel who can say they don't learn something new every day. | 14:24 |
masuberu | agree | 14:25 |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible: Ansible-ize OpenStack Designate https://review.openstack.org/408465 | 14:25 |
portdirect | v1k0d3n: true dat! i try not to though - I'm worried that my mind may run out of space for thundercats trivia :/ | 14:25 |
sbezverk | portdirect: brain upgrades are coming ;-) | 14:26 |
v1k0d3n | can't sacrifice the important things! lol | 14:26 |
*** shardy has joined #openstack-kolla | 14:28 | |
sbezverk | portdirect: I checked zuul and it looks like this fix holds, so please be ready to act it.. | 14:30 |
zhubingbing | hello guys | 14:31 |
portdirect | hey zhubingbing o/ | 14:31 |
*** masuberu has quit IRC | 14:33 | |
zhubingbing | hi | 14:34 |
*** Jeffrey4l_ has joined #openstack-kolla | 14:35 | |
zhubingbing | sup Jeffrey4l | 14:35 |
srwilkers | morning | 14:35 |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible: Ansible-ize OpenStack Designate https://review.openstack.org/408465 | 14:36 |
*** lrensing has joined #openstack-kolla | 14:40 | |
portdirect | sup srwilkers | 14:40 |
*** portdirect is now known as portdirect_lunch | 14:41 | |
sbezverk | portdirect: srwilkers: need you mighty +2 :0 | 14:42 |
sbezverk | https://review.openstack.org/#/c/416981/ | 14:42 |
*** lamt has joined #openstack-kolla | 14:42 | |
*** Serlex has joined #openstack-kolla | 14:45 | |
srwilkers | done :) | 14:46 |
openstackgerrit | Merged openstack/kolla-ansible: Move ironic precheck into its own role https://review.openstack.org/415481 | 14:46 |
openstackgerrit | Merged openstack/kolla-ansible: Move trove precheck into its own role https://review.openstack.org/416032 | 14:46 |
sbezverk | srwilkers: portdirect: thanks hopefully it will unblock us | 14:46 |
*** n3v3rm0r3r has joined #openstack-kolla | 14:46 | |
srwilkers | sbezverk, hopefully so | 14:46 |
openstackgerrit | Merged openstack/kolla-kubernetes: Nailing ansible version at the kube gate https://review.openstack.org/416981 | 14:49 |
*** jtriley has joined #openstack-kolla | 14:49 | |
*** prameswar has quit IRC | 14:49 | |
*** adrian_otto has joined #openstack-kolla | 14:50 | |
*** salv-orlando has quit IRC | 14:50 | |
*** Jeffrey4l_ has quit IRC | 14:52 | |
sean-k-mooney | Jeffrey4l_: no i was offline/login but away im online for the next few hours if you still want me | 14:53 |
*** satyar has joined #openstack-kolla | 14:53 | |
*** Pavo has joined #openstack-kolla | 14:56 | |
*** mdnadeem has quit IRC | 14:57 | |
*** yingjun has joined #openstack-kolla | 14:58 | |
*** absubram has joined #openstack-kolla | 14:58 | |
*** msimonin has quit IRC | 15:00 | |
*** bjolo has quit IRC | 15:00 | |
*** prameswar has joined #openstack-kolla | 15:03 | |
*** Jeffrey4l_ has joined #openstack-kolla | 15:05 | |
openstackgerrit | Marcus Williams proposed openstack/kolla: [WIP] Add OpenDaylight Container to kolla https://review.openstack.org/416369 | 15:06 |
satyar | Hi sdake | 15:08 |
*** msimonin has joined #openstack-kolla | 15:10 | |
*** sp__ has joined #openstack-kolla | 15:11 | |
*** britthouser has quit IRC | 15:12 | |
*** mattmceuen has joined #openstack-kolla | 15:12 | |
*** britthouser has joined #openstack-kolla | 15:13 | |
*** lamt has quit IRC | 15:15 | |
*** Mech422 has quit IRC | 15:15 | |
*** Mech422 has joined #openstack-kolla | 15:16 | |
satyar | hi sdake_ | 15:16 |
*** portdirect_lunch is now known as portdirect | 15:16 | |
*** lamt has joined #openstack-kolla | 15:16 | |
*** sdake has joined #openstack-kolla | 15:20 | |
sdake | morning peeps | 15:20 |
*** sdake_ has joined #openstack-kolla | 15:23 | |
srwilkers | morning sdake | 15:23 |
*** mattmceuen has quit IRC | 15:23 | |
sbezverk | morning sdake | 15:24 |
sbezverk | srwilkers: could you please review this one: https://review.openstack.org/#/c/416729/ | 15:24 |
*** imcsk8 has joined #openstack-kolla | 15:24 | |
*** sdake has quit IRC | 15:25 | |
srwilkers | sbezverk, of course | 15:26 |
*** mattmceuen has joined #openstack-kolla | 15:28 | |
*** sp__ has quit IRC | 15:32 | |
*** salv-orlando has joined #openstack-kolla | 15:34 | |
openstackgerrit | Merged openstack/kolla-kubernetes: Add default searchPath to almost all templates. https://review.openstack.org/416729 | 15:35 |
*** Jeffrey4l_ has quit IRC | 15:36 | |
*** jtriley has quit IRC | 15:36 | |
*** yingjun has quit IRC | 15:38 | |
*** yingjun has joined #openstack-kolla | 15:41 | |
*** papacz has quit IRC | 15:45 | |
jemcevoy | morning everyone... | 15:45 |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible: Unify haproxy prechecks condition order https://review.openstack.org/415510 | 15:47 |
egonzalez90 | morning jemcevoy | 15:47 |
jemcevoy | Yesterday on my development cluster... I made an openviswitch change that took down the whole network at the office so I had to ask to have my cluster manually powered off. I have it all back up and running except for the database... which only came 2/3 back... | 15:47 |
*** Jeffrey4l_ has joined #openstack-kolla | 15:48 | |
jemcevoy | after running the recover_mariadb thing... | 15:48 |
jemcevoy | This is the state of galera now: http://paste.openstack.org/show/594009/ | 15:48 |
jemcevoy | Anyone know how I can get the third node to rejoin the galera cluster? | 15:49 |
egonzalez90 | jemcevoy: the container in third node is running? | 15:50 |
*** yingjun has quit IRC | 15:51 | |
jemcevoy | egomzalez98: Yes it is running and I tried rebooting it this morning | 15:53 |
*** salv-orl_ has joined #openstack-kolla | 15:53 | |
jemcevoy | The mariadb container is in a restarting mode now... | 15:54 |
jemcevoy | 5960884f69f kolla4echo.scm.penguincomputing.com:4000/kolla/centos-binary-mariadb:3.0.1 "kolla_start" 15 hours ago Restarting (1) 6 minutes ago mariadb | 15:54 |
egonzalez90 | jemcevoy: can you share logs? | 15:55 |
*** salv-orlando has quit IRC | 15:56 | |
jemcevoy | sure.... If you can tell me which ones and where to find them... nothing shows up in kibana | 15:56 |
egonzalez90 | docker logs mariadb and logs from /var/lib/docker/volumes/kolla_logs/_data/mariadb | 15:57 |
jemcevoy | The last kibana error is from 07:25 (half hour ago) when I did the host reboot | 15:57 |
jemcevoy | egomzalez98: The log is 1.5 MB Where can I upload it? | 16:00 |
Jeffrey4l_ | egonzalez90, could u review this? https://review.openstack.org/414687 | 16:00 |
*** sdake_ has quit IRC | 16:02 | |
*** sdake has joined #openstack-kolla | 16:03 | |
SamYaple | morning | 16:05 |
egonzalez90 | Jeffrey4l_: why is needed -H in sudo, is not running as root in his $HOME? | 16:05 |
jemcevoy | egomzalez98: I copied the log to nginx on my laptop: https://salemherbals.com/Kolla/mariadb.log | 16:06 |
Jeffrey4l_ | egonzalez90, Request that the security policy set the HOME environment variable to the home directory specified by the target user's password database entry. Depending on | 16:06 |
Jeffrey4l_ | the policy, this may be the default behavior. | 16:06 |
Jeffrey4l_ | copyied from man sudo | 16:06 |
Jeffrey4l_ | the issue is: if we do not add -H, .ccache fodler in jenkins home folder will ownered by root. | 16:06 |
SamYaple | jmccarthy: just to be clear, you are able to access mariadb, but its only running on 2/3 nodes, yes? | 16:07 |
SamYaple | jemcevoy: ^^ | 16:07 |
SamYaple | jmccarthy: ignore | 16:07 |
*** aric49 has joined #openstack-kolla | 16:08 | |
Jeffrey4l_ | next one egonzalez90 https://review.openstack.org/415504 ;) | 16:08 |
jemcevoy | Yes mysql is working.. but on 2/3 of the nodes | 16:08 |
Jeffrey4l_ | SamYaple, could u give some review/advise for above patch? ^^ | 16:08 |
egonzalez90 | Jeffrey4l_: too big ;) | 16:08 |
jemcevoy | SamYaple: ^^ | 16:09 |
SamYaple | jemcevoy: on the third node, stop mariadb container and remove the mariadb data volume. then remove the /var/log/kolla/mysql/* files (they are log bin files) | 16:09 |
Jeffrey4l_ | a little. the mainly is kolla_toolbox.py module. others are repeated. | 16:09 |
SamYaple | jemcevoy: _then_ run the playbooks again to start all the nodes | 16:09 |
*** TxGirlGeek has joined #openstack-kolla | 16:10 | |
Jeffrey4l_ | there is a child patch about kolla_toolbox.py module. https://review.openstack.org/416046 it is still WIP, but almost done. i will fix other roles. it make the playbooks better to review ;) | 16:10 |
jemcevoy | SamYaple: by running the playbooks you mean deploy? and removing the volumes is that deleting the hundreds of files in /var/lib/docker/volumes/mariadb/_data/ | 16:13 |
*** rmart04 has quit IRC | 16:13 | |
SamYaple | jemcevoy: if there is any node running in the galera cluster, you should always run deploy | 16:15 |
SamYaple | jemcevoy: removing the volume is `docker vol rm mariadb` or whatever teh volume remove syntax is | 16:15 |
*** sayantan_ has joined #openstack-kolla | 16:16 | |
jemcevoy | Thanks SamYaple... | 16:16 |
*** sdake has quit IRC | 16:16 | |
*** inc0 has joined #openstack-kolla | 16:17 | |
inc0 | good morning | 16:17 |
Jeffrey4l_ | morning | 16:17 |
jemcevoy | morning inc0 | 16:17 |
Jeffrey4l_ | please review this inc0 https://review.openstack.org/415504 ;) | 16:18 |
SamYaple | o/ | 16:18 |
*** msimonin has quit IRC | 16:19 | |
sbezverk | inc0 good morning | 16:19 |
sbezverk | inc0 we end up nailing 4.0.0.b2 as ansible version at the gate to move forward | 16:19 |
jmccarthy | Bit behind the curve, what is a good doc/starting point to check out kolla-kubernetes ? | 16:19 |
sbezverk | last night fix, did not fix the issue :9 | 16:20 |
sbezverk | ;( | 16:20 |
inc0 | sbezverk, cool, you pushed images to hub? | 16:20 |
inc0 | ahh | 16:20 |
inc0 | ok, I get it now | 16:20 |
inc0 | anyway, this fix is merged | 16:20 |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla: TEST NEWTON BRANCH https://review.openstack.org/417033 | 16:21 |
inc0 | can I have eyes on this too? | 16:21 |
inc0 | https://review.openstack.org/#/c/404435/ | 16:21 |
*** mliima has quit IRC | 16:21 | |
kfox1111 | morning. | 16:22 |
