*** spatel has joined #openstack-ansible | 00:01 | |
*** spatel has quit IRC | 00:06 | |
*** DanyC has quit IRC | 00:10 | |
*** DanyC has joined #openstack-ansible | 00:11 | |
*** DanyC has quit IRC | 00:16 | |
*** waxfire has joined #openstack-ansible | 00:27 | |
*** waxfire has quit IRC | 00:32 | |
*** waxfire has joined #openstack-ansible | 00:33 | |
*** spatel has joined #openstack-ansible | 00:34 | |
*** waxfire has quit IRC | 00:34 | |
*** waxfire has joined #openstack-ansible | 00:35 | |
*** waxfire has quit IRC | 00:36 | |
*** waxfire has joined #openstack-ansible | 00:37 | |
*** openstackgerrit has joined #openstack-ansible | 01:03 | |
openstackgerrit | Merged openstack/openstack-ansible stable/train: Bump ansible version to 2.8.8 https://review.opendev.org/703390 | 01:03 |
---|---|---|
*** nicolasbock has quit IRC | 02:28 | |
*** oligau has quit IRC | 02:29 | |
*** oligau has joined #openstack-ansible | 02:43 | |
*** waxfire has quit IRC | 03:08 | |
*** gyee has quit IRC | 03:11 | |
*** waxfire has joined #openstack-ansible | 03:15 | |
*** spatel has quit IRC | 03:23 | |
*** waxfire has quit IRC | 03:30 | |
*** waxfire has joined #openstack-ansible | 03:31 | |
*** waxfire has quit IRC | 03:33 | |
*** waxfire has joined #openstack-ansible | 03:33 | |
*** waxfire has quit IRC | 03:34 | |
*** waxfire has joined #openstack-ansible | 03:34 | |
*** waxfire has quit IRC | 03:35 | |
*** waxfire has joined #openstack-ansible | 03:36 | |
*** waxfire has quit IRC | 03:37 | |
*** waxfire has joined #openstack-ansible | 03:38 | |
*** udesale has joined #openstack-ansible | 04:19 | |
*** goldyfruit has quit IRC | 04:49 | |
*** vesper has quit IRC | 05:18 | |
*** vesper11 has joined #openstack-ansible | 05:23 | |
*** raukadah is now known as chkumar|rover | 05:28 | |
*** evrardjp has quit IRC | 05:34 | |
*** evrardjp has joined #openstack-ansible | 05:34 | |
*** errr has quit IRC | 05:55 | |
*** errr has joined #openstack-ansible | 05:56 | |
*** Neurognostic has joined #openstack-ansible | 06:15 | |
*** redkrieg has quit IRC | 06:22 | |
*** elenalindq has joined #openstack-ansible | 06:22 | |
*** redkrieg has joined #openstack-ansible | 06:24 | |
*** rgogunskiy has joined #openstack-ansible | 06:24 | |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_swift master: Use py2 shebang for centos distro installs https://review.opendev.org/705247 | 07:14 |
*** cshen has joined #openstack-ansible | 07:19 | |
*** pcaruana has joined #openstack-ansible | 07:23 | |
*** rgogunskiy_ has joined #openstack-ansible | 07:32 | |
*** rgogunsk_ has joined #openstack-ansible | 07:33 | |
*** rgogunskiy has quit IRC | 07:34 | |
*** rgogunskiy_ has quit IRC | 07:37 | |
jrosser | cshen: https://review.opendev.org/#/c/706874/ and https://review.opendev.org/#/c/706823/ these look very similar and are proposed pretty much together | 07:38 |
jrosser | any thoughts on which is preferred? | 07:39 |
*** udesale has quit IRC | 07:46 | |
*** udesale has joined #openstack-ansible | 07:47 | |
openstackgerrit | Mohammed Naser proposed openstack/openstack-ansible-os_manila master: Start using uWSGI role https://review.opendev.org/704935 | 08:03 |
openstackgerrit | Mohammed Naser proposed openstack/openstack-ansible-os_manila master: Start using uWSGI role https://review.opendev.org/704935 | 08:03 |
openstackgerrit | Mohammed Naser proposed openstack/openstack-ansible-os_manila master: Add image download for manila https://review.opendev.org/704972 | 08:04 |
