*** gyee has quit IRC | 00:10 | |
*** tmartins has joined #openstack-ansible | 00:19 | |
*** ThiagoCMC has quit IRC | 00:20 | |
*** ThiagoCMC has joined #openstack-ansible | 00:30 | |
*** tmartins has quit IRC | 00:31 | |
*** macz has joined #openstack-ansible | 00:44 | |
*** macz has quit IRC | 00:49 | |
*** hwoarang has quit IRC | 01:04 | |
*** hwoarang has joined #openstack-ansible | 01:06 | |
*** BjoernT has joined #openstack-ansible | 01:53 | |
*** BjoernT_ has joined #openstack-ansible | 01:56 | |
*** BjoernT has quit IRC | 01:58 | |
*** BjoernT_ has quit IRC | 02:38 | |
*** BjoernT has joined #openstack-ansible | 02:38 | |
*** BjoernT has quit IRC | 02:38 | |
*** macz has joined #openstack-ansible | 02:45 | |
*** spatel has joined #openstack-ansible | 02:49 | |
*** macz has quit IRC | 02:50 | |
*** spatel has quit IRC | 02:53 | |
*** tmartins has joined #openstack-ansible | 03:12 | |
*** ThiagoCMC has quit IRC | 03:14 | |
*** ThiagoCMC has joined #openstack-ansible | 03:34 | |
*** tmartins has quit IRC | 03:36 | |
*** hwoarang has quit IRC | 03:44 | |
*** hwoarang has joined #openstack-ansible | 03:45 | |
*** macz has joined #openstack-ansible | 04:46 | |
*** macz has quit IRC | 04:51 | |
*** markvoelker has joined #openstack-ansible | 04:52 | |
*** markvoelker has quit IRC | 04:58 | |
*** hwoarang has quit IRC | 06:34 | |
*** hwoarang has joined #openstack-ansible | 06:41 | |
*** arbrandes has quit IRC | 07:00 | |
*** arbrandes has joined #openstack-ansible | 07:01 | |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: [DNM] test ansible 2.9 https://review.opendev.org/681003 | 07:31 |
---|---|---|
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: [DNM] test ansible 2.9 https://review.opendev.org/681003 | 07:31 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible master: Imported Translations from Zanata https://review.opendev.org/688255 | 07:50 |
*** jawad_axd has joined #openstack-ansible | 08:19 | |
*** LONorell has joined #openstack-ansible | 08:51 | |
*** LONorell has quit IRC | 08:52 | |
*** markvoelker has joined #openstack-ansible | 08:55 | |
*** LONorell has joined #openstack-ansible | 08:56 | |
*** markvoelker has quit IRC | 09:00 | |
LONorell | Is it ok to ask for assistance in this channel? I am trying to set up an OS-Ansible AiO installation with no luck. | 09:37 |
jrosser | LONorell: its fine to ask, it'll be much more active here on weekdays though | 09:45 |
LONorell | Got some appointments now, but will get back later when I can focus on this. Let's hope there is someone here to help later :) | 09:48 |
jrosser | LONorell: i assume you are following this? https://docs.openstack.org/openstack-ansible/stein/user/aio/quickstart.html | 10:02 |
*** brtknr has quit IRC | 10:18 | |
*** brtknr has joined #openstack-ansible | 10:21 | |
*** hwoarang has quit IRC | 10:23 | |
*** hwoarang has joined #openstack-ansible | 10:29 | |
*** brtknr_ has joined #openstack-ansible | 10:32 | |
*** brtknr_ has quit IRC | 10:35 | |
*** brtknr_ has joined #openstack-ansible | 10:36 | |
*** ivve has joined #openstack-ansible | 10:56 | |
*** dave-mccowan has quit IRC | 11:55 | |
*** dave-mccowan has joined #openstack-ansible | 11:57 | |
*** ianychoi has joined #openstack-ansible | 11:59 | |
*** markvoelker has joined #openstack-ansible | 12:18 | |
*** markvoelker has quit IRC | 12:23 | |
*** radeks_ has joined #openstack-ansible | 13:01 | |
LONorell | @jrosser, yes, byt on the "latest" link instead of "stein" I believe my problems might be related to the recent update of centos7 and maybe I'm not using the latest code from the git repository. | 13:01 |
LONorell | I pasted the errors I got when trying to run setup-hosts playbook on https://etherpad.openstack.org/p/LONorell | 13:04 |
LONorell | the files referenced cannot be found, but there are files with similar names but different hashes (?) in that location. | 13:05 |
LONorell | Makes me think the files were updated with latest Centos 7 update, and the code I cloned is unaware of this change. | 13:06 |