*** adrian_otto has quit IRC | 16:22 | |
kfox1111 | jmccarthy: unfortunately, we're nearing the end of refactoring the project and the documentation hasn't kept up. :/ | 16:22 |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla: NO_MERGE: test mitaka branch https://review.openstack.org/393877 | 16:23 |
sbezverk | kfox1111: good morning | 16:24 |
openstackgerrit | zhubingbing proposed openstack/kolla: Change freezer user to root https://review.openstack.org/416889 | 16:25 |
aric49 | Hi Everyone --- question about Kolla philosophy. I was reviewing the documentation and didn't find very much about maintaining Kolla installations and troubleshooting issues. Are there any best practices for restarting misbehaving containers, or is it as simple as "docker restart container" ? | 16:25 |
*** mliima has joined #openstack-kolla | 16:25 | |
jemcevoy | SamYaple: that fixed it... Galera see all three nodes again... :-) http://paste.openstack.org/show/594018/ | 16:28 |
portdirect | inc0: just looking at that review - and being lazy - where does the registry get started? | 16:28 |
inc0 | https://review.openstack.org/#/c/404435/21/tools/gate_run.sh :83 | 16:28 |
inc0 | no, hold on | 16:29 |
portdirect | found it: https://github.com/openstack/kolla/blob/master/tools/setup_gate.sh#L148-L152 | 16:29 |
inc0 | ahh so I guess something else got merged | 16:29 |
portdirect | sorry to send you looking | 16:29 |
*** jtriley has joined #openstack-kolla | 16:32 | |
inc0 | Jeffrey4l_, did you figure out gate issues? | 16:36 |
Jeffrey4l_ | inc0, which one? | 16:37 |
inc0 | libvirt one I mean | 16:37 |
Jeffrey4l_ | after configure cpu_mode=none, gate is green. | 16:37 |
inc0 | kk makes sense | 16:37 |
Jeffrey4l_ | i can not explain this. seem some bug/issue in libvirt 2.0 or qemu. | 16:37 |
Jeffrey4l_ | there are some upstream bug track this. | 16:37 |
inc0 | I think cpu model in xen-spawned vm is some xen model, which makes qemu stupid | 16:38 |
inc0 | anyway, I'm glad we got this doen | 16:38 |
openstackgerrit | Merged openstack/kolla: Run sudo with -H parameter to set the proper home dir https://review.openstack.org/414687 | 16:38 |
Jeffrey4l_ | no only xen. i saw other related issue on physical machine. | 16:38 |
Jeffrey4l_ | yep. | 16:38 |
inc0 | guys can we merge packing registry please? | 16:38 |
inc0 | so we can get deploy gates voting before release? | 16:39 |
*** rhallisey has quit IRC | 16:39 | |
sean-k-mooney | hi guys. has there been any futer discussion about storing example template override files in the contrib dir? | 16:45 |
inc0 | not really sean-k-mooney | 16:46 |
inc0 | we'll need one thing to do this, not sure if we already done it | 16:46 |
*** sayantan_ has quit IRC | 16:46 | |
*** sayanta__ has joined #openstack-kolla | 16:46 | |
inc0 | and that is support for multiple override files | 16:46 |
sean-k-mooney | inc0: ok im reviewing one of the odl patches. for my dpdk patch i was storing it in the docs temparaily since we had not made any dission | 16:47 |
*** unicell1 has quit IRC | 16:47 | |
SamYaple | jemcevoy: im a magician, what can i say | 16:47 |
sean-k-mooney | haveing a directory for all example config overrides makes sense to. perhaps under contrib or at the top level | 16:48 |
SamYaple | aric49: docker restart container | 16:48 |
*** zhubingbing has quit IRC | 16:48 | |
sean-k-mooney | inc0: i assume the stance that we dont allow source builds of infra containers still stands | 16:48 |
inc0 | well, not by default | 16:49 |
sean-k-mooney | that is the reason for the template override in both the odl and ovs-dpdk case for os,s that dont ship a binary package or for those who want to use a different verion then the os provides | 16:49 |
inc0 | I'm ok with having override file with source compilation | 16:49 |
*** saneax-_-|AFK is now known as saneax | 16:49 | |
openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Fix the mariadb/rabbitmq elements and add searchPaths. https://review.openstack.org/416744 | 16:49 |
inc0 | you know what? If you want to upload override file for dpdk, feel free to | 16:50 |
inc0 | let's create new directory for overrides | 16:50 |
inc0 | and add first one - dpdk | 16:50 |
inc0 | in the meantime I'll work on multiple override file support | 16:50 |
jemcevoy | SamYaple: I agree... But I just killed the corp network again... guess the neutron part of the redeploy caused another openvswitch loop. I need to change the topology so that I am not on the same L2 as everyone else... | 16:50 |
SamYaple | sean-k-mooney: i think when came up before, the stance was non-openstack service == no source builds | 16:50 |
inc0 | jemcevoy, that's usually good idea;) | 16:51 |
SamYaple | sean-k-mooney: but I wouldn't mind seeing a looser approach to container building | 16:51 |
SamYaple | _if_ we can ensure multiple peopel get the same result | 16:51 |
inc0 | let's start with override file | 16:51 |
sean-k-mooney | inc0: sure i can do that as part of the next patch set. | 16:51 |
inc0 | as dpdk or ODL isn't exactly most typical deployment | 16:51 |
*** vhosakot has joined #openstack-kolla | 16:51 | |
sean-k-mooney | SamYaple: i think the approch i am takeing for dpdk should be reporducable https://review.openstack.org/#/c/408866/1/doc/networking-guide.rst | 16:52 |
sean-k-mooney | SamYaple: im using the kolla build plugin support to download the source then using a tempalte override to compile and install it | 16:53 |
inc0 | you're upgrading kernel from inside container? that's...uhh | 16:54 |
inc0 | especially that it wont work;) | 16:54 |
SamYaple | sean-k-mooney: yep. none of this templating stuff existed when we talked abbout this previously | 16:54 |
*** sdake has joined #openstack-kolla | 16:54 | |
sean-k-mooney | we will have the same issue with ovn also as most distros dont ship it as a binary package. | 16:54 |
sean-k-mooney | inc0: am i i dont think i am | 16:55 |
inc0 | % if base_distro in ['centos', 'fedora', 'oraclelinux', 'rhel'] %} | 16:55 |
inc0 | {% set ovs_dpdk_packages = [ | 16:55 |
inc0 | 'kernel-devel-$(uname -r)', | 16:55 |
sean-k-mooney | inc0: im pulling the kernel headers there | 16:55 |
sean-k-mooney | i am not upgrading the kernel | 16:55 |
inc0 | ok, well if you pull different header than target node | 16:56 |
sean-k-mooney | later im also not using the dpdk build kernel module and using the ones from the host instead | 16:56 |
inc0 | will it work? | 16:56 |
inc0 | ok, my bad then | 16:56 |
inc0 | anyway, this shouldn't be in networking guide imho | 16:56 |
sean-k-mooney | yes so long as the kernel did not break usespace it will work fine. | 16:56 |
inc0 | let's create contrib and add dpdk there | 16:57 |
inc0 | like all configs+override+dedicated doc | 16:57 |
sean-k-mooney | inc0: yep the docks patch is actully also my todo list/notes at the moment | 16:57 |
openstackgerrit | Mauricio Lima proposed openstack/kolla-ansible: Allow cinder-volume to be configured to use HNAS nfs https://review.openstack.org/417094 | 16:57 |
inc0 | kk | 16:57 |
*** pcaruana has quit IRC | 16:58 | |
sean-k-mooney | but yes ill submit a patch to just create the directory and then the dpdk patches and odl patches can both depend on that so that neither is blocked by the other | 16:58 |
inc0 | cool | 16:58 |
*** jtriley has quit IRC | 16:58 | |
inc0 | we'll need to do subdirectory in docs for stuff like that | 16:59 |
egonzalez90 | the problem with template-overrides is that with actual gates we can't test if build and deploy work | 16:59 |
*** hyakuhei has quit IRC | 16:59 | |
*** hyakuhei has joined #openstack-kolla | 16:59 | |
*** hyakuhei has quit IRC | 16:59 | |
*** hyakuhei has joined #openstack-kolla | 16:59 | |
inc0 | egonzalez90, yeah, but contrib not being gated at isn't killer for me tbh | 16:59 |
inc0 | not now anyway | 16:59 |
*** matrohon has quit IRC | 17:00 | |
inc0 | and for dedicated gates we can build relevant images in gates | 17:00 |
sean-k-mooney | ya i think this is the kind of stuff that third party cis would be usefull for | 17:00 |
inc0 | yeah | 17:00 |
inc0 | contrib should be super easy to contribute to | 17:00 |
inc0 | and add new stuff that might be less supported | 17:01 |
sean-k-mooney | when this works we could have the intel-nfv ci run on change to the kolla repo and use the template-override to build and deploy the container | 17:01 |
inc0 | speaking of overrides, who wants to write calico support for kolla? | 17:01 |
egonzalez90 | there was a calico PS IIRC | 17:01 |
inc0 | yeah | 17:02 |
inc0 | but person responsible left | 17:02 |
inc0 | and it should be override really | 17:02 |
inc0 | as it's fairly magical thing | 17:02 |
sean-k-mooney | haha i think pike/ocata will be a release of neutron backends. calico,dpdk,odl,ovn,sriov | 17:02 |
inc0 | extramally cool, but far from default | 17:02 |
* inc0 is afraid the he released the kraken | 17:03 | |
sean-k-mooney | actully inc0 has anything happended with out stratgy pluging that would allow extenible ansible? | 17:03 |
* inc0 hides from anyone with storage vendor tshirt | 17:03 | |
SamYaple | inc0: you released a new version of ceph? | 17:03 |
egonzalez90 | im afraid of who is going to maintain all those backends, I have no idea about calico | 17:03 |
*** Serlex has quit IRC | 17:04 | |
inc0 | sean-k-mooney, license killed the effort to get it into kolla:/ I'm trying to tap into our RH friends to provide incentives to merge it into ansible | 17:04 |
inc0 | SamYaple, we're running Jewel now | 17:04 |
SamYaple | so it was mentioned that if someone wanted to make a puppet kolla, it should be puppet-kolla (under puppet namespace) rather than kolla-puppet | 17:04 |
sean-k-mooney | SamYaple: is that related to the tripple0 kolla support? | 17:05 |
inc0 | I disagree | 17:05 |
SamYaple | inc0: no i was joking, new version of ceph is kraken | 17:05 |
inc0 | ahh | 17:05 |
inc0 | ok | 17:05 |
inc0 | didn't follow that | 17:05 |
SamYaple | sean-k-mooney: nah its in a ML thread talking about adding kolla-salt | 17:05 |
sean-k-mooney | ah ok | 17:05 |
inc0 | so puppet-kolla I'd assume is puppet module to install kolla;) | 17:05 |