openstackgerrit | Mohammed Naser proposed openstack/openstack-ansible-os_manila master: Add Manila key generation and distribution https://review.opendev.org/705019 | 08:04 |
cshen | jrosser: you could take 706874 as preferred. I was teaching my colleague to contribute a patch. | 08:12 |
jrosser | cshen: awesome thanks, new contributors++ :) | 08:13 |
cshen | no problem. I guess he needs a second patch. I will tell him. | 08:13 |
bhyrted | im trying to add cloudkitty role to my openstack lxc containers based enviroment, i added at /etc/openstack_deploy/env.d/cloudkitty.yml http://paste.openstack.org/show/789415 | 08:17 |
bhyrted | added to /opt/openstack-ansible/ansible-role requirements name: os_cloudkitty | 08:18 |
bhyrted | scm: git | 08:18 |
bhyrted | src: https://opendev.org/openstack/openstack-ansible-os_cloudkitty | 08:18 |
bhyrted | version: master | 08:18 |
bhyrted | trackbranch: master | 08:18 |
bhyrted | and when i get playbook lxc-containers-create i get no cloudkitty containers | 08:18 |
bhyrted | my inventory seems to be half empty | 08:19 |
bhyrted | cat /etc/openstack_deploy/openstack_inventory.json | grep cloudkitty | 08:19 |
bhyrted | "cloudkitty_containers", | 08:19 |
bhyrted | "cloudkitty-hosts": { | 08:19 |
bhyrted | "cloudkitty-infra_hosts": { | 08:19 |
bhyrted | "cloudkitty_all": { | 08:19 |
bhyrted | "cloudkitty_api", | 08:19 |
bhyrted | "cloudkitty_engine" | 08:19 |
bhyrted | "cloudkitty_api": { | 08:19 |
bhyrted | "cloudkitty_container": { | 08:19 |
bhyrted | "cloudkitty_containers": { | 08:19 |
bhyrted | "cloudkitty_engine": { | 08:19 |
bhyrted | "cloudkitty_hosts": { | 08:19 |
bhyrted | "cloudkitty_hosts", | 08:19 |
bhyrted | what am I doing wrong?;-) | 08:19 |
bhyrted | also on /etc/openstack_deploy/openstack_user_config.yml i have cloudkitty-hosts: *infrastructure_hosts | 08:20 |
bhyrted | I know Im missing something but i dont know what.. | 08:26 |
*** tosky has joined #openstack-ansible | 08:27 | |
*** jcosmao has joined #openstack-ansible | 08:40 | |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-openstack_hosts master: Use python3 to drop hosts file script where possible https://review.opendev.org/705849 | 08:53 |
*** rpittau|afk is now known as rpittau | 08:55 | |
*** DanyC has joined #openstack-ansible | 09:05 | |
*** maharg101 has joined #openstack-ansible | 09:08 | |
*** DanyC has quit IRC | 09:10 | |
*** DanyC has joined #openstack-ansible | 09:13 | |
*** DanyC has quit IRC | 09:14 | |
*** DanyC has joined #openstack-ansible | 09:15 | |
*** DanyC has quit IRC | 09:18 | |
*** DanyC has joined #openstack-ansible | 09:19 | |
*** miloa has joined #openstack-ansible | 09:25 | |
*** sshnaidm|afk is now known as sshnaidm | 09:25 | |
openstackgerrit | yedhu proposed openstack/openstack-ansible master: Added regexp to avoid multiple entry of lxc.start.order in os-lxc-container-setup.yml https://review.opendev.org/706874 | 09:44 |
*** ivve has joined #openstack-ansible | 09:53 | |
openstackgerrit | yedhu proposed openstack/openstack-ansible master: Fix idempotency for os-lxc-container-setup.yml https://review.opendev.org/706874 | 10:01 |
mnaser | o | 10:04 |
mnaser | is world broken again? | 10:04 |
mnaser | i think this is the virtualenv 20.0.0 breakages | 10:05 |
chkumar|rover | mnaser, yes, we are fixing this way https://review.opendev.org/#/c/707062/13/install-deps.sh | 10:06 |
noonedeadpunk | In constraints it seems that it's bumped to 16.7.9 | 10:07 |
