*** zhurong has quit IRC | 00:01 | |
*** godleon has joined #openstack-kolla | 00:10 | |
*** hieulq_ has joined #openstack-kolla | 00:39 | |
*** huikang has joined #openstack-kolla | 00:40 | |
*** duonghq has joined #openstack-kolla | 00:45 | |
*** zhurong has joined #openstack-kolla | 01:12 | |
openstackgerrit | Li Yingjun proposed openstack/kolla: Add option to support dhcp/l3 agent ha https://review.openstack.org/352180 | 01:23 |
---|---|---|
*** rhallisey has quit IRC | 01:25 | |
*** RA_ has joined #openstack-kolla | 01:26 | |
*** RA_ is now known as RossKrumbeck | 01:29 | |
*** RossKrumbeck is now known as rkrum | 01:35 | |
*** rkrum has left #openstack-kolla | 01:39 | |
*** rkrum has joined #openstack-kolla | 01:39 | |
*** rkrum has quit IRC | 01:40 | |
*** rkrum has joined #openstack-kolla | 01:40 | |
openstackgerrit | Hiroki Ito proposed openstack/kolla: Fix prechecks of a vip address https://review.openstack.org/351588 | 01:49 |
*** huikang has quit IRC | 01:58 | |
*** huikang has joined #openstack-kolla | 02:00 | |
*** signed8bit_Zzz is now known as signed8bit | 02:01 | |
*** signed8bit has quit IRC | 02:02 | |
openstackgerrit | Ryan Hallisey proposed openstack/kolla-kubernetes: Ansible support for Mariadb https://review.openstack.org/334255 | 02:36 |
openstackgerrit | Ryan Hallisey proposed openstack/kolla-kubernetes: Add an --all-in-one flag to the CLI https://review.openstack.org/352138 | 02:36 |
*** fragatina has joined #openstack-kolla | 02:38 | |
*** fragatina has quit IRC | 02:43 | |
*** huikang has quit IRC | 02:46 | |
*** zhubingbing has joined #openstack-kolla | 02:48 | |
*** zhugaoxiao has quit IRC | 02:57 | |
openstackgerrit | zhubingbing proposed openstack/kolla: Add Barbibcan container dockerfile https://review.openstack.org/351822 | 03:09 |
*** daneyon has joined #openstack-kolla | 03:15 | |
*** daneyon has quit IRC | 03:19 | |
*** huikang has joined #openstack-kolla | 03:28 | |
*** zhugaoxiao has joined #openstack-kolla | 03:36 | |
*** ayoung has quit IRC | 03:39 | |
*** dave-mccowan has quit IRC | 03:47 | |
*** sdake has joined #openstack-kolla | 03:58 | |
openstackgerrit | Ross Annetts proposed openstack/kolla: Fix designate bootstrap commands. https://review.openstack.org/352206 | 03:59 |
openstackgerrit | Ross Annetts proposed openstack/kolla: Fix designate bootstrap commands. https://review.openstack.org/352206 | 04:03 |
sdake | evening ffolks | 04:03 |
sdake | god was hoping when i woke up it was monday | 04:04 |
*** zhurong has quit IRC | 04:04 | |
*** zhurong has joined #openstack-kolla | 04:05 | |
zhubingbing | - - | 04:18 |
sdake | anyone know if sbezverk had success? | 04:23 |
*** senk_ has joined #openstack-kolla | 04:24 | |
zhubingbing | i don't know | 04:25 |
*** bootsha has quit IRC | 04:27 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/kolla: Updated from global requirements https://review.openstack.org/351901 | 04:32 |
*** evewind has joined #openstack-kolla | 04:35 | |
*** evewind has left #openstack-kolla | 04:41 | |
*** evewind has joined #openstack-kolla | 04:43 | |
*** coolsvap has joined #openstack-kolla | 04:46 | |
*** huikang has quit IRC | 04:57 | |
*** huikang has joined #openstack-kolla | 04:57 | |
*** huikang has quit IRC | 05:02 | |
*** mdnadeem_ has joined #openstack-kolla | 05:02 | |
*** daneyon has joined #openstack-kolla | 05:03 | |
*** evewind has left #openstack-kolla | 05:06 | |
*** fragatina has joined #openstack-kolla | 05:07 | |
*** daneyon has quit IRC | 05:08 | |
*** senk_ has quit IRC | 05:09 | |
*** senk_ has joined #openstack-kolla | 05:13 | |
*** fragatina has quit IRC | 05:13 | |
*** senk_ has quit IRC | 05:26 | |
*** bootsha has joined #openstack-kolla | 05:28 | |
*** bootsha has quit IRC | 05:33 | |
openstackgerrit | Merged openstack/kolla: Updated from global requirements https://review.openstack.org/351901 | 05:47 |
*** duonghq has quit IRC | 05:48 | |
*** duonghq has joined #openstack-kolla | 06:01 | |
*** mewald has joined #openstack-kolla | 06:10 | |
*** senk_ has joined #openstack-kolla | 06:28 | |
*** bootsha has joined #openstack-kolla | 06:29 | |
*** bootsha has quit IRC | 06:35 | |
*** dmueller-evoila has joined #openstack-kolla | 06:35 | |
*** sdake has quit IRC | 06:38 | |
*** duonghq has quit IRC | 06:40 | |
*** zhubingbing has quit IRC | 06:44 | |
*** zhubingbing_ has joined #openstack-kolla | 06:46 | |
*** fragatina has joined #openstack-kolla | 06:47 | |
*** fragatina has quit IRC | 06:52 | |
openstackgerrit | zhubingbing proposed openstack/kolla: Add Barbican container dockerfile https://review.openstack.org/351822 | 06:54 |
*** Mr_Broken has joined #openstack-kolla | 07:12 | |
*** duonghq has joined #openstack-kolla | 07:20 | |
*** rkrum has quit IRC | 07:23 | |
*** Serlex has joined #openstack-kolla | 07:35 | |
*** ad_rien_ has joined #openstack-kolla | 07:37 | |
*** Mr_Broken has quit IRC | 07:43 | |
*** daneyon has joined #openstack-kolla | 07:45 | |
*** ad_rien_ has quit IRC | 07:47 | |
*** diga has joined #openstack-kolla | 07:50 | |
*** daneyon has quit IRC | 07:50 | |
*** bootsha has joined #openstack-kolla | 07:51 | |
*** matrohon has joined #openstack-kolla | 07:58 | |
*** zhubingbing__ has joined #openstack-kolla | 08:00 | |
*** zhubingbing_ has quit IRC | 08:03 | |
*** zhubingbing__ has quit IRC | 08:06 | |
*** ad_rien_ has joined #openstack-kolla | 08:13 | |
*** athomas has joined #openstack-kolla | 08:13 | |
*** ad_rien_1 has joined #openstack-kolla | 08:14 | |
*** ad_rien_ has quit IRC | 08:18 | |
*** matrohon has quit IRC | 08:19 | |
*** sean-k-mooney has quit IRC | 08:24 | |
*** b_bezak has joined #openstack-kolla | 08:27 | |
*** athomas has quit IRC | 08:34 | |
*** athomas has joined #openstack-kolla | 08:35 | |
*** matrohon has joined #openstack-kolla | 08:36 | |
*** matrohon has quit IRC | 08:36 | |
-openstackstatus- NOTICE: Gerrit is going to be restarted | 08:39 | |
*** gfidente has joined #openstack-kolla | 08:39 | |
openstackgerrit | Duong Ha-Quang proposed openstack/kolla: Port rabbitmq bootstrap to Ansbile 2.0 https://review.openstack.org/352275 | 08:43 |
*** fragatina has joined #openstack-kolla | 08:50 | |
*** bmace has joined #openstack-kolla | 08:51 | |
duonghq | coolsvap: are you deploy Kolla in VM or directly on machine for dev Kolla? | 08:51 |
coolsvap | mostly directly on a test machine | 08:52 |
duonghq | has you tried every base distro config? | 08:52 |
*** matrohon has joined #openstack-kolla | 08:52 | |
coolsvap | duonghq: anything particular blocking you? | 08:54 |
*** fragatina has quit IRC | 08:55 | |
duonghq | coolsvap: I got strange behavior, I described in #1609502 | 08:57 |
duonghq | https://bugs.launchpad.net/kolla/+bug/1609502 | 08:57 |
openstack | Launchpad bug 1609502 in kolla "Container bootstrap_mariadb exits due to Galera failed" [Undecided,New] | 08:57 |
duonghq | It happen in centos base image | 08:57 |
duonghq | but when I switch to ubuntu it disappear for 1st run, and appear again in 2nd,... | 08:58 |
duonghq | but when I re-execute kolla-ansible deploy is disappear again | 08:58 |
coolsvap | duonghq: alright I will confirm and get back to you | 08:59 |
duonghq | thank you | 08:59 |
*** dmueller-evoila_ has joined #openstack-kolla | 09:03 | |
*** dmueller-evoila has quit IRC | 09:07 | |
*** matrohon has quit IRC | 09:08 | |
*** b_bezak_ has joined #openstack-kolla | 09:19 | |
*** b_bezak has quit IRC | 09:20 | |
*** matrohon has joined #openstack-kolla | 09:24 | |
*** Guest52040 has joined #openstack-kolla | 09:28 | |
openstackgerrit | Paul Bourke proposed openstack/kolla: Customizations for Cinder https://review.openstack.org/350587 | 09:30 |
*** diga has quit IRC | 09:32 | |
*** daneyon has joined #openstack-kolla | 09:34 | |
*** rmart04 has joined #openstack-kolla | 09:37 | |
*** daneyon has quit IRC | 09:38 | |
*** athomas has quit IRC | 09:45 | |
openstackgerrit | Paul Bourke proposed openstack/kolla: Document the Dockerfile customisation mechanism https://review.openstack.org/351123 | 09:46 |
*** rkrum has joined #openstack-kolla | 09:46 | |
openstackgerrit | Paul Bourke proposed openstack/kolla: Customizations for Cinder https://review.openstack.org/350587 | 09:49 |
*** fragatina has joined #openstack-kolla | 09:51 | |
*** athomas has joined #openstack-kolla | 09:53 | |
*** fragatina has quit IRC | 09:56 | |
*** matrohon has quit IRC | 10:01 | |
*** zhurong has quit IRC | 10:03 | |
*** bootsha has quit IRC | 10:04 | |
openstackgerrit | Paul Bourke proposed openstack/kolla: Customizations for Nova https://review.openstack.org/351676 | 10:10 |
openstackgerrit | Paul Bourke proposed openstack/kolla: Document the Dockerfile customisation mechanism https://review.openstack.org/351123 | 10:13 |
*** mnasiadka has joined #openstack-kolla | 10:14 | |
openstackgerrit | Ross Krumbeck proposed openstack/kolla: Fix designate bootstrap. https://review.openstack.org/352309 | 10:15 |
*** matrohon has joined #openstack-kolla | 10:17 | |
openstackgerrit | Ross Krumbeck proposed openstack/kolla: Fix designate bootstrap. https://review.openstack.org/352206 | 10:18 |
*** duonghq has left #openstack-kolla | 10:24 | |
*** hieulq_ has quit IRC | 10:27 | |
*** rkrum has quit IRC | 10:38 | |
pbourke | hey coolsvap | 10:56 |
*** matrohon has quit IRC | 10:57 | |
pbourke | coolsvap: I think we're good to go for the rally/tempest testing in osic | 10:57 |