inc0 | not install openstack;) | 17:05 |
inc0 | with kolla | 17:05 |
SamYaple | inc0: im just relating what just popped into the ML thread | 17:06 |
inc0 | to which I'll reply | 17:06 |
SamYaple | idk. im just not a fan of kolla-ansible and kolla-kubernetes (and kolla-salt) voting on accepting a kolla-chef repo. I think only kolla core should care about that | 17:07 |
sean-k-mooney | but ya if we had the extenible ansible support we could handel the vendor backend out of the main kolla tree or in it via contrib repos. | 17:08 |
sean-k-mooney | sometimes i think kolla core is becomeing a replacement for apt/rpm packages | 17:09 |
sean-k-mooney | all hail kolla-bash | 17:09 |
SamYaple | sean-k-mooney: more like kolla-ssh amirite | 17:09 |
inc0 | SamYaple, yeah, I'm goign to reply that it doesn't make any sense for kolla-ansible cores to vote on accepting salt | 17:09 |
inc0 | especially that if we put this requirement, the more engines we have the harder it will be to add new one | 17:10 |
SamYaple | sean-k-mooney: if we really kick kolla build progress into high gear we might be able to build a base neutron layer with overlay lyers for dpdk,odl,ovn etc | 17:10 |
inc0 | which is exactly what we don't want | 17:10 |
SamYaple | inc0: also, lets be honest, you could easily get sour people blocking for no reason | 17:10 |
inc0 | sean-k-mooney, I keep saying, to me docker images in the way we're using them are little more than glorified packages | 17:10 |
SamYaple | though i dont think we have that problem _now_, kolla is still young | 17:10 |
erhudy | inc0: we (bloomberg) want to do calico support for kolla probably | 17:11 |
sean-k-mooney | SamYaple: actully since ovn ins in the ovs git repo my dpdk template override is actully already building ovn i just does not run it | 17:11 |
SamYaple | inc0: as it should be! glorified packages forever! | 17:11 |
inc0 | well looking at our current state I don't think ti would be a case, but yeah, religions are real | 17:11 |
erhudy | or kolla support for calico, putting words in order is hard | 17:11 |
inc0 | erhudy, did we talk on summit about that?:) | 17:11 |
erhudy | probably not | 17:11 |
inc0 | I talked with someone from bloomberg about that | 17:12 |
erhudy | in barcelona? | 17:12 |
inc0 | anyways, yeah, would be awesome guys, I know many people asked about calico with a bit of reserve | 17:12 |
inc0 | yeah | 17:12 |
erhudy | would've been our team lead then, he was the only one who went | 17:12 |
erhudy | but there'll be more of us in boston | 17:12 |
inc0 | cool, I'd be happy to meet you guys | 17:12 |
erhudy | we're still trying to rustle up support internally for actually using kolla in the first place, so this ship could still hit the rocks and sink | 17:12 |
inc0 | yeah I know it's tough sell | 17:13 |
erhudy | but as long as that comes through, calico is something we want to do | 17:13 |
inc0 | anyway, let me know if you need any help | 17:13 |
hawi | hi. has anyone used browbeat with kolla installation? just searching some way to benchmark setup... | 17:13 |
sean-k-mooney | inc0: by the way pci-passhtoruh and sriov almost work with kolla today. we might contribute the support for that early next cycle if we get back around to it. we need it for our ci down the road | 17:13 |
inc0 | I'd be happy to have bloomberg in the fold too:) | 17:13 |
inc0 | hawi, we did some rally testing on our OSIC cluster, 130+ node openstack | 17:14 |
inc0 | so far I haven't seen any proof of kolla affecting perfomrance of openstack in any way | 17:14 |
inc0 | SamYaple, didn't you run kolla with pci-passthrough alreayd? | 17:15 |
SamYaple | inc0: yea, but i had to manually hack on the pci binds | 17:16 |
sean-k-mooney | inc0: we have done it in our lab but the libvirt container needs to be tweeked to allow access to /dev/vfio | 17:16 |
SamYaple | nothing dynamic and flexible | 17:16 |
sean-k-mooney | for sriov you also need to deploy the neutron sriov nic agent | 17:16 |
sean-k-mooney | so that is another minor tweek to introduce a new container based of off neutron-base and the relevent ansible code | 17:17 |
SamYaple | sean-k-mooney: which i had (because source) | 17:17 |
SamYaple | i ran this before tokyo | 17:17 |
SamYaple | not to much manual tweaking, but it certainly wasnt merge worth | 17:18 |
sean-k-mooney | ya it was like the ironic support 98% works out of the box | 17:18 |
sean-k-mooney | ironic should work now out of the box | 17:18 |
*** bmace has quit IRC | 17:18 | |
*** bmace has joined #openstack-kolla | 17:19 | |
sean-k-mooney | but still one or two edgecase to take care of for pcipassthrough/sriov but nothing majour | 17:19 |
SamYaple | agreed. it wasnt hard to get working | 17:19 |
sean-k-mooney | the intel-nfv ci run ontop of a kolla could so we need to add sriov support when we migrate sriov testing to the intel-nfv ci | 17:19 |
*** athomas has quit IRC | 17:20 | |
portdirect | sean-k-mooney: erhudy: great! - getting OVN and calico running on kolla-k8s is something I'd really like to see happen | 17:20 |
SamYaple | for you portdirect, anything! | 17:21 |
portdirect | SamYaple: OVN makes loads more sense in a k8s enviroment | 17:21 |
portdirect | (and I assume calico) | 17:21 |
portdirect | as you can hava single sdn layer for both - I have pods and vms coesisting on the same networks - and that realli is a game changer | 17:22 |
*** jtriley has joined #openstack-kolla | 17:22 | |
sean-k-mooney | portdirect: once i finish with dpdk ill take a look at an ovn container if no one else has got to it. its really easy to create a tempalte overide(basica remove dpdk stuff for dpdk-ovs one) to support ovn | 17:23 |
SamYaple | agreed | 17:23 |
sean-k-mooney | portdirect: yes that is definetly a nic addvalue | 17:24 |
portdirect | yeah - its really easy - ive got some here: https://github.com/portdirect/harbor/tree/latest/docker/openvswitch that could be useful for reference | 17:24 |
*** sdake has quit IRC | 17:24 | |
sean-k-mooney | interesting was going to look at the devstack plugin for ovn but ya i dont think it would be too hard. if i dont get dpdk support complete though soon ill be shot so im going to finish that up first | 17:26 |
sean-k-mooney | k8 recommend not sharing the etcd store that k8s uses with anything else right? | 17:27 |
portdirect | yeah thats a massive no-no | 17:27 |
portdirect | fine for developement | 17:27 |
sean-k-mooney | for calicao and optionally ovn we would have to spin up a seccond instance of etcd. any issue runing etcd on k8s | 17:28 |
inc0 | sean-k-mooney, none | 17:28 |
sean-k-mooney | i assume as long as the data is stored on a shared volume its fine form a ha point of view | 17:28 |
openstackgerrit | sayantani proposed openstack/kolla: Update Bifrost documentation https://review.openstack.org/417114 | 17:28 |
inc0 | in fact, this was the first example of operator that coreos published;0 | 17:28 |
*** sdake has joined #openstack-kolla | 17:28 | |
inc0 | etcd on top of k8s that is | 17:28 |
sean-k-mooney | oh ya i did actully read(skim) that. i shoudl re read that article | 17:29 |
portdirect | sean-k-mooney: slightly loosing you i think - in a k8s context we should use calicos cni mainifests I think - whihc sets up etcd as part of that | 17:29 |
inc0 | oh btw sayanta__ , meet sean-k-mooney , Sean is from Intel too, he wrote initial bifrost support | 17:30 |
inc0 | so he would be good person to ask if you run into issues:) | 17:30 |
sayanta__ | Hey sean-k-mooney, inc0 | 17:30 |
inc0 | Sean, Sayantani is in my team, back in Texas | 17:30 |
openstackgerrit | Sam Yaple proposed openstack/kolla-ansible: Add zookeeper role https://review.openstack.org/412423 | 17:30 |
sayanta__ | sean-k-mooney will reach out to you if I need help :) | 17:31 |
portdirect | not sure how to deal with other kolla projects, but with kolla-k8s we would also want to use the kubernetes-ovn projects work i think | 17:31 |
sean-k-mooney | portdirect: oh ya that would make sense just was courious re etcd on k8s and if that was an issue given both calico and potenatil ovn might use etcd for there db in the future | 17:31 |
sean-k-mooney | sayanta__: no worries im not always on irc but if im not just send me an email and ill jump on | 17:31 |
*** msimonin has joined #openstack-kolla | 17:32 | |
sayanta__ | sean-k-mooney thanks. Will do so :) | 17:32 |
SamYaple | portdirect: the more external-to-kolla projects you can use the better | 17:32 |
dmsimard | Hey, can we merge this ? Missing one more +2 in kolla-ansible, it literally does not work for CentOS: https://review.openstack.org/#/c/413911/ | 17:32 |
portdirect | +(sideways 8) SamYaple :) | 17:32 |
SamYaple | portdirect: if we could get to the point of no infra services and just kolla-builds-openstack... oh boy | 17:32 |
dmsimard | mandre, sean-k-mooney ^ | 17:32 |
inc0 | Jeffrey4l_, pbourke ;) | 17:33 |
inc0 | ^ | 17:33 |
Jeffrey4l_ | yes? | 17:33 |
sean-k-mooney | oh yes that issue. that is a result of a change in docker 1.12 | 17:33 |
portdirect | SamYaple: with the openstack on k8s projects I'm trying to get there, first ceph, then mariadb... | 17:33 |
pbourke | done | 17:33 |
Jeffrey4l_ | thats my patch ;) | 17:34 |
*** adrian_otto has joined #openstack-kolla | 17:34 | |
*** msimonin has quit IRC | 17:34 | |
dmsimard | Jeffrey4l_: that's my bug :P | 17:34 |
portdirect | SamYaple: one day I'd love to see no need for Kolla at all - but think thats a long way off | 17:34 |
Jeffrey4l_ | lol | 17:34 |
SamYaple | portdirect: yea im trying to make all infra services truly optional in kolla-salt | 17:35 |
SamYaple | portdirect: but its a balance of flexibility and readability | 17:35 |
*** jtriley has quit IRC | 17:36 | |
openstackgerrit | Merged openstack/kolla-ansible: Remove -H fd:// for docker.service in the baremetal role https://review.openstack.org/413911 | 17:37 |
*** jtriley has joined #openstack-kolla | 17:37 | |
*** egonzalez90 has quit IRC | 17:37 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla: TEST NEWTON BRANCH https://review.openstack.org/417033 | 17:37 |
*** lrensing has quit IRC | 17:37 | |
sean-k-mooney | by the way does kolla update docker on upgrade? im guessing no? | 17:38 |