mnaser | urgh | 10:10 |
mnaser | today on things i dont want to deal with | 10:10 |
mnaser | oh we dont log python_venv_build | 10:13 |
mnaser | so i think what's happening here | 10:14 |
mnaser | python_venv_build tries to upgrade setuptools to latest | 10:14 |
mnaser | and everything breaks :-) | 10:14 |
mnaser | which means every single deployment is probably broken now | 10:15 |
chkumar|rover | mnaser, yes tripleo is totally RED | 10:15 |
mnaser | https://github.com/openstack/openstack-ansible/blob/5516771028899c6b89990ff8da6026ff8254fa88/inventory/group_vars/all/all.yml#L146-L147 | 10:15 |
mnaser | https://github.com/openstack/openstack-ansible/blob/5516771028899c6b89990ff8da6026ff8254fa88/global-requirement-pins.txt | 10:16 |
mnaser | we can probably pin there | 10:16 |
mnaser | and that should fix things | 10:16 |
jrosser | wow really updates to latest? | 10:16 |
mnaser | :) | 10:16 |
mnaser | we're just CUTTING EDGE | 10:16 |
jrosser | i wonder if on osa thats actaully pinned though? | 10:16 |
mnaser | it aint | 10:16 |
mnaser | see global-requirement-pins.txt | 10:16 |
mnaser | and two links above | 10:16 |
mnaser | we pin setuptools/pip/wheel | 10:16 |
mnaser | but not virtualenv | 10:16 |
*** bhyrted has quit IRC | 10:17 | |
mnaser | wait hold up | 10:17 |
jrosser | ooh | 10:17 |
mnaser | but we dont upgrade virtualenv | 10:17 |
*** bhyrted has joined #openstack-ansible | 10:17 | |
jrosser | right, we use the distro one *always* don't we? | 10:17 |
mnaser | and pip setuptools and wheel are both pinned | 10:17 |
mnaser | yeah | 10:17 |
mnaser | waait | 10:17 |
mnaser | must be something else | 10:17 |
jrosser | we leanred the hard way never to mess with virtualevn from the host | 10:18 |
mnaser | wait tripleo solves this by removing the system virtualenv | 10:18 |
mnaser | so something else is conflicting then? | 10:18 |
mnaser | the failure is | 10:20 |
mnaser | fatal: [aio1]: FAILED! => {"msg": "The conditional check '((_virtualenv_version.stdout | trim) == 'none') or ((_virtualenv_version.stdout | trim) is version('1.10', '<'))' failed. The error was: Version comparison: 'LooseVersion' object has no attribute 'version'\n\nThe error appears to be in '/home/zuul/src/opendev.org/openstack/ansible-role-python_venv_build/tasks/python_venv_preflight.yml': line 33, column | 10:20 |
mnaser | 3, but may\nbe elsewhere in the file depending on the exact syntax problem.\n\nThe offending line appears to be:\n\n\n- name: Fail when required virtualenv version is not present\n ^ here\n"} | 10:20 |
mnaser | inside: "python_venv_build : Fail when required virtualenv version is not present" | 10:20 |
mnaser | wait this is a metal job | 10:20 |
mnaser | where do we get virtualenv in metal jobs | 10:20 |
mnaser | i wonder if infra ships one | 10:21 |
mnaser | https://github.com/sapcc/manila/commit/f5e3ce3f4802353ba86a1e42cd56bd02b22ec5ec | 10:25 |
mnaser | maybe bumping six? | 10:25 |
noonedeadpunk | Right now deploying aio from master and by far it's ok | 10:26 |
noonedeadpunk | (already on nove build) | 10:26 |
noonedeadpunk | *nova | 10:26 |
mnaser | https://github.com/pre-commit/pre-commit/issues/1325#issuecomment-584280357 | 10:26 |
mnaser | Seems six is pulled in from the global site-packages of Ubuntu folder and hence pip ignores the six requirement constraint... One can either bump six from Ubuntu or install pre-commit into an isolated environment as I see. Generally, it's not supported to mix Ubuntu python packages with packages from PyPi in the same python environment. Use virtual environments to create separation, or install pre-commit from | 10:26 |