*** brtknr has quit IRC | 13:09 | |
LONorell | That is the files yum try to download from http://download.fedoraproject.org/pub/epel/7/x86_64/repodata/ | 13:15 |
*** jawad_axd has quit IRC | 14:33 | |
*** pcaruana has joined #openstack-ansible | 15:25 | |
*** ivve has quit IRC | 15:35 | |
jrosser | LONorell: the OSA code is tested constantly against Centos 7.7 | 16:04 |
LONorell | @jrosser, as I'd expect, but still, it (the yum commands from the playbook?) references nonexistent files in an existing centos repo | 16:15 |
jrosser | can you link to the think in the platbook which you think is wrong? | 16:16 |
jrosser | *thing | 16:16 |
LONorell | the urls seem to redirect to mirrors, like http://mirror.nsc.liu.se/fedora-epel/7/x86_64/repodata/ | 16:16 |
jrosser | can i just check that you've not set up the EPEL repo before running the OSA install, or it wasnt there already somehow on the server? | 16:17 |
LONorell | I did not set up the EPEL repo before running the OSA installation. | 16:18 |
LONorell | What kind of link are you suggesting? I posted a link to a etherpad where I pasted the failing parts of my output. | 16:23 |
jrosser | you are suggesting the playbook references non existant files | 16:25 |
LONorell | The playbook makes yum try to dl files from an existing directory in the mirrored repository. | 16:33 |
jrosser | the way the EPEL repositories redirect and what they do with hashed names is really outside the scope of OSA | 16:34 |
jrosser | we add the repo, and use ansible to install some packages | 16:34 |
jrosser | so it's more question of if the mirrors are broken in some way, or is there an actual bug with the ansible that sets up and then uses the repos | 16:35 |
LONorell | I cen see the centos repo is out of scope for the project. But is there something I may have missed in my installation? Or will others get the same problem? | 16:35 |
LONorell | But right now I am considering trying to use Ubuntu as host for my installation instead, eventhough it hurts... :) | 16:36 |
LONorell | As far as I can understand, the OSA project is rooted in Ubuntu environment? Makes sense to use that to minimise problems? | 16:38 |
jrosser | every patch is tested on lots of OS, they are all supported, see here for an example https://review.opendev.org/#/c/681003/ | 16:40 |
jrosser | there are folk using centos fine | 16:40 |
jrosser | similarly ubuntu and others | 16:40 |
LONorell | Which makes me question my line of work :D | 16:40 |
LONorell | Why won't it work for me. Gotta be something I'm doing the wrong way. | 16:41 |
LONorell | Well, I'll try a change of environment. If it works better, good for me. If it don't, Ill be back here soon. | 16:42 |
jrosser | if you get stuck do ask | 16:43 |
LONorell | Thanks! | 16:43 |
jrosser | i have no answer for the centos repo thing other than wonder if the mirrors are broken/inconsistent | 16:43 |
ThiagoCMC | Hey guys! I'm trying to deploy OSA Stein, the etcd-install.yml is failing. | 16:44 |
ThiagoCMC | I'm actually surprised that the binary `/usr/local/bin/etcd` exists both inside of its container (vosctrl-1_etcd_container-fa410ee2), as well as inside of the bare-metal (vosctrl-1)!!! Is this expected? | 16:45 |
ThiagoCMC | The "TASK [systemd_service : Load service]" fail. | 16:46 |
ThiagoCMC | There is even etcd messages on /var/log/syslog of the bare metal! O_O | 16:46 |
ThiagoCMC | Like: "vosctrl-1 etcd[16306]: couldn't find local name "vosctrl-1" in the initial cluster configuration" | 16:47 |
ThiagoCMC | I wasn't expecting to see etcd binaries and log messages outside of its container. | 16:47 |
ThiagoCMC | I'm wondering about how to clean it up... :-/ | 16:48 |
ThiagoCMC | And how to proper deploy etcd (which seems to be a Magnum requirement, correct me if I'm wrong). | 16:48 |
ThiagoCMC | Here is the Ansible task error: https://pastebin.com/imgzQSQ2 | 16:49 |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible stable/stein: Update etcd playbook to target etcd from etcd_all https://review.opendev.org/688275 | 16:52 |
ThiagoCMC | I'm guessing that I'll have to clean up the inventory... | 16:52 |
