| *** admin0 has quit IRC | 00:11 | |
| *** cjloader has joined #openstack-ansible | 00:12 | |
| *** cjloader has quit IRC | 00:55 | |
| *** cjloader has joined #openstack-ansible | 00:55 | |
| *** cjloader has quit IRC | 00:55 | |
| *** gregoryo has joined #openstack-ansible | 00:55 | |
| *** cjloader has joined #openstack-ansible | 00:56 | |
| *** cjloader has quit IRC | 01:04 | |
| openstackgerrit | weizj proposed openstack/openstack-ansible master: Remove the unnecessary space https://review.openstack.org/582814 | 01:09 |
|---|---|---|
| *** tux_ has joined #openstack-ansible | 01:09 | |
| openstackgerrit | weizj proposed openstack/openstack-ansible-galera_server master: Remove the unnecessary space https://review.openstack.org/582816 | 01:24 |
| openstackgerrit | weizj proposed openstack/openstack-ansible-galera_server master: Remove the unnecessary space https://review.openstack.org/582817 | 01:25 |
| openstackgerrit | weizj proposed openstack/openstack-ansible-os_tacker master: Remove the unnecessary space https://review.openstack.org/582818 | 01:33 |
| *** vnogin has joined #openstack-ansible | 01:54 | |
| *** vnogin has quit IRC | 01:59 | |
| *** poul has quit IRC | 03:05 | |
| *** poul has joined #openstack-ansible | 03:10 | |
| *** poul has quit IRC | 03:15 | |
| *** poul_ has joined #openstack-ansible | 03:17 | |
| *** nurdie has joined #openstack-ansible | 03:18 | |
| *** idlemind has quit IRC | 03:20 | |
| *** armaan has quit IRC | 03:20 | |
| *** poul_ has quit IRC | 03:21 | |
| *** nurdie has quit IRC | 03:23 | |
| *** tux_ has quit IRC | 03:26 | |
| *** poul has joined #openstack-ansible | 03:27 | |
| *** poul has quit IRC | 03:32 | |
| *** nurdie has joined #openstack-ansible | 03:39 | |
| *** nurdie has quit IRC | 03:42 | |
| *** poul has joined #openstack-ansible | 03:46 | |
| *** poul_ has joined #openstack-ansible | 03:48 | |
| *** poul has quit IRC | 03:48 | |
| *** TxGirlGeek has joined #openstack-ansible | 03:49 | |
| *** poul_ has quit IRC | 03:52 | |
| *** TxGirlGeek has quit IRC | 03:56 | |
| *** poul has joined #openstack-ansible | 03:58 | |
| *** poul has quit IRC | 04:03 | |
| *** poul has joined #openstack-ansible | 04:08 | |
| *** Bhujay has joined #openstack-ansible | 04:12 | |
| *** poul has quit IRC | 04:13 | |
| *** Bhujay has quit IRC | 04:13 | |
| *** Bhujay has joined #openstack-ansible | 04:14 | |
| *** Bhujay has quit IRC | 04:16 | |
| *** udesale has joined #openstack-ansible | 04:16 | |
| *** poul has joined #openstack-ansible | 04:19 | |
| *** nurdie has joined #openstack-ansible | 04:20 | |
| *** poul has quit IRC | 04:24 | |
| *** poul has joined #openstack-ansible | 04:39 | |
| *** poul has quit IRC | 04:43 | |
| *** nurdie has quit IRC | 04:45 | |
| *** nurdie has joined #openstack-ansible | 04:46 | |
| *** poul has joined #openstack-ansible | 04:49 | |
| *** nurdie has quit IRC | 04:50 | |
| *** poul has quit IRC | 04:54 | |
| *** poul has joined #openstack-ansible | 05:00 | |
| *** poul has quit IRC | 05:04 | |
| *** poul has joined #openstack-ansible | 05:10 | |
| *** poul has quit IRC | 05:15 | |
| *** poul has joined #openstack-ansible | 05:20 | |
| *** poul has quit IRC | 05:25 | |
| *** poul has joined #openstack-ansible | 05:31 | |
| *** poul has quit IRC | 05:36 | |
| *** fdsa has quit IRC | 05:40 | |
| *** poul has joined #openstack-ansible | 05:41 | |
| *** poul has quit IRC | 05:45 | |
| *** vnogin has joined #openstack-ansible | 05:51 | |
| *** poul has joined #openstack-ansible | 05:52 | |
| *** vnogin has quit IRC | 05:52 | |
| *** poul has quit IRC | 05:56 | |
| *** vnogin has joined #openstack-ansible | 05:57 | |
| *** poul has joined #openstack-ansible | 06:02 | |
| *** tomtom002 has quit IRC | 06:05 | |
| *** tomtom001 has joined #openstack-ansible | 06:05 | |
| *** poul has quit IRC | 06:06 | |
| *** poul has joined #openstack-ansible | 06:12 | |
| *** poul has quit IRC | 06:17 | |
| *** poul_ has joined #openstack-ansible | 06:23 | |
| *** vnogin has quit IRC | 06:24 | |
| *** poul_ has quit IRC | 06:27 | |
| *** poul has joined #openstack-ansible | 06:33 | |
| *** pcaruana has joined #openstack-ansible | 06:36 | |
| *** poul has quit IRC | 06:38 | |
| ioni | is there a way to only deploy l3 agent on a specific host ? i tried placing in neutron_l3_agent group the host but didn't install any l3 agent | 06:39 |
| *** cshen has joined #openstack-ansible | 06:51 | |
| *** nurdie has joined #openstack-ansible | 06:53 | |
| evrardjp | ansmith: I think there is something wrong with our oslomsg story. We defined rpc|notify users/vhosts in group vars, the password is in secrets. We generate different password for rpc_password and notify_password. | 07:00 |
| evrardjp | 2 problems: 1) the username and vhost for oslomsg_rpc|notify are the same -- which makes the tuple userid,password,vhost not unique (as passwords for rpc/notify are different but the rest is the same) | 07:02 |
| evrardjp | 2) We don't migrate from old var _rabbitmq_ to _oslomsg_ which will kill the upgrades. | 07:02 |
| evrardjp | ioni: interesting | 07:03 |
| evrardjp | ioni: which branch? | 07:03 |
| evrardjp | for Queens and above you can have your own static inventory in /etc/openstack_deploy/inventory.ini to make things simple : ) | 07:03 |
| ioni | evrardjp i'm on pike | 07:07 |
| evrardjp | yeah so dynamic inventory it is. | 07:07 |
| evrardjp | use env.d | 07:07 |
| evrardjp | check your /etc/openstack_deploy/openstack_inventory.json results | 07:07 |
| *** sandipan has joined #openstack-ansible | 07:10 | |
| *** poul has joined #openstack-ansible | 07:10 | |
| sandipan | Hi ! I have a query | 07:10 |
| ioni | evrardjp well if i use the script to list hosts and groups, it's not listed in l3 group | 07:11 |
| sandipan | can default nova-compute setup with kvm & qemu co exists with nova-ironic implementation ? | 07:12 |
| *** poul_ has joined #openstack-ansible | 07:13 | |
| sandipan | as per the latest document -- https://docs.openstack.org/openstack-ansible-os_ironic/latest/configure-ironic.html#openstack-ansible-deployment " it is not possible | 07:13 |
| evrardjp | ioni: check the json to make 100% sure. That inventory manage is not 100% reliable to my tastes. | 07:16 |
| evrardjp | But it's generally a good hint that something doesn't work as expected. | 07:16 |
| *** poul has quit IRC | 07:16 | |
| evrardjp | sandipan: good question, I don't know | 07:17 |
| evrardjp | sandipan: although what you linked is not explaining that | 07:18 |
| evrardjp | sandipan: what I can tell you is that many ppl around here are running a cloud with nova and ironic. | 07:19 |
| evrardjp | nova + kvm && nova + ironic | 07:19 |
| sandipan | thanks evrardjp , Yes I also want to do the same | 07:20 |
| sandipan | but not getting any document on that | 07:20 |
| evrardjp | This doc seem outdated | 07:20 |
| sandipan | Maybe you are right | 07:20 |
| evrardjp | I'd love to have a user story around it for your use case | 07:21 |
| *** epalper has joined #openstack-ansible | 07:21 | |
| evrardjp | kvm + ironic | 07:21 |
| evrardjp | giving an example on how to do it | 07:21 |
| evrardjp | would you mind helping? | 07:21 |
| sandipan | no not at all | 07:21 |
| sandipan | but I have to set it first | 07:21 |
| evrardjp | ofc | 07:21 |
| evrardjp | we'll find you help | 07:22 |
| evrardjp | on this chan | 07:22 |
| evrardjp | just not at this hour : ) | 07:22 |
| sandipan | thanks , any help is welcome | 07:22 |
| evrardjp | maybe come later in the day, see if you can catch cloudnull or prometheanfire for example. | 07:22 |
| prometheanfire | wat | 07:22 |
| sandipan | I can wait for some time evrardjp | 07:22 |
| evrardjp | prometheanfire: go to bed, it's too late for ironic questions now : p | 07:23 |
| prometheanfire | it's always too late | 07:23 |
| evrardjp | prometheanfire: omg why are you still on this channel :) It's like 3 AM for you? | 07:23 |
| prometheanfire | 2:23 | 07:23 |
| evrardjp | hahah | 07:23 |
| *** admin0 has joined #openstack-ansible | 07:26 | |
| evrardjp | odyssey4me: are you there? | 07:30 |
| hwoarang | good morning | 07:30 |
| evrardjp | or hwoarang that works too | 07:30 |
| evrardjp | anyone to bounce ideas :) | 07:30 |
| evrardjp | hwoarang: you have 2 minutes? | 07:31 |
| hwoarang | depends | 07:31 |
| evrardjp | hwoarang: if true, have a look at this: https://github.com/openstack/openstack-ansible/blob/stable/queens/inventory/group_vars/nova_all.yml#L33-L42 | 07:31 |
| evrardjp | that's queens | 07:31 |
| evrardjp | no oslomsg work | 07:31 |
| evrardjp | that's the group vars | 07:31 |
| evrardjp | as you can see | 07:31 |
| evrardjp | we have userid nova, a password generated from secrets in nova_rabbitmq_password unless it's overriden by a user. | 07:32 |
| evrardjp | nowadays we have this: | 07:32 |
| evrardjp | https://github.com/openstack/openstack-ansible/blob/master/inventory/group_vars/all/oslo-messaging.yml#L250-L251 | 07:32 |
| evrardjp | https://github.com/openstack/openstack-ansible/blob/master/inventory/group_vars/all/oslo-messaging.yml#L257-L258 | 07:32 |
| evrardjp | no wiring of the password, so we are using the generated password in : | 07:33 |
| evrardjp | https://github.com/openstack/openstack-ansible/blob/master/etc/openstack_deploy/user_secrets.yml#L115-L116 | 07:33 |
| evrardjp | which WILL be different for rpc and notify | 07:33 |
| evrardjp | but the user is the same so we break stuff. | 07:33 |
| evrardjp | NOW | 07:33 |
| evrardjp | what do you think we should do. IMO we should rollback to have one pair of secrets: nova_rabbitmq_password and refer to it. If someone wants its own separation, he should generate the appropriate user/password | 07:35 |
| evrardjp | what's your opinion? | 07:35 |
| evrardjp | alternatively, we can by default have a different username for rpc and notify | 07:35 |
| *** hachi has joined #openstack-ansible | 07:36 | |
| evrardjp | tell me what you prefer and I will apply it (think about upgrades too) | 07:36 |
| *** aludwar has quit IRC | 07:38 | |
| *** lvdombrkr has joined #openstack-ansible | 07:38 | |
| hwoarang | maybe first option | 07:38 |
| evrardjp | ok. | 07:40 |
| evrardjp | thanks | 07:40 |
| *** tosky has joined #openstack-ansible | 07:40 | |
| *** DanyC has joined #openstack-ansible | 07:41 | |
| *** aludwar has joined #openstack-ansible | 07:43 | |
| *** nurdie has quit IRC | 07:44 | |
| *** rpittau has joined #openstack-ansible | 07:51 | |
| *** shardy has joined #openstack-ansible | 07:54 | |
| *** dmsimard has quit IRC | 07:54 | |
| *** dmsimard has joined #openstack-ansible | 07:55 | |
| *** shardy has quit IRC | 07:55 | |
| *** shardy has joined #openstack-ansible | 07:57 | |