pbourke | coolsvap: can you let us know whats your status on that? | 10:57 |
*** zhubingbing__ has joined #openstack-kolla | 10:57 | |
*** senk__ has joined #openstack-kolla | 10:59 | |
*** senk_ has quit IRC | 11:01 | |
coolsvap | pbourke: I have not checked the status over the weekend, I can start the tests tonight | 11:02 |
pbourke | coolsvap: cool let us know if you need help, im sure there'll be others on in a few hours also | 11:02 |
pbourke | coolsvap: are you ok with the vpn/tmux etc? | 11:03 |
coolsvap | pbourke: no, I need to configure that as well, was working on my system + a test system from friday | 11:04 |
coolsvap | just getting relaxed after I got everything up and runnig as expected | 11:04 |
coolsvap | pbourke: do you have steps for the vpn/tmux config? | 11:05 |
pbourke | coolsvap: Most of what you need is in the email we got | 11:06 |
coolsvap | pbourke: alright I will check, will ask you if I need any help regarding it | 11:07 |
pbourke | coolsvap: I run the vpn client in docker if you prefer to the browser plugin thing in the email | 11:07 |
pbourke | https://github.com/brk3/docker-f5fpc | 11:07 |
pbourke | coolsvap: once on the vpn, the deploy node is root@172.21.8.253 password 'cobbler' | 11:07 |
pbourke | coolsvap: when you get there, run 'tmux att' | 11:08 |
pbourke | coolsvap: https://etherpad.openstack.org/p/kolla-N-midcycle-osic for notes | 11:08 |
coolsvap | yeah I was reading the notes in etherpad | 11:08 |
*** sdake has joined #openstack-kolla | 11:08 | |
*** sdake has quit IRC | 11:08 | |
*** sdake has joined #openstack-kolla | 11:08 | |
*** godleon has quit IRC | 11:10 | |
coolsvap | sdake: hey | 11:11 |
*** coolsvap is now known as coolsvap_ | 11:12 | |
*** matrohon has joined #openstack-kolla | 11:13 | |
*** zhubingbing_ has joined #openstack-kolla | 11:13 | |
*** zhubingbing__ has quit IRC | 11:17 | |
zhubingbing_ | 1 | 11:17 |
zhubingbing_ | ping | 11:17 |
*** daneyon has joined #openstack-kolla | 11:22 | |
*** daneyon has quit IRC | 11:26 | |
*** sdake has quit IRC | 11:36 | |
*** dwalsh has joined #openstack-kolla | 11:36 | |
*** bootsha has joined #openstack-kolla | 11:39 | |
*** sdake has joined #openstack-kolla | 11:39 | |
*** Guest52040 has quit IRC | 11:40 | |
*** sdake has quit IRC | 11:41 | |
*** sdake has joined #openstack-kolla | 11:43 | |
sdake | zhubingbing_ what can i do for you | 11:44 |
sdake | hey coolsvap_ | 11:44 |
zhubingbing_ | help me review mycode | 11:44 |
zhubingbing_ | --“) | 11:44 |
sdake | hey coolsvap_ how much competition do you have for requirements repo | 11:44 |
sdake | zhubingbing_ just woke up - i dont review until atleast 1 hour after i'm up | 11:45 |
zhubingbing_ | ok | 11:45 |
*** matrohon has quit IRC | 11:45 | |
sdake | pbourke around? | 11:55 |
pbourke | sdake: hey | 11:56 |
sdake | pbourke did you see my review for documentation of the osic cluster | 11:56 |
pbourke | sdake: I did, looks good | 11:56 |
sdake | i saw you volunteered to be emea coordinator | 11:56 |
pbourke | sdake: yeah, good with you? | 11:56 |
sdake | it looks good but is incompoelte - i'd like to keep it up to date as we proceed | 11:56 |
sdake | yup- just was going to ask you to help enforce the documentation along the way concept :) | 11:57 |
pbourke | sdake: we can do our best, an alternate is to just have people update the etherpad then we'll compose into the review towards the end | 11:57 |
pbourke | sdake: otherwise there will be huge churn on the review | 11:57 |
pbourke | what you reckon? | 11:57 |
sdake | that will be alot of pain (etherpad) - data will get lost - i dont mind churn in he review | 11:57 |
pbourke | ok no prob | 11:57 |
sdake | i would like coauthor list to be long on that review :) | 11:57 |
sdake | etherpad is a fantastic tool | 11:58 |
sdake | but I feel in this case critical data will become lost and there will be no way to point it out | 11:59 |
pbourke | sdake: ive updated the pad with big text pointing to the review | 11:59 |
openstackgerrit | Christian Berendt proposed openstack/kolla: Add script to validate syntax of Jinja2 template files https://review.openstack.org/352378 | 12:00 |
sdake | wonder how good thi ssafe is nd reproduce the problem with strace without a direct segfault of ovs-ddb | 12:01 |
sdake | https://www.amazon.com/dp/B00ZYHR412?psc=1 | 12:02 |
openstackgerrit | Paul Bourke proposed openstack/kolla: Move haproxy to control nodes in sample inventories https://review.openstack.org/352379 | 12:02 |
*** matrohon has joined #openstack-kolla | 12:02 | |
*** senk__ is now known as senk | 12:04 | |
*** senk has quit IRC | 12:05 | |
*** mliima has joined #openstack-kolla | 12:09 | |
*** jtriley has joined #openstack-kolla | 12:14 | |
*** rhallisey has joined #openstack-kolla | 12:15 | |
rhallisey | sbezverk, sdake, so mounting /dev fixed that problem | 12:21 |
rhallisey | that's awesome | 12:21 |
sdake | rhallisey i think so - but sbezverk never confirmed before i ptfo'ed last night | 12:21 |
rhallisey | ok | 12:21 |
rhallisey | what is ovs doing with /dev? | 12:22 |
*** matrohon has quit IRC | 12:23 | |
*** rkrum has joined #openstack-kolla | 12:31 | |
rkrum | I've written the ansible role for Designate, currently testing but should be complete in the next few days. Can this be assigned to me https://blueprints.launchpad.net/kolla/+spec/ansible-designate. | 12:36 |
*** matrohon has joined #openstack-kolla | 12:39 | |
*** bootsha has quit IRC | 12:46 | |
*** dave-mccowan has joined #openstack-kolla | 12:52 | |
*** fragatina has joined #openstack-kolla | 12:52 | |
rhallisey | rkrum, assigned it | 12:53 |
rkrum | thanks | 12:53 |
*** fragatina has quit IRC | 12:58 | |
*** matrohon has quit IRC | 13:01 | |
sbezverk | sdake ping | 13:06 |
sdake | sbezverk shoot | 13:06 |
sbezverk | what happened with -2??? | 13:06 |
sdake | i was blocking temporarily until we gotconfirmation that /dev:/dev was the correct solution - rather then sleep 1 | 13:07 |
sbezverk | sdake: ok you have confirmation, but I want that people understand dev mount fixed only sleep 1, we still need this script for Daemonset | 13:08 |
sdake | cool | 13:09 |
sdake | busy - atm - give me 5 minutes | 13:09 |
sbezverk | sdake I am going to push final version for review.. | 13:10 |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla: Introduce a script to launch ovsdb-server process https://review.openstack.org/351784 | 13:14 |
*** fragatina has joined #openstack-kolla | 13:15 | |
*** matrohon has joined #openstack-kolla | 13:17 | |
*** fragatina has quit IRC | 13:20 | |
openstackgerrit | Christian Berendt proposed openstack/kolla: Add script to validate syntax of Jinja2 template files https://review.openstack.org/352378 | 13:21 |
openstackgerrit | Christian Berendt proposed openstack/kolla: Fix service_type of mistral endpoint https://review.openstack.org/352120 | 13:24 |
sdake | sbezverk back | 13:27 |
*** zhurong has joined #openstack-kolla | 13:27 | |
sdake | sbezverk i think rhallisey had some concern that the abi shouldbe used | 13:27 |
sdake | but the issue there is the abi doesn't currently support multiple process runs | 13:27 |
rhallisey | my point was that the command is customizable | 13:28 |
rhallisey | where as the config stuff was not | 13:28 |
rhallisey | that's why I'm a little hesitant | 13:29 |
rhallisey | you can change to command to execute a file | 13:29 |
*** zhurong has quit IRC | 13:31 | |
sbezverk | sdake: well in this case we need to decide, if we want to use kube features to the maximum, then we need to change a bit the strategy. but if we want to use old aproach of controlling each step in ansible, that is fine too, but we will get get kube advantages.. | 13:31 |
sdake | ok rhallisey well can you sort out a proposal for the nedium term | 13:31 |
sdake | and we can just merge this in the short term since it seems to fix the problem | 13:31 |
sbezverk | sdake: s/get/not/ | 13:32 |
sdake | what i'm missing in this patch is where is the dev bindmount? | 13:32 |
sbezverk | sdake: dev mount in daemonset definition | 13:32 |
*** zhurong has joined #openstack-kolla | 13:32 | |
sbezverk | I am about to push it to kolla-kubernetes | 13:33 |
*** banix has joined #openstack-kolla | 13:33 | |
*** ayoung has joined #openstack-kolla | 13:34 | |
*** matrohon has quit IRC | 13:35 | |
*** ayoung has quit IRC | 13:37 | |
*** ayoung has joined #openstack-kolla | 13:43 | |
*** rkrum has quit IRC | 13:46 | |
*** matrohon has joined #openstack-kolla | 13:51 | |
rhallisey | sbezverk, can you add todo comment in that patch | 13:52 |
rhallisey | just so we track this temp fix | 13:52 |
openstackgerrit | Christian Berendt proposed openstack/kolla: Set up libvirt to use huge pages https://review.openstack.org/352429 | 13:53 |
*** mntdevops has joined #openstack-kolla | 13:54 | |
sbezverk | rhallisey: sure thing, as soon as I push out my DaemonSet :-) | 13:54 |
rhallisey | kk thanks | 13:54 |
mntdevops | hello there, Kolla folks. I've been trying for a few days now to install an all-in-one from the current master branch on bare metal | 13:55 |
mntdevops | and I think I'm running into several problems stemming from my lack of a working globals.yml | 13:55 |
mntdevops | does anyone have a currently working globals.yml, say from the devstack instance? | 13:56 |