mnaser | Ubuntu package manager also. In the future, pypa/pip#988 would throw an exception, but for now, it silently ignores it. If this is your developer machine consider using https://pipxproject.github.io/pipx/installation/ to automatically create these isolated environments for tools. | 10:26 |
*** ansmith has joined #openstack-ansible | 10:26 | |
mnaser | what os, noonedeadpunk ? | 10:27 |
noonedeadpunk | centos | 10:27 |
noonedeadpunk | oh, wait, I do distro deploy, so disregard... | 10:27 |
mnaser | so i guess in that case what we really need to do is bump six | 10:27 |
noonedeadpunk | six is also present in u-c... | 10:28 |
*** ansmith_ has quit IRC | 10:28 | |
noonedeadpunk | so if we use it properly.... | 10:29 |
*** cshen_ has joined #openstack-ansible | 10:34 | |
* noonedeadpunk doing ubuntu source aio | 10:34 | |
*** cshen has quit IRC | 10:37 | |
jrosser | mnaser: it's pretty sad whats being shown in #infra, its like were not actually testing on the distros we think we are | 10:51 |
*** rpittau is now known as rpittau|bbl | 11:21 | |
*** NewJorg has quit IRC | 11:26 | |
*** cshen_ has quit IRC | 11:28 | |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_swift master: Use py2 shebang for centos distro installs https://review.opendev.org/705247 | 11:30 |
noonedeadpunk | ++ | 11:30 |
noonedeadpunk | ubuntu source aio deployes fine by far | 11:30 |
*** NewJorg has joined #openstack-ansible | 11:32 | |
*** sluna has joined #openstack-ansible | 11:35 | |
mnaser | jrosser: thats a bummer. | 11:43 |
*** cshen has joined #openstack-ansible | 11:45 | |
jrosser | yes it is, and it makes it really tough becasue we're not testing on the thing we deploy with/to | 11:46 |
jrosser | you'd never see that virtualenv 20 thing in the wild, and equally we'd miss something that might crop up with the actual distro versions | 11:46 |
openstackgerrit | Sam Choraria proposed openstack/openstack-ansible-ops master: Allow specified http ports to be ignored by packetbeat https://review.opendev.org/707124 | 11:50 |
openstackgerrit | Sam Choraria proposed openstack/openstack-ansible-ops master: Allow specified http ports to be ignored by packetbeat https://review.opendev.org/707124 | 11:54 |
*** nicolasbock has joined #openstack-ansible | 12:02 | |
openstackgerrit | Merged openstack/openstack-ansible-ops master: Allow beat processors to be defined through configuration data https://review.opendev.org/705965 | 12:09 |
*** udesale_ has joined #openstack-ansible | 12:11 | |
*** udesale has quit IRC | 12:13 | |
openstackgerrit | Sam Choraria proposed openstack/openstack-ansible-ops master: Allow specified http ports to be ignored by packetbeat https://review.opendev.org/707124 | 12:18 |
*** ansmith has quit IRC | 12:26 | |
*** Neurognostic has quit IRC | 12:36 | |
*** Neurognostic has joined #openstack-ansible | 12:36 | |
*** cshen has quit IRC | 12:40 | |
mgariepy | gillesMo, do you have more detail on the 16.04/18.04 issue ? i would like to track down what the problem is. | 12:44 |
mgariepy | also not having issue on 18.1.13 | 12:48 |