jrosser | ThiagoCMC: ^ | 16:52 |
jrosser | and you don't need etcd deployed as part of your openstack infrastructure for magnum | 16:53 |
ThiagoCMC | Oh, really? | 16:53 |
ThiagoCMC | Thanks! | 16:53 |
jrosser | magnum creates it's own in the vm it builds | 16:53 |
ThiagoCMC | Hmmm... | 16:53 |
ThiagoCMC | Well, Magnum was deployed okay but, when I try to access the "Container Infra -> Clusters | 16:54 |
ThiagoCMC | ... on Horizon, the following error messages appear: | 16:54 |
ThiagoCMC | "Error: Unable to retrieve the clusters." and "Error: Unable to retrieve the stats." | 16:54 |
ThiagoCMC | So, I thought that Magnum was failing because my conf.d/etcd.yml wasn't there (it wasn't enabled). | 16:55 |
jrosser | it'll be a game of digging though the magnum logs, and possibly the heat lots too | 16:56 |
jrosser | *logs | 16:56 |
*** jawad_axd has joined #openstack-ansible | 16:57 | |
ThiagoCMC | ok | 16:57 |
ThiagoCMC | I must confess that I'm a bit lost with the logs after this systemd thing... hehe | 16:58 |
ThiagoCMC | Where are the logs? lol | 16:58 |
guilhermesp | ThiagoCMC: centos? | 16:58 |
ThiagoCMC | Ubuntu 18.04 | 16:58 |
ThiagoCMC | I do have rsyslog container | 16:58 |
guilhermesp | try journalctl -u magnum-{api,conductor} -f | 16:58 |
guilhermesp | of journalctl -u magnum-{api,conductor} --since "" if you want to travel back in time | 16:59 |
ThiagoCMC | Nice! Thanks! | 16:59 |
guilhermesp | horizon logs are too general... I'd see also what --debug with cli can tell me | 17:00 |
guilhermesp | but best way is to be watching those logs | 17:00 |
*** jawad_axd has quit IRC | 17:02 | |
ThiagoCMC | AHA! Got the Magnum error: https://pastebin.com/zmkVAKPM | 17:03 |
ThiagoCMC | Creepy 404 lol | 17:03 |
ThiagoCMC | On magnum-api | 17:04 |
ThiagoCMC | logs... | 17:04 |
guilhermesp | what about conductor? | 17:05 |
ThiagoCMC | conductor doesn't show any error, it's just running. | 17:07 |
guilhermesp | journalctl -u mangum-conductor --since "30 minutes ago" | grep Error | 17:07 |
guilhermesp | or error, or ERROR, well do'nt remember now :P | 17:08 |
ThiagoCMC | ^_^ | 17:08 |
ThiagoCMC | Nothing... =P | 17:08 |
guilhermesp | heat logs? | 17:08 |
guilhermesp | well, you could try restart the services ( if you are not in a prod env ), try to list clusters from cli, or even create clusters templates | 17:09 |
guilhermesp | and watch the logs | 17:09 |
guilhermesp | open 2 terminals, one for api and one for conductor | 17:09 |
ThiagoCMC | Yep, I did that with tmux | 17:09 |
ThiagoCMC | Heat looks fine | 17:10 |
jrosser | watch out for multiple controllers too - the log entries end up randomly around them | 17:10 |
guilhermesp | if you have them under haproxy, let only one instance to run, so you're gonna have centralized logs | 17:10 |
guilhermesp | yeah | 17:10 |
guilhermesp | what jrosser just said | 17:10 |
ThiagoCMC | Yeah, I have 3 controllers, I stopped 2. | 17:10 |
ThiagoCMC | lol | 17:10 |
jrosser | and the conductor is RPC so i've tended to just shut down 2 magnum containers | 17:10 |
ThiagoCMC | Got it... | 17:11 |
ThiagoCMC | Guys, I appreciate the help! Gonna go eat something... ^_^ | 17:11 |
ThiagoCMC | See ya! | 17:11 |
*** markvoelker has joined #openstack-ansible | 17:15 | |
*** jawad_axd has joined #openstack-ansible | 17:38 | |
*** jawad_axd has quit IRC | 17:43 | |
*** pcaruana has quit IRC | 17:43 | |
*** jawad_axd has joined #openstack-ansible | 17:59 | |
*** jawad_axd has quit IRC | 18:04 | |
*** redrobot has quit IRC | 18:11 | |
*** markvoelker has quit IRC | 18:50 | |
LONorell | Looking for advice in preparing aio install on ubuntu. for starters some git. | 19:33 |
LONorell | Following guide on https://docs.openstack.org/openstack-ansible/latest/user/aio/quickstart.html#building-an-aio | 19:34 |
LONorell | I have cloned the openstack-ansible repo, but after that? | 19:35 |
LONorell | If I run git describe --abbrev=0 --tags I get 19.0.0.0rc1 while git tag -l shows tags up to 19.0.4 | 19:37 |
LONorell | Does that mean I am using an old version of the stein installation? | 19:38 |