*** huikang has joined #openstack-kolla | 13:56 | |
mntdevops | I notice the bootstrap code for the vagrant file does a few seds to replace some thigns | 13:56 |
mntdevops | so is that the recommended method? | 13:57 |
mntdevops | take the current globals.yml from master | 13:57 |
pbourke | mntdevops: the one that ships in the repo works, but you will need to tweak some things for your environment as outlined in the docs | 13:57 |
mntdevops | and apply those seds by hand? | 13:57 |
mntdevops | I think that's where I'm getting into trouble | 13:57 |
mntdevops | some of the plays fail with empty items (options being passed to kolla_docker.py) | 13:58 |
mntdevops | and that stems from misconfiguration | 13:58 |
mntdevops | I guess what I'm saying is | 13:58 |
mntdevops | there are a lot of options that all have to be right :-( | 13:58 |
*** coolsvap_ is now known as coolsvap | 13:58 | |
pbourke | if you like you could pastebin your globals and some of the errors you're seeing. then we might be able to see where the problems are | 13:58 |
coolsvap | sdake: i have tough competition | 13:58 |
mntdevops | and I was hoping you all had one that worked | 13:58 |
mntdevops | so I could compare mine to yours that way | 13:58 |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Introducing DaemonSet for Neutron Openvswitch components https://review.openstack.org/352433 | 13:58 |
coolsvap | think they sent the poll to participating projects PTLs | 13:59 |
mntdevops | thanks pbourke | 13:59 |
mntdevops | tell you what | 13:59 |
mntdevops | I'll share my github | 13:59 |
mntdevops | repo | 13:59 |
mntdevops | just give me a min | 14:00 |
mntdevops | very much appreciate the extra eyes | 14:00 |
*** Liuqing has joined #openstack-kolla | 14:00 | |
mntdevops | https://github.com/mntdevops/netpxemirror-ac/blob/master/ansible/roles/os-aio/files/globals.yml\ | 14:01 |
mntdevops | https://github.com/mntdevops/netpxemirror-ac/blob/master/ansible/roles/os-aio/files/globals.yml | 14:01 |
*** huikang has quit IRC | 14:02 | |
mntdevops | my key issue at the moment surrounds cinder and nova | 14:02 |
mntdevops | I can start instances provided no volume needs to be created | 14:02 |
mntdevops | I've created the cinder-volumes vg | 14:02 |
mntdevops | and I just this morning enabled LVM2 backed cinder | 14:03 |
mntdevops | which I hope resolves things | 14:03 |
pbourke | mntdevops: so if you're getting errors during deploy we should look into those first | 14:03 |
mntdevops | it's kolla-ansible building now | 14:03 |
pbourke | mntdevops: otherwise all bets are off for runtime | 14:03 |
pbourke | mntdevops: do you have output for the parts of the deploy that are failing? | 14:03 |
mntdevops | the deploy works cleanly atm | 14:04 |
mntdevops | it's when I go to start a new instance with a volume | 14:04 |
pbourke | ok | 14:04 |
mntdevops | I'm in the process of rebuilding from scratch | 14:04 |
mntdevops | with the iscsi and tgt stuff now in there | 14:04 |
mntdevops | enable_cinder_backend_lvm: "yes" cinder_backend_iscsi: "{{ enable_iscsi }}" cinder_backend_ceph: "{{ enable_ceph }}" # Cinder's iSCSI backend !!!REQUIRES!!! two parameters: # 1 - IP address of the server hosting LVM Volume group # 2 - The name of Volume group which Cinder will use. cinder_volume_group: cinder-volumes | 14:05 |
mntdevops | ah | 14:05 |
mntdevops | enable_cinder_backend_lvm: "yes" | 14:05 |
mntdevops | cinder_backend_iscsi: "{{ enable_iscsi }}" | 14:05 |
mntdevops | (which is yes) | 14:05 |
mntdevops | cinder_volume_group: cinder-volumes | 14:05 |
mntdevops | I think this is the area of issue | 14:05 |
mntdevops | these specific settings | 14:06 |
*** matrohon has quit IRC | 14:06 | |
mntdevops | I'm not sure what the setting has to be for "# 1 - IP address of the server hosting LVM Volume group" | 14:06 |
mntdevops | #2 should be correct | 14:06 |
mntdevops | again, thank you for looking into this | 14:07 |
mntdevops | I'll post the logs/errors when I get them | 14:07 |
mntdevops | another question: If I have a registry on my network, how/when do I need to re-run kolla-ansible build --push? | 14:08 |
mntdevops | in other words | 14:08 |
mntdevops | is absolutely zero configuration baked into the containers? | 14:09 |
mntdevops | such that I can build them once and just reconfigure as needed? | 14:09 |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla: Introduce a script to launch ovsdb-server process https://review.openstack.org/351784 | 14:09 |
sbezverk | rhallisey: done with note about tmp workaound.. | 14:09 |
mntdevops | or do some configuration changes require certain containers to have to be rebuild | 14:09 |
mntdevops | t | 14:09 |
mntdevops | I think what I'm struggling with is "if I make a change to globals.yml, what build/pull/deploy/reconfigure related things must be done to make that take effect?" | 14:11 |
rhallisey | sbezverk, k wfm | 14:12 |
*** fragatina has joined #openstack-kolla | 14:17 | |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Introducing DaemonSet for Neutron Openvswitch components https://review.openstack.org/352433 | 14:18 |
coolsvap | pbourke: which browser did you use for the sslvpn plugin installation? | 14:19 |
coolsvap | I am not able te do it with either chrome or firefox | 14:19 |
pbourke | coolsvap: I couldn't get the plugin to work with any browser | 14:19 |
pbourke | coolsvap: that's why I made the dockerfile | 14:19 |
sdake | mntdevops - zero configuration is built into the containers - one push is all you need unless there is a security update on a maintenance branch - then you would want to push again | 14:21 |
*** fragatina has quit IRC | 14:21 | |
sdake | coolsvap i used safari | 14:21 |
coolsvap | ah. let me check that | 14:22 |
sdake | the sslvpn plugin does not work with chrome | 14:22 |
sdake | chrome took out the nssapiv2 iirc | 14:22 |
sdake | firefox may have done the same hting | 14:22 |
*** matrohon has joined #openstack-kolla | 14:22 | |
coolsvap | sdake: hmm chrome blocked it so did firefox | 14:22 |
sdake | i dont know if blocked is the right word - thiink defeatured is more accurate | 14:23 |
sdake | not sure aobut firefox - i didnt tr ythat | 14:23 |
*** matrohon has quit IRC | 14:24 | |
*** mnasiadka has quit IRC | 14:25 | |
wirehead_ | sdake: "the afternoons are typically downtime for our community" Yeah. I've had to completely alter my routine to keep up with the community. :P | 14:25 |
coolsvap | hmm blocked may be a bit harsh, chrome discontinued support for NPAPI plugins :) | 14:26 |
pbourke | coolsvap: try the dockerfile you'll be connected in no time | 14:26 |
coolsvap | pbourke: building it | 14:26 |
mntdevops | ok, thanks sdake. That makes sense | 14:30 |
sdake | hey core reviewers | 14:31 |
sdake | one of the glare cats is coming by at 16:00 utc to chat | 14:31 |
sdake | we need to understand how this impacts openstack - because rom my parsing of jay's comments, itseems like glare will eventually replace glance entirely | 14:32 |
*** ad_rien_1 has quit IRC | 14:32 | |
sdake | so lets be nice and just ask the technical questions that need to be sked | 14:32 |
sdake | main concern being how do we upgrade | 14:33 |
openstackgerrit | Merged openstack/kolla-kubernetes: Neutron-network services split out into discrete pods https://review.openstack.org/350260 | 14:33 |
wirehead_ | So, we can't just sit around and glare at them? | 14:33 |
sdake | you can do whaever you like as long sa it is polite :) | 14:37 |
*** mewald1 has joined #openstack-kolla | 14:39 | |
*** zhubingbing_ has quit IRC | 14:39 | |
*** matrohon has joined #openstack-kolla | 14:40 | |
*** mewald has quit IRC | 14:41 | |
*** fragatina has joined #openstack-kolla | 14:43 | |
*** b_bezak_ has quit IRC | 14:45 | |
mntdevops | well I'll be... the iscsi/lvm2 bits were the trick | 14:45 |
mntdevops | enable_cinder_backend_lvm: "yes" was the key | 14:46 |
sbezverk | wirehead_ could you please check this PS: https://review.openstack.org/352433 | 14:46 |
mntdevops | I guess it just took the extra set of eyes! Thanks again | 14:47 |
wirehead_ | yeah. | 14:47 |
*** zhubingbing has joined #openstack-kolla | 14:50 | |
*** mewald has joined #openstack-kolla | 14:50 | |
zhubingbing | hi | 14:50 |
zhubingbing | sdake, | 14:50 |
*** huikang has joined #openstack-kolla | 14:51 | |
*** mewald1 has quit IRC | 14:51 | |
*** huikang has quit IRC | 14:55 | |
zhubingbing | 1 | 14:56 |
*** mliima has quit IRC | 14:58 | |
zhubingbing | - - | 15:02 |
sdake | pbourke re coordination, any updates on where we are currently? | 15:02 |
sdake | (related to osic) | 15:02 |
*** egonzalez90 has joined #openstack-kolla | 15:03 | |
pbourke | sdake: going by the etherpad (https://etherpad.openstack.org/p/kolla-N-midcycle-osic) we're ready for stage 1.2 which is running tempest/rally | 15:03 |
*** fragatina has quit IRC | 15:03 | |
*** duonghq has joined #openstack-kolla | 15:04 | |
pbourke | coolsvap is looking into it and has informed me he'll shout when he needs us :) | 15:04 |
sdake | nice | 15:04 |
coolsvap | sdake: i am configuring rally | 15:04 |
sdake | i dont know that we specifically need rally or tempest reports | 15:04 |
*** fragatina has joined #openstack-kolla | 15:04 | |
sdake | but i've never run rally so i don't nkow what type of reportit produces | 15:04 |
sdake | perhaps the master doc should include those reports | 15:04 |
sdake | and then we can break them out later as separate docs | 15:04 |