openstackgerrit | sayantani proposed openstack/kolla: Update Bifrost documentation https://review.openstack.org/417114 | 17:38 |
sean-k-mooney | the -H fd:// issue will be hit by anyone upgrading form docker 1.11 to 1.12+ | 17:39 |
Jeffrey4l_ | sean-k-mooney, hrm, seems yes. should remove docker.socket file too, if it exist. | 17:40 |
*** unicell has joined #openstack-kolla | 17:40 | |
*** matrohon has joined #openstack-kolla | 17:40 | |
inc0 | sean-k-mooney, no, kolla doesn't upgrade docker | 17:40 |
*** unicell has quit IRC | 17:40 | |
*** unicell has joined #openstack-kolla | 17:40 | |
dmsimard | sean-k-mooney: the baremetal role pins to 1.11 though https://github.com/openstack/kolla-ansible/blob/master/ansible/roles/baremetal/defaults/main.yml#L22 | 17:41 |
dmsimard | why ? | 17:41 |
inc0 | dmsimard, docker has bad habit of breaking things between versions | 17:41 |
inc0 | so we keep it safe | 17:41 |
sean-k-mooney | to prevent the issue with -h fd:// and because the gsg at the time said to uses 1.11 explcitly | 17:41 |
Jeffrey4l_ | wait. first of all, 1.11 has docker.socket. | 17:42 |
sean-k-mooney | Jeffrey4l_: yes | 17:42 |
Jeffrey4l_ | remove -H fd:// may cause some issue. | 17:42 |
*** sayanta__ has quit IRC | 17:42 | |
Jeffrey4l_ | but keep docker.socket is a bad idea. | 17:43 |
sean-k-mooney | the -H fd:// issue should only happen with 1.12+ when it was removed | 17:43 |
Jeffrey4l_ | so how about "push a new patch to remove docker.socket file too" | 17:43 |
Jeffrey4l_ | keep docker.socket will cause the docker won't be running when boot the host os. | 17:43 |
sean-k-mooney | am well the baremetal role is intended to be run on a clean system so docker is likely not installed | 17:44 |
sean-k-mooney | fedora enables it by default but i dont think centos does | 17:44 |
sean-k-mooney | ubuntu has lxc enable not docker so there is not issue on that side | 17:44 |
Jeffrey4l_ | docker.socket is a lazy load service tech. | 17:45 |
dmsimard | There's no mention of 1.11 explicitely in http://docs.openstack.org/developer/kolla/quickstart.html | 17:45 |
Jeffrey4l_ | service won't run until the socket get the first package. | 17:45 |
Jeffrey4l_ | packet* | 17:46 |
sean-k-mooney | well there is a simple way to test. anyone have a vm they can install docker in and reboot to see what happens with our -H fd:// | 17:46 |
Jeffrey4l_ | i reproduce this in my env. anyone can confirm this again? | 17:47 |
dmsimard | There's that darn influxdb key that I'm unable to get again | 17:50 |
dmsimard | I'm inside the container that's trying to build with bash -- I can curl anything but the curl on the influxdb key is not working. | 17:51 |
dmsimard | it's like a deity somewhere doesn't want me to use kolla | 17:52 |
sean-k-mooney | dmsimard: my mistake. the quickstart used to give commands to install specific versions https://github.com/openstack/kolla/blob/fb5773f1000b5552888ab723843247b8a68dda57/doc/quickstart.rst i taught i saw one for 1.10 when i was writing the baremental role but i guess not. 1.11 was proabley the most recent docker that was out at the time | 17:53 |
*** saneax is now known as saneax-_-|AFK | 17:53 | |
sean-k-mooney | dmsimard: are you pehind a proxy | 17:54 |
openstackgerrit | Merged openstack/kolla-ansible: Fix code style in start_compute.yml https://review.openstack.org/416583 | 17:54 |
dmsimard | sean-k-mooney: no, and I get this kind of issue from different environments | 17:54 |
pbourke | Jeffrey4l_: can you do string substitution in python with a $ ? | 17:54 |
sean-k-mooney | dmsimard: oh ok. your just haunted so | 17:54 |
Jeffrey4l_ | hrm. string.replace? pbourke | 17:55 |
dmsimard | from inside the container, I can curl plenty of things, networking works and stuff .. but I keep seeing issues getting thing like the influxdb key or the percona repository rpm | 17:55 |
pbourke | Jeffrey4l_: this is wrt to your review https://review.openstack.org/#/c/396830 | 17:55 |
Jeffrey4l_ | ah, this is oslo.config 's feature. | 17:55 |
dmsimard | pretty darn frustrating | 17:55 |
pbourke | weird | 17:56 |
pbourke | ok if it works | 17:56 |
Jeffrey4l_ | pbourke, check this https://github.com/openstack/oslo.config/blob/master/oslo_config/cfg.py#L251 | 17:56 |
Jeffrey4l_ | it works. i tested. | 17:56 |
sean-k-mooney | pbourke: it kindof confusing though | 17:56 |
sean-k-mooney | it would be more clear to just use % | 17:56 |
pbourke | oslo has plenty of oddity like that | 17:57 |
Jeffrey4l_ | other openstack project, like nova or cidner use this feature. | 17:57 |
pbourke | too late to fight it now | 17:57 |
sean-k-mooney | ok i gues that is fine if its used in other openstack pojects i guess it quilifies as domain knolage | 17:58 |
sean-k-mooney | by the way is there any way in oslo config that would allow us to swap between tarbal sources and git sources by default? | 17:59 |
sean-k-mooney | that was a feature that was asked for in the austin summit by someone at the operators meetup | 18:00 |
Jeffrey4l_ | i have no idea on that point ;( | 18:00 |
Jeffrey4l_ | yep. it is a feature nice to have | 18:00 |
sean-k-mooney | ya but the only way i can think of doing it would be to have 2 differnt source dictionaly and some how switch on a conditional when generating the config | 18:01 |
Jeffrey4l_ | yep. | 18:02 |
sean-k-mooney | that seamed like a lot of work for not such a huge gain. you could proably do something with awk/sed to trahsform one into the other. but again that is hacky. | 18:04 |
Jeffrey4l_ | awk/sed is easy to wrong. some projects have different tarball name and repo name. | 18:05 |
*** jtriley has quit IRC | 18:05 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible: Optimize reconfigure action for nova https://review.openstack.org/407416 | 18:05 |
jrich523 | hey guys... is there a way to query and delete (pipe) with the cli? openstack server list --name whatever | openstack delete | 18:06 |
jrich523 | seems like that should be something that i can do | 18:06 |
dmsimard | Jeffrey4l_: so I probably forgot the conclusion, were we going to update to ansible 2.1 for ocata ? | 18:06 |
jrich523 | or maybe im just coming up with reasons to work on the powershell module :) | 18:06 |
Jeffrey4l_ | seems we didn't get the conclusion ;( but i hope to. | 18:06 |
dmsimard | jrich523: openstack server list -f value -c ID |xargs openstack server delete ? | 18:08 |
Jeffrey4l_ | dmsimard, anyway, push a patch and collect opinion ;) | 18:08 |
sean-k-mooney | jrich523: why not openstack server delete name? | 18:10 |
jrich523 | that scares me, but i'll try it :) | 18:10 |
jrich523 | i want all of my test servers | 18:10 |
jrich523 | so anythign with jr in it | 18:10 |
sean-k-mooney | ah ok | 18:10 |
jrich523 | openstack server list --name .*jr.* | remove! | 18:10 |
jrich523 | lets see if dmsimard is my hero! | 18:11 |
jrich523 | i'll test with show :) | 18:11 |
kfox1111 | I think https://review.openstack.org/#/c/416744/ is ready | 18:11 |
Jeffrey4l_ | jrich523, openstack server list -f value -c Name -c ID | awk '/^jr/{print $2}' | xargs -n1 openstack server delete | 18:11 |
jrich523 | that is one serious hot mess of data | 18:11 |
jrich523 | oh, hmm hows that different from using --name? | 18:12 |
Jeffrey4l_ | hrm. i never used --name ;( | 18:12 |
Jeffrey4l_ | personally, i like awk/sed lol | 18:12 |
jrich523 | im a windows guy, so.... you know... i like good gui's, but not happening here! :) | 18:12 |
jrich523 | powershell would be ideal, and one of these days, i'll do it | 18:13 |
jrich523 | so i could just do | 18:13 |
dmsimard | bash/linux CLI is all about learning how to pipe things :) | 18:13 |
jrich523 | Get-POServer -name *jr* | remove-psserver -force | 18:13 |
sean-k-mooney | jrich523: but there is a minecraft mod for openstack and microsfot did by them sot that is totally the office windows gui lol | 18:13 |
Jeffrey4l_ | i am a windows hater | 18:13 |
jrich523 | which i learned with powershell dmsimard :) | 18:13 |
jrich523 | thats fair | 18:14 |
Jeffrey4l_ | sean-k-mooney, wow? where is the mod? | 18:14 |
jrich523 | but, the last 2-3 years have been game changes | 18:14 |
*** ntpttr_ has quit IRC | 18:14 | |
Jeffrey4l_ | sound interesting | 18:14 |
jrich523 | finally grasping opensource | 18:14 |
sdake | kfox1111 ping | 18:14 |
jrich523 | lol im not sure i full grasped that.. what is this minecraft mode doing? | 18:14 |
jrich523 | suppose a link would do :) | 18:15 |
dmsimard | jrich523: I don't see the problem with "openstack server list -f value -c ID |xargs openstack server delete" -- xargs is a way to iterate through lines or other separators.. it's equivalent of doing "for server in $(openstack server list -f value -c ID)"; do openstack server delete $server; done" | 18:15 |
sean-k-mooney | Jeffrey4l_: https://www.youtube.com/watch?v=JDmyefJo7bs | 18:15 |
dmsimard | it's just that the openstackclient CLI doesn't accept values as STDIN so you can't pipe directly to it | 18:15 |
Jeffrey4l_ | thanks. | 18:15 |
jrich523 | oh | 18:16 |
jrich523 | well thats pretty cool... | 18:16 |
jrich523 | im getting better at nix, but im still jr at best :( | 18:16 |
jrich523 | i have graduated from the "google is my linux knowledge" tho :) | 18:16 |
jrich523 | so, is it safe to assume you guys all hate horizon? | 18:17 |
sean-k-mooney | jrich523: i love horizon | 18:18 |
sean-k-mooney | it means i dont have to rember long xarg/sed/awk commands | 18:18 |
sean-k-mooney | but i use what ever is simpler at the time | 18:18 |
Jeffrey4l_ | the mode is very cool. | 18:18 |
Jeffrey4l_ | mod* | 18:19 |
jrich523 | ugh the version i've been using (Cisco metapod) has been really slow/buggy/and mostly doesnt do many 'neat' things | 18:19 |
sean-k-mooney | Jeffrey4l_: totally not pratical but it certenly has prommise | 18:19 |
jrich523 | for chuckles, anyone use a managed service openstack, did you like it, would you recommend it? | 18:19 |