| openstackgerrit | Hugh Saunders proposed openstack/openstack-ansible-ops master: Revert "Make space for Cinder/Swift nodes if using nspawn" https://review.openstack.org/582884 | 08:00 |
| evrardjp | hello hughsaunders | 08:01 |
| jrosser | hughsaunders: there might already be a fix for that? https://review.openstack.org/#/c/582808/ | 08:04 |
| evrardjp | good morning odyssey4me | 08:09 |
| evrardjp | :) | 08:09 |
| evrardjp | and jrosser | 08:09 |
| jrosser | o/ | 08:09 |
| odyssey4me | o/ evrardjp - just arrived in Chichester to visit hughsaunders for the day ;) | 08:09 |
| evrardjp | hahaha | 08:09 |
| evrardjp | is that true? | 08:09 |
| evrardjp | it looks you are in hughsaunders land | 08:09 |
| evrardjp | not sure if it's only figuratively | 08:10 |
| *** cshen has quit IRC | 08:10 | |
| evrardjp | i am in sed land, and sed is winning. | 08:10 |
| openstackgerrit | Merged openstack/openstack-ansible-ops master: mnaio: Only resize Swift & Cinder machines00 LV when using nspawn https://review.openstack.org/582808 | 08:22 |
| *** czunker has joined #openstack-ansible | 08:22 | |
| *** tux_ has joined #openstack-ansible | 08:27 | |
| *** lirt has joined #openstack-ansible | 08:28 | |
| *** tux_ has quit IRC | 08:32 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible master: Prevent incorrect credentials https://review.openstack.org/582891 | 08:32 |
| evrardjp | odyssey4me: hwoarang critical fix here: https://review.openstack.org/#/c/582891/1 | 08:33 |
| evrardjp | (else nova has a few errors in its logs and have trouble spawning up) | 08:33 |
| ioni | evrardjp it doesn't have the host in neutron_l3_agent in json inventory | 08:38 |
| evrardjp | ioni: that's the issue :p | 08:38 |
| evrardjp | how did you modify things? | 08:39 |
| ioni | evrardjp i just added neutron_l3_agent: in openstack_user_config.yml | 08:39 |
| ioni | so i have | 08:40 |
| ioni | https://paste.xinu.at/aOcs/ | 08:40 |
| evrardjp | ioni: that's the issue | 08:41 |
| evrardjp | what are you trying to do? | 08:41 |
| ioni | do use another host only to install l3 agent for neutron | 08:41 |
| ioni | without dhcp or other agents | 08:41 |
| evrardjp | ok so you want to isolate l3 agent | 08:42 |
| ioni | yes | 08:42 |
| evrardjp | this is an example for extra containers: https://github.com/openstack/openstack-ansible/blob/master/etc/openstack_deploy/env.d/extra_container.yml.example | 08:42 |
| evrardjp | and this is neutron definitions | 08:43 |
| evrardjp | https://github.com/openstack/openstack-ansible/blob/master/inventory/env.d/neutron.yml | 08:43 |
| evrardjp | you should really make sure you're trying on a dev env | 08:43 |
| evrardjp | before trying on prod | 08:43 |
| *** ssbarnea1 has joined #openstack-ansible | 08:43 | |
| evrardjp | that is _by far_ easier since I brought the static inventories in OSA in Queens. | 08:44 |
| ioni | alright, it seems simpler to just deploy everything in network_hosts and disable the services i don't want on specific hosts from openstack client | 08:45 |
| *** ssbarnea1 has joined #openstack-ansible | 08:45 | |
| ioni | i should really upgrade to queens | 08:46 |
| evrardjp | not sure what you're trying to achieve at the end | 08:46 |
| ioni | what's the status of queens? | 08:46 |
| ioni | production ready? | 08:46 |
| evrardjp | well it's a stable branch | 08:46 |
| evrardjp | :p | 08:46 |
| ioni | yep | 08:46 |
| evrardjp | we try to not break stuff whatever the branch :) | 08:46 |
| ioni | i do have my dev on queens but i had some issues with some containers | 08:47 |
| evrardjp | ofc master is a moving target, while stable branches aren't | 08:47 |
| evrardjp | queens is different -- less containers | 08:47 |
| evrardjp | try the upgrade in a lab before doing the upgrade | 08:47 |
| evrardjp | check the current upgrade status in gates | 08:47 |
| evrardjp | and fix bugs :) | 08:47 |
| evrardjp | if any | 08:47 |
| evrardjp | that's my approach at least | 08:47 |
| evrardjp | but from a clean greenfield deploy everything is fine | 08:48 |
| ioni | what i'm trying to achieve is more hosts with l3 | 08:48 |
| evrardjp | ok -- network_hosts doesn't do the trick? | 08:48 |
| ioni | it does deploy too many agents :) | 08:49 |
| evrardjp | that's how I started to scale, and it was just fine for me | 08:49 |
| evrardjp | I had only two network hosts though | 08:49 |
| evrardjp | for starters | 08:49 |
| evrardjp | then moved to 3 | 08:49 |
| ioni | do you also have distributed routers? | 08:49 |
| evrardjp | nope | 08:49 |
| evrardjp | I had them HA | 08:49 |
| evrardjp | which was good enough | 08:50 |
| ioni | ok | 08:50 |
| evrardjp | but my network nodes had very good nics. | 08:50 |
| ioni | 10G or 40G? | 08:50 |
| evrardjp | 100 | 08:50 |
| ioni | nice | 08:50 |
| evrardjp | I was in a telco too. | 08:51 |
| evrardjp | :p | 08:51 |
| evrardjp | changes your vision of BW/jitter. | 08:51 |
| evrardjp | anyway, it was good enough for most of our openstack use cases | 08:51 |
| evrardjp | how is distributed routers helping you ? | 08:52 |
| evrardjp | (I suppose you're using OVS) | 08:52 |
| *** gregoryo has quit IRC | 08:53 | |
| ioni | evrardjp i don't use OVS, maybe i was using the wrong term, but i want redundancy for routers, maybe with VRRP | 08:59 |
| evrardjp | ioni: l3ha | 08:59 |
| evrardjp | so what's the issue by adding more network nodes? | 08:59 |
| *** aludwar has quit IRC | 08:59 | |
| evrardjp | what are the agents you find "too much" ? | 09:00 |
| *** armaan has joined #openstack-ansible | 09:00 | |
| ioni | dhcp i prefer to keep only on specific hosts | 09:00 |
| ioni | and i don't know what happens with i introduce more dhcp agents | 09:00 |
| evrardjp | it will distribute things better per tenant? | 09:01 |
| evrardjp | it will distribute tenants/networks better* ? | 09:01 |
| evrardjp | I'd do so. | 09:01 |
| evrardjp | if you want more scale, just throw more network-nodes on the issue. It generally works :p | 09:02 |
| ioni | yes, that's what i'm trying to achive, to distribute better the traffic for l3 | 09:02 |
| evrardjp | (depends on what "scale" means) | 09:02 |
| jrosser | dedicated network nodes seems to be working nicely for me | 09:02 |
| evrardjp | voila! | 09:02 |
| jrosser | and a plan to add more as i need then | 09:02 |
| evrardjp | yeah and it's easy to manage | 09:02 |
| evrardjp | because you understand easily what they do | 09:02 |
| evrardjp | :) | 09:02 |
| ioni | ok | 09:02 |
| *** armaan_ has joined #openstack-ansible | 09:02 | |
| evrardjp | if you try to disable things ioni , it will get messy | 09:02 |
| evrardjp | for your ops | 09:02 |
| evrardjp | "did I disable this on host x or host y?" | 09:03 |
| jrosser | and if you care about such things it keeps your external connectivity in a very well defined place | 09:03 |
| evrardjp | "when we do upgrades, do we need to manage x or y" | 09:03 |
| evrardjp | "if I disable x, what's the VRRP priority's state if y dies " | 09:03 |
| evrardjp | that's more complex | 09:03 |
| ioni | right | 09:03 |
| evrardjp | jrosser: I did the same. network nodes being super simple to figure out. | 09:04 |
| evrardjp | and easy to monitor | 09:04 |
| *** aludwar has joined #openstack-ansible | 09:04 | |
| jrosser | evrardjp: agree - theres quite a lot going on with all the bridges, namespaces and whatnot that gets created having a dedicated node is really a plus | 09:05 |
| evrardjp | ioni: We probably don't want to influence you -- but keep in mind our code was written in a way to help deployers generally -- it makes sense to do so :) | 09:05 |
| evrardjp | jrosser: yeah and easier to scale, as "they are their own stateful thing" | 09:05 |
| *** armaan has quit IRC | 09:05 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible master: Prevent incorrect credentials https://review.openstack.org/582891 | 09:10 |
| *** cronix has quit IRC | 09:10 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_tempest master: Trivial: Fix the pep8 warning https://review.openstack.org/576789 | 09:13 |
| *** electrofelix has joined #openstack-ansible | 09:14 | |
| czunker | Hi there | 09:31 |
| czunker | we used openstack-ansible to install gnocchi with ceph backend and redis as incoming storage. That needed some changes to osa variables. Would that be worth a commit to the documentation? | 09:31 |
| evrardjp | czunker: yes. A user story. | 09:33 |
| evrardjp | you can wire it here: https://github.com/openstack/openstack-ansible/blob/master/doc/source/user/index.rst | 09:34 |
| evrardjp | and create a new folder with your story | 09:34 |
| czunker | evrardjp: Will do that. Thanks for the hint | 09:35 |
| evrardjp | czunker: "Deploying with telemetry + redis and ceph example" | 09:36 |
| evrardjp | :) | 09:36 |
| evrardjp | You can probably add a note into https://docs.openstack.org/openstack-ansible/latest/user/ceph/full-deploy.html to point to your specific story. | 09:36 |
| evrardjp | czunker: or extend that one, if you think they are close in terms of architecture | 09:37 |
| evrardjp | But I think it's probably worth having another story to focus on your points | 09:37 |
| evrardjp | And explain how and why it differs from first point | 09:38 |
| czunker | evrardjp: Main focus are the gnocchi tweaks to use the redis cluster. We use ceph-ansible and not the ceph integration auf osa, so nothing I can contribute there. | 09:40 |
| evrardjp | czunker: mmm I'd like to have the ceph story split in two: One to deploy with ceph-ansible inside OSA, and the other is to use the external ceph cluster (whatever the method is used) | 09:41 |
| evrardjp | I think your story is unique due to redis | 09:41 |
| evrardjp | if you see what I mean | 09:41 |
| evrardjp | that's the part of your user story you should explain, the rest should just point to it | 09:42 |
| evrardjp | so maybe "Telemetry with Gnocchi and Redis example" ? | 09:42 |
| evrardjp | anyway | 09:42 |
| evrardjp | we'll discuss in the patch : ) | 09:42 |
| evrardjp | Just giving ideas :D | 09:42 |
| czunker | evrardjp: Ideas are always welcome | 09:43 |
| evrardjp | I want to avoid the repetitions if possible :) | 09:43 |
| odyssey4me | interesting ansible 2.7 change wrt import/include role: https://github.com/ansible/ansible/pull/41330/files | 09:44 |
| evrardjp | odyssey4me: that's gonna be a nightmare | 09:46 |
| evrardjp | :) | 09:47 |
| evrardjp | but "interesting" is the right word | 09:48 |
| *** armaan_ has quit IRC | 09:53 | |
| *** armaan has joined #openstack-ansible | 09:58 | |
| *** electrofelix has quit IRC | 09:59 | |
| *** vnogin has joined #openstack-ansible | 10:09 | |
| *** electrofelix has joined #openstack-ansible | 10:11 | |
| *** prasen has joined #openstack-ansible | 10:12 | |