*** rh-jelabarre has joined #openstack-ansible | 13:07 | |
*** goldyfruit has joined #openstack-ansible | 13:18 | |
*** rpittau|bbl is now known as rpittau | 13:22 | |
*** ansmith has joined #openstack-ansible | 13:24 | |
*** Neurognostic_ has joined #openstack-ansible | 13:24 | |
*** Neurognostic has quit IRC | 13:28 | |
*** cshen has joined #openstack-ansible | 13:29 | |
*** Neurognostic__ has joined #openstack-ansible | 13:31 | |
*** Neurognostic has joined #openstack-ansible | 13:33 | |
*** Neurognostic_ has quit IRC | 13:34 | |
*** Neurognostic__ has quit IRC | 13:35 | |
*** Neurognostic_ has joined #openstack-ansible | 13:40 | |
*** Neurognostic has quit IRC | 13:43 | |
*** Neurognostic__ has joined #openstack-ansible | 13:43 | |
*** Neurognostic has joined #openstack-ansible | 13:47 | |
*** Neurognostic_ has quit IRC | 13:47 | |
*** Neurognostic_ has joined #openstack-ansible | 13:48 | |
*** Neurognostic__ has quit IRC | 13:50 | |
*** Neurognostic__ has joined #openstack-ansible | 13:50 | |
*** Neurognostic has quit IRC | 13:51 | |
*** Neurognostic_ has quit IRC | 13:52 | |
*** pcaruana has quit IRC | 14:00 | |
bhyrted | ERROR murano TypeError: You cannot set Response.body to a text object (use Response.text) | 14:18 |
bhyrted | any hint? | 14:18 |
bhyrted | ;-) | 14:18 |
jrosser | bhyrted: i don't know about that..... but did you get anywhere with the cloudkitty thing from earlier? | 14:20 |
*** goldyfruit has quit IRC | 14:38 | |
*** goldyfruit has joined #openstack-ansible | 14:38 | |
*** cshen has quit IRC | 14:42 | |
*** chkumar|rover is now known as raukadah | 14:50 | |
*** pcaruana has joined #openstack-ansible | 14:54 | |
*** cshen has joined #openstack-ansible | 15:07 | |
*** cshen_ has joined #openstack-ansible | 15:14 | |
*** cshen has quit IRC | 15:16 | |
*** Neurognostic_ has joined #openstack-ansible | 15:16 | |
*** rgogunsk_ has quit IRC | 15:18 | |
*** Neurognostic__ has quit IRC | 15:19 | |
*** Neurognostic__ has joined #openstack-ansible | 15:22 | |
*** spatel has joined #openstack-ansible | 15:24 | |
*** Neurognostic has joined #openstack-ansible | 15:26 | |
*** Neurognostic_ has quit IRC | 15:26 | |
*** Neurognostic_ has joined #openstack-ansible | 15:27 | |
*** Neurognostic_ has quit IRC | 15:28 | |
*** Neurognostic_ has joined #openstack-ansible | 15:28 | |
*** Neurognostic__ has quit IRC | 15:29 | |
*** Neurognostic__ has joined #openstack-ansible | 15:30 | |
*** Neurognostic has quit IRC | 15:31 | |
*** cshen_ has quit IRC | 15:31 | |
*** Neurognostic_ has quit IRC | 15:34 | |
*** cshen has joined #openstack-ansible | 15:43 | |
*** Neurognostic_ has joined #openstack-ansible | 15:52 | |
*** udesale_ has quit IRC | 15:52 | |
*** udesale_ has joined #openstack-ansible | 15:53 | |
*** cshen has quit IRC | 15:55 | |
*** Neurognostic__ has quit IRC | 15:56 | |
*** KeithMnemonic has quit IRC | 15:56 | |
*** Neurognostic has joined #openstack-ansible | 15:56 | |
*** Neurognostic_ has quit IRC | 15:59 | |
*** Neurognostic has quit IRC | 16:01 | |
noonedeadpunk | #startmeeting openstack_ansible_meeting | 16:01 |
openstack | Meeting started Tue Feb 11 16:01:44 2020 UTC and is due to finish in 60 minutes. The chair is noonedeadpunk. Information about MeetBot at http://wiki.debian.org/MeetBot. | 16:01 |
openstack | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 16:01 |
*** openstack changes topic to " (Meeting topic: openstack_ansible_meeting)" | 16:01 | |
openstack | The meeting name has been set to 'openstack_ansible_meeting' | 16:01 |
noonedeadpunk | #topic office hours | 16:01 |
*** openstack changes topic to "office hours (Meeting topic: openstack_ansible_meeting)" | 16:01 | |
jrosser | hello | 16:03 |
