*** brokencycle has quit IRC | 00:19 | |
*** gyee has quit IRC | 00:22 | |
*** markvoelker has joined #openstack-ansible | 00:25 | |
*** mpjetta has quit IRC | 00:27 | |
*** markvoelker has quit IRC | 00:29 | |
*** dave-mccowan has quit IRC | 00:47 | |
*** macz has quit IRC | 00:52 | |
*** dave-mccowan has joined #openstack-ansible | 00:52 | |
*** markvoelker has joined #openstack-ansible | 01:00 | |
cjloader | https://review.opendev.org/#/c/682446/ and https://review.opendev.org/#/c/680284/ passed gates (Ironic packages/Ironic Inspector) noonedeadpunk jrosser please look in the morning =) | 02:13 |
---|---|---|
*** markvoelker has quit IRC | 02:47 | |
*** BjoernT has joined #openstack-ansible | 02:51 | |
*** nicolasbock has quit IRC | 02:52 | |
*** BjoernT_ has joined #openstack-ansible | 02:54 | |
*** BjoernT has quit IRC | 02:56 | |
*** udesale has joined #openstack-ansible | 04:04 | |
*** vakuznet has quit IRC | 04:16 | |
*** pcaruana has joined #openstack-ansible | 04:42 | |
*** gillesMo has quit IRC | 05:07 | |
openstackgerrit | Merged openstack/openstack-ansible-os_ironic master: Remove UCS packages from os_ironic https://review.opendev.org/682446 | 05:08 |
cjloader | ty d34dh0r53 for pushing it through | 05:09 |
cjloader | now need to focus on my inspector patch noonedeadpunk jrosser | 05:09 |
cjloader | its passed gates | 05:14 |
*** radeks has joined #openstack-ansible | 05:28 | |
ChosSimbaOne | Good morning. So finally got the lb working. Had to purge my containers as it added the bond interrface to the container, and f'ed conectivity to the host :-) | 05:38 |
ChosSimbaOne | Would you guys recommend an dedicated neutron node on metal for queens? This is for a pilot deployment of openstack with max 10 compute nodes. Connectivity is 2x25Gbit, so I was thinking it could offload the rest of the service on the infra node by having a dedicated neutron server? | 05:40 |
*** BjoernT_ has quit IRC | 05:45 | |
jrosser | ChosSimbaOne: queens? really? :) | 06:00 |
ChosSimbaOne | jrosser: Yeah, well long story short, we would have integrated with ACI and our network consultant told us that the ACI module from cisco was only supported in Queens so that is why :-) | 06:01 |
jrosser | this is why i have avoided any kind of vendor magic in networking completely | 06:02 |
jrosser | cisco constantly try to get us doing ACI for stuff and we just say no every time | 06:03 |
ChosSimbaOne | Yeah, I am starting to regread that decision as well, but to late now :-S | 06:03 |
ChosSimbaOne | I wish we had that kind of will power :-) | 06:03 |
jrosser | well anyway, i have dedicated network nodes, and did that back on queens too | 06:04 |
jrosser | so it can be done and does work | 06:04 |
jrosser | i would strongly suggest doing "whatever it takes" to make the any interface names you have to specify to neutron to be uniform across your compute and network nodes | 06:05 |
ChosSimbaOne | Okay super. Well i guess it should just be connected to the br-mgmt and br-vxlan right? and the set the is_metal: true in an env.d file right? | 06:08 |
jrosser | the neutron agent is on metal anyway for queens afaik? | 06:13 |
jrosser | imho env.d should be fiddled to put the neutron api server on the infra node and the l3 agent only on the network node | 06:14 |
ChosSimbaOne | Cool I will dig into that, and let you know how it goes :-) | 06:17 |