| *** czunker has left #openstack-ansible | 10:13 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-ceph_client master: Fix usage of "|" for tests https://review.openstack.org/582226 | 10:20 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-rabbitmq_server master: Fix usage of "|" for tests https://review.openstack.org/582222 | 10:20 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-galera_client master: Fix usage of "|" for tests https://review.openstack.org/582213 | 10:20 |
| evrardjp | odyssey4me: oh great, look at the failure of https://review.openstack.org/#/c/582246/ | 10:23 |
| evrardjp | for xenial | 10:23 |
| evrardjp | a nice rsync permission issue now. | 10:23 |
| evrardjp | I suppose the latest change is problematic | 10:23 |
| *** vnogin has quit IRC | 10:23 | |
| evrardjp | I will manually rebase see if everything is alright, just in case. | 10:23 |
| *** udesale has quit IRC | 10:24 | |
| *** vnogin has joined #openstack-ansible | 10:25 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-os_cinder master: Fix usage of "|" for tests https://review.openstack.org/582246 | 10:26 |
| *** vnogin has quit IRC | 10:29 | |
| *** electrofelix has quit IRC | 10:34 | |
| *** Bhujay has joined #openstack-ansible | 10:36 | |
| openstackgerrit | Markos Chandras (hwoarang) proposed openstack/openstack-ansible master: Add support for using distribution packages for OpenStack services https://review.openstack.org/579770 | 10:39 |
| *** laztoum has joined #openstack-ansible | 10:46 | |
| *** poul_ has quit IRC | 11:03 | |
| *** poul has joined #openstack-ansible | 11:03 | |
| openstackgerrit | Christian Zunker proposed openstack/openstack-ansible-os_heat master: Explain how to configure public endpoint https://review.openstack.org/582919 | 11:09 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible master: Prevent incorrect credentials https://review.openstack.org/582891 | 11:17 |
| *** suggestable has joined #openstack-ansible | 11:17 | |
| *** electrofelix has joined #openstack-ansible | 11:23 | |
| openstackgerrit | Markos Chandras (hwoarang) proposed openstack/openstack-ansible master: Add support for using distribution packages for OpenStack services https://review.openstack.org/579770 | 11:28 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible master: Prevent incorrect credentials https://review.openstack.org/582891 | 11:30 |
| openstackgerrit | Markos Chandras (hwoarang) proposed openstack/openstack-ansible master: Add support for using distribution packages for OpenStack services https://review.openstack.org/579770 | 11:30 |
| *** qwebirc64668 has joined #openstack-ansible | 11:32 | |
| qwebirc64668 | hello! it is for a while that i am trying to integrate openstack ansible to openbaton, but always i face different error, i was wondering would you help me | 11:35 |
| qwebirc64668 | i'm pretty sure the problem is not in openbaton side | 11:36 |
| *** udesale has joined #openstack-ansible | 11:41 | |
| qwebirc64668 | the controller ip address is 172.29.236.11, behind the HAPROXY the address is 172.29.236.10, in registering new pop in openbaton, we set the url as 172.29.236.11:5000/v3 but when ever openbaton wants to get the images from glance, it sends an API to openstack and the openstack translate the API and send it to 172.29.236.10 as a result we get NULL, this is my main problem up now and please help me | 11:42 |
| guilhermesp | odyssey4me: just receive this update os include_role yesterday when was merged :) | 11:42 |
| guilhermesp | well, I see as a positive merge, I wont need to worry about it in the future, in face of the problems that I've had | 11:43 |
| guilhermesp | I think I posted here yesterday at night | 11:43 |
| Bhujay | qwebirc64668: no idea about openbaton but some thing like this controls the endpoint ....whenever openstack client is making a call to keystone , there is a parameter to set interface=public or internal | 11:47 |
| Bhujay | when a call is made from say openbaton server there shd be a file such as cloud.yml etc where the interface or endpoint_type is defined | 11:51 |
| *** hachi has quit IRC | 12:06 | |
| qwebirc64668 | Bhujay thank you for your answer but i'm not sure i got it completely, you mean that there is a parameter that i can set when i am sending api from openbaton to openstack ? | 12:07 |
| openstackgerrit | Merged openstack/openstack-ansible-os_neutron stable/queens: Install openvswitch packages on 'ml2.ovs.dvr' configuration https://review.openstack.org/581040 | 12:08 |
| *** hachi has joined #openstack-ansible | 12:09 | |
| qwebirc64668 | bhujay is there any way to set default value for this parameter whether the parameter is not set in cloud.yml ? | 12:14 |
| guilhermesp | odyssey4me evrardjp if you please ca confirm to me: does shade covers monasca api interaction (alarms, notifications, measurement..)? I think it doens't https://bit.ly/2Jp76sE | 12:15 |
| guilhermesp | in that case, port the modules that we have will require a bit more of work... I was thinking about do it now so fix the last issue with https://review.openstack.org/#/c/578721/19 | 12:16 |
| Bhujay | qwebirc64668: there are three values for os_interface , public , internal and admin , consult a bit with the openstack client help it should give you helpful clues | 12:19 |
| Bhujay | in cloud.yml in key name is interface | 12:20 |
| *** tux_ has joined #openstack-ansible | 12:29 | |
| *** armaan has quit IRC | 12:31 | |
| *** tux_ has quit IRC | 12:34 | |
| qwebirc64668 | i think my problem is my openstack ansible publicURL, when i change os_interface (in openrc file )to publicURL i miss the access to images and i face error :"Unable to establish connection to http://172.29.236.10:9292/v2/images: ('Connection aborted.', BadStatusLine("''",))" | 12:37 |
| *** laztoum has quit IRC | 12:38 | |
| qwebirc64668 | bhujay | 12:38 |
| Bhujay | not sure how it is related to openstack ansible ... but whatever it is why don't u change os_interface=internalURL in openrc | 12:41 |
| Dmitriy | could please someone review it again? https://review.openstack.org/#/c/581459/ | 12:45 |
| *** armaan has joined #openstack-ansible | 12:47 | |
| openstackgerrit | Merged openstack/openstack-ansible master: Disable keepalived ping tests by default https://review.openstack.org/580379 | 12:47 |
| qwebirc64668 | Bhujay the problem is no one can access to different part of my openstack from public URL and openbaton uses this kind of url | 12:47 |
| *** qwebirc56949 has joined #openstack-ansible | 12:52 | |
| qwebirc56949 | Hello! guys | 12:52 |
| *** throwsb1 has joined #openstack-ansible | 12:52 | |
| *** qwebirc28776 has joined #openstack-ansible | 12:52 | |
| qwebirc56949 | i have the same problem with public url and i don't know what to do!!!could anyone help me plz | 12:53 |
| openstackgerrit | Markos Chandras (hwoarang) proposed openstack/openstack-ansible master: Add support for using distribution packages for OpenStack services https://review.openstack.org/579770 | 12:53 |
| *** qwebirc64668 has quit IRC | 12:54 | |
| openstackgerrit | Christian Zunker proposed openstack/openstack-ansible master: Example setup for gnocchi with ceph and redis https://review.openstack.org/582946 | 12:55 |
| evrardjp | guilhermesp: I am not sure we need to port things over to shade right now | 12:56 |
| evrardjp | guilhermesp: it would be good to make it work with the existing | 12:56 |
| evrardjp | then improve stuff | 12:56 |
| evrardjp | Dmitriy: I've reviewed -- do you need help? | 12:59 |
| guilhermesp | ok evrardjp trying to find a way right now | 13:02 |
| *** electrofelix has quit IRC | 13:07 | |
| Dmitriy | evrardjp: Actually yes. I'd like to see it merged (and it seems, that without approve it's not possible). And after this change https://review.openstack.org/#/c/582300/1/tasks/masakari_service_setup.yml functional test fails on the "Setup the service" state | 13:08 |
| Dmitriy | And I really don't know how to fix it, without rollback to the previous version of the file | 13:08 |
| tosky | evrardjp, odyssey4me: still no luck with os_sahara; https://review.openstack.org/#/c/582538/ fixes some issues, but the ubuntu jobs fails on multipathd installation | 13:12 |
| evrardjp | guilhermesp: Dmitriy tosky I have to finish something real quick, then will answer your question in the reverse order. | 13:15 |
| tosky | sure, no rush | 13:16 |
| tosky | it was just an update | 13:16 |
| mgariepy | can I get another vote on this please ? https://review.openstack.org/#/q/I1bd872a50ee506081b2d8274533a954a58a76fde | 13:20 |
| guilhermesp | np evrardjp :) | 13:24 |
| hwoarang | hmm | 13:25 |
| hwoarang | what is up with fixes like this one https://review.openstack.org/#/c/582770/ | 13:25 |
| hwoarang | it feels it's more like a fix to improve personal number than something worth pursuing on every single role :/ | 13:25 |
| evrardjp | cores... I am fine if you don't listen to me, but please stop doing silly things like: https://review.openstack.org/#/c/580379/ | 13:26 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible master: Revert "Disable keepalived ping tests by default" https://review.openstack.org/582953 | 13:26 |
| evrardjp | that's never gonna work. | 13:26 |
| hwoarang | why is it silly? the revert does not explain it | 13:26 |
| evrardjp | I just clicked on revert -- but I will just add a patch on top | 13:27 |
| evrardjp | it's easier | 13:27 |
| evrardjp | You will see. | 13:27 |
| hwoarang | ok | 13:27 |
| *** epalper has quit IRC | 13:28 | |
| *** epalper has joined #openstack-ansible | 13:28 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible master: Fix haproxy checks https://review.openstack.org/582954 | 13:29 |
| evrardjp | ^ | 13:29 |
| evrardjp | hwoarang: ^ | 13:29 |
| hwoarang | "In thee precedent patch" . what precedent patch? | 13:31 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible master: Fix haproxy checks https://review.openstack.org/582954 | 13:32 |
| evrardjp | updated. | 13:32 |
| mgariepy | evrardjp, you get my +2 | 13:33 |
| evrardjp | mgariepy: thanks. Sorry I complained. | 13:34 |
| *** armaan has quit IRC | 13:35 | |
| evrardjp | I have the impression we have thousand of things to fix until the end of the cycle, and today I worked on things that should have worked the whole time. | 13:35 |
| openstackgerrit | Merged openstack/openstack-ansible-openstack_openrc stable/queens: Add OS_AUTH_TYPE env var in openrc file https://review.openstack.org/582548 | 13:35 |
| evrardjp | Just because of lack of attention in the reviews | 13:35 |
| *** armaan has joined #openstack-ansible | 13:35 | |
| evrardjp | sometimes by me, which frustrates me even more. | 13:35 |
| evrardjp | we are loosing time that way, and we shouldn't. | 13:36 |