sean-k-mooney | if by managed service you meend the kolla/devstack cloud i set up on my devs server then no i think its a terible idea lol but if you mean managed by other people so i dont have keep fixing thing then maybe :) | 18:21 |
sean-k-mooney | but no i havent used any managed openstack offerings | 18:21 |
jrich523 | no, i mean like, cisco metapod, mirantis | 18:21 |
jrich523 | seems scary to me, even more so after working with metapod | 18:22 |
jrich523 | didnt figure you guys would have, but didnt hurt to ask :) | 18:22 |
sean-k-mooney | i think the more basic an vanilla the managed service the better your experice will be | 18:22 |
*** jascott1 has joined #openstack-kolla | 18:22 | |
sean-k-mooney | im sure someone else here has | 18:22 |
jrich523 | yeah probably, but even jsut the over all stability of of it worries me for some clients that think its magic | 18:22 |
sean-k-mooney | they may even run one | 18:23 |
openstackgerrit | Merged openstack/kolla: Fix a typo in releasenotes https://review.openstack.org/416853 | 18:23 |
sean-k-mooney | jrich523: by the way if you are on window and you want to get more comfortable with linux or bash you should try cygwin. | 18:25 |
jrich523 | im on it now for the openstack cli | 18:26 |
jrich523 | however i only use it for this... im about 100x faster with powershell :) | 18:26 |
jrich523 | plus powershell works on nix now.. so... i'll continue with that for now... its a lot of fun on nix... really powerful and quick to get stuff done (for me) | 18:26 |
SamYaple | jrich523: kolla-powershell | 18:27 |
SamYaple | go! | 18:27 |
sean-k-mooney | i like that powershell can interact with .net but i have only ever realy used that the other way around e.g. form c# to powershell | 18:27 |
jrich523 | lol | 18:27 |
jrich523 | i dont think there would need to be an kolla specific cmdlets tho? | 18:27 |
* srwilkers groans | 18:27 | |
jrich523 | i mean, if you can think of some i'd be more than happy to do it (once i sort out the damn token issue) | 18:27 |
SamYaple | i agree with you though jrich523, powershell > bash (at least when compared to thier primary environments) | 18:28 |
SamYaple | so powershell on windows > bash on linux | 18:28 |
jrich523 | well, objects + pipe, how can you dislike it? :) | 18:28 |
SamYaple | thats what sold me for sure | 18:28 |
jrich523 | so crazy powerful and so much easier to work with | 18:28 |
SamYaple | but as i told you before, my windows admin days are a decade ago | 18:28 |
sean-k-mooney | you do have pipes on windows too | 18:28 |
*** cliles has quit IRC | 18:28 | |
jrich523 | rough learning curve tho | 18:28 |
jrich523 | sorta sean-k-mooney | 18:28 |
SamYaple | sean-k-mooney: pipes on linux chop information | 18:29 |
jrich523 | windows has.... matured... in the last 10 years :) | 18:29 |
SamYaple | pipes in powershell can pass information that isn't used through to more pipes | 18:29 |
jrich523 | its not the horrible mess it use to be | 18:29 |
*** msimonin has joined #openstack-kolla | 18:29 | |
jrich523 | its a different horrible mess | 18:29 |
SamYaple | jrich523: i know. im not a windows hater | 18:29 |
jrich523 | but a mess geared towards scripting and automation | 18:29 |
sean-k-mooney | i used to run tcp dump on remote servers via ssh and pip that to wireshark installed locally on windows | 18:29 |
jrich523 | nice :) | 18:30 |
jrich523 | i wrote powershell script that did something sorta similar.. basically built and ran network tests between windows and solaris boxes.. was kinda neat... took all the results and generated a report... | 18:30 |
jrich523 | i really love me some powershell | 18:30 |
sean-k-mooney | to be extra windows gui frenly i turned the command into a desktop shortcut. now i just ssh in and use tshark in the terminal its just quicker | 18:31 |
jrich523 | lol nice | 18:31 |
jrich523 | is your workstation windows? | 18:31 |
sean-k-mooney | ya | 18:31 |
SamYaple | so dhellmann brought up a good point on the ML, why does kolla-salt or kolla-puppet (or kolla-k8s or kolla-ansible) need to exist under the kolla namespace | 18:32 |
jrich523 | lol.. ctrl+d in cygwin is close.. in powershell ISE its "focus back on command line" | 18:32 |
jrich523 | ... i lost my work :( | 18:32 |
jrich523 | cuz its specifically for configuring kolla with those tools? | 18:32 |
jrich523 | just a random guess | 18:32 |
SamYaple | if they are simply consuming the kolla container images, and really have little relation to each other past that, why should kolla be building this heierachy of projects | 18:32 |
SamYaple | i have to say, that is a position I hadn't considered, but I think he is correct | 18:33 |
jrich523 | well thats the point of kolla, to build the heierachy of projects.. because after that, you've jsut got openstack :) | 18:33 |
*** msimonin has quit IRC | 18:33 | |
SamYaple | jrich523: no, the point of kolla is to build openstack images, kolla-ansible is to deploy them | 18:33 |
SamYaple | kolla-k8s is to deploy them | 18:34 |
jrich523 | hmmm | 18:34 |
SamYaple | but kolla is seperate | 18:34 |
jrich523 | fair point, but, what good is an image if you cant easily deploy it | 18:34 |
SamYaple | what good is keystone if you cant auth to it? | 18:34 |
SamYaple | like keystone standalone, what does it really do? | 18:34 |
SamYaple | it needs to be consumed | 18:34 |
jrich523 | i guess i see kolla as two parts, the images and the deployment, which is really why i love it | 18:34 |
jrich523 | right | 18:34 |
SamYaple | i agree. that is correct | 18:34 |
sean-k-mooney | SamYaple: well they dont but if you want to make sure the image work for all the deployment tools it makes sense to group them | 18:35 |
SamYaple | but, the question was posed, why is the deployment under the kolla namespace, to what advantage is that? | 18:35 |
jrich523 | i like the kolla-*, partly because i can easily find cmd options | 18:35 |
jrich523 | i assume there is a find for commands? i thnk i used it once | 18:35 |
openstackgerrit | Merged openstack/kolla: Indent all binary part https://review.openstack.org/416819 | 18:35 |
SamYaple | jrich523: im not sure I follow | 18:35 |
jrich523 | but... lol... powershell would be, get-command kolla-* | 18:35 |
jrich523 | which, im sure there is a linux version of that, which where im going is... without reading docs i can gain a lot of info about whats available to me | 18:36 |
SamYaple | sean-k-mooney: ok, what if a deploy tool just consumes kolla containers? | 18:36 |
SamYaple | sean-k-mooney: what if openstack-ansible started consuming kolla images | 18:36 |
sean-k-mooney | well fule and trippleo have projects under there namespace that do that | 18:36 |
SamYaple | sean-k-mooney: should it be kolla-openstack-ansible then? | 18:36 |
jrich523 | i'll be back in a bit guys... always enjoy the chats :) | 18:36 |
sean-k-mooney | e.g. fule-ccp | 18:37 |
SamYaple | sean-k-mooney: tripleo is a project in its own right, and the puppet-tripleo exists | 18:37 |
SamYaple | so if anything, it should be ansible-kolla | 18:37 |
SamYaple | salt-kolla | 18:37 |
sean-k-mooney | yes but untill the repo split it was just kolla | 18:37 |
SamYaple | which is why this is a question _now_ | 18:37 |
SamYaple | weve known for a long time the repo will split | 18:38 |
SamYaple | i first talked about it before merging the first ansible code | 18:38 |
SamYaple | now that it has... well this is a valid converation | 18:38 |
SamYaple | we are all under the _openstack_ namespace, not the kolla namespace | 18:38 |
sean-k-mooney | ya thats true | 18:39 |
sean-k-mooney | but what is the mailing list tread about | 18:39 |
sean-k-mooney | i havent been reading it | 18:39 |
SamYaple | kolla-salt as a kolla delievable | 18:40 |
SamYaple | sean-k-mooney: "[openstack-dev] [tc][kolla] Adding new deliverables" | 18:40 |
sean-k-mooney | ok and what are the implication of that vs kolla-salt as a salt deliverabel i gues in this case? | 18:41 |
sean-k-mooney | or kolla-puppet as a deliverable of the openstack puppet modules | 18:41 |
sean-k-mooney | is it just a nameing question | 18:41 |
SamYaple | no, thats the point, it doesnt need to be a delieverable as a subset | 18:42 |
SamYaple | it just needs to exist in openstack | 18:42 |
sean-k-mooney | ah ok ya i get that | 18:42 |
SamYaple | so <random_project_name> that uses salt and kolla containers to deploy openstack | 18:42 |
sean-k-mooney | i dont see why it cant be standalobe | 18:42 |
sean-k-mooney | networking-ovs-dpdk is standalone, its not part of neutron, or the neutron staidium though it is a neutron afiliated project | 18:43 |
SamYaple | it can be, I just had be assuming it would be advantagous to be kolla-salt, but the more i think about it, the less it seems like it would be advantagous | 18:43 |
britthouser | Seems like if there comes a day when a major change in containers come, it would be need to be synchronized with all the deployment repos | 18:43 |
britthouser | that would be hard(er) if they are under different governance | 18:43 |
SamYaple | britthouser: kind of like when keystone adds domains and projects to something that only had tenants? | 18:43 |
britthouser | Yeah I think there is a need to keep things in sync, but mostly one way. | 18:44 |
SamYaple | at this point, kolla images produce an api with standard ways of communication (config.json, extend_start) any major changes _would_ be breaking the api and thats why gates and testing existsts | 18:44 |
sean-k-mooney | britthouser: unless we dropped docker support and moved to lxd there are not many ways to break the images consoumed but the kolla-* deployment tools | 18:45 |
sean-k-mooney | well no that is not quite true | 18:45 |
britthouser | I’m not saying unexpected change, I’m saying, Kolla needs to make the change. and deployment tools need to be updated. If everyone is in their silo, its hard to coordinate. If we’re all in the same kolla silo, then we can say, “This phase X happens" | 18:45 |
britthouser | at least that’s how I see it. | 18:45 |
britthouser | Its a very rare occurrence though, so give it the proper weight. | 18:46 |