coolsvap | sdake: rally has nice reports | 15:05 |
sdake | but lookingto folks doingthe work to comment on whether we should or shouldn't include reports for each scneario | 15:05 |
pbourke | coolsvap: if/when you're building or deploying stuff, can you time the commands | 15:06 |
pbourke | we can do more specific benchmarking also but nice to have an overall set of times for general operations | 15:06 |
coolsvap | pbourke: sure | 15:07 |
openstackgerrit | Li Yingjun proposed openstack/kolla: Add option to support dhcp/l3 agent ha https://review.openstack.org/352180 | 15:07 |
*** srwilkers has joined #openstack-kolla | 15:09 | |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla: Customizations for Ceph https://review.openstack.org/351823 | 15:09 |
*** matrohon has quit IRC | 15:10 | |
*** srwilkers_ has joined #openstack-kolla | 15:11 | |
*** srwilkers has quit IRC | 15:13 | |
*** srwilkers_ is now known as srwilkers | 15:13 | |
*** srwilkers has quit IRC | 15:15 | |
*** egonzalez90 has quit IRC | 15:16 | |
*** srwilkers has joined #openstack-kolla | 15:16 | |
*** zhurong has quit IRC | 15:18 | |
*** mliima has joined #openstack-kolla | 15:19 | |
*** matrohon has joined #openstack-kolla | 15:27 | |
*** rmart04 has quit IRC | 15:31 | |
*** matrohon has quit IRC | 15:32 | |
duonghq | rhallisey: according to your sugestion, I will head to bp swift-kubernetes first, so who is working on this bp? | 15:34 |
rhallisey | duonghq, nobody is. I can assign it to you | 15:34 |
duonghq | okay, thank you | 15:35 |
wirehead_ | Yay! :) | 15:36 |
rhallisey | wirehead_, hey | 15:37 |
wirehead_ | duonghq: There's a pod that was created for Swift a long time ago and hasn't been touched since. | 15:37 |
duonghq | wirehead_: interesting | 15:38 |
sdake | say wirehead_ - any idea on the state of nova booting in kolla-kubernetes | 15:38 |
wirehead_ | sdake: Very very very close. :) | 15:39 |
*** dmueller-evoila_ has quit IRC | 15:39 | |
wirehead_ | sdake: As in, the processes are up, and I was starting to actually poke at it. | 15:39 |
rhallisey | nice | 15:39 |
sdake | need any guidnace with compute? | 15:39 |
sdake | I did that work initially and have extensive understanding of all the things that need to go toghether to make computehappen | 15:39 |
wirehead_ | sdake: hm. Not yet. Will probably need some help shortly. | 15:41 |
sdake | ping if needed | 15:41 |
*** ppowell has joined #openstack-kolla | 15:42 | |
*** harlowja_at_home has joined #openstack-kolla | 15:42 | |
*** Liuqing has quit IRC | 15:43 | |
duonghq | I think Hiroki Ito has nice work at https://review.openstack.org/#/c/351588/ but he just got one +2, can somebody verify his work? | 15:44 |
patchbot | duonghq: patch 351588 - kolla - Fix prechecks of a vip address | 15:44 |
*** srwilkers has quit IRC | 15:44 | |
*** Serlex has quit IRC | 15:46 | |
*** fragatina has quit IRC | 15:49 | |
*** matrohon has joined #openstack-kolla | 15:49 | |
*** srwilkers has joined #openstack-kolla | 15:51 | |
*** daneyon has joined #openstack-kolla | 15:52 | |
*** mewald has quit IRC | 15:53 | |
duonghq | coolsvap: can you take a look at Hiroki Ito's work | 15:54 |
*** jrich523 has joined #openstack-kolla | 15:55 | |
coolsvap | duonghq: done | 15:56 |
duonghq | coolsvap: I saw, thanks | 15:57 |
*** michauds has joined #openstack-kolla | 15:57 | |
*** daneyon has quit IRC | 15:57 | |
*** inc0 has joined #openstack-kolla | 15:58 | |
inc0 | hey | 15:59 |
inc0 | any news my dear kollagues? | 15:59 |
inc0 | any movement on osic cluster? pbourke sdake | 15:59 |
*** matrohon has quit IRC | 15:59 | |
*** haplo37__ has joined #openstack-kolla | 16:00 | |
*** mfedosin has joined #openstack-kolla | 16:00 | |
sdake | inc0 yes | 16:00 |
*** srwilkers_ has joined #openstack-kolla | 16:00 | |
pbourke | inc0: inc0 http://eavesdrop.openstack.org/irclogs/%23openstack-kolla/latest.log.html#t2016-08-08T15:03:31 | 16:01 |
sdake | we have yet to have a US coordinator though | 16:01 |
inc0 | count me in | 16:01 |
sdake | not sure who that would be | 16:01 |
sdake | I am too busy atm to take on that role | 16:01 |
mfedosin | sdake: hey :) you may ask me any questions about Glare | 16:01 |
sdake | fortunately our emea coordinator (pbourke) is here | 16:01 |
sdake | mfedosin cool | 16:02 |
sdake | mfedosin did I understand the thread correctly that glare will eventually take on image storage? | 16:02 |
jrich523 | hey guys, trying to do the kolla-build and im getting some failures, but, its failing and im not really getting any details other than which packages failed.. | 16:02 |
sdake | jrich523 we are having an impromptu community meeting related to an important dependency in openstack | 16:02 |
jrich523 | however it doesnt say why, or how they are being installed (which packages) | 16:02 |
jrich523 | ahh i'll wait :) | 16:03 |
*** srwilkers has quit IRC | 16:03 | |
*** srwilkers_ is now known as srwilkers | 16:03 | |
sdake | jrich523 can you wait a bit - and we will be with you in about 30 mins? | 16:03 |
jrich523 | sounds good! | 16:03 |
mfedosin | sdake: correct :) It is my greatest wish | 16:03 |
sdake | mfedosin cool - is an upgrade approach being defined | 16:03 |
sdake | mfedosin the reason I ask is nova and cinder both have provided upgrade approaches for their software | 16:03 |
inc0 | sdake, I can take on it | 16:03 |
sdake | mfedosin and other big tent projects as well | 16:03 |
inc0 | I won't be online 100% of time but thats one of my priorities for this week | 16:04 |
mfedosin | sdake: at least we support microversions | 16:04 |
mfedosin | it's for API | 16:04 |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Introducing DaemonSet for Neutron Openvswitch components https://review.openstack.org/352433 | 16:04 |
mfedosin | for artifact type we follow standard oslo_versionedobjects techniques | 16:05 |
sdake | mfedosin more concerned about upgrade approach from glance to glare for the image specific portions | 16:05 |
sdake | the rest we can handle since there is no history of glare implementation in ODMs | 16:05 |
mfedosin | ah, that one... | 16:05 |
mfedosin | it's under discussions, but Glare is able to work with Glance images in parallel | 16:06 |
mfedosin | so, you can create an image with Glare and use it in Glance, and vice versa | 16:06 |
sdake | mfedosin i could probably read the code and find out - but mind expanding in more detail? | 16:06 |
sdake | ok so a occata deployment would include both glance and glare? | 16:07 |
sdake | and have endponits for both? | 16:07 |
sdake | and glare would sort out working with glance in that model? | 16:07 |
Mech422 | Morning | 16:08 |
mfedosin | okay, so, first of all, Glare has pluggable db api , which means that images artifact type is able to work with Glance tables | 16:08 |
sdake | ah | 16:09 |
sdake | cool | 16:09 |
*** srwilkers has quit IRC | 16:09 | |
mfedosin | we wrote an adapter that transforms one format to ours | 16:09 |
sdake | so in occata in a pefect world, glance would not be needed, just the database contents for an upgrade? | 16:09 |
mfedosin | I hope so | 16:10 |
sdake | cool different topic | 16:10 |
mfedosin | then we want to add Glare support to Nova | 16:11 |
sdake | glance has been pretty negative about supporting file backed storage with multi-controller model (for ha) | 16:11 |
mfedosin | there are no big differences in APIs | 16:11 |
sdake | any plans to solve that problem? | 16:11 |
sdake | we need it to make cinder work properly | 16:11 |
sdake | with multiple controllers with iscsi | 16:12 |
sdake | cinder integrates many hardware based storage solutions via iscsi | 16:12 |
mfedosin | initially we will use glance_store library to communicate with storage backends | 16:12 |
sdake | the last statement on the review was that glance should create a whole slew of plugins for harwdware vendor gear | 16:13 |
mfedosin | https://github.com/openstack/glance_store/tree/master/glance_store | 16:13 |
sdake | mfedosin not an expert in all this stuff - thanks for the link :) | 16:13 |
sdake | mind giving me a TLDR :) | 16:13 |
sdake | i want to keep it at 30 minutes - rather then reading that doc now | 16:13 |
mfedosin | at least there are cinder driver to work with cinder https://github.com/openstack/glance_store/blob/master/glance_store/_drivers/cinder.py | 16:14 |
sdake | yes - that is not the issue | 16:14 |
sdake | moment let me find a link | 16:14 |
sdake | https://bugs.launchpad.net/cinder/+bug/1571211 | 16:15 |
openstack | Launchpad bug 1571211 in Cinder "In multinode setup VM fails to launch due to cinder not checking with all glance api servers" [Medium,In progress] - Assigned to John Griffith (john-griffith) | 16:15 |
sdake | anyone else from kolla community have any questions for Mike? | 16:16 |
coolsvap | inc0: i have configured rally | 16:16 |
openstackgerrit | Merged openstack/kolla: Fix prechecks of a vip address https://review.openstack.org/351588 | 16:16 |
inc0 | coolsvap, great, I need to work out VPN connection here | 16:16 |
inc0 | and will join you guys later | 16:16 |
*** huikang has joined #openstack-kolla | 16:17 | |
coolsvap | i am taking a look at benchmarking tests to run | 16:17 |
mfedosin | sdake: ha, it's a know bug - they got the same in Nova | 16:18 |
mfedosin | known | 16:18 |
sdake | we worked aorudn it in nova by using the nova api list | 16:18 |