| evrardjp | We should increase test coverage if we can. | 13:36 |
| mgariepy | yeah but the keepalived stuff is kinda hard to test i think. | 13:37 |
| mgariepy | and there are a lot of thing that are not tested yet. | 13:37 |
| evrardjp | yeah but the other mistakes I was fixing today could have. | 13:39 |
| evrardjp | we should basically have a if "ERROR" is in logs -> fail | 13:39 |
| *** dave-mccowan has joined #openstack-ansible | 13:39 | |
| evrardjp | if other ppl want to tackle some of my reviews: https://review.openstack.org/#/q/owner:jean-philippe%2540evrard.me+status:open+is:mergeable | 13:39 |
| *** armaan has quit IRC | 13:39 | |
| evrardjp | they don't all work, so please feel free to help too. :D | 13:40 |
| evrardjp | ok next -- tosky :) | 13:40 |
| evrardjp | so suse and centos work, but ubuntu doesn't | 13:40 |
| openstackgerrit | Merged openstack/openstack-ansible-openstack_openrc stable/pike: Add OS_AUTH_TYPE env var in openrc file https://review.openstack.org/582547 | 13:41 |
| evrardjp | interesting... issue in the modprobe. That should be due to recent series of odyssey4me 's patches. | 13:41 |
| openstackgerrit | Michael Vollman proposed openstack/ansible-config_template master: Compare dict vars to determine changed https://review.openstack.org/582532 | 13:43 |
| odyssey4me | evrardjp: I suspect nova's play will need something like https://review.openstack.org/582579 too | 13:43 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-tests master: [WIP] Add become to load kernel module https://review.openstack.org/582957 | 13:46 |
| *** jwitko has joined #openstack-ansible | 13:46 | |
| *** tux_ has joined #openstack-ansible | 13:46 | |
| evrardjp | odyssey4me: I am trying with this ^ | 13:47 |
| evrardjp | tosky: do you mind if I try my hack? | 13:47 |
| tosky | evrardjp: of course not | 13:47 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-os_sahara master: Add few more services to the inventory (nova, memcached) https://review.openstack.org/582538 | 13:47 |
| odyssey4me | evrardjp: oh, it's its own play - then better to use connection: local and become: true, and get rid of remote_user | 13:47 |
| openstackgerrit | Michael Vollman proposed openstack/ansible-config_template master: Compare dict vars to determine changed https://review.openstack.org/582532 | 13:47 |
| evrardjp | not sure if it's gonna work | 13:47 |
| *** armaan has joined #openstack-ansible | 13:47 | |
| evrardjp | odyssey4me: feel free to edit the patch | 13:48 |
| evrardjp | I have no strong opinion | 13:48 |
| evrardjp | I can do connection: local yes | 13:48 |
| prometheanfire | evrardjp: still have ironic questions? | 13:48 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-tests master: [WIP] Add become to load kernel module https://review.openstack.org/582957 | 13:49 |
| evrardjp | prometheanfire: myself I do not, but other ppl around here may have. It was to know if (and how) it was possible to have a cloud with ironic and kvm. | 13:49 |
| evrardjp | I didn't say how it was possible | 13:50 |
| openstackgerrit | Michael Vollman proposed openstack/ansible-config_template master: Compare dict vars to determine changed https://review.openstack.org/582532 | 13:50 |
| guilhermesp | yes it is | 13:50 |
| ebbex_ | seriously, wt-actual-f is wrong with 'openstack-ansible-functional-centos-7' on stable/pike? | 13:50 |
| guilhermesp | I have one | 13:50 |
| evrardjp | ebbex_: everything? | 13:50 |
| evrardjp | :D | 13:50 |
| evrardjp | guilhermesp: :) | 13:50 |
| guilhermesp | actually, nova-lxd, custom kvm and ironic ( will be part of my speech at openstack days ) | 13:50 |
| evrardjp | ebbex_: nobody really cared at some point in P. Till more contributors came over and fixed it for their case: Queens and above. | 13:51 |
| guilhermesp | but just a note, the time I did the deployment was stable/pike | 13:51 |
| evrardjp | ebbex_: sorry if you're in the "beta" (as listed in the documentation!) | 13:51 |
| evrardjp | Dmitriy: still there? | 13:51 |
| prometheanfire | yes, the way it's done is to use image metadata | 13:52 |
| prometheanfire | and tie that to flavor metadata | 13:52 |
| Dmitriy | evrardjp Yep) | 13:52 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-tests master: Use local connection and become for nova_compute https://review.openstack.org/582957 | 13:52 |
| prometheanfire | node gets tagged with a resource class https://github.com/openstack/openstack-ansible-ops/blob/master/scripts/ironic_register-nodes.sh#L46 | 13:53 |
| ebbex_ | I'm trying to get https://review.openstack.org/#/c/581041/ to pass and merge so I can get on with my life. And it's running out of mirrors, or can't resolve domains. It passed once, but ofcourse then suze breaks on resolve or some other crap. | 13:53 |
| evrardjp | Dmitriy: ok let's focus on your thing now :) | 13:53 |
| prometheanfire | flavor gets tagged with a resource class https://github.com/openstack/openstack-ansible-ops/blob/master/scripts/ironic_register-nodes.sh#L61 (at least one) | 13:53 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_magnum master: Execute service setup against a delegated host using Ansible built-in modules https://review.openstack.org/582296 | 13:53 |
| prometheanfire | image gets tagged with ironic https://github.com/openstack/openstack-ansible-ops/blob/master/scripts/ironic_generate-images.sh#L44 | 13:53 |
| prometheanfire | those three should do it | 13:54 |
| prometheanfire | evrardjp: ^ | 13:54 |
| evrardjp | cool | 13:54 |
| evrardjp | good to know prometheanfire | 13:54 |
| Dmitriy | evrardjp: I'd be really very pleased if we could to figure it out, as I have already written part for installing masakari-monitors. | 13:54 |
| prometheanfire | I know it wasn't you asking :P | 13:54 |
| evrardjp | I will repeat it when the contributor wants it | 13:54 |
| evrardjp | Dmitriy: you are working on 581459 right? | 13:55 |
| evrardjp | Dmitriy: not sure what the issue is | 13:55 |
| evrardjp | I will add the funcitonal test if you don't mind | 13:55 |
| Dmitriy | Yes. But as https://review.openstack.org/#/c/582300/ has been added - functional test fails. | 13:55 |
| Dmitriy | You may see zuul failure in patch set 5 | 13:56 |
| antonym | coreywright: thanks for taking care of that, yeah the shrinking of the volume is only done on the nspawn to make room for the cinder and swift volumes, if using lxc, it just nukes the lxc volume and uses that space | 13:56 |
| Dmitriy | It fails on this task (and due to this change ) https://review.openstack.org/#/c/582300/1/tasks/masakari_service_setup.yml | 13:56 |
| *** tux__ has joined #openstack-ansible | 13:56 | |
| evrardjp | Dmitriy: I see the failure, I don't see the logs | 13:57 |
| evrardjp | I've abandonned and reopened to retrigger a testing | 13:57 |
| ebbex_ | evrardjp: Would you mind looking over this https://review.openstack.org/#/c/581041 and override zuul if it seems legit to do so? | 13:58 |
| evrardjp | do you think humans can be "cooked" by the sun? I wonder how much exposure of the sun I can get right where I am. I am pretty sure I won't last a minute! | 14:00 |
| *** tux_ has quit IRC | 14:00 | |
| Dmitriy | So in my test environment it failed with the following problem http://paste.openstack.org/show/725845/ | 14:00 |
| evrardjp | ebbex_: it is not possible to override zuul. You need to write a patch to bypass it | 14:00 |
| evrardjp | ebbex_: your patch seems acceptable | 14:01 |
| ebbex_ | evrardjp: could you point me in the direction as to which files I need to look at to fix the 'openstack-ansible-functional-centos-7' issues? | 14:02 |
| evrardjp | ebbex_: I think it would be better if you speak to mgariepy and mnaser who have both been following all of that closer than me. | 14:02 |
| evrardjp | Dmitriy: when was that done? | 14:03 |
| ebbex_ | Ok, thanks :) | 14:03 |
| evrardjp | ebbex_: yw :) | 14:03 |
| Dmitriy | evrardjp In friday, when I was working on https://review.openstack.org/#/c/582675/ | 14:03 |
| *** Bhujay has quit IRC | 14:03 | |
| evrardjp | Dmitriy: I think a few patches merged since then. odyssey4me did a great job lately on those changes -- he pushed a lot of patches during the weekend | 14:04 |
| evrardjp | Dmitriy: That's why a recheck should help | 14:04 |
| *** openstackgerrit has quit IRC | 14:04 | |
| Dmitriy | Ok, let's wait. I'll monitor the result and update you about them, ok? | 14:06 |
| *** openstackgerrit has joined #openstack-ansible | 14:06 | |
| openstackgerrit | Andy Smith proposed openstack/openstack-ansible-os_cinder master: Setup oslo.messaging extra packages for optional drivers https://review.openstack.org/582632 | 14:06 |
| evrardjp | Dmitriy: sure, thanks! | 14:07 |
| *** sandipan has quit IRC | 14:07 | |
| evrardjp | guilhermesp: ok so where are we now? How can I help? It's the module that can't due his work, right? | 14:08 |
| guilhermesp | yep evrardjp | 14:08 |
| guilhermesp | let me share my ideas | 14:08 |
| guilhermesp | https://bit.ly/2JpZHZY is that possible to call our monasca_bin variable to toggle this hardcoded path? | 14:09 |
| guilhermesp | that's what I'm trying to figure out | 14:09 |
| evrardjp | I think we shouldn't do this at all | 14:09 |
| evrardjp | let me patch that quick | 14:09 |
| guilhermesp | k | 14:09 |
| guilhermesp | but any particular reason that we shouldn't do this? | 14:10 |
| evrardjp | guilhermesp: Yeah, it's hidden behavior | 14:10 |
| evrardjp | we should rely on what's standard :) | 14:10 |
| guilhermesp | k, makes sense | 14:11 |
| evrardjp | could you remind me why the patch 15 worked and the patch here doesn't anymore? | 14:11 |
| evrardjp | it's due to the changes external to the fixing of the role, like the move to the db and keystone modules inside the role, right? | 14:11 |
| guilhermesp | openrc was removed from meta/main.yml | 14:11 |
| guilhermesp | https://github.com/openstack/openstack-ansible-os_monasca/commit/4b46ce34a709791bac55afb67d214c8a40050065 | 14:13 |
| guilhermesp | your workaround that didn't work https://review.openstack.org/#/c/578721/19/tasks/monasca_alarms_setup.yml | 14:13 |
| openstackgerrit | Markos Chandras (hwoarang) proposed openstack/openstack-ansible master: Add support for using distribution packages for OpenStack services https://review.openstack.org/579770 | 14:14 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-os_monasca master: Fix monasca https://review.openstack.org/578721 | 14:14 |
| evrardjp | oh yeah if openrc isn't there that's not gonna work | 14:14 |
| evrardjp | I don't see a point of adding this client in all cases though, so let me just bring the openrc back | 14:15 |
| openstackgerrit | Andy Smith proposed openstack/openstack-ansible master: [WIP] Add qdrouterd role for messaging backend deployment https://review.openstack.org/570261 | 14:15 |