LONorell | And, if so, how do I change to 19.0.4? | 19:38 |
*** dmsimard has quit IRC | 19:55 | |
ThiagoCMC | I'm facing the following problem with Magnum (OSA Stein on Ubuntu 18.04): https://ask.openstack.org/en/question/117228/unable-to-use-magnum-on-packstack/ I tried the packstack changes on magnum.conf and it worked! So, apparently, OSA Stein magnum.conf.j2 is broken. | 20:17 |
ThiagoCMC | The magnum.conf needs the "www_authenticate_uri with /v3" on its end. | 20:19 |
ThiagoCMC | Now, how to launch a Kubernetes Cluster in 5 minutes? LOL | 20:19 |
jrosser | LONorell: the git describe line you use there returns you the most recent tag | 20:20 |
jrosser | it's important to understand the the master branch is the work-in-progress for whatever the *next* release of openstack will be | 20:21 |
jrosser | so if you want to deploy stein, checkout the most recent 19.x tag, or stable/stein is the head of the branch which will become the next tagged stein point release | 20:22 |
jrosser | LONorell: to build a stein AIO do the following | 20:25 |
jrosser | clone the openstack-ansible repo, run scripts/bootstrap-ansible.sh, run scripts/bootstrap-aio.sh | 20:25 |
jrosser | then move on to the 'run playbooks' section, it should be no more complicated than that | 20:26 |
LONorell | @jrosser, ok. No special setup necessary, then. Weird that it always has some kind of problems for me, then. | 20:27 |
*** markvoelker has joined #openstack-ansible | 20:28 | |
jrosser | the quickstart guide does detail quite a lot of possible customisations / options | 20:28 |
LONorell | Just dled ubuntu 18.04.3 server and installed on one partition 240 GB, and just added ssh server options. Ran update/distupgrade and cloned the repo. I'll run the scripts now. | 20:29 |
jrosser | and you really don't need them for a very standard AIO setup | 20:29 |
jrosser | it's good to try to get your test environment somehow disposable / easily repeatable | 20:30 |
ThiagoCMC | Is there a ste-by-step guide to deploy Kubernetes in 5 minutes with Stein/Magnum? https://thenewstack.io/openstack-stein-focuses-on-kubernetes-bare-metal-nfv-and-more/ | 20:33 |
ThiagoCMC | I was in Denver at the summit and this was super cool to watch! But now, I wanna do it. ^_^ | 20:33 |
LONorell | Currently running script. Getting some warnings on missing .ymls, probably just missing customization files | 20:34 |
ThiagoCMC | I honestly don't remember if what I saw, was exactly this (the 5 min Kubernetes cluster with Magnum)... | 20:34 |
*** markvoelker has quit IRC | 20:38 | |
LONorell | So, the bootstrap-ansible.sh script fails when trying to git clone openstack-ansible-os_tempest. see partial output at https://etherpad.openstack.org/p/LONorell | 20:45 |
LONorell | Also errors over lack of btrfs in /var/lib/lxc et al | 20:47 |
LONorell | With what looks like an git error message "fatal: reference is not a tree:", I don't know if there is anything I might have done wrong to cause it. My network setup is simply the server inside my nat-gateway, using DHCP IPv4. | 20:49 |
jrosser | LONorell: you are still running from the master branch of openstack-ansible though? | 20:55 |
jrosser | if you want a deployment of stein you much checkout the stein tag, or branch of openstack-ansible | 20:56 |
jrosser | the openstack-ansible repo contains the git SHA of all the ansible roles, and all the openstack services that will be used | 20:56 |
ThiagoCMC | I'm curious, what in OSA is actually using etcd? | 20:57 |
jrosser | calico | 20:57 |
ThiagoCMC | Is it enabled by default on Neutron playbooks? | 20:59 |
ThiagoCMC | Because I don't have a proper etcd right now and Neutron seens to be working fine. | 20:59 |
ThiagoCMC | Oh, I can see that it isn't enabled by default. | 21:00 |
LONorell | So in addition to "https://opendev.org/openstack/openstack-ansible" I should use "git checkout stein"? | 21:01 |
ThiagoCMC | I'm actually more interested in Networking OVN to scaly my Neutron... Calico seems very similar with current Neutron Agents. | 21:01 |