sdake | cinder has a same api lsit feature | 16:18 |
sdake | but cinder reviewers dont want to ping all the api servers | 16:18 |
mfedosin | sdague fixed it there | 16:18 |
sdake | kolla core cats | 16:20 |
sdake | glare is meant to be an artifact storage system - and one of the artifacts it will be storing is images | 16:21 |
sdake | any quesetions related to upgrade or reconfigure? | 16:21 |
sdake | or anything else? | 16:21 |
sdake | (thanks Mike btw, this is greatly appreciated so we can plan properly) | 16:21 |
mfedosin | You're welcome :) | 16:22 |
*** srwilkers has joined #openstack-kolla | 16:22 | |
mfedosin | if there are any other questions feel free to ping me in IRC or by email: mfedosin@gmail.com | 16:22 |
sdake | mfedosin irc is better - recorded that way | 16:23 |
sdake | glare channel is #openstack-glare? | 16:23 |
mfedosin | it's not created yet | 16:24 |
mfedosin | but will be in a couple of days | 16:24 |
sdake | whats its planned name? | 16:24 |
mfedosin | #openstack-glare for sure | 16:24 |
sdake | cool | 16:28 |
sdake | i'll post a folllowup link to this irc conversation so the other ODM folks can learn from it | 16:28 |
sdake | thanks for your time :) | 16:28 |
sdake | i guess the rest of the kolla community found the answers useful or they would have asked more :) | 16:28 |
*** harlowja_at_home has quit IRC | 16:29 | |
*** david-lyle_ has joined #openstack-kolla | 16:29 | |
inc0 | pbourke, coolsvap sdake anyone downloaded f5fpc by any chance? | 16:31 |
sdake | inc0 no idea dude, i just downloaded a plugin into safari | 16:31 |
sdake | some f5 plugin | 16:31 |
pbourke | inc0: https://github.com/brk3/docker-f5fpc | 16:31 |
inc0 | pbourke, woot thanks | 16:31 |
*** jem_ has joined #openstack-kolla | 16:31 | |
*** jem_ is now known as Guest99846 | 16:32 | |
inc0 | one of us made it? | 16:32 |
*** david-lyle has quit IRC | 16:33 | |
coolsvap | i think pbourke did | 16:33 |
Guest99846 | I'll sign back in with the longer nicname I was using last weel | 16:33 |
Guest99846 | I'll sign back in with the longer nicname I was using last week | 16:33 |
*** Guest99846 has left #openstack-kolla | 16:33 | |
*** srwilkers has quit IRC | 16:33 | |
*** jemcevoy has joined #openstack-kolla | 16:34 | |
*** rmart04 has joined #openstack-kolla | 16:35 | |
*** unicell has quit IRC | 16:36 | |
sdake | jrich523 ok - i'm from the federal government and I'm here to help | 16:37 |
sdake | jrich523 could you repeat your question | 16:37 |
*** jemcevoy has quit IRC | 16:38 | |
jrich523 | im trying to run the build-kolla and its having a problem with a few of the packages | 16:39 |
jrich523 | mostly around Watcher it seems | 16:40 |
sdake | ok watcher isn't mandatory | 16:40 |
sdake | if watcher fails, ignore it | 16:40 |
sdake | anything else failing? | 16:40 |
*** srwilkers has joined #openstack-kolla | 16:40 | |
jrich523 | but senlin | 16:40 |
sdake | jrich523 are you doing an eval or planning to do development? | 16:40 |
jrich523 | kuryr | 16:40 |
sdake | senlin also not relevant yet | 16:40 |
jrich523 | just trying to get my feet under me :) | 16:40 |
sdake | kuryr BDWNF | 16:40 |
sdake | kuryr needs keystonev3 support before we support it in kolla | 16:41 |
jrich523 | yeah just those 3, watcher, senlin, kuryr | 16:41 |
sdake | good - your golden on your build then | 16:41 |
sdake | the other 3 are new services | 16:41 |
sdake | i'd highly recommend if you are doing an eval, to use a stable branch | 16:41 |
jrich523 | ok then! back to following directions :) | 16:41 |
jrich523 | oh, well im just following the step by step directions | 16:42 |
sdake | (which are not in the instructions unfortunately) | 16:42 |
jrich523 | kind of a bummer :) | 16:42 |
sdake | what you are deploying (master) is the master of all openstack projects | 16:42 |
sdake | and really meant for developers | 16:42 |
sdake | please oh please would omeone fix that in our documentation! | 16:42 |
jrich523 | well i didnt specify, so i assume master, unless someone changed the default branch | 16:42 |
coolsvap | okay so I have started tempest verification with rally | 16:42 |
sdake | the default branch is master | 16:42 |
jrich523 | yeah, thought so | 16:43 |
jrich523 | so shouldnt master be stable? | 16:43 |
coolsvap | I am still getting hold of benchmark scenarios most of them are failing | 16:43 |
coolsvap | not failing as faling on kolla but cannot be validated by rally itself :) | 16:44 |
sdake | jrich523 master deploys all openstack services from master | 16:45 |
sdake | jrich523 do you think all oepnstack services from master are stable? | 16:45 |
jrich523 | ... should be | 16:45 |
sdake | interesting perspective :) | 16:46 |
jrich523 | normally you make a branch to dev on, and then once its stable merge it... but apparently there is a different workflow here :) | 16:46 |
sdake | master of openstack is broken o na constant basis | 16:46 |
jrich523 | lol good to know | 16:46 |
jrich523 | is there a "stable" branch or something i should use? | 16:46 |
sdake | ye | 16:46 |
sdake | s | 16:46 |
sdake | git checout stable/mitaka | 16:46 |
sdake | git checkout stable/mitaka | 16:46 |
sdake | pip remove (or uniinstanll forget which) kolla | 16:46 |
sdake | then pip install -e . from the kolla directory | 16:47 |
sdake | also you will need ansible 1.9.4 | 16:47 |
*** b_bezak has joined #openstack-kolla | 16:48 | |
*** athomas has quit IRC | 16:48 | |
jrich523 | looks like current is 2.1.0.0, roll that back to 1.9.4? | 16:48 |
sdake | right | 16:48 |
sdake | pip install ansible==1.9.4 should do the trick | 16:49 |
jrich523 | yup, that did it | 16:49 |
*** fragatina has joined #openstack-kolla | 16:50 | |
*** diogogmt has joined #openstack-kolla | 16:50 | |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Introducing DaemonSet for Neutron Openvswitch components https://review.openstack.org/352433 | 16:51 |
openstackgerrit | zhubingbing proposed openstack/kolla: Add Barbican ansible role Partially-Implements: blueprint barbican-ansible https://review.openstack.org/352515 | 16:51 |
*** jemcevoy has joined #openstack-kolla | 16:51 | |
coolsvap | sdake: we are missing scenario for swift in osic | 16:51 |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla: Introduce a script to launch ovsdb-server process https://review.openstack.org/351784 | 16:51 |
*** chopmann has joined #openstack-kolla | 16:51 | |
zhubingbing | I completed the Barbican ansible role, we can go to help me review under the code | 16:53 |
zhubingbing | thinks all | 16:53 |
sdake | coolsvap you mean in the etherpad? | 16:53 |
sdake | we are missing scnearios for lots of services | 16:53 |
openstackgerrit | zhubingbing proposed openstack/kolla: Add Barbican ansible role https://review.openstack.org/352515 | 16:53 |
coolsvap | sdake: yes | 16:54 |
openstackgerrit | Paul Bourke proposed openstack/kolla: Customizations for Cinder https://review.openstack.org/350587 | 16:54 |
sdake | lets tackle what we have | 16:54 |
coolsvap | yes | 16:55 |
sdake | then we can add if we have more time | 16:55 |
sdake | i want to give sean-k-mooney a full week as promise to sort out bifrost | 16:55 |
sdake | that takes priority imo over invidiual service sccale testing | 16:55 |
sdake | but open to arguents against :) | 16:55 |
*** inc0 has quit IRC | 16:55 | |
coolsvap | agree, just something I saw failed because service not configured in tests | 16:56 |
coolsvap | so :) | 16:57 |
sdake | so rally wont work at all? | 16:57 |
jemcevoy | Morning inc0, I got further over the weekend with my install when I switched to the mitaka branch... Now I am a bit confused by the Ceph disk labeling... Do the partition names need to be unique across the cluster of for each host? I am using _1 _2 _3 on all my hosts and bootstrapping OSD's failed. I have 5 nodes in my cluster and an example of how I am labeling the disks is in this host_vars file: http://paste.openstack | 16:58 |
sdake | jrich523 just to duble check, can you past e outpu of pip show kolla please | 16:59 |
coolsvap | sdake: it works but I think it has got some issues with v3 in general | 16:59 |
coolsvap | keystone v3 | 16:59 |
sdake | v3 of which - keystone? | 16:59 |
jrich523 | sure, once the build finishes | 17:00 |
*** rmart04 has quit IRC | 17:00 | |
duonghq | coolsvap: finally I got kolla up and running today afternoon with ubuntu/source (my local time), I just run deploy twice and I ran, still cannot figure out why it didn't run. I'll investigate more on this case when I have time. It seems that strange behavior | 17:01 |
sdake | duonghq cool - nice work! | 17:02 |
coolsvap | duonghq: cool | 17:04 |
openstackgerrit | Paul Bourke proposed openstack/kolla: Customizations for Cinder https://review.openstack.org/350587 | 17:05 |
*** unicell has joined #openstack-kolla | 17:05 | |
zhubingbing | sdake, | 17:06 |
zhubingbing | Can you assign BP trove to me | 17:06 |
zhubingbing | ? | 17:06 |
sdake | zhubingbing otp atm - coolsvap can you do so or rhallisey ? | 17:06 |
sdake | zhubingbing orjust work on it and i'll ge tto it in 30 ins | 17:07 |
duonghq | I need to take a rest, bye bye | 17:09 |
*** duonghq has left #openstack-kolla | 17:09 | |
*** sdake_ has joined #openstack-kolla | 17:13 | |
*** sdake has quit IRC | 17:15 | |
*** daneyon has joined #openstack-kolla | 17:22 | |
*** daneyon_ has joined #openstack-kolla | 17:23 | |
*** daneyon has quit IRC | 17:27 | |
jrich523 | sdake_ i assume you want to see the version from that output, 2.0.3.dev10 | 17:28 |