openstackgerrit | Merged openstack/kolla: Add support for installing binary packages of watcher https://review.openstack.org/416426 | 18:46 |
openstackgerrit | Merged openstack/kolla: Add tarballs-base parameter for kolla-build https://review.openstack.org/396830 | 18:47 |
SamYaple | britthouser: but thats the point of openstack as a whole | 18:48 |
SamYaple | britthouser: there isnt a central authority. thats not the model openstack follows | 18:48 |
*** mcnanci has joined #openstack-kolla | 18:48 | |
SamYaple | you are refering to a benevolant dictator model | 18:48 |
*** msimonin has joined #openstack-kolla | 18:49 | |
*** jtriley has joined #openstack-kolla | 18:51 | |
sean-k-mooney | for apis im in favor of the auortiative stance but with the deployment tools its not an api boundry. | 18:52 |
*** sayantan_ has joined #openstack-kolla | 18:52 | |
sean-k-mooney | the one situation i would not like to see happen is that we cant update images because we have to firs tupdate 5 differnet deployment tools that are in 5 different languages for the chage to function | 18:53 |
*** LiberalSquash has joined #openstack-kolla | 18:55 | |
*** LiberalSquash has left #openstack-kolla | 18:55 | |
SamYaple | so sean-k-mooney what you are refering to is solved (in the openstack world) by inter project collaboration | 18:55 |
SamYaple | sean-k-mooney: the container images themselves are absolutely an api boundry | 18:55 |
SamYaple | that was esstablished prior to mitaka | 18:56 |
sean-k-mooney | yes im thinking more of how neutron works | 18:56 |
SamYaple | thats why seperating ansible was so important | 18:56 |
sean-k-mooney | when you add a new feature to the neutron api you only have to provide an implentaiton in one of the reference backends not all of the for it to be added | 18:56 |
sean-k-mooney | its up to each backend then to add support if it relevent to there comunity. | 18:57 |
SamYaple | sean-k-mooney: thats the same for kolla though | 18:57 |
SamYaple | ubuntu and centos dont have to be implemetned at the same time | 18:57 |
SamYaple | containers images themselves dont have to have deployment tool counterparts to be merged | 18:57 |
SamYaple | we have far more services containerized than we do deployable with kolla-ansible | 18:58 |
sean-k-mooney | all true | 18:58 |
SamYaple | the interface into the container _itself_ is not subject to change without collaboration | 18:58 |
SamYaple | but the fact is, that interface was born of collabroation | 18:58 |
SamYaple | albeit with kolla-mesos, but its been working great for kolla-salt and kolla-k8s without need to change | 18:59 |
sean-k-mooney | e.g. e.g. the extend start and command.json files? | 18:59 |
SamYaple | yes | 18:59 |
sean-k-mooney | k8s does not use them currently though | 18:59 |
sean-k-mooney | or atleast not for everything | 18:59 |
SamYaple | can you give an example? | 18:59 |
sean-k-mooney | there is special case kode in places to support k8s. am yes on sec | 19:00 |
SamYaple | but mind you, _that_ was the agreed upon api, not subject to change without appropriate deprecation time | 19:00 |
dmsimard | sean-k-mooney: still trying to troubleshoot my crappy curls on stuff like percona and influx.. I think I have a hint. It looks related to SSL (some docker issues mention lowering the MTU through docker opts, wtf?) | 19:00 |
sean-k-mooney | SamYaple: https://github.com/openstack/kolla/blob/master/docker/openvswitch/openvswitch-db-server/start_ovsdb_server.sh | 19:02 |
SamYaple | so sean-k-mooney you are talking about those helper scripts in the container | 19:02 |
SamYaple | my personal prefernce would be to see those removed and moved to either a central config repo (which has been dicussed) or stripped out to thier respective deployment repos | 19:03 |
SamYaple | i consider than an artifact of kolla-ansible being in kolla repo | 19:03 |
SamYaple | whether others agree or not, im not sure. but i do know there was never a discussion of those scripts being reusable | 19:04 |
sean-k-mooney | yes perhaps | 19:04 |
*** goldyfruit has quit IRC | 19:04 | |
SamYaple | even the baked in extend_start scripts themselves have kolla-ansible artifacts | 19:04 |
SamYaple | but since the api is config.json and extend_start interface, you can override them | 19:04 |
sean-k-mooney | but if you can override them why was it nessicary to bake in the k8s support into the image? | 19:05 |
SamYaple | sean-k-mooney: i wasnt here for that. i would have disgreed with it | 19:05 |
SamYaple | i would have also disagreed with k8s infecting the ansible config with k8s code too, though its a bit more reasonable when you remember initially they were thinking of _using_ ansible with k8s | 19:06 |
SamYaple | nowadays its a wierd cross-project entaglement that has broken k8s several times | 19:06 |
*** Bico_Fino has quit IRC | 19:06 | |
TxGirlGeek | inc0 ? | 19:07 |
SamYaple | sean-k-mooney: my personal opinion here, we can provide script "libraries" that just have importable functions and call those apis, then have the deploy projects source those functions and bind in thier own extend_start | 19:07 |
SamYaple | but that hasnt been discussed, just what I would like to see happen | 19:07 |
*** rhallisey has joined #openstack-kolla | 19:08 | |
sean-k-mooney | ya actully for the dpdk patches i might handel this a little differently. that said im not targeting k8s in the first version so i was just going to supprot ansible and extend | 19:08 |
sean-k-mooney | SamYaple: but yes that sound reasonable. | 19:09 |
SamYaple | sean-k-mooney: look here: https://github.com/SamYaple/kolla-salt/blob/master/salt/kolla/mariadb/bootstrap.sls#L42 | 19:09 |
sean-k-mooney | i kind of get the felling that either kolla should not have been split or it should have been split more. kolla-images,kolla-docs,kolla-config | 19:09 |
SamYaple | i can bind in the bootstrap script over extend_start (this was the original design, btw) | 19:10 |
SamYaple | or use config.json to copy it in, same result | 19:10 |
SamYaple | sean-k-mooney: i think well see kolla-config | 19:10 |
*** MasterOfBugs has joined #openstack-kolla | 19:10 | |
SamYaple | but splitting kolla was definetely the right call | 19:10 |
SamYaple | idk about kolla-docs. but i guess that would be ok | 19:10 |
SamYaple | we can have this discussion only _because_ kolla was split | 19:11 |
sean-k-mooney | thats true. anyway i better run soon | 19:11 |
SamYaple | great chat as always sean-k-mooney | 19:12 |
SamYaple | sean-k-mooney: hit me up if you want me to look over your dpdk stuff | 19:13 |
SamYaple | or just add me as a reviewer | 19:13 |
*** lrensing has joined #openstack-kolla | 19:13 | |
openstackgerrit | Sam Yaple proposed openstack/kolla-ansible: Add zookeeper role https://review.openstack.org/412423 | 19:13 |
*** harlowja has joined #openstack-kolla | 19:16 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible: optimize register.yml file https://review.openstack.org/416046 | 19:17 |
jrich523 | SamYaple: zookeeper role huh, someone needs to feed the kollas? :) | 19:18 |
SamYaple | jrich523: its for monasca, but good joke! | 19:19 |
*** prameswar has quit IRC | 19:20 | |
jrich523 | i know what none of that is, but thats ok :) | 19:22 |
jrich523 | im always amused at the names i come across for tech... running out of things to use lol | 19:22 |
SamYaple | yep | 19:23 |
SamYaple | the next release of ceph is called "kraken" | 19:23 |
jrich523 | lol should be some sort of software name tracker, after it hasnt been updated in 2 years you're allowed to take the name :) | 19:23 |
jrich523 | ugh, just a code name right? | 19:23 |
SamYaple | mind you there is already a ceph project unrelated called "kraken" | 19:23 |
SamYaple | nah | 19:23 |
jrich523 | oh geesh | 19:23 |
SamYaple | they do the a b c d naming just like openstack | 19:23 |
jrich523 | sounds like an awesome and extremely straight forward approach :) | 19:23 |
jrich523 | lol might as well stick with the military codes in that case then lol | 19:24 |
*** celo_r has joined #openstack-kolla | 19:26 | |
dmsimard | Just confirmed my issue with retrieving keys and repositories in the base kolla image | 19:27 |
dmsimard | curl would always fail on either percona repository setup or influxdb key | 19:28 |
dmsimard | Turns out the default MTU of eth0 in our (OpenStack) VMs is 1400 and both docker0 and eth0 inside the container were set to 1500 | 19:28 |
openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Fix service fork of ceph_workflow https://review.openstack.org/417140 | 19:29 |
dmsimard | What a shitty issue, really | 19:29 |
jrich523 | wow... why... | 19:29 |
jrich523 | those are some odd choices | 19:29 |
dmsimard | Well I never specified a default MTU of 1400 anywhere, it must come from somewhere | 19:30 |
jrich523 | lol interesting.. was curious so started to google, came across this... http://www.tp-link.com/us/FAQ-190.html | 19:31 |
jrich523 | i've never really seen a method for picking the 'ideal' MTU, i was fairly certain that it needed to just be real consistent and all was good | 19:31 |
kfox1111 | 1400 is an odd mtu. | 19:32 |
jrich523 | yup | 19:32 |
kfox1111 | I don't think that's a number a computer would have generated. | 19:32 |
kfox1111 | seems like a human probably did. | 19:32 |
*** msimonin has quit IRC | 19:32 | |
jrich523 | agreed | 19:32 |
jrich523 | you must have done it dmsimard, how dare you | 19:33 |
kfox1111 | still, yeah, I don't think there is anything in docker that tunes its mtu to match the network interfaces mtu's. | 19:33 |
jrich523 | yeah i've never come across anything like that | 19:33 |
kfox1111 | docker leaves that as an exersie to the reader. | 19:33 |
jrich523 | hey SamYaple, it seems like xargs is more of a concat? its basically tacking all the values on to the end, which, i think is ok for delete since it will take multiple, space seperated values, but i think the explaination for how it works is off :) | 19:34 |
kfox1111 | xargs? it takes a list and makes it a set of args. | 19:35 |