*** udesale_ has quit IRC | 16:06 | |
*** udesale_ has joined #openstack-ansible | 16:07 | |
noonedeadpunk | o/ | 16:10 |
noonedeadpunk | I feel like we'll have tons of work with ansible 2.10... | 16:11 |
*** ivve has quit IRC | 16:13 | |
noonedeadpunk | also wondering who will be managing rabbitmq collections, as modules have comunity support and seems not so active.... | 16:14 |
*** udesale_ has quit IRC | 16:16 | |
noonedeadpunk | and what our next steps will be if nobody is... But this discussion is more for SIG actually | 16:16 |
noonedeadpunk | jrosser: has everything for train tempest fix got merged? | 16:24 |
jrosser | hi | 16:26 |
jrosser | noonedeadpunk: yes i was trying to get someone to bite with helping move that rabbitmq stuff forward | 16:26 |
noonedeadpunk | Yeah, I saw that... | 16:27 |
jrosser | i think that this is the last bit of train/tempest https://review.opendev.org/#/c/706983/ | 16:27 |
jrosser | and then we need a SHA bump to pick up the changes to os_tempest in the train release | 16:28 |
jrosser | i need to go and look back across all the stable branches to double check that some requirements URL pin didnt get left behind | 16:28 |
jrosser | different topic but we should talk about this virtualenv 20.x breaking a ton of stuff | 16:29 |
noonedeadpunk | yeah... | 16:31 |
noonedeadpunk | Maybe we can drop virtualenv with integrated repo? | 16:31 |
noonedeadpunk | so do right the same thing triplo does | 16:32 |
*** gyee has joined #openstack-ansible | 16:32 | |
jrosser | you mean downgrade it? | 16:33 |
noonedeadpunk | or downgrade.... Actually I think we do install it later | 16:33 |
noonedeadpunk | as in blank aio I didn't catch any problems | 16:34 |
jamesdenton | o/ | 16:34 |
noonedeadpunk | as I get in infra image virtualenv is already installed via pip? | 16:34 |
jrosser | yes, and they put in 'latest', which is very much not like what you'd find on a real bionic cloud image, for example | 16:35 |
noonedeadpunk | yeah | 16:35 |
noonedeadpunk | wait... | 16:35 |
noonedeadpunk | does bionic have venv installed out of the box? | 16:36 |
noonedeadpunk | so the correct solution will be probably building CI images with corresponding to u-c... | 16:36 |
noonedeadpunk | _probably_ | 16:37 |
jrosser | we put it there from distro package https://github.com/openstack/openstack-ansible/blob/master/scripts/bootstrap-ansible.sh#L95 | 16:37 |
noonedeadpunk | I think we should drop this package - it's for py2 iirc | 16:38 |
noonedeadpunk | as well as python-dev, python-minimal, python-minimal | 16:39 |
jrosser | i don't think u-c is the solution here | 16:40 |
jrosser | becasue from an OSA perspective, and end user starts with a stock install of their OS of choice | 16:40 |
jrosser | and that comes with a distro python and distro virtualenv, of some version | 16:41 |
noonedeadpunk | so we may just drop venv module with bootstrap-host role | 16:41 |
jrosser | and we decouple ourselved from the distro with the ansible-runtime venv created with the distro virtualenv, then upgrade setuptools and pip and whatever inside that to be uniform | 16:42 |
jrosser | thats totally not what we see in the CI runs, becasue there isn't a distro virtualenv in the same way | 16:42 |
jrosser | it's been baked into the image in a totally CI specific way | 16:42 |
noonedeadpunk | yeah, that's true. I've just checked and ubuntu do not have venv py3 module out of the box | 16:43 |