| *** chyka has joined #openstack-ansible | 14:15 | |
| evrardjp | ansmith: FYI I had to do this: https://review.openstack.org/#/c/582891/ | 14:15 |
| guilhermesp | and I think we haven't had this problem in the tests before coz we didn't go into execpt clause https://github.com/openstack/openstack-ansible-os_monasca/blob/402c012b9f77141d418485e2afd5c4f3fec35c7c/library/monasca_notification_method.py#L121 | 14:15 |
| evrardjp | so you're not surprised. | 14:16 |
| *** nurdie has joined #openstack-ansible | 14:16 | |
| *** chyka has joined #openstack-ansible | 14:16 | |
| evrardjp | guilhermesp: that's what my patch is trying to fix | 14:16 |
| evrardjp | ansible_python_interpreter should point ot the venv which has the client | 14:16 |
| evrardjp | else I'd build a node with that role, and see where it fails | 14:17 |
| evrardjp | let me just add openrc real quick :) | 14:17 |
| guilhermesp | k, so you're avoinding the code to get that hardcoded path | 14:19 |
| ansmith | evrardjp: thanks, i will examine the change | 14:19 |
| mgariepy | ebbex_, https://review.openstack.org/#/c/581041 the error on centos is mirror error, recheck it should pass. | 14:19 |
| *** lbragstad has joined #openstack-ansible | 14:20 | |
| *** kstev has joined #openstack-ansible | 14:22 | |
| ebbex_ | mgariepy: I have 8 rechecks on it. And that f'n centos thing breaks every single time, either on mirrors or dns-resolves. | 14:24 |
| ansmith | evrardjp: not sure if you can take a quick look but is there anything additional i need to consider for https://review.openstack.org/#/c/582632/? | 14:25 |
| *** kstev has quit IRC | 14:25 | |
| *** prasen has quit IRC | 14:27 | |
| *** cjloader_away has joined #openstack-ansible | 14:28 | |
| *** cjloader_away is now known as cjloader | 14:31 | |
| evrardjp | guilhermesp: I have another patch pending, but I'd rather watch this one first | 14:32 |
| evrardjp | please don't push anything :) | 14:32 |
| guilhermesp | k evrardjp I'm monitoring them :) | 14:37 |
| *** coreywright has quit IRC | 14:38 | |
| evrardjp | ansmith: so... we have a mechanism for that... named the optional packages, like this: https://github.com/openstack/openstack-ansible-os_neutron/blob/master/vars/source_install.yml#L82-L84 | 14:38 |
| evrardjp | so here you'd say for what optional part your want it | 14:38 |
| evrardjp | put the package name in it | 14:38 |
| evrardjp | (which would be oslo.messaging[amqp] or something like that) | 14:38 |
| evrardjp | so that it would be built on the rpeo | 14:39 |
| evrardjp | repo | 14:39 |
| evrardjp | then | 14:39 |
| evrardjp | in your tasks | 14:39 |
| evrardjp | (almost the same way you did it) | 14:39 |
| ansmith | ahh, great, let me try it out | 14:39 |
| *** exodusftw has quit IRC | 14:39 | |
| evrardjp | you can have things like this: https://github.com/openstack/openstack-ansible-os_neutron/blob/master/tasks/providers/dragonflow_config.yml#L16-L33 | 14:39 |
| evrardjp | so that it works also outside the role tests | 14:40 |
| *** exodusftw has joined #openstack-ansible | 14:40 | |
| *** armaan has quit IRC | 14:41 | |
| *** armaan has joined #openstack-ansible | 14:41 | |
| *** qwebirc28776 has quit IRC | 14:43 | |
| *** kstev has joined #openstack-ansible | 14:44 | |
| *** armaan has quit IRC | 14:45 | |
| mgariepy | ebbex_, that's annoying i but one check did pass yesterday http://logs.openstack.org/41/581041/2/check/openstack-ansible-functional-centos-7/ab6c756/job-output.txt.gz | 14:45 |
| *** vakuznet has joined #openstack-ansible | 14:51 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_keystone master: Execute service setup against a delegated host using Ansible built-in modules https://review.openstack.org/579993 | 14:53 |
| cloudnull | mornings | 14:54 |
| *** pcaruana has quit IRC | 15:02 | |
| *** lirt has quit IRC | 15:05 | |
| *** kelv_ has joined #openstack-ansible | 15:06 | |
| evrardjp | morning cloudnull | 15:07 |
| cloudnull | good games this weekend evrardjp! | 15:08 |
| evrardjp | haha | 15:08 |
| evrardjp | it's over now | 15:08 |
| cloudnull | it is, | 15:09 |
| evrardjp | it was quite a welcome of Belgian team in Brussels, but I didn't got the chance to attend | 15:09 |
| evrardjp | get* | 15:09 |
| cloudnull | now we wait ... four more years | 15:09 |
| evrardjp | 2 more for the Euro | 15:09 |
| evrardjp | but that's not as big | 15:09 |
| evrardjp | but still big | 15:09 |
| evrardjp | :) | 15:09 |
| cloudnull | indeed ! | 15:09 |
| evrardjp | and confederates cup | 15:09 |
| evrardjp | if that's the proper ter | 15:10 |
| evrardjp | term | 15:10 |
| evrardjp | https://en.wikipedia.org/wiki/FIFA_Confederations_Cup | 15:10 |
| cloudnull | yup. hopefully USMNT can get it together for the tournament | 15:12 |
| cloudnull | also the Woman's world Cup https://www.fifa.com/womensworldcup/index.html - next eyar | 15:14 |
| cloudnull | **year | 15:14 |
| cloudnull | so much futbol :) | 15:14 |
| cloudnull | evrardjp https://review.openstack.org/#/c/574006/ - i think it needs yet another update | 15:16 |
| cloudnull | trying to test that locally and nova refuses to staRT | 15:16 |
| cloudnull | IDK if others are seeing the same ? | 15:16 |
| evrardjp | cloudnull: https://review.openstack.org/#/c/582891/ | 15:17 |
| evrardjp | cloudnull: in an AIO, we deploy designate, so designate integration is enabled for nova. So notifications have to be turned on. But notifications have wrong credentials without patch ^ | 15:18 |
| *** cshen has joined #openstack-ansible | 15:20 | |
| *** FrankZhang has joined #openstack-ansible | 15:21 | |
| tux__ | cloudnull: I am adding cinder in my running openstack cloud and i am seeing this error when running os-cinder-install.yml | 15:28 |
| tux__ | http://paste.openstack.org/show/726014/ | 15:28 |
| tux__ | I am not seeing any cinder package (tarball) | 15:28 |
| tux__ | who will update repo in OSA ? | 15:29 |
| Dmitriy | tux__: is lsyncd working in the "main" repo container? | 15:34 |
| tux__ | Yes, i have check all 3 repo container they have same directory and file view | 15:35 |
| tux__ | cinder is nowhere | 15:35 |
| *** epalper has quit IRC | 15:35 | |
| tux__ | Do i need to run repo-build.yml ? | 15:36 |
| tosky | fun, now 582538 failed also for centos | 15:36 |
| Dmitriy | tux__: if I were you, I'd try to run it | 15:38 |
| tux__ | let me do then | 15:38 |
| *** lvdombrkr has quit IRC | 15:40 | |
| cloudnull | evrardjp my testing is in a hardware lab | 15:47 |
| odyssey4me | tux__: whenever the host groups change, it's advisable to re-run repo-build so that any new component venvs are built | 15:48 |
| cloudnull | I've rebased things this morning and will try again in a few | 15:48 |
| *** coreywright has joined #openstack-ansible | 15:50 | |
| tux__ | odyssey4me: how we can put that in documentation :) | 15:53 |
| tux__ | you guys are expert but new people like me struggle a lot to find this kind of small thing.. | 15:53 |
| evrardjp | cloudnull: not sure about the link | 15:53 |
| evrardjp | yeah try with my patch : ) | 15:54 |
| cloudnull | with https://review.openstack.org/#/c/582891/ | 15:54 |
| cloudnull | ? | 15:54 |
| evrardjp | yeah | 15:54 |
| cloudnull | will do | 15:54 |
| evrardjp | should work | 15:54 |
| evrardjp | "should" | 15:54 |
| evrardjp | I am trying it right now | 15:54 |
| evrardjp | it's repo building atm | 15:54 |
| cloudnull | im not deploying designate, however I will give this a try now | 15:55 |
| Dmitriy | evrardjp: test are "ready" for https://review.openstack.org/#/c/581459/ | 15:58 |
| Dmitriy | some new error, which is not even related to this role | 15:59 |
| tux__ | Dmitriy: & odyssey4me: got this error during running repo-build.yml http://paste.openstack.org/show/726020/ | 16:01 |
| tux__ | look like github issue | 16:02 |
| tux__ | https://github.com/SPICE/spice-html5 (404 error) | 16:02 |
| Dmitriy | tux__: yep, spice-html5 has moved from github | 16:03 |
| tux__ | damn it!!! so what is the fix here? | 16:03 |
| evrardjp | tux__: we have a release note for that | 16:03 |
| evrardjp | take new code otherwise :) | 16:04 |
| evrardjp | newer* | 16:04 |
| Dmitriy | tux__: I'd recommend you to switch to novnc console. | 16:04 |
| tux__ | I am using pike 16.0.14 | 16:04 |
| guilhermesp | evrardjp: your PR worked :) | 16:04 |
| evrardjp | novnc rulez | 16:04 |
| tux__ | how to switch novnc and where ? | 16:04 |
| evrardjp | guilhermesp: cool | 16:05 |
| guilhermesp | is executing tempest tests now | 16:05 |
| tux__ | can you give me hit or direction ? | 16:05 |
| evrardjp | let's make it better | 16:05 |
| Dmitriy | nova_console_type: novnc in user_variables | 16:05 |
| evrardjp | guilhermesp: I didn't even push my uber patch | 16:05 |
| evrardjp | that's sad | 16:05 |
| tux__ | do i need to run any playbook related that component ? | 16:05 |
| evrardjp | I had an awesome one | 16:05 |
| Dmitriy | evrardjp: spice is not the best console, especially if you plan to run Windows, novnc is a bit better IMO | 16:06 |
| guilhermesp | but I'm just curious about that one that you'd push, so If there's a way to show me ... | 16:06 |
| evrardjp | like remotely delegating to the local node and all | 16:06 |
| evrardjp | guilhermesp: yeah I can push it and we revert later? | 16:06 |
| guilhermesp | yes you can evrardjp | 16:06 |
| evrardjp | I mean we revert in the commit | 16:06 |
| evrardjp | let's wait for official results first | 16:06 |
| guilhermesp | just want to see the tempest execution here | 16:06 |
| evrardjp | so that we have a proof it's working :) | 16:06 |
| guilhermesp | evrardjp yep :) | 16:06 |
| tux__ | Dmitriy: going to do "nova_console_type: novnc" and running repo-build | 16:07 |
| evrardjp | Dmitriy: that's what I meant. Novnc rules! Like novnc rocks! | 16:07 |
| guilhermesp | I still believe that the single one tempest test will fail, but it's something that I will fix (monasca related ) | 16:07 |
| evrardjp | :) | 16:07 |
| evrardjp | guilhermesp: sounds good | 16:07 |
| evrardjp | I trust you | 16:07 |
| evrardjp | Dmitriy: I don't like this error -- I don't know yet the reason but I expect a keystone failure with rabbitmq as I see the task didn't run | 16:08 |
| evrardjp | let me dig into that issue | 16:08 |
| FrankZhang | evrardjp barbican seems only required 'pbr' >= 2.0 and no cap version limitation: https://github.com/openstack/barbican/blob/stable/queens/requirements.txt#L23 while I'm still unclear what this complains for: http://logs.openstack.org/84/579284/4/experimental/openstack-ansible-deploy-barbican-ubuntu-xenial/266a344/job-output.txt.gz#_2018-07-13_15_28_32_752513 | 16:09 |