jrich523 | right, the explaination was more like foreach, rather than join and dump | 19:36 |
kfox1111 | cat filelist | xargs ls -l | 19:36 |
kfox1111 | its like foreach, but also does batching. | 19:36 |
kfox1111 | a foreach would be the same as: | 19:36 |
kfox1111 | cat filelist | xargs -i ls -l {} | 19:36 |
*** unicell has quit IRC | 19:37 | |
jrich523 | its works ok with delete, because delete will take multiple names, but for show it doesnt work | 19:37 |
jrich523 | err is the {} to denote "current item"? | 19:37 |
kfox1111 | yeah. | 19:37 |
jrich523 | ah cool | 19:37 |
portdirect | o/ | 19:37 |
kfox1111 | so, say you have a file "1\n2\n3\n4\n" | 19:37 |
* portdirect reads scrollback | 19:37 | |
kfox1111 | cat file | xargs echo | 19:37 |
kfox1111 | could end up: echo 1 2; echo 3 4 | 19:38 |
kfox1111 | or echo 1 2 3 4; | 19:38 |
jrich523 | openstack server list --name pat -f value -c ID | xargs openstack delete <- is what i did, and thats ok, but if i change to | 19:38 |
jrich523 | openstack server list --name pat -f value -c ID | xargs openstack show <- fails, but im going to try the -i | 19:38 |
kfox1111 | theres a -L option too that says how many lines to collaps at a time. | 19:38 |
kfox1111 | -L 1 I think would have the same effect. | 19:39 |
jrich523 | oh neat, hmm | 19:39 |
kfox1111 | I kind of prefer the -i {} thing though, as you can put the args in different places then always the end... | 19:40 |
jrich523 | grr ctrl+d | 19:40 |
jrich523 | yeah, if im going to remember one, its that one :) | 19:40 |
kfox1111 | so like: cat filelist | xargs -i echo foo: {} bar:.... | 19:40 |
kfox1111 | works. | 19:40 |
jrich523 | ugh, so damn slow... i wonder if its the VPN or openstack :-/ | 19:41 |
*** salv-orl_ has quit IRC | 19:41 | |
kfox1111 | but I usually use a completely different pattern for single line processing. | 19:41 |
kfox1111 | cat file | while read line; do echo $line; done | 19:42 |
jrich523 | that seems overly wordy for the same effect | 19:42 |
kfox1111 | you can easily do multiple things per line then. | 19:42 |
kfox1111 | cat file | while read line; do echo processing $line; md5sum $line; rm $line; done | 19:42 |
jrich523 | ah true | 19:42 |
kfox1111 | etc. | 19:42 |
SamYaple | kfox1111: hey are you using kollas rabbitmq image? | 19:42 |
kfox1111 | SamYaple: yeah. | 19:43 |
SamYaple | i was thinking of switching the kolla-salt to the official docker rabbitmq image. its better built | 19:43 |
portdirect | sean-k-mooney/SamYaple: noticed the kolla-k8s specific script discussion in kolla-k8s, I'd really like to see this gone - with k8s there is no reason for us to have anything backed into the container other than the binaries we need | 19:43 |
portdirect | *baked | 19:43 |
SamYaple | portdirect: agreed, thats what im doing for kolla-salt | 19:43 |
kfox1111 | SamYaple: yeah. we've been talking about optionally doing that too. | 19:43 |
SamYaple | kfox1111: in this case, the rabbitmq image didnt exist in a usable form. but looking at its history, its been solid for a year or more now | 19:43 |
SamYaple | https://hub.docker.com/r/library/rabbitmq/tags/ | 19:44 |
SamYaple | kfox1111: complete with drop-in plugin support | 19:44 |
kfox1111 | SamYaple: we'd probably just recommend folks use: https://github.com/kubernetes/charts/tree/master/stable/rabbitmq | 19:44 |
portdirect | ^^ +1 | 19:45 |
SamYaple | kfox1111: yea that makes sense for k8s | 19:45 |
kfox1111 | I think there's still benifit to allowing using kolla container for it though. | 19:45 |
kfox1111 | rdo packages are tested against rdo's rabbit, not other rabbits. | 19:46 |
kfox1111 | ubuntu with their own, etc. | 19:46 |
SamYaple | kfox1111: soooooo have you looked at the images? | 19:47 |
kfox1111 | so, really depends on what source the operator trusts more. | 19:47 |
SamYaple | kfox1111: https://github.com/openstack/kolla/blob/master/docker/rabbitmq/Dockerfile.j2#L21 | 19:47 |
SamYaple | not as tested as you might think | 19:47 |
kfox1111 | heh. | 19:47 |
kfox1111 | thats unfortunate. | 19:47 |
sean-k-mooney | SamYaple: https://bugs.launchpad.net/kolla-ansible/+bug/1654375 just incase this bites you in the future though i know you dont use cache teirs anymore. | 19:47 |
openstack | Launchpad bug 1654375 in kolla-ansible "ceph role does not set target_max_bytes for cach pools" [Undecided,New] | 19:47 |
SamYaple | sean-k-mooney: yea i think we talked about this right? | 19:48 |
sean-k-mooney | ya i confirmed tha twas the issue | 19:48 |
SamYaple | ah cool. thanks for that | 19:48 |
sean-k-mooney | you can correct it by setting the value on the pools manually and then temperally decresasing the replicase in the cach from 3 to 2 if the cache is already full | 19:49 |
SamYaple | kfox1111: i think the reason for this is the RDO pacakge was so old the clusterer plugin didnt work. and clsuterer plugin is amazing and needed | 19:49 |
sean-k-mooney | ill try to get back to it next week when i have a little more time. | 19:49 |
kfox1111 | SamYaple: ah. | 19:50 |
kfox1111 | I just did manual clustering. felt it was safer. | 19:50 |
SamYaple | youd be wrong there! | 19:50 |
SamYaple | oh the partitioning | 19:50 |
sean-k-mooney | portdirect: even for kolla-ansibel these helper scripts could be impleneted by ansible either as a bind or just in the ansible code itself | 19:50 |
kfox1111 | I know what all my nodes were going to be so just hardcoding the members wasn't hard. | 19:50 |
SamYaple | kfox1111: clusterer was for handling partitioning and hard cluster downs | 19:51 |
SamYaple | without clusterer if your cluster stops, you might not get it back up | 19:51 |
*** sean-k-mooney is now known as sean-k-mooneyAFK | 19:51 | |
kfox1111 | I never had a problem. | 19:51 |
kfox1111 | mostly with the partitioning not working as advertized. :/ | 19:51 |
*** sdake has quit IRC | 19:52 | |
kfox1111 | rabbit clustering is nasty. :/ | 19:52 |
*** Pavo has quit IRC | 19:52 | |
SamYaple | kfox1111: conversly, ive handled hundreds of environemnts and clusterer and autopartition healing thats in kolla is far superior | 19:52 |
kfox1111 | when the warning shows up in their docs, "don't let your network partition", you konw they don't understand how clustering works. | 19:52 |
SamYaple | both of these are subjective of course | 19:52 |
kfox1111 | cool. I'll have to give it a shot then. :) | 19:53 |
SamYaple | clusterer is amazing for upgrades too | 19:53 |
portdirect | sean-k-mooney - great :) Ideally I think each project should be able to run any container really - I've been running openstack-helm with my alpine based containers - that are totaly unrelated to kolla's and would love to get that flexibility into kolla-k8s | 19:53 |
SamYaple | basically if you stop the "gospel" node last, you dont have to worry about start order | 19:53 |
SamYaple | amazing | 19:53 |
*** sp__ has joined #openstack-kolla | 19:54 | |
kfox1111 | heh. | 19:54 |
inc0 | portdirect, on that note, I'd love to have alpine as one of distros in kolla | 19:54 |
portdirect | inc0 was working on it - though its really hard with kolla atm | 19:54 |
kfox1111 | inc0: portdirect +1 | 19:55 |
kfox1111 | portdirect: tried kolla-kubernetes with your alpine containers? | 19:56 |
kfox1111 | would be really interested to hear if it works. :) | 19:56 |
portdirect | kfox1111: wouldnt work atm :( | 19:56 |
kfox1111 | why not? | 19:56 |
kfox1111 | oh. is it how the configmaps get coppied in? | 19:56 |
portdirect | inc0: i'll come to you in a min :) | 19:56 |
portdirect | kfox1111: yes | 19:57 |
kfox1111 | ah. | 19:57 |
kfox1111 | hmm.... | 19:57 |
kfox1111 | could ajusting the container_... | 19:57 |
kfox1111 | container_config_directory setting make it work? | 19:57 |
srwilkers | o/ | 19:58 |
kfox1111 | hi srwilkers | 19:58 |
portdirect | kfox1111: not really - we would need to rewrite the entrypoint for each pod - this is what openstack-helm does - over there we only use the binaries in the kolla images | 19:58 |
portdirect | and none of the kolla scrips in them | 19:59 |
*** bjolo has joined #openstack-kolla | 19:59 | |
inc0 | portdirect, so reason we didn't do alpine yet is because infra services aren't source, are hard to come by in alpine repo and I don't trust these bins there | 19:59 |
inc0 | but if we change our arch in kolla-k8s to "just oipenstack, we deploy mariadb from helm" | 19:59 |
inc0 | and only care about openstack services | 20:00 |
inc0 | I think alpine starts making sense | 20:00 |
portdirect | inc0: right back on track | 20:00 |
kfox1111 | portdirect: I think that setting specifies where the configmap stuff shows up though, so could bind it to /etc/nova/ for example? | 20:00 |
portdirect | the issue with kolla-build and alpine is beasue of the number of theings we need to build for openstack | 20:01 |
kfox1111 | I guess there are a few templates that call kolla_start explicitly though.... | 20:01 |
jascott1 | inc0 we are in room by Rusty's cube if you want to join | 20:01 |
portdirect | esiest way to explain is looking at one of my docker containers for an example | 20:01 |
inc0 | ok, jascott1 will be there in a sec, others - meeting time;( | 20:01 |
inc0 | portdirect, let's talk about alipine more soon, I'd love to see it moving | 20:02 |
inc0 | brb | 20:02 |
portdirect | roger | 20:02 |
portdirect | kfox1111: yeah - not a probelm at the moment - this is *kolla*-kubernetes afterall - but would be a good goal to have for a later release | 20:03 |
kfox1111 | portdirect: +1 :) | 20:04 |
*** TxGirlGeek has quit IRC | 20:06 | |
portdirect | inc0: for when you get back - primary issue with alpine is that kolla has too many layers - and to keep the images small we need to have a way of effeciently adding and removing build tools and libs from containers | 20:06 |
*** TxGirlGeek has joined #openstack-kolla | 20:06 | |
portdirect | as when I made a quick alpine poc a few monts ago the images ended up being only about 50mb smaller than ubuntu - and thats just not worth it | 20:07 |
kfox1111 | portdirect: https://review.openstack.org/#/c/416744/ please | 20:08 |