srwilkers | good afternoon everyone — a coworker and I are following the quickstart guide for kolla-kubernetes for developing, not the full install for evaluation. when we go to bootstrap mariadb, we get the following: ‘ansible_processor_vcpus’ is undefined. is there something we’re missing, or are we just making a silly mistake? | 17:28 |
rhallisey | srwilkers, no that is a config issue | 17:29 |
rhallisey | srwilkers, https://github.com/openstack/kolla/blob/master/ansible/group_vars/all.yml#L188 | 17:29 |
srwilkers | rhallisey, thanks | 17:30 |
rhallisey | if you change that to be openstack_service_workers: "" and run the kolla-ansible genconfig it will fix the issue | 17:30 |
rhallisey | srwilkers, that will work for now. May need to change it when we add rabbitmq | 17:30 |
srwilkers | great, thanks | 17:31 |
*** timss has quit IRC | 17:33 | |
openstackgerrit | Duane DeCapite proposed openstack/kolla: Change cleanup to remove as cleanup is a misnomer https://review.openstack.org/351869 | 17:34 |
*** ppowell has quit IRC | 17:37 | |
openstackgerrit | Duane DeCapite proposed openstack/kolla: Change cleanup to destroy as cleanup is a misnomer https://review.openstack.org/351869 | 17:37 |
sdake_ | rhallisey ping - can you review the above ^ | 17:40 |
*** senk has joined #openstack-kolla | 17:42 | |
sdake_ | jrich523 looksgood | 17:42 |
senk | can someone review this? https://review.openstack.org/#/c/352030/ | 17:42 |
patchbot | senk: patch 352030 - kolla - Change Quickstart to follow code conventions | 17:42 |
*** david-lyle_ has quit IRC | 17:42 | |
sdake_ | senk i'm struggling to understand the rationale | 17:44 |
sdake_ | senk iff one were provided, I think it may help tha treview along | 17:44 |
senk | sdake_: official code conventions @ http://docs.openstack.org/contributor-guide/writing-style/code-conventions.html say it's bad practice.. i mentioned it in the commit? | 17:45 |
*** david-lyle has joined #openstack-kolla | 17:45 | |
sdake_ | it doesn't say why | 17:46 |
senk | i guess it's better to turn brain on when installing something than to simply copy paste commands and hoping it works :D | 17:46 |
senk | and of course anyone agreed to that convention and consistency is king :D | 17:47 |
senk | you get no information about the dependencies that get installed along with the package | 17:51 |
*** ppowell has joined #openstack-kolla | 17:52 | |
mntdevops | senk, may I suggest also adding items for disabling/configuring firewalld and setting SELinux to permissive/disabled? | 17:52 |
mntdevops | on centos, at least | 17:52 |
mntdevops | both of those items were a "duh" moment for me | 17:53 |
mntdevops | but just a hint about them would be beneficial, IMO | 17:54 |
mntdevops | the SELinux bit was silently preventing nova-compute from accessing /dev/kvm | 17:54 |
senk | feel free to push a review on that | 17:54 |
mntdevops | me? | 17:54 |
*** sdake has joined #openstack-kolla | 17:55 | |
senk | if you want | 17:55 |
mntdevops | I'm just a new user here who *just* got it working :) | 17:55 |
senk | docker and SELinux caused me some sleepless nights :D | 17:55 |
sdake | coolsvap results look good | 17:55 |
sdake | 10% of test cases fail it looks like | 17:55 |
sdake | but who knows if hat is tempest or kolla | 17:56 |
mntdevops | senk, not sure how to "push a review" | 17:56 |
mntdevops | sign up for launchpad? | 17:56 |
mntdevops | then do the equivalent of a pull request? | 17:56 |
coolsvap | yes we can analyse that | 17:56 |
sdake | mntdevops no its a bit more involved then that | 17:56 |
mntdevops | lol | 17:56 |
mntdevops | I just suggested a couple tweaks to the getting started guide | 17:57 |
sdake | mntdevops if your interested in becoming an openstack developer, I cna gie you a crash course in how to setup your environment | 17:57 |
mntdevops | well | 17:57 |
mntdevops | baby steps first | 17:57 |
sdake | roger | 17:57 |
mntdevops | not new to linux | 17:57 |
sdake | offer stands :) | 17:57 |
mntdevops | not new to ansible | 17:57 |
mntdevops | new to openstack | 17:57 |
mntdevops | :) | 17:57 |
*** sdake_ has quit IRC | 17:57 | |
mntdevops | tell you what | 17:57 |
mntdevops | I'll stick with it some more | 17:57 |
mntdevops | and see if I start picking up on it | 17:57 |
mntdevops | it's a fire hose now | 17:57 |
sdake | right | 17:58 |
mntdevops | all I was saying to senk was that a mention to firewalld and selinux in the getting started would be gootd | 17:58 |
mntdevops | good | 17:58 |
senk | mntdevops: http://docs.openstack.org/infra/manual/developers.html | 17:58 |
mntdevops | e.g. disabling them | 17:58 |
mntdevops | until you know what the heck you're doing, at least :) | 17:58 |
coolsvap | sdake: I will try some hands on benchmark tests before handing over for scenario 2, is it fine? | 17:58 |
sdake | all yours | 17:59 |
senk | mntdevops: you can follow that from top to bottom and test your setup at http://docs.openstack.org/infra/manual/sandbox.html#sandbox | 17:59 |
sdake | a table summary in the documentation we are producing of the pass/fail/skip or tempet would be helpful | 17:59 |
sdake | coolsvap can you pull down the review, modify it, and git review it again? | 18:00 |
coolsvap | sdake: will do | 18:00 |
mntdevops | senk, thanks | 18:00 |
mntdevops | I'll look into that | 18:00 |
sdake | that is probably the first time anyone has run tempest against kolla | 18:00 |
rhallisey | sdake, how did it go? | 18:01 |
sdake | coolsvap is master being deployed - or mitaka? | 18:01 |
sdake | coolsvap can you get a link up somewhere for people to download the html | 18:01 |
coolsvap | sdake: master | 18:01 |
rhallisey | I think pbourke has done it | 18:01 |
coolsvap | sdake: yes I will do | 18:01 |
sdake | oh well then those results aren't all that good | 18:01 |
sdake | i mean valuable | 18:01 |
senk | mntdevops: always hard to get started. if you need help, feel free to ping me | 18:01 |
rhallisey | not sure though | 18:01 |
*** chopmann has quit IRC | 18:03 | |
*** timss has joined #openstack-kolla | 18:03 | |
coolsvap | sdake: rhallisey https://paste.fedoraproject.org/404471/14706794 | 18:04 |
coolsvap | https://paste.fedoraproject.org/404471/14706794/raw/ | 18:05 |
coolsvap | raw looks good | 18:05 |
sdake | nap time | 18:10 |
sdake | bbiaf | 18:10 |
irtermite | @all you don't happen to see anything going on with iLo device 10.15.243.222 ? I am reviewing an old ticket I had before handing it off to this group. | 18:12 |
*** b_bezak has quit IRC | 18:12 | |
*** gfidente has quit IRC | 18:19 | |
openstackgerrit | David Wang proposed openstack/kolla-kubernetes: Added support for template variables https://review.openstack.org/351408 | 18:23 |
sdake | irtermite - need me to login and check it out? | 18:25 |
sdake | I don't really know the ilo command set | 18:25 |
sdake | pbourke would know but he is in EU tz | 18:25 |
sdake | inc0 may know but he isn't around | 18:25 |
sdake | britthouser woulddefinatey know but he is on pto for a week | 18:25 |
coolsvap | sdake: I am off to bed, will update the review tomorrow morning | 18:27 |
sdake | coolsvap rq, does tempest output in rst? | 18:27 |
sdake | that is what we really need in terms of output | 18:27 |
*** ayoung has quit IRC | 18:27 | |
sdake | html and json isn't particularly helpful | 18:27 |
sdake | i guess we could convert the json to rst | 18:28 |
coolsvap | sdake: its not in rst as per say but we an convert as per req | 18:28 |
coolsvap | because the paste i showed you is too big, I am not too inclined to include that after every scenario | 18:29 |
sdake | we could have a file per scenario | 18:29 |
coolsvap | I will truncate it to keep only info part | 18:29 |
coolsvap | yes | 18:30 |
sdake | would be good data to haveo n what failed atleat in one place | 18:34 |
sdake | so people could eval if they care about those scenarios | 18:35 |
*** senk has quit IRC | 18:44 | |
openstackgerrit | Patrick Powell proposed openstack/kolla: Add Nova sanity check https://review.openstack.org/347416 | 18:49 |
*** Kimmo___ has joined #openstack-kolla | 18:50 | |
*** tuvenen has joined #openstack-kolla | 18:50 | |
*** qbjolof has joined #openstack-kolla | 18:50 | |
*** ntpttr has quit IRC | 18:50 | |
*** bjolo_ has quit IRC | 18:50 | |
*** macsz has quit IRC | 18:50 | |
*** tuvenen_ has quit IRC | 18:50 | |
*** Kimmo__ has quit IRC | 18:50 | |
*** ntpttr- has joined #openstack-kolla | 18:50 | |
openstackgerrit | Patrick Powell proposed openstack/kolla: Add Heat sanity check https://review.openstack.org/350071 | 18:50 |
*** macsz has joined #openstack-kolla | 18:50 | |
*** mliima has quit IRC | 18:56 | |
jemcevoy | I got further over the weekend with my install when I switched to the mitaka branch... Now I am a bit confused by the Ceph disk labeling... Do the partition names need to be unique across the cluster of for each host? I am using _1 _2 _3 on all my hosts and bootstrapping OSD's failed. I have 5 nodes in my cluster and an example of how I am labeling the disks is in this host_vars file: http://paste.openstack.org/show/5520 | 19:21 |
sdake | jemcevoy iwould think they wouldn't but i am not certain | 19:25 |