jrosser | so i've been watching what they're doing for hacking up devstack to fix this today | 16:44 |
jrosser | and all of this is firmly in the category of "messing with the distro python with stuff from pip" <- this is not cool | 16:44 |
noonedeadpunk | agree with that | 16:47 |
noonedeadpunk | but not sure what to do with that... As I think we won't be given another images... | 16:48 |
* jrosser lurking in #infra, this is all being discussed right now | 16:48 | |
jrosser | i note how they say 'virtualenv' is old tech to be replaced with 'venv' | 16:49 |
jrosser | for py3 | 16:49 |
noonedeadpunk | so we install it here for hosts https://opendev.org/openstack/openstack-ansible-openstack_hosts/src/branch/master/vars/ubuntu-18.04.yml#L54 | 16:50 |
jrosser | right | 16:52 |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: Drop py2 deb packages for bootstrap-ansible https://review.opendev.org/707201 | 16:53 |
*** waxfire has quit IRC | 16:53 | |
*** waxfire has joined #openstack-ansible | 16:54 | |
*** rpittau is now known as rpittau|afk | 17:00 | |
jrosser | tbh i'm not quite following how / what is being fixed | 17:02 |
jrosser | there some devstack specific stuff and some changes to the images | 17:02 |
noonedeadpunk | tbh me too | 17:06 |
*** ivve has joined #openstack-ansible | 17:16 | |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: Drop virtualenv pip package for CI https://review.opendev.org/707212 | 17:16 |
noonedeadpunk | #endmeeting | 17:17 |
*** openstack changes topic to "Launchpad: https://launchpad.net/openstack-ansible || Weekly Meetings: https://wiki.openstack.org/wiki/Meetings/openstack-ansible || Review Dashboard: http://bit.ly/2xA1eZC" | 17:17 | |
openstack | Meeting ended Tue Feb 11 17:17:06 2020 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 17:17 |
openstack | Minutes: http://eavesdrop.openstack.org/meetings/openstack_ansible_meeting/2020/openstack_ansible_meeting.2020-02-11-16.01.html | 17:17 |
openstack | Minutes (text): http://eavesdrop.openstack.org/meetings/openstack_ansible_meeting/2020/openstack_ansible_meeting.2020-02-11-16.01.txt | 17:17 |
openstack | Log: http://eavesdrop.openstack.org/meetings/openstack_ansible_meeting/2020/openstack_ansible_meeting.2020-02-11-16.01.log.html | 17:17 |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: Drop py2 deb packages for bootstrap-ansible https://review.opendev.org/707201 | 17:17 |
*** kopecmartin is now known as kopecmartin|afk | 17:27 | |
*** evrardjp has quit IRC | 17:34 | |
*** evrardjp has joined #openstack-ansible | 17:34 | |
*** spatel has quit IRC | 17:50 | |
*** miloa has quit IRC | 17:51 | |
*** errr has quit IRC | 17:55 | |
*** errr has joined #openstack-ansible | 17:57 | |
*** jcosmao has left #openstack-ansible | 18:13 | |
*** macz has joined #openstack-ansible | 18:15 | |
*** DanyC_ has joined #openstack-ansible | 18:21 | |
*** DanyC_ has quit IRC | 18:24 | |
*** DanyC has quit IRC | 18:24 | |
*** DanyC has joined #openstack-ansible | 18:25 | |
*** DanyC has quit IRC | 18:29 | |
*** alvinstarr has joined #openstack-ansible | 19:31 | |
*** nsmeds has left #openstack-ansible | 20:23 | |
*** owalsh has quit IRC | 20:30 | |
*** rgogunskiy has joined #openstack-ansible | 20:33 | |
*** rgogunskiy has quit IRC | 20:37 | |
*** owalsh has joined #openstack-ansible | 20:44 | |
*** masterpe has joined #openstack-ansible | 21:00 | |
masterpe | I have three controllers, when I run galera-install.yml --limit controller1* the max_connections of galera is set to 200 | 21:00 |