kfox1111 | portdirect: yeah. +1 from removing build tools all around. | 20:08 |
kfox1111 | seems silly to have build tools in binary distro's too. | 20:08 |
portdirect | though its kinda hilarious how much quicker apk is than yum, on my workstation this entire process: https://github.com/portdirect/harbor/blob/latest/docker/openstack/openstack-nova/openstack-nova-alpine/Dockerfile#L23-L40 takes less time than yum to install the gcc and python-dev! | 20:09 |
kfox1111 | flexability vs speed. | 20:09 |
portdirect | think in this case you get both actually | 20:10 |
portdirect | but i think about everything else gets thrown out of the window to get there... | 20:10 |
kfox1111 | pkg stuff in python is very flexible. :) | 20:10 |
kfox1111 | plugins are slick. :) | 20:10 |
kfox1111 | but slow... :/ | 20:11 |
*** adrian_otto has quit IRC | 20:11 | |
*** goldyfruit has joined #openstack-kolla | 20:11 | |
portdirect | yeah - though just to empasise it further this is the script thats being run in the middle of that: https://github.com/portdirect/harbor/blob/latest/docker/openstack/openstack-nova/common-assets/opt/harbor/build/dockerfile.sh#L24-L33 | 20:12 |
kfox1111 | ah | 20:13 |
*** jascott1 has quit IRC | 20:18 | |
*** unicell has joined #openstack-kolla | 20:25 | |
*** jascott1 has joined #openstack-kolla | 20:25 | |
*** Pavo has joined #openstack-kolla | 20:25 | |
*** msimonin has joined #openstack-kolla | 20:27 | |
dmsimard | jrich523: well there's the culprit http://codesearch.openstack.org/?q=dhcp-option-force%3D26%2C1400&i=nope&files=&repos= | 20:27 |
dmsimard | re: MTU 1400 | 20:27 |
dmsimard | tl;dr of the explanation revolves around GRE/VXLAN overhead compensation | 20:28 |
portdirect | yeah - they do that to ensure operation inside a vm for testing (assuming that the host has a 1500mtu) | 20:28 |
dmsimard | portdirect: I spent way too much time troubleshooting SSL transaction errors within kolla containers because the host was at 1500, the VM was at 1400 and the container inside the VM was at 1500 | 20:29 |
dmsimard | because obviously like 95% of the network related things work and these apparently random ones don't work :) | 20:30 |
*** TxGirlGeek has quit IRC | 20:30 | |
*** jascott1 has quit IRC | 20:30 | |
portdirect | :( that sucks - if its any consolation I've been though the EXACT same issue | 20:30 |
dmsimard | docker doesn't seem to automatically match the host (in this case my VM) MTU, there's an issue about that https://github.com/docker/docker/issues/22297 | 20:31 |
dmsimard | funnily enough, openstack is plastered everywhere across that issue :) | 20:31 |
portdirect | yeah - its been brought up with them well before that as well | 20:32 |
dmsimard | oh yeah, this is just one of the 15 issues around the topic of SSL/MTU inside docker containers :) | 20:32 |
*** tovin07 has joined #openstack-kolla | 20:33 | |
* dmsimard writes a blog post about this | 20:33 | |
* portdirect is sad that it's 2007 and we still hit mtu issues | 20:34 | |
dmsimard | It's like DNS, it's never DNS :) | 20:34 |
portdirect | oh gawd dont start me on docker DNS... | 20:35 |
dmsimard | I thought docker was production ready, come on now :P | 20:35 |
* srwilkers wonders how portdirect is living 10 years in the past | 20:35 | |
* portdirect has weird lag issues with his wifi at the moment - hows the future looking - you get that hoverboard yet srwilkers? | 20:36 | |
*** unicell has quit IRC | 20:37 | |
SamYaple | portdirect: did... did you just ask srwilkers a question in your head? | 20:39 |
* portdirect thinks other people can hear the voices too. then wonders what that means. | 20:40 | |
portdirect | SamYaple: of course not - that would be silly. | 20:40 |
SamYaple | I CAN TALK IN COLORS | 20:41 |
*** sdake has joined #openstack-kolla | 20:41 | |
openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Switch python deployments to global api. https://review.openstack.org/417181 | 20:43 |
*** sp__ has quit IRC | 20:44 | |
*** sayantan_ has quit IRC | 20:51 | |
*** MasterOfBugs has quit IRC | 20:57 | |
*** msimonin has quit IRC | 20:59 | |
*** satyar has quit IRC | 20:59 | |
*** sayantan_ has joined #openstack-kolla | 20:59 | |
*** Bico_Fino has joined #openstack-kolla | 21:00 | |
*** sayantan_ has quit IRC | 21:04 | |
inc0 | back | 21:04 |
*** TxGirlGeek has joined #openstack-kolla | 21:06 | |
dmsimard | Jeffrey4l_, sean-k-mooneyAFK: I filed another bug for the baremetal role (sorry) https://bugs.launchpad.net/kolla-ansible/+bug/1654395 | 21:07 |
openstack | Launchpad bug 1654395 in kolla-ansible "The baremetal role does not properly extend/override the docker systemd unit file" [Undecided,New] | 21:07 |
dmsimard | I'll hack something in the meantime to address my MTU problem. | 21:07 |
*** sayantan_ has joined #openstack-kolla | 21:08 | |
*** bjolo has quit IRC | 21:11 | |
openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Switch svc to use global api https://review.openstack.org/417190 | 21:12 |
*** salv-orlando has joined #openstack-kolla | 21:14 | |
*** Bico_Fino has quit IRC | 21:21 | |
*** jascott1 has joined #openstack-kolla | 21:23 | |
*** JoseMello has quit IRC | 21:25 | |
openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Switch python deployments to global api. https://review.openstack.org/417181 | 21:29 |
openstackgerrit | Marcus Williams proposed openstack/kolla: [WIP] Add OpenDaylight Container to kolla https://review.openstack.org/416369 | 21:34 |
*** Jeffrey4l_ has quit IRC | 21:35 | |
openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Switch svc to use global api https://review.openstack.org/417190 | 21:35 |
*** tovin07 has left #openstack-kolla | 21:42 | |
*** dave-mccowan has quit IRC | 21:47 | |
*** Jeffrey4l_ has joined #openstack-kolla | 21:48 | |
*** portdirect is now known as portdirect_away | 21:49 | |
*** salv-orl_ has joined #openstack-kolla | 21:53 | |
*** salv-orlando has quit IRC | 21:56 | |
*** mgiles has quit IRC | 21:57 | |
*** srwilkers is now known as srwilkers_away | 21:59 | |
*** masuberu has joined #openstack-kolla | 22:04 | |
*** msimonin has joined #openstack-kolla | 22:04 | |
openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla: Support for multiple override files https://review.openstack.org/417213 | 22:07 |
*** saneax-_-|AFK is now known as saneax | 22:09 | |
*** msimonin has quit IRC | 22:09 | |
*** MasterOfBugs has joined #openstack-kolla | 22:09 | |
*** masuberu has quit IRC | 22:10 | |
*** msimonin has joined #openstack-kolla | 22:14 | |
*** msimonin1 has joined #openstack-kolla | 22:16 | |
*** msimonin has quit IRC | 22:16 | |
*** inc0 has quit IRC | 22:17 | |
*** sayantan_ has quit IRC | 22:19 | |
*** matrohon has quit IRC | 22:21 | |
*** saneax is now known as saneax-_-|AFK | 22:26 | |
*** shardy_ has joined #openstack-kolla | 22:27 | |
*** shardy_ has quit IRC | 22:28 | |
*** jtriley has quit IRC | 22:30 | |
*** jheroux has quit IRC | 22:32 | |
*** salv-orl_ has quit IRC | 22:33 | |
*** saneax-_-|AFK is now known as saneax | 22:34 | |
*** dave-mccowan has joined #openstack-kolla | 22:36 | |
*** salv-orlando has joined #openstack-kolla | 22:36 | |
*** eaguilar has joined #openstack-kolla | 22:39 | |
*** sdake has quit IRC | 22:39 | |
*** fguillot has quit IRC | 22:40 | |
*** SamYaple has quit IRC | 22:41 | |
*** SamYaple has joined #openstack-kolla | 22:42 | |
*** aric49 has quit IRC | 22:47 | |
*** dave-mccowan has quit IRC | 22:48 | |
*** saneax is now known as saneax-_-|AFK | 22:54 | |
*** sayantan_ has joined #openstack-kolla | 22:58 | |
*** jascott1 has quit IRC | 22:58 | |
*** jascott1 has joined #openstack-kolla | 23:00 | |
*** sayantan_ has quit IRC | 23:02 | |
*** sdake has joined #openstack-kolla | 23:03 | |
*** TxGirlGeek has quit IRC | 23:04 | |
*** saneax-_-|AFK is now known as saneax | 23:05 | |
*** jascott1 has quit IRC | 23:05 | |
*** jrobinson has joined #openstack-kolla | 23:05 | |
*** jrobinson_ has joined #openstack-kolla | 23:06 | |
*** selena_091 has joined #openstack-kolla | 23:08 | |
openstackgerrit | Joshua Harlow proposed openstack/kolla: Allow for image additions archives (as well as plugins) https://review.openstack.org/411534 | 23:09 |
selena_091 | hi i have a question for dev..when using cmdline `kolla-build` to build images matching the passed string like for example `kolla-build keystone ` builds all images matching keystone..so is keystone registered as a positional arg of oslo config ? does it actual get assigned to `regex` opt.?? | 23:10 |
selena_091 | im confused about this line https://github.com/openstack/kolla/blob/master/kolla/common/config.py#L544 | 23:11 |
*** selena_091 has quit IRC | 23:13 | |
*** sayantan_ has joined #openstack-kolla | 23:14 | |
*** selena_091 has joined #openstack-kolla | 23:14 | |
*** goldyfruit has quit IRC | 23:16 | |
*** saneax is now known as saneax-_-|AFK | 23:16 | |
openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Switch keystone jobs to global api https://review.openstack.org/417234 | 23:20 |
*** saneax-_-|AFK is now known as saneax | 23:26 | |
*** jascott1 has joined #openstack-kolla | 23:26 | |
*** absubram has quit IRC | 23:28 | |
*** shardy has quit IRC | 23:34 | |
*** unicell has joined #openstack-kolla | 23:34 | |
*** MasterOfBugs has quit IRC | 23:36 | |
*** unicell has quit IRC | 23:36 | |
*** msimonin1 has quit IRC | 23:36 | |
*** msimonin has joined #openstack-kolla | 23:36 | |
*** mattmceuen has quit IRC | 23:37 | |
*** msimonin has quit IRC | 23:40 | |
*** unicell has joined #openstack-kolla | 23:42 | |
*** saneax is now known as saneax-_-|AFK | 23:43 | |
*** salv-orlando has quit IRC | 23:45 | |
*** jrobinson_ has quit IRC | 23:46 | |
*** jrobinson has quit IRC | 23:46 | |
*** unicell1 has joined #openstack-kolla | 23:48 | |
*** unicell has quit IRC | 23:49 | |
*** selena_091 has quit IRC | 23:50 | |
*** jascott1 has quit IRC | 23:53 | |
*** saneax-_-|AFK is now known as saneax | 23:54 | |
*** selena_091 has joined #openstack-kolla | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!