sdake | jemcevoy your system databae is not setup properly -did you run kolla-ansible -i inventory-file clenaup? | 19:26 |
jrich523 | hey guys getting the following error when i try to do kolla-ansible pull : msg: Unknown error message: Tag 2.0.3 not found in repository docker.io/kolla/centos-binary-kolla-toolbox | 19:26 |
sdake | jrich523 yes - your using a git checkout instead of a git tag | 19:26 |
sdake | to make it work add "openstack_release: 2.0.3' to globals.yml | 19:26 |
jrich523 | ok | 19:27 |
sdake | jrich523 fwiw we are working on sorting out the best way to solve this problem but without a clear solution atm | 19:27 |
sdake | its al pbr related | 19:27 |
*** mfedosin has quit IRC | 19:28 | |
jrich523 | well, i tried starting with devstack, and that went really badly... talked to the openstack folks and they said my best bet was to start here since its the most stable/best documented | 19:28 |
jrich523 | which, even tho its not perfect, its still a lot easier :) | 19:28 |
sdake | which openstack folks did you talk to | 19:28 |
jrich523 | and also you guys respond, which helps a lot :) | 19:28 |
jrich523 | #openstack | 19:28 |
jrich523 | it was tossing a ton of errors when i followed the docs and basically i got "yup, thats what it does right now" | 19:29 |
jrich523 | and was suggest to try this one, or one other, forget the other flavor | 19:29 |
jemcevoy | sdake, Between installs I have been cobbler kickstarting the cluster then my ansible scripts that add a bunch os ssh keys tweak the network and label the disks for ceph... Then I run kolla-ansible deploy. | 19:32 |
sdake | jemcevoy you need to cleanup between those i suspect | 19:33 |
sdake | which other flavor | 19:33 |
sdake | devstack? | 19:33 |
sdake | oh nm didn't read your full thread | 19:33 |
jrich523 | nah, let me see if i have the logs | 19:33 |
jemcevoy | A kickstart is pretty clean... | 19:34 |
sdake | i think what hey wer esuggesting is to use an actual deployment tool not a development tool such as devstack ;) | 19:34 |
*** dwalsh has quit IRC | 19:34 | |
jrich523 | OSA | 19:34 |
jrich523 | right | 19:35 |
openstackgerrit | Merged openstack/kolla-kubernetes: 3/4 of the Neutron Compute services start correctly https://review.openstack.org/350285 | 19:35 |
*** zhubingbing has quit IRC | 19:42 | |
*** macsz has quit IRC | 19:47 | |
*** macsz has joined #openstack-kolla | 19:47 | |
*** v1k0d3n has joined #openstack-kolla | 19:52 | |
jemcevoy | jrich523, I am doing my first install using Kolla... My first problem was that I was installing master which is a moving target. Make sure that you are using a stable branch. | 19:56 |
*** v1k0d3n has quit IRC | 19:57 | |
jemcevoy | I am using origin/stable/mitaka | 19:57 |
*** dwalsh has joined #openstack-kolla | 20:02 | |
*** ppowell has quit IRC | 20:02 | |
Mech422 | jemcevoy: partition names for ceph are machine local - each machine CAN have a _1 _2 _3 | 20:02 |
*** dwalsh has quit IRC | 20:03 | |
Mech422 | jemcevoy: btw - I rebuild my stuff pretty much the same way, but I use foreman instead of cobbler | 20:03 |
Mech422 | jemcevoy: oh - not sure the ceph partition stuff got backported to Mitaka ? | 20:04 |
Mech422 | jemcevoy: you might need to use full disks on Mitaka | 20:04 |
jemcevoy | What is the label pattern for full disks? | 20:04 |
Mech422 | jemcevoy: errgg...some long string...hang on, let me check... | 20:05 |
Mech422 | jemcevoy: shell: sgdisk --change-name=4:KOLLA_CEPH_OSD_BOOTSTRAP_1 /dev/sda | 20:05 |
Mech422 | jemcevoy: shell: sgdisk --change-name=3:KOLLA_CEPH_OSD_BOOTSTRAP_1_J /dev/sda | 20:06 |
Mech422 | jemcevoy: remember, they need to be GPT partition tables | 20:06 |
openstackgerrit | Christian Berendt proposed openstack/kolla: Add cloudkitty ansible role https://review.openstack.org/349305 | 20:08 |
jemcevoy | I have 3 ceph disks on my vmhosts Would then be named KOLLA_CEPH_OSD_BOOTSTRAP_1 KOLLA_CEPH_OSD_BOOTSTRAP_2 KOLLA_CEPH_OSD_BOOTSTRAP_3 and and the journals on the same disk? | 20:08 |
*** jtriley has quit IRC | 20:09 | |
Mech422 | jemcevoy: theoretically...not sure if mitaka supports partitions for ceph | 20:09 |
Mech422 | jemcevoy: I run off master, which had a patch applied | 20:10 |
Mech422 | jemcevoy: I have KOLLA_CEPH_OSD_BOOTSTRAP_1 and KOLLA..._BOOTSTRAP_1_J on my boot drive | 20:11 |
Mech422 | jemcevoy: but unpatched, I think you need a whole disk for ceph use | 20:11 |
jemcevoy | Mech422, so where do I put the journals? | 20:12 |
Mech422 | jemcevoy: if you use a whole disk, kolla does that for ya... | 20:13 |
rhallisey | dcwangmit01, one more thing we need in the CLI is the ability to check that status of a pod that was started by a job/rc/daemon-set | 20:16 |
rhallisey | the status you get from pods are descriptive enough to know if the service is healthy | 20:17 |
rhallisey | for example a job resource won't tell you the health of a pod | 20:17 |
*** srwilkers has quit IRC | 20:22 | |
sbezverk | dcwangmit01: it looks like kolla-kubernetes does not return status of job | 20:25 |
sbezverk | rhallisey: but it does check the status ofservice | 20:25 |
rhallisey | sbezverk, we need the pod status | 20:25 |
rhallisey | or we will never know if it's healthy | 20:26 |
sbezverk | rhallisey: it return pod status | 20:26 |
rhallisey | you can't check that status of a mariadb bootstrap pod | 20:27 |
rhallisey | from the cli | 20:27 |
sbezverk | dcwangmit01: please check this: http://paste.openstack.org/show/552084/ | 20:27 |
sbezverk | rhallisey: in all cases bootstrap is a job, right, if pod keeps crashing job status should reflect it | 20:27 |
rhallisey | sbezverk, it doesn't | 20:28 |
rhallisey | because kubernetes views the job as created | 20:28 |
rhallisey | so it's kubernetes thinks it's healthy | 20:28 |
rhallisey | same goes for rc | 20:28 |
rhallisey | when you query a pod it will return that status of the containers | 20:28 |
rhallisey | which is what were after | 20:28 |
rhallisey | but we need the pods attached to the bootstrap resource | 20:29 |
rhallisey | job/rc/daemonset | 20:29 |
sbezverk | rhallisey: I see what you mean but there is no easy way to get pod name for rc and jor job as they are auto-generated | 20:30 |
rhallisey | you have to do a string search | 20:30 |
rhallisey | https://github.com/ansible/ansible-modules-extras/pull/2466/commits/3b2fa43dedfac30cf37fa367cdff1cdf2fcbb673#diff-76158272904e427e5435304fb425a193R346 | 20:30 |
*** sdake has quit IRC | 20:33 | |
*** inc0 has joined #openstack-kolla | 20:33 | |
*** Jeffrey4l has quit IRC | 20:35 | |
*** Jeffrey4l has joined #openstack-kolla | 20:35 | |
*** bootsha has joined #openstack-kolla | 20:36 | |
*** vhosakot has joined #openstack-kolla | 20:37 | |
*** inc0 has quit IRC | 20:38 | |
sbezverk | rhallisey: if job has 3-4 pods do you want to see individual ok for each pod or one consolidated? | 20:38 |
rhallisey | sbezverk, good question | 20:38 |
*** mark-casey has joined #openstack-kolla | 20:40 | |
rhallisey | just a second | 20:40 |
rhallisey | sbezverk, can you get them individually based on name? | 20:43 |
rhallisey | sbezverk, what pods show up for you when you do a keystone bootstrap | 20:43 |
rhallisey | mine isn't working atm | 20:44 |
openstackgerrit | Merged openstack/kolla-kubernetes: Fix Nova bootstrap https://review.openstack.org/346452 | 20:46 |
openstackgerrit | Merged openstack/kolla-kubernetes: Register Nova with Keystone https://review.openstack.org/351957 | 20:46 |
dcwangmit01 | rhallisey: If kolla-kube creates a job, it can check the status of a job. If it creates an RC, it can check the status of the RC. | 20:47 |
rhallisey | dcwangmit01, right, but we also need to status of the pods associated with the job/rc | 20:48 |
dcwangmit01 | rhallisey: sbezverk The arg to kolla-kube "job" is just a label. It could be "foobar", at which point it would check all of the kube resources defined under services_resources.yml, which could be of any types | 20:48 |
dcwangmit01 | rhallisey: Yes ^ true. | 20:48 |
dcwangmit01 | So job isn't a type in service_resources.yml. If you mean job, it probably is under the label bootstrap | 20:49 |
dcwangmit01 | kolla-kubernetes resource status job keystone -v --debug | 20:49 |
dcwangmit01 | resource_type [job] not in valid resource_types [configmap|disk|pv|pvc|svc|bootstrap|pod] | 20:49 |
dcwangmit01 | sbezverk: Try kolla-kube resource status bootstrap keystone | 20:49 |
*** jeh has quit IRC | 20:51 | |
*** vhosakot has quit IRC | 20:51 | |
dcwangmit01 | What that does is look under service_resources.yml, finds service keyston, finds all files under the bootstrap resource_type. Then, one file at a time run through jinja, then feeds each individual kube yaml snippet to kubectl describe, testing the output on the "kind". E.g. If it's an RC, it ensures pod# actual==desired. If it's a SVC, it ensures the svc has an IP, etc. | 20:51 |
*** vhosakot has joined #openstack-kolla | 20:52 | |
openstackgerrit | Ken Wronkiewicz proposed openstack/kolla: Nova kubernetes changes https://review.openstack.org/346451 | 20:53 |
dcwangmit01 | Also, the naming of the job to be "bootstrap" is an unfortunate legacy. kfox1111 Wants me to patch and rename bootstrap to be job, to match the kube type. I'll do so after a bunch of the kolla-kube patches/merge conflicts go through | 20:53 |
wirehead_ | well, also, we should probably change the -pod to -rc, except we should wait on that till after we turn them into deployments, so people’s head doesnt spin. | 20:55 |