| Dmitriy | tux__: you have to run at least repo-build and os-nova roles. But generally you may run setup-infrastrucvture and setup-openstack safely | 16:09 |
| tux__ | setup-openstack will take really long time to finish :( | 16:10 |
| evrardjp | Dmitriy: oh great we have kernel stacktraces. | 16:10 |
| tux__ | will do that anyway | 16:10 |
| FrankZhang | btw master barbican's requirement of 'pbr' is same, greater or equal to 2.0: https://github.com/openstack/barbican/blob/master/requirements.txt#L24 | 16:10 |
| *** hachi has quit IRC | 16:11 | |
| evrardjp | Dmitriy: I don't see errors in keystone though: http://logs.openstack.org/59/581459/5/check/openstack-ansible-functional-ubuntu-xenial/7e79335/logs/openstack/infra1/keystone/keystone.log.txt.gz | 16:12 |
| evrardjp | recheck? | 16:12 |
| Dmitriy | let's try, but it failed with the same error for all operating systems | 16:13 |
| Dmitriy | evrardjp: I've also commented your notices here https://review.openstack.org/#/c/580591/1 | 16:14 |
| Dmitriy | the point is, that in current state is zookeeper role is not functional at all | 16:14 |
| *** udesale has quit IRC | 16:15 | |
| *** DanyC has quit IRC | 16:18 | |
| evrardjp | FrankZhang: http://logs.openstack.org/84/579284/4/experimental/openstack-ansible-deploy-barbican-ubuntu-xenial/266a344/logs/openstack/aio1_repo_container-9312b122/repo/wheel_build.log | 16:19 |
| evrardjp | problem with your groups I think | 16:20 |
| evrardjp | or with the git repo | 16:20 |
| evrardjp | to be digged a little further | 16:20 |
| *** cshen has quit IRC | 16:20 | |
| evrardjp | Dmitriy: that's a role that's used by openstack-infra, not openstack-ansible, we are just consumers. I think it's under development by pabelanger | 16:21 |
| evrardjp | odyssey4me: might give us insights about that role | 16:21 |
| evrardjp | FrankZhang: I think the first task you see failing is ignored and the last failure matters. That's the building of the wheels | 16:22 |
| evrardjp | so this needs to be fixed | 16:22 |
| Dmitriy | evrardjp: ok, got it | 16:23 |
| evrardjp | I don't feel well, so I trust you the keys of the castle while I go for some sleep | 16:23 |
| FrankZhang | cool, have a good night! | 16:23 |
| evrardjp | Dmitriy: that's why I was also careful on defaulting scope to "all" imagine deploying it to all the infra nodes! :D | 16:23 |
| evrardjp | FrankZhang: just a small rest. I will be back today. | 16:23 |
| guilhermesp | evrardjp: go to rest! You show me after the PR that you'd push. I will focus on the tempest error now | 16:23 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/ansible-role-systemd_service master: Correct log warning for unknown lvalue https://review.openstack.org/582997 | 16:24 |
| Dmitriy | evrardjp: good night. May I try to katch you tomorrow about masakari role? Or someone else:) | 16:24 |
| Dmitriy | *catch | 16:24 |
| evrardjp | guilhermesp: is the result published? | 16:25 |
| evrardjp | I can just push the patch | 16:25 |
| guilhermesp | yes evrardjp | 16:25 |
| guilhermesp | feel free :) | 16:25 |
| evrardjp | oh yeah | 16:25 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-os_monasca master: Fix monasca https://review.openstack.org/578721 | 16:25 |
| evrardjp | Dmitriy: don't hesitate :) | 16:25 |
| evrardjp | Dmitriy: that's a curious (and new) issue we need to sort out | 16:26 |
| Dmitriy | Yep, especially for such newbie as me) | 16:26 |
| *** qwebirc56949 has quit IRC | 16:29 | |
| tux__ | Dmitriy: Now i am getting this new error in repo-build.yml error is super long i cut some output http://paste.openstack.org/show/726025/ | 16:33 |
| *** lbragstad has quit IRC | 16:33 | |
| *** poul has quit IRC | 16:33 | |
| tux__ | 'python_version == \"3.4\"' don't match your environment\nIgnoring ruamel.yaml: markers | 16:34 |
| tux__ | my git tag is 16.0.13-7-g56099d7 | 16:35 |
| Dmitriy | Ignoring python_version should be safe and shouldn't cause role failure | 16:38 |
| openstackgerrit | Merged openstack/openstack-ansible-tests master: Use local connection and become for localhost plays https://review.openstack.org/582359 | 16:40 |
| *** epalper has joined #openstack-ansible | 16:40 | |
| tux__ | Dmitriy: but it failed playbook but wondering what will happened to next task in that playbook? | 16:43 |
| Dmitriy | The problem seems to be in switching to some git commit, which is not present. But I actually can't say, why this is happening. | 16:44 |
| tux__ | currently i am running "16.0.13-7-g56099d7" git | 16:44 |
| tux__ | if i switch to 16.0.14 does that create any issue ? | 16:44 |
| tux__ | Dmitriy: ^^ | 16:45 |
| Dmitriy | tux__: do you have operational cluster? | 16:45 |
| *** epalper has quit IRC | 16:45 | |
| tux__ | no its Development | 16:45 |
| Dmitriy | Than you may really try to upgrade to 16.0.15 or even stable/pike branch by this guide https://docs.openstack.org/openstack-ansible/pike/user/minor-upgrade.html | 16:48 |
| tux__ | Dmitriy: sounds like a good plan | 16:49 |
| tux__ | Dmitriy: is this safe to do on production also ? | 16:50 |
| Dmitriy | It is supposed to be safe, but I'd tested all upgrades on some sandbox, as it also upgrades openstack components | 16:56 |
| *** cshen has joined #openstack-ansible | 16:56 | |
| Dmitriy | You can never be sure, that any software upgrade won't break your running infratructure. | 16:58 |
| *** suggestable has quit IRC | 16:58 | |
| *** shardy has quit IRC | 16:59 | |
| *** cshen has quit IRC | 17:02 | |
| *** cshen has joined #openstack-ansible | 17:18 | |
| *** lbragstad has joined #openstack-ansible | 17:24 | |
| *** cshen has quit IRC | 17:27 | |
| *** cshen has joined #openstack-ansible | 18:00 | |
| *** poopcat has joined #openstack-ansible | 18:01 | |
| *** cshen has quit IRC | 18:05 | |
| *** poul has joined #openstack-ansible | 18:07 | |
| *** poul_ has joined #openstack-ansible | 18:08 | |
| *** poul has quit IRC | 18:12 | |
| openstackgerrit | Markos Chandras (hwoarang) proposed openstack/openstack-ansible-galera_server master: vars: SUSE: Drop python-pycrypto from SUSE packages https://review.openstack.org/583013 | 18:15 |
| hwoarang | mnaser: ^^ any particular reason you didn't drop it for suse? | 18:15 |
| mnaser | hwoarang: at the time it didnt seem to break anything in suse (but was breaking in centos) and i don't know the suse env as well | 18:16 |
| hwoarang | ok i will check | 18:16 |
| hwoarang | but if you modify vars/$distro please ping me if it's easy so i can check for suse and keep distros in sync when possible | 18:16 |
| hwoarang | and i can also check for ubuntu | 18:17 |
| *** hachi has joined #openstack-ansible | 18:17 | |
| openstackgerrit | Markos Chandras (hwoarang) proposed openstack/openstack-ansible master: Add support for using distribution packages for OpenStack services https://review.openstack.org/579770 | 18:19 |
| mnaser | hwoarang: i will do that in the future | 18:22 |
| mnaser | hwoarang: also.. i've been a bit mia but awesome progress on distro installs | 18:22 |
| mnaser | i see centos 7 just had 5 tempest failures but that's already amazing progress | 18:22 |
| mgariepy | mnaser, http://logs.openstack.org/41/581041/2/check/openstack-ansible-functional-centos-7/280aa0d/job-output.txt.gz#_2018-07-16_15_38_33_781112 | 18:23 |
| mgariepy | any idea on what's happening there? | 18:23 |
| mnaser | looks like an infra issue? | 18:24 |
| mnaser | or is it happening everytime? | 18:24 |
| mgariepy | almost every time | 18:25 |
| mgariepy | on this patch it passed once yesterday | 18:25 |
| mgariepy | failed since. sometime only 1 container sometimes more.. | 18:25 |
| mnaser | thats so confusing | 18:26 |
| mgariepy | race condition maybe ? | 18:26 |
| mnaser | "Install packages required to build neutron python package" being a task is already something that doesnt seeem right | 18:27 |
| mnaser | the role should install anything it needs? | 18:27 |
| cjloader | anyone know why i'm seeing this? | 18:27 |
| mnaser | https://github.com/openstack/openstack-ansible-tests/blob/master/test-install-neutron.yml#L60-L70 | 18:27 |
| cjloader | TASK [Copy global requirement pins file to host] AttributeError: 'Connection' object has no attribute 'container_user’ | 18:27 |
| cloudnull | http://paste.openstack.org/show/726034/ - anyone run into this ? | 18:28 |
| cjloader | in master | 18:28 |
| mnaser | mgariepy: https://github.com/openstack/openstack-ansible-os_neutron/blob/master/bindep.txt | 18:28 |
| *** KeithMnemonic has joined #openstack-ansible | 18:28 | |
| mnaser | it looks like bindep already takes care of this for us? | 18:28 |
| cloudnull | seeing this with rocky in w/ nova, seems haproxy isn't getting out of maintenance mode ? | 18:28 |
| mnaser | mgariepy: oh but this is inside the containers, hmm. | 18:28 |
| *** KeithMnemonic has quit IRC | 18:29 | |
| mnaser | mgariepy: honestly, not sure, i do believe this seems like a network issue tho. has it always been seeing that issue at rax? | 18:29 |
| hwoarang | mnaser: yeah the distro jobs on AIO repo is giving me 500 error on keystone. i am going to look at it again tomorrow but it's a bit of a mess to debug that | 18:29 |
| mnaser | cjloader: are you sure everything is up to date? | 18:29 |
| mgariepy | not only rax | 18:29 |
| mnaser | i know there was some changes recently in the plugins repo | 18:29 |
| mgariepy | seen limestone and ?inap? also | 18:30 |
| cjloader | i'm using master proper | 18:30 |
| cjloader | of osa | 18:30 |
| mnaser | cjloader: version of ansible up to date? | 18:30 |
| mnaser | cloudnull: in gate or locally? | 18:30 |
| mnaser | what does `nova-status upgrade check` yield locally? | 18:30 |
| cjloader | cloned down this morning | 18:31 |
| *** KeitMnemonic has joined #openstack-ansible | 18:31 | |
| -tux__- [root@ostack-osa /opt/openstack-ansible]# git checkout 16.0.15 | 18:32 | |
| -tux__- error: pathspec '16.0.15' did not match any file(s) known to git. | 18:32 | |
| tux__ | Try to do this https://docs.openstack.org/openstack-ansible/pike/user/minor-upgrade.html | 18:33 |
| tux__ | is that documentation issue because "git checkout -b 16.0.15" works | 18:34 |
| *** KeithMnemonic has joined #openstack-ansible | 18:39 | |
| tux__ | why its still showing older version | 18:39 |
| -tux__- [root@ostack-osa /opt/openstack-ansible]# git describe --tags | 18:39 | |
| -tux__- 16.0.13-7-g56099d7 | 18:39 | |
| *** KeithMnemonic has quit IRC | 18:40 | |
| tux__ | I am new in git and may asking stupid question | 18:40 |
| cloudnull | mnaser testing this on a 25 node cluster. | 18:48 |
| openstackgerrit | Andy Smith proposed openstack/openstack-ansible-os_neutron master: Setup oslo.messaging extra packages for optional drivers https://review.openstack.org/583021 | 18:54 |