ThiagoCMC | Thanks for the info! | 21:02 |
jrosser | ThiagoCMC: here is where etcd gets enables in CI jobs for calico scenarios https://github.com/openstack/openstack-ansible/blob/master/tests/roles/bootstrap-host/vars/main.yml#L36-L38 | 21:02 |
jrosser | LONorell: the openstack-ansible repo contains branches and tags for several releases of openstack, stein, rocky, queens and so on | 21:03 |
ThiagoCMC | Got it! =) | 21:03 |
jrosser | you must check out the branch or tag you need corresponding to the version of openstack you want to deploy | 21:03 |
jrosser | so to deploy stein, you want the stable/stein branch | 21:04 |
jrosser | if you want to test bleeding edge or do development work on the next release, use master <- you probably don't want to do this | 21:05 |
LONorell | Ok, that information is sorely missing in the instruction :( Or is it me not taking in that info? | 21:08 |
jrosser | i did point you to the stein branch documentation earlier https://docs.openstack.org/openstack-ansible/stein/user/aio/quickstart.html | 21:09 |
jrosser | where it tells you to checkout stable/stein | 21:09 |
jrosser | the documentation is also branched just like the rest of the code | 21:10 |
*** tbarron_ has joined #openstack-ansible | 21:10 | |
LONorell | You sure did that. My bad (What Samuel L Jackson said about assumptions). | 21:13 |
jrosser | git branch -r | 21:13 |
jrosser | ^ that will show you the available branches | 21:13 |
jrosser | about every two weeks or so there is a "minor release", where a new tag on each of those branches is created | 21:13 |
jrosser | so for stein thats where the 19.x.y tags come from | 21:14 |
jrosser | so if you have a real production deployment you can upgrade it incrementally from one tag to the next, on a particular branch | 21:14 |
*** radeks_ has quit IRC | 21:17 | |
LONorell | Thanks, @jrosser I hope I will get further with less problems now. There are a lot of things for me to take in in a short time now. Have to turn in a report on my openstack endeavors during next week. | 21:18 |
LONorell | Planned to be further along by now :D | 21:18 |
LONorell | I'd be happy to credit you in my report, @jrosser | 21:19 |
jrosser | if you stick to the stock AIO config, you should get a deployment done in a hour or two | 21:20 |
ThiagoCMC | This freaking systemd!!! When I run `apt update ; apt full-upgrade` in my HAProxy node, and systemd is in the update list for APT, it kills the VIPs and Keepalived doesn't bring them back! Why the heck systemd is in production?! | 21:24 |
ThiagoCMC | Now, after ~20 years trusting in Debian-style system upgrades, because of systemd, I don't have the same level of trust in my systems anymore. This is nuts. | 21:27 |
ThiagoCMC | :-/ | 21:27 |
ThiagoCMC | https://github.com/acassen/keepalived/issues/836 :-@ | 21:32 |
ThiagoCMC | Daammnn... https://bugs.launchpad.net/netplan/+bug/1815101 - OSA is affected by this as well. | 21:38 |
openstack | Launchpad bug 1815101 in systemd (Ubuntu Eoan) "[master] Restarting systemd-networkd breaks keepalived, heartbeat, corosync, pacemaker (interface aliases are restarted)" [Medium,In progress] - Assigned to Rafael David Tinoco (rafaeldtinoco) | 21:38 |
jrosser | you can still use ifupdown | 21:42 |
ThiagoCMC | Sure... But I got rid of ifupdown a long time ago and my Netplan files are really nice and kinda complex already... | 21:46 |
ThiagoCMC | It's good to know why the heck my VIPs were disapearing, at least now I know what to do and when. It looks like that the Bionic backport is on its way. | 21:47 |
*** dave-mccowan has quit IRC | 22:07 | |
*** markvoelker has joined #openstack-ansible | 22:39 | |
*** markvoelker has quit IRC | 22:44 | |
ThiagoCMC | Where can I get a image for Magnum? So I can create a CCluster Template...? | 23:14 |
ThiagoCMC | I do have Ubuntu 18.04 and CentOS 7 images but, they don't show up under Container Infra menu... | 23:14 |
*** dave-mccowan has joined #openstack-ansible | 23:14 | |
*** dmsimard has joined #openstack-ansible | 23:32 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!