masterpe | The current connections to that galera node is more then that. | 21:01 |
admin0 | you need to override that value | 21:04 |
admin0 | masterpe, galera_max_connections: 6000 ( user_variables.yml) as an example | 21:05 |
masterpe | On other clusers with the same config, this value is set to 1600 | 21:06 |
admin0 | was giving an example | 21:06 |
admin0 | you can set it to what you think works for you | 21:06 |
admin0 | 200 i think is default which is a bit low for busy platforms | 21:06 |
masterpe | yes | 21:07 |
masterpe | it is it uses at least 900 | 21:07 |
*** elenalindq has quit IRC | 21:07 | |
masterpe | But I understand why is is calculated to 200 and not 1600 what I would expect. | 21:08 |
masterpe | Of cource the easyest sollution is to set galera_max_connections | 21:08 |
jrosser | masterpe: https://opendev.org/openstack/openstack-ansible-galera_server/src/branch/master/templates/my.cnf.j2#L56-L59 | 21:12 |
jrosser | seems to be vcpu * 100 if you don't specify otherwise, how many CPU do you have? | 21:12 |
masterpe | root@controller1-galera-container-a7cca5e1:~# cat /proc/cpuinfo |grep "processor"|wc -l | 21:14 |
masterpe | 40 | 21:14 |
jrosser | https://opendev.org/openstack/openstack-ansible-galera_server/src/branch/master/templates/my.cnf.j2#L1-L7 | 21:15 |
jrosser | whats suspicious there is that there is a default in that calculation of 2 | 21:15 |
jrosser | which perhaps suggests that vcpu is undefined (somehow) here https://opendev.org/openstack/openstack-ansible-galera_server/src/branch/master/templates/my.cnf.j2#L4 | 21:16 |
*** spatel has joined #openstack-ansible | 21:18 | |
* masterpe sent a long message: < https://matrix.org/_matrix/media/r0/download/matrix.org/uMSzxJnKhaVPTunezTkStWyl > | 21:19 | |
masterpe | on a test cluser: | 21:20 |
* masterpe sent a long message: < https://matrix.org/_matrix/media/r0/download/matrix.org/UHPrYCqphQLSXCJeiRTmqnUa > | 21:20 | |
jrosser | templates are hard to debug | 21:22 |
*** spatel has quit IRC | 21:22 | |
jrosser | adding some more lines, maybe trying to include another comment line with calculated_max_connections in it might help | 21:22 |
jrosser | and work back from that | 21:22 |
jrosser | does it behave properly without the --limit? | 21:23 |
masterpe | I made a test j2 template and loaded with: | 21:40 |
* masterpe sent a long message: < https://matrix.org/_matrix/media/r0/download/matrix.org/XgcEBXWsvarAAEKPmByOOMGk > | 21:40 | |
masterpe | and the j2 has the calculation and the output of the calculation | 21:41 |
masterpe | then I get as output 1600 | 21:41 |
jrosser | there’s something wierd with long messages from matrix.org | 22:01 |
jrosser | maybe paste | 22:01 |
jrosser | paste.openstack.org or similar for code snippets | 22:02 |
*** ahosam has joined #openstack-ansible | 22:04 | |
*** DanyC has joined #openstack-ansible | 22:07 | |
masterpe | Sorry | 22:08 |
*** ansmith has quit IRC | 22:09 | |
masterpe | It is working now, I still don't know why it was failing | 22:10 |
jrosser | it could be that facts had not been gathered | 22:10 |
jrosser | so the cpu count was somehow missing, but I might expect the template to totally fail in that case.... | 22:11 |
masterpe | maby | 22:11 |
*** ahosam has quit IRC | 22:15 | |
*** rh-jelabarre has quit IRC | 22:31 | |
*** ansmith has joined #openstack-ansible | 23:27 | |
*** nicolasbock has quit IRC | 23:36 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!