| tux__ | why there is not tag 16.0.15 ??? http://paste.openstack.org/show/726039/ | 18:55 |
| tux__ | cloudnull: & Dmitriy: ? any thought | 18:56 |
| tux__ | or i am missing something | 18:57 |
| *** KeitMnemonic has quit IRC | 19:00 | |
| *** KeithMnemonic has joined #openstack-ansible | 19:01 | |
| jrosser | tux__: look here at branch->tags you see them all listed https://github.com/openstack/openstack-ansible | 19:07 |
| tux__ | my git pointing to https://git.openstack.org/openstack/openstack-ansible | 19:08 |
| tux__ | how do i switch it to github ? | 19:09 |
| tux__ | Can someone fix this document also ? https://docs.openstack.org/openstack-ansible/newton/newton/developer-docs/quickstart-aio.html | 19:09 |
| mgariepy | anyone seen issue with repo server on pike not updating ownership of the directory ? | 19:11 |
| tux__ | jrosser: why i am not able to see them in "git tag" ? | 19:15 |
| *** jproulx1 has joined #openstack-ansible | 19:15 | |
| jproulx1 | I'm taking a first run at OSA (from mitaka-eol for $REASONS) do I need to pre-create & configure br-mgmt (and other bridges) and related address or is that handled as part of the playbooks? | 19:17 |
| *** hachi has quit IRC | 19:18 | |
| *** girish has joined #openstack-ansible | 19:18 | |
| vakuznet | tux__, try: git checkout master; git pull; git tag | 19:18 |
| girish | hi can anybody help me configure ceph with openstack ansible | 19:19 |
| girish | i have /dev/sda5 on 6 blades which i need to use for all the storage requirements | 19:20 |
| girish | but when i configure ceph with openstack ansible i get error saying that keyring file is not generated by ansible-ceph | 19:20 |
| girish | An exception occurred during task execution. To see the full traceback, use -vvv. The error was: AnsibleFileNotFound: Could not find or access '/etc/opetack_deploy/ceph-fetch//c24b571b-894a-4b12-9c8a-774d26549b91/var/lib/ceph/bootstrap-osd/ceph.keyring' failed: [infra1] (item={u'name': u'/var/lib/ceph/bootstrap-osd/ceph.keyring', u'copy_key': True}) => {"changed": false, "failed": true, "item": {"copy_k": true, "name": "/va | 19:21 |
| tux__ | vakuznet: Thakns now i can see but this is what i am getting http://paste.openstack.org/show/726045/ | 19:21 |
| tux__ | girish: http://paste.openstack.org/ for output | 19:26 |
| *** vakuznet has quit IRC | 19:28 | |
| jrosser | tux__: you created a new branch with "git checkout -b 16.0.15" | 19:31 |
| jrosser | so you now have a branch called 16.0.15 and a tag, so thats why it complains about it | 19:31 |
| *** jproulx has joined #openstack-ansible | 19:31 | |
| *** girish has quit IRC | 19:32 | |
| jrosser | i would recommend a tool like gitk to help visualise this | 19:32 |
| tux__ | am i all set or do i need to do anything else jrosser:? | 19:32 |
| jrosser | from your description i *think* you have just created a new branch called 16.0.15 at the point you previously had checked out | 19:33 |
| tux__ | Can i delete everything and checkout 16.0.15 again.. to make it clean | 19:33 |
| jrosser | i don't think that is what you want | 19:33 |
| *** hachi has joined #openstack-ansible | 19:33 | |
| tux__ | currently i am running 16.0.13 and i want to upgrade to 16.0.15 | 19:33 |
| jrosser | git checkout master <- move off your new branch | 19:34 |
| jrosser | git branch -D 16.0.15 <- delete the branch you made | 19:34 |
| tux__ | done | 19:34 |
| jrosser | git checkout -b local_branch_name origin/remote_tag_name <- whatever you need | 19:35 |
| jrosser | this will all be much clearer with gitk or similar showing you what's going on | 19:35 |
| jrosser | you can then check that everything looks right by comparing the sha/log of the point you checked out with the branches on github | 19:36 |
| tux__ | "local_branch_name" what actually should i put? | 19:36 |
| tux__ | jrosser: i will try gitk (once this issue got resolved) | 19:36 |
| openstackgerrit | Andy Smith proposed openstack/openstack-ansible-os_cinder master: Setup oslo.messaging extra packages for optional drivers https://review.openstack.org/582632 | 19:36 |
| jrosser | i don't believe the branch name matters, someone will no doubt say if it does :) | 19:37 |
| openstackgerrit | Andy Smith proposed openstack/openstack-ansible master: [WIP] Add qdrouterd role for messaging backend deployment https://review.openstack.org/570261 | 19:37 |
| tux__ | can i do "git checkout -b 16.0.15" again? | 19:38 |
| tux__ | jrosser: ^^ | 19:38 |
| jrosser | personally i wouldnt, as thats the tag name and will just be confusing | 19:38 |
| tux__ | :( | 19:38 |
| tux__ | all i want checkout feaking 16.0.15 | 19:38 |
| tux__ | should i do "rm -rf /opt/openstack-ansible" and check out clean 16.0.15 ?? | 19:39 |
| jrosser | you shouldnt need to do that | 19:39 |
| tux__ | jrosser: then what are the option i have? | 19:40 |
| jrosser | 16.0.15 is a tag, just a pointer to a place in the repo, i showed you the syntax to check out a tag just before | 19:40 |
| jrosser | alternatively, as 16.0.15 is the current stable relase, you can checkout the branch stable/pike | 19:40 |
| tux__ | git checkout -b local_branch_name origin/remote_tag_name <--- what is local_brance_name here | 19:41 |
| tux__ | or should i just do "git checkout -b origin/16.0.15 | 19:41 |
| jrosser | no, becasue the field after -b signifies the local branch, that would not be a good choice | 19:41 |
| -tux__- [root@ostack-osa /opt/openstack-ansible]# git checkout stable/pike | 19:42 | |
| -tux__- Already on 'stable/pike' | 19:42 | |
| -tux__- Your branch is behind 'origin/stable/pike' by 23 commits, and can be fast-forwarded. | 19:42 | |
| -tux__- (use "git pull" to update your local branch) | 19:42 | |
| -tux__- [root@ostack-osa /opt/openstack-ansible]# | 19:42 | |
| jrosser | use lab_deploy or stable-pike or whatever fits for you | 19:42 |
| jrosser | right, so that tells you you have a *local* branch currently checked out called stable/pike | 19:43 |
| jrosser | and that the upstream code (origin) is ahead | 19:43 |
| jrosser | and it shows you how to move your locally checked out branch up to that point | 19:43 |
| tux__ | (use "git pull" to update your local branch) >>> | 19:43 |
| *** vakuznet has joined #openstack-ansible | 19:43 | |
| tux__ | should i do git pull? | 19:44 |
| tux__ | http://paste.openstack.org/show/726046/ | 19:44 |
| jrosser | if you want your checked out stable/pike to move forward to the opstream stable/pike, yes | 19:44 |
| jrosser | right good | 19:45 |
| -tux__- [root@ostack-osa /opt/openstack-ansible]# git describe | 19:45 | |
| -tux__- 16.0.15-7-gab73d49 | 19:45 | |
| tux__ | is it safe to run this tag in production? | 19:45 |
| jrosser | bingo :) | 19:45 |
| jrosser | well that is for you to decide, there is a new point release every couple of weeks | 19:46 |
| tux__ | jrosser: Thanks!! man ( I am old school guy using svn since born and git is new so trying to my head around :() | 19:46 |
| jrosser | and for the stable branches it's pretty much only bugfixes that get backported | 19:46 |
| jrosser | were you having trouble with spice console earlier? | 19:47 |
| tux__ | yes | 19:47 |
| tux__ | i am getting all short of issue so thinking to upgrade minor version | 19:47 |
| *** xoxefdp_ has joined #openstack-ansible | 19:47 | |
| jrosser | right, so you see there is a release note for that added create mode 100644 releasenotes/notes/spice-html5-repo-change-b4ad1c25ec94ee99.yaml | 19:47 |
| jrosser | in general i think it is a good plan to keep an eye on the point releases | 19:48 |
| jrosser | if nothing else it will show you bugs that have been fixed/backported and may be important to you | 19:48 |
| tux__ | Yup! now i realize | 19:48 |
| tux__ | for minor upgrade do i need to copy /etc/openstack_deploy back or ignore that step? | 19:50 |
| tux__ | jrosser: ^^ | 19:50 |
| jrosser | it's all in the documentation iirc | 19:51 |
| jrosser | https://docs.openstack.org/openstack-ansible/pike/user/minor-upgrade.html | 19:51 |
| tux__ | https://docs.openstack.org/openstack-ansible/pike/user/minor-upgrade.html | 19:51 |
| tux__ | Damn!! both on same page | 19:51 |
| tux__ | got it | 19:51 |
| tux__ | jrosser: is queen ready for production yet or still under construction ? | 19:52 |
| jrosser | queens is released and similar to pike just getting bugfixes, rocky is under construction | 19:53 |
| tux__ | should i use queen instead of pike before going in production? | 19:53 |
| *** chyka has quit IRC | 19:54 | |
| *** KeithMnemonic has quit IRC | 19:57 | |
| *** kberger has joined #openstack-ansible | 19:57 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Add additional ES cluster tuning https://review.openstack.org/583039 | 19:58 |
| cloudnull | jrosser ^ | 19:58 |
| jrosser | cloudnull: will look | 19:59 |
| jrosser | cloudnull: does interface speed deal with bonds? :) | 19:59 |
| *** cshen has joined #openstack-ansible | 20:00 | |
| cloudnull | jproulx1 If the setup is a production environment (real hardware) OSA will not setup the bridges for you | 20:00 |
| cloudnull | if you are testing an AIO or something similar the AIO scripts will do all of the interface setup needed | 20:00 |
| cloudnull | jrosser it should | 20:00 |
| cloudnull | http://paste.openstack.org/show/726049/ | 20:01 |
| cloudnull | thats a quick playbook i used to test the new dynamic variable | 20:02 |
| jproulx | cloudnull thanks | 20:02 |
| *** xoxefdp_ has quit IRC | 20:03 | |
| *** chyka has joined #openstack-ansible | 20:12 | |
| *** chyka has quit IRC | 20:14 | |
| *** chyka has joined #openstack-ansible | 20:14 | |
| guilhermesp | is there a "wizard" button on gerrit to revert one patch set to a previous one or do I do this via cli submmiting another patch set? | 20:15 |
| guilhermesp | in this case from 21 to 20 | 20:15 |
| guilhermesp | https://review.openstack.org/#/c/578721/21 | 20:15 |
| *** jproulx has quit IRC | 20:26 | |
| cloudnull | guilhermesp you can fetch your old pr, amend the commit, and then re-review it | 20:27 |
| cloudnull | under teh download button there's a checkout option | 20:27 |
| cloudnull | copy and paste that link into the terminal to checkout the previous version of your commit | 20:28 |
| guilhermesp | cloudnull thanks man, I already use the fetch commands under download ( a lot in this PR hahah ). Great, so answered, I will need to submmit a new patch set ;) | 20:29 |
| *** tux__ has quit IRC | 20:30 | |
| openstackgerrit | Guilherme Steinmuller Pimentel proposed openstack/openstack-ansible-os_monasca master: Fix monasca https://review.openstack.org/578721 | 20:30 |
| guilhermesp | much better then wizards actually (wizards are unpredictable sometimes) | 20:31 |
| *** DanyC has joined #openstack-ansible | 20:39 | |
| *** devx has quit IRC | 20:40 | |