*** ayoung has joined #openstack-kolla | 20:55 | |
*** jeh has joined #openstack-kolla | 20:56 | |
*** jeh has quit IRC | 20:56 | |
*** jeh has joined #openstack-kolla | 20:56 | |
dcwangmit01 | Yes! | 20:57 |
rhallisey | dcwangmit01, job or bootstrap works | 20:57 |
rhallisey | job might be better since that is the kube object name | 20:57 |
dcwangmit01 | Yes! | 20:57 |
rhallisey | dcwangmit01, just need some clarification from above. Are you agreeing that we don't get that status of the pods from an rc/job | 20:58 |
openstackgerrit | Ken Wronkiewicz proposed openstack/kolla-kubernetes: Remove hostnetworking for Nova control plane and convert to RC https://review.openstack.org/347572 | 20:58 |
*** egonzalez90 has joined #openstack-kolla | 20:58 | |
rhallisey | currently I'm not seeing, but I may need a pointer :) | 20:58 |
dcwangmit01 | We don't get the status if one Kube resource spawns others. For example, an RC spawns Pods. A deployment spawns pods. Right now, if kolla-kube creates an RC, it can check the status of the RC, and not the subpods. However, it does get some high-level info from the RC itself, such as the number of pods actual==desired. | 20:59 |
dcwangmit01 | ^ so yes I agree, and yes, we need to figure out the linkage | 21:00 |
rhallisey | gotcha | 21:00 |
dcwangmit01 | However, will that require kolla-kube interpreting labels and selectors? | 21:00 |
dcwangmit01 | We'll be re-implementing part of kubernetes! | 21:00 |
dcwangmit01 | Which makes me reluctant | 21:00 |
dcwangmit01 | So kolla-kube resource status commands (currently) check only the status of the kube resources that it directly creates | 21:01 |
dcwangmit01 | Not any of the spawned sub-pods | 21:01 |
rhallisey | how about rest calls | 21:01 |
*** coolsvap has quit IRC | 21:01 | |
dcwangmit01 | rhallisey: That's where we would have to go next. Basically, do not rely on kubectl describe. Go directly to the API. | 21:02 |
rhallisey | hm ok | 21:02 |
rhallisey | may put that on hold for this week | 21:02 |
dcwangmit01 | I did not go that way for the first implementation, because it's a lot of work. We have to worry about kube configuration, proper authentication, etc. | 21:02 |
dcwangmit01 | Although, we must go that way eventually | 21:02 |
rhallisey | true | 21:02 |
rhallisey | cli v2.0 | 21:03 |
dcwangmit01 | It's just, when do we want to break that seal and go down that path of lots of work (but much better information). | 21:03 |
rhallisey | well it's once we tackle the most urgent matter | 21:03 |
rhallisey | which is a solid demo | 21:04 |
dcwangmit01 | And for now, can we just rely on a combination of "kolla-kube resource status" which only check the status of kube resources that we create, and a "kolla-kube resource health mariadb" which does actual papplcation level port checking, etc. The latter doesn't exist | 21:04 |
rhallisey | probably won't show deploment, but features of using kube | 21:04 |
wirehead_ | Hm. I could use a few patches +workflow’d. | 21:04 |
dcwangmit01 | In kolla? Can't help you. | 21:05 |
rhallisey | dcwangmit01, I'll bring it up wed at the meeting | 21:05 |
*** vhosakot_ has joined #openstack-kolla | 21:05 | |
wirehead_ | Kolla-kube | 21:05 |
rhallisey | and make a bp | 21:05 |
*** bootsha has quit IRC | 21:05 | |
wirehead_ | https://review.openstack.org/#/c/347572/ | 21:05 |
patchbot | wirehead_: patch 347572 - kolla-kubernetes - Remove hostnetworking for Nova control plane and c... | 21:05 |
*** vhosakot has quit IRC | 21:05 | |
dcwangmit01 | wirehead_: A bunch of your patches just went through minutes ago | 21:05 |
wirehead_ | Yeah, that one needed to be rebased after the earlier patches went through | 21:05 |
*** bootsha has joined #openstack-kolla | 21:06 | |
dcwangmit01 | wirehead_: Just did it. I see you rebased | 21:06 |
dcwangmit01 | Do you guys know why the CI system isn't running for this: https://review.openstack.org/#/c/351408/ | 21:07 |
patchbot | dcwangmit01: patch 351408 - kolla-kubernetes - Added support for template variables | 21:07 |
*** vhosakot_ has quit IRC | 21:07 | |
dcwangmit01 | ^ I posted a "recheck" which didn't do anything. Right now, it should be a merge conflict, but still I don't get why CI hasn't run | 21:07 |
wirehead_ | Jenkins has not run on that one. | 21:08 |
dcwangmit01 | I'm wondering why? Jenkins hasn't run for any of the 3 patchsets | 21:08 |
*** haplo37__ has quit IRC | 21:08 | |
rhallisey | needs a reverify | 21:09 |
dcwangmit01 | does recheck==reverify? | 21:12 |
dcwangmit01 | rhallisey: Last comment here says it's recheck: https://bugs.launchpad.net/openstack-ci/+bug/1061929 | 21:13 |
openstack | Launchpad bug 1061929 in OpenStack Core Infrastructure "Reduce confusion between recheck/reverify" [Wishlist,Invalid] | 21:13 |
dcwangmit01 | I'll do both from now on :) | 21:14 |
*** bootsha has quit IRC | 21:15 | |
rhallisey | dcwangmit01, recheck will re run the gate. Reverify will re run the approval to see if it's good to be merged | 21:15 |
wirehead_ | Wow. | 21:17 |
*** daneyon has joined #openstack-kolla | 21:17 | |
*** inc0 has joined #openstack-kolla | 21:20 | |
dcwangmit01 | rhallisey: Thanks for the explanation. | 21:21 |
rhallisey | no problem | 21:21 |
*** huikang has quit IRC | 21:22 | |
*** daneyon has quit IRC | 21:22 | |
wirehead_ | Like, I can’t go full snark and call gerrit hipster-made, artisinal, small-batch custom software run at openstack scale, because it is used by other large projects that hate gitjub. | 21:22 |
wirehead_ | But. Sometimes.. :P | 21:23 |
rhallisey | :) | 21:23 |
*** haplo37__ has joined #openstack-kolla | 21:23 | |
rhallisey | wirehead_, dcwangmit01 if you guys have a few minutes, scan through the review q and see if anything is good to go | 21:24 |
wirehead_ | I also can’t go full snark at hipster-made artisinal, small-batch anything, given my known cullinary preferences, but.. | 21:24 |
rhallisey | there are a few patches that have a +2 | 21:24 |
*** inc0 has quit IRC | 21:24 | |
rhallisey | I g2g for today. I'll ttyl | 21:24 |
*** rhallisey has quit IRC | 21:25 | |
openstackgerrit | Merged openstack/kolla-kubernetes: Added support for template variables https://review.openstack.org/351408 | 21:42 |
*** inc0 has joined #openstack-kolla | 21:44 | |
*** egonzalez90 has quit IRC | 21:47 | |
*** mark-casey has quit IRC | 21:55 | |
*** banix has quit IRC | 22:03 | |
openstackgerrit | Ryan Hallisey proposed openstack/kolla-kubernetes: Ansible support for Mariadb https://review.openstack.org/334255 | 22:04 |
openstackgerrit | David Wang proposed openstack/kolla-kubernetes: Add Ceph support for Mariadb https://review.openstack.org/350312 | 22:11 |
inc0 | anyone working on OSIC cluster atm? | 22:16 |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Glance bootstrap to create service/project/role https://review.openstack.org/346215 | 22:26 |
*** haplo37__ has quit IRC | 22:32 | |
*** michauds has quit IRC | 22:37 | |
*** bmace has quit IRC | 22:47 | |
inc0 | networking question guys, how to check with tcpdump if I'm having mtu issue? | 22:47 |
inc0 | and fragmetation happends | 22:47 |
inc0 | ? | 22:47 |
jemcevoy | inc0, I usually use tracepath to checkout mtu's | 22:50 |
jemcevoy | inc0, You can also force ping not to fragment like this : ping -M do -s 8972 -c 1 $addr | 22:53 |
jemcevoy | that is for jumbo frames. You have to subtract 28 from the mtu size for the headers | 22:54 |
inc0 | my network is fleaky and I wonder if that's because of mtu | 22:55 |
inc0 | but now I added few v's into tcpdump and see lot's of incorrect packets | 22:55 |
inc0 | so I guess retransmissions happens | 22:55 |
*** banix has joined #openstack-kolla | 23:00 | |
*** bmace has joined #openstack-kolla | 23:04 | |
openstackgerrit | Xinyuan Huang proposed openstack/kolla: Trivial typo fix in doc https://review.openstack.org/352631 | 23:10 |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Add persistent volume to Glance https://review.openstack.org/348304 | 23:22 |
*** inc0 has quit IRC | 23:24 | |
openstackgerrit | Christian Berendt proposed openstack/kolla: Improve script to check used source versions https://review.openstack.org/352640 | 23:27 |
*** inc0 has joined #openstack-kolla | 23:29 | |
*** rkrum has joined #openstack-kolla | 23:32 | |
openstackgerrit | Christian Berendt proposed openstack/kolla: Do not iterate over listing in the sanity checks https://review.openstack.org/352642 | 23:39 |
*** bmace has quit IRC | 23:42 | |
*** inc0 has quit IRC | 23:42 | |
openstackgerrit | Xinyuan Huang proposed openstack/kolla: Fix typo in doc https://review.openstack.org/352631 | 23:43 |
openstackgerrit | Christian Berendt proposed openstack/kolla: Replace horizon default config with custom config https://review.openstack.org/306928 | 23:45 |
*** zhurong has joined #openstack-kolla | 23:46 | |
openstackgerrit | Christian Berendt proposed openstack/kolla: Customizations for kibana https://review.openstack.org/346252 | 23:49 |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Moving mariadb-disk.sh.j2 script to become common/common-disk.sh.j2 https://review.openstack.org/352646 | 23:55 |
*** zhurong has quit IRC | 23:55 | |
*** sdake has joined #openstack-kolla | 23:56 | |
sdake | hey cats | 23:57 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!