| *** devx has joined #openstack-ansible | 20:41 | |
| openstackgerrit | Markos Chandras (hwoarang) proposed openstack/openstack-ansible-os_tempest master: DNM - Testing OBS Master https://review.openstack.org/567160 | 20:46 |
| jrosser | cloudnull: http://paste.openstack.org/show/726052/ | 20:51 |
| jrosser | cloudnull: also http://paste.openstack.org/show/726053/ | 20:57 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Add additional ES cluster tuning https://review.openstack.org/583039 | 20:57 |
| cloudnull | jrosser looks like we need to only apply that setting via the api | 21:00 |
| jrosser | and the second one either renamed or removed | 21:01 |
| openstackgerrit | Logan V proposed openstack/openstack-ansible-plugins master: tests: Do not use ansible_host to pass container IP https://review.openstack.org/582329 | 21:02 |
| openstackgerrit | Logan V proposed openstack/openstack-ansible-plugins master: Allow container detection when host matches container_name https://review.openstack.org/582330 | 21:02 |
| jrosser | i still think the speed calculation only sees half the interfaces | 21:02 |
| jrosser | but opening it up to 75% of 10G is no doubt fine | 21:02 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Add additional ES cluster tuning https://review.openstack.org/583039 | 21:03 |
| *** hachi has quit IRC | 21:05 | |
| openstackgerrit | Andy Smith proposed openstack/openstack-ansible-os_neutron master: Setup oslo.messaging extra packages for optional drivers https://review.openstack.org/583021 | 21:09 |
| cloudnull | jrosser that should make it happy | 21:10 |
| cloudnull | I'm updating my elk cluster with this all now | 21:10 |
| jrosser | mine is really hurting | 21:12 |
| jrosser | load avg of 10+ on the data nodes | 21:13 |
| openstackgerrit | Victor Palma proposed openstack/openstack-ansible-ops master: adding kolide fleet https://review.openstack.org/583046 | 21:13 |
| cloudnull | hum... | 21:13 |
| cloudnull | same issue? | 21:13 |
| cloudnull | with the stack trace? | 21:13 |
| jrosser | no, i commented those two settings out for now | 21:13 |
| jrosser | this isnt great [log1-elastic-logstash-container-26065c09] [gc][301] overhead, spent [503ms] collecting in the last [1s] | 21:14 |
| cloudnull | something to do with that last change? | 21:15 |
| cloudnull | or something else ? | 21:15 |
| jrosser | in fact load avg of 17 now on the elected master - thats going to end badly i suspect | 21:15 |
| jrosser | theres not much really in the last change apart from relaxing the network trottle | 21:15 |
| jrosser | i expect it's working through a backlog of the last ~10mins of queued up data | 21:17 |
| jrosser | and i have a bunch of shards in recovery | 21:19 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Add additional ES cluster tuning https://review.openstack.org/583039 | 21:19 |
| cloudnull | ok. so those last two settings have to be applied to the indexes | 21:20 |
| cloudnull | and are no longer configured on the nodes. | 21:21 |
| cloudnull | after I removed those settings from the config my cluster state went back to green | 21:22 |
| jrosser | it looks like there is some threshold above which it starts making those gc log messages | 21:23 |
| jrosser | they are probably worth looking out for, as it's an indicator that everyhting is getting bogged down with jvm memory managment gone nuts | 21:23 |
| jenglisch_ | hm, whats the 'best' way to configure neutron to use 1500 mtu for vxlan tunnel networks when the underlying network already uses jumboframes? | 21:25 |
| jenglisch_ | there is no possiblity to set any neutron mtu settings using ansible variables and also no 'additional settings' option available | 21:25 |
| jenglisch_ | i'm considering a custom ml2 template, but i'm curious if noone else had that problem before | 21:26 |
| cloudnull | jrosser http://paste.openstack.org/show/726055/ - i have a mess of those GC log messages, idk if that's expected or something that needs to be tuned ? | 21:26 |
| *** tux_ has joined #openstack-ansible | 21:26 | |
| logan- | jenglisch_: https://github.com/continuouspw/openstack-ci-cloud/blob/a6b54afc23ffe26547da9df38b029ff00a734f27/openstack_deploy/user_variables.yml#L37-L40 | 21:26 |
| logan- | try applying that and then create a new vxlan network | 21:27 |
| logan- | it should have 1500 tenant mtu | 21:27 |
| jrosser | cloudnull: yeah i got a whole bunch of them when i restarted the cluster - the load avg went mad then it's all settled down | 21:27 |
| jrosser | however i have a feeling we're missing collecting the elasticsearch logs | 21:27 |
| cloudnull | jrosser you are right | 21:28 |
| jenglisch_ | logan-: perfect, ty, i crawled through the templates and forgot that ansible has a module for modifying ini files directly | 21:28 |
| cloudnull | http://paste.openstack.org/show/726056/ | 21:29 |
| cloudnull | the elk logs are not being bind mounted to the host | 21:29 |
| jrosser | we could filebeat from the elk container | 21:30 |
| jrosser | detect /var/log/elasticsearch | 21:30 |
| cloudnull | that we could | 21:31 |
| jenglisch_ | logan-: is setting the path_mtu in the ml2 config optional ? https://docs.openstack.org/newton/networking-guide/config-mtu.html | 21:31 |
| *** tux_ has quit IRC | 21:31 | |
| logan- | jenglisch_: are you running newton? if so .. it is probably required | 21:31 |
| logan- | all of the mtu stuff was reworked at some point after newton, in ocata or pike | 21:32 |
| cloudnull | jrosser http://paste.openstack.org/show/726057/ would bind mount /var/log onto the host like we do in the rest of OSA however that cloud/would make the container very unhappy if we do it to an existing environment. | 21:32 |
| logan- | my example is running on a pike cloud | 21:32 |
| logan- | so I think if you are running on pike it will not be required, earlier than that and I am not sure :) | 21:32 |
| jenglisch_ | nope. running queens, and read newton as neutron :/ | 21:33 |
| *** nurdie has quit IRC | 21:33 | |
| *** jwitko has quit IRC | 21:33 | |
| *** nurdie has joined #openstack-ansible | 21:34 | |
| jrosser | cloudnull: kibana seems to log to the journal - i wonder if thats a better approach | 21:34 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Add additional ES cluster tuning https://review.openstack.org/583039 | 21:34 |
| cloudnull | jrosser That's a good idea | 21:35 |
| jrosser | then theres no magic wiring above that thats already there | 21:35 |
| *** nurdie has quit IRC | 21:38 | |
| *** hachi has joined #openstack-ansible | 21:42 | |
| *** ianw_pto is now known as ianw | 21:44 | |
| *** cshen has quit IRC | 21:44 | |
| *** lbragstad has quit IRC | 21:47 | |
| *** FrankZhang has quit IRC | 21:48 | |
| openstackgerrit | Logan V proposed openstack/openstack-ansible-plugins master: Allow container detection when host matches container_name https://review.openstack.org/582330 | 21:52 |
| openstackgerrit | Merged openstack/ansible-role-systemd_service master: Correct log warning for unknown lvalue https://review.openstack.org/582997 | 21:53 |
| *** throwsb1 has quit IRC | 21:53 | |
| *** admin0 has quit IRC | 21:53 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Add config overrides for systemd for better auditing https://review.openstack.org/583057 | 22:00 |
| *** lbragstad has joined #openstack-ansible | 22:04 | |
| *** tux_ has joined #openstack-ansible | 22:06 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Add config overrides for systemd for better auditing https://review.openstack.org/583057 | 22:07 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Add config overrides for systemd for better auditing https://review.openstack.org/583057 | 22:09 |
| cloudnull | jrosser ok, so that ^ makes our elastic-stack use the journal | 22:09 |
| *** hachi has quit IRC | 22:09 | |
| *** tux_ has quit IRC | 22:11 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Add config overrides for systemd for better auditing https://review.openstack.org/583057 | 22:11 |
| *** cshen has joined #openstack-ansible | 22:20 | |
| *** cshen has quit IRC | 22:24 | |
| *** tux_ has joined #openstack-ansible | 22:28 | |
| *** jproulx has joined #openstack-ansible | 22:29 | |
| *** kstev has quit IRC | 22:32 | |
| *** dave-mccowan has quit IRC | 22:33 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Add additional ES cluster tuning https://review.openstack.org/583039 | 22:42 |
| *** jproulx has left #openstack-ansible | 22:42 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Add additional ES cluster tuning https://review.openstack.org/583039 | 22:45 |
| openstackgerrit | Logan V proposed openstack/openstack-ansible-plugins master: Allow container detection when host matches container_name https://review.openstack.org/582330 | 22:48 |
| openstackgerrit | Logan V proposed openstack/openstack-ansible-plugins master: Allow container detection when host matches container_name https://review.openstack.org/582330 | 22:51 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Add config overrides for systemd for better auditing https://review.openstack.org/583057 | 22:51 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Add additional ES cluster tuning https://review.openstack.org/583039 | 22:52 |
| *** pmannidi has joined #openstack-ansible | 22:54 | |
| *** DanyC has quit IRC | 22:56 | |
| *** DanyC has joined #openstack-ansible | 22:57 | |
| *** DanyC has quit IRC | 23:01 | |
| *** tosky has quit IRC | 23:15 | |
| *** poul_ has quit IRC | 23:18 | |
| tux_ | Folks any idea what is going on with Install PIP http://paste.openstack.org/show/726058/ | 23:23 |
| tux_ | global-requirement-pins.txt has pip==9.0.3 | 23:26 |
| *** poul has joined #openstack-ansible | 23:27 | |
| tux_ | I am following https://docs.openstack.org/openstack-ansible/pike/user/minor-upgrade.html | 23:28 |
| tux_ | At ./scripts/bootstrap-ansible.sh got this warning "You are using pip version 9.0.3, however version 10.0.1 is available." | 23:29 |
| tux_ | is it a good idea to switch pip from 9.0.3 to 10.0.1 ? | 23:29 |
| *** poul has quit IRC | 23:31 | |
| *** poul has joined #openstack-ansible | 23:36 | |
| *** poul has quit IRC | 23:41 | |
| openstackgerrit | Andy Smith proposed openstack/openstack-ansible-os_cinder master: Setup oslo.messaging extra packages for optional drivers https://review.openstack.org/582632 | 23:41 |
| *** Adri2000 has quit IRC | 23:43 | |
| *** Adri2000 has joined #openstack-ansible | 23:47 | |
| *** poul has joined #openstack-ansible | 23:53 | |
| tux_ | does someone ran is this issue ? | 23:54 |
| tux_ | cloudnull: ^^ | 23:54 |
| tux_ | Dmitriy: ^^ | 23:54 |
| *** chyka has quit IRC | 23:55 | |
| *** poul has quit IRC | 23:58 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!