*** kopecmartin|off is now known as kopecmartin | 06:18 | |
*** prometheanfire has quit IRC | 06:23 | |
*** prometheanfire has joined #openstack-ansible | 06:24 | |
*** markvoelker has joined #openstack-ansible | 06:27 | |
*** markvoelker has quit IRC | 06:31 | |
*** cshen has joined #openstack-ansible | 06:52 | |
*** udesale has quit IRC | 06:57 | |
*** udesale has joined #openstack-ansible | 06:58 | |
*** arbrandes has quit IRC | 07:00 | |
*** arbrandes has joined #openstack-ansible | 07:01 | |
*** rpittau|afk is now known as rpittau | 07:05 | |
cshen | Good morning. Does anybody deep know about rpc_conn_pool_size used in neutron.conf and nova.conf? Is this parameter shared with all rpc workers, or each worker has its own conn pool? | 07:15 |
*** udesale has quit IRC | 07:18 | |
*** udesale has joined #openstack-ansible | 07:19 | |
*** tosky has joined #openstack-ansible | 07:23 | |
*** luksky has joined #openstack-ansible | 07:37 | |
*** ivve has joined #openstack-ansible | 07:45 | |
*** kaiokmo has quit IRC | 07:57 | |
*** zbr is now known as zbr|ruck | 07:57 | |
*** kaiokmo has joined #openstack-ansible | 08:02 | |
*** arxcruz|ruck is now known as arxcruz|rover | 08:20 | |
*** markvoelker has joined #openstack-ansible | 08:28 | |
*** markvoelker has quit IRC | 08:33 | |
*** kaiokmo has quit IRC | 08:36 | |
*** kaiokmo has joined #openstack-ansible | 08:37 | |
CeeMac | morning | 08:38 |
jrosser | CeeMac: morning..... do i remember right you had dedicated network nodes and moved the neutron API off them? | 08:40 |
CeeMac | hi jrosser | 08:47 |
CeeMac | correct | 08:47 |
jrosser | could you share what you did to env.d? | 08:47 |
CeeMac | let me poke in my notes | 08:47 |
CeeMac | its been a while and i've been distracted with non-osa stuff (boo) | 08:47 |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_barbican master: Add global override for service bind address https://review.opendev.org/683062 | 08:49 |
CeeMac | ChosSimbaOne: jrosser http://paste.openstack.org/show/777611/ | 08:56 |
CeeMac | thats what I needed to do to move neutron-server from network nodes to infra nodes | 08:57 |
jrosser | excellent thanks, i'm going to do the same | 08:57 |
CeeMac | :) | 08:58 |
ChosSimbaOne | CeeMac: Thank you so much! This is a really great help :-) | 08:58 |
ChosSimbaOne | Will try and deploy right away. | 08:59 |
CeeMac | sure thing, good luck :) | 08:59 |
ChosSimbaOne | So is the neutron service setup on directly on the infra host normally or is everything in the container neutron_server? | 08:59 |
CeeMac | container for neutron_server on infra | 08:59 |
CeeMac | neutron agents on bare metal on compute nodes for me | 09:00 |
CeeMac | i'm on rocky though, not sure if that makes a difference | 09:00 |
CeeMac | also, if you run into issues with network and compute nodes having differnet physical interface names there is a workaround for that, although I don't know if that works in queens | 09:01 |
jrosser | i think that was only added later | 09:01 |
CeeMac | yeah, think it was backported to rocky | 09:01 |
jrosser | but you can work around it by adding neutron.conf overrides to group_vars/<things> rather than the documented setup in openstack_user_config->provider_networks | 09:02 |
jrosser | but that gets messy | 09:02 |
jrosser | ChosSimbaOne: with the standard setup if you make a dedicated network node the neutron API container and the L3 agent will run on the network node | 09:03 |
jrosser | i want to move the API container back to the infra nodes | 09:03 |
ChosSimbaOne | Okay, but if i have already run the osa playbooks where my network host was the infra node, do i need to do some cleanup or should purging the container be enough. | 09:04 |
ChosSimbaOne | God advice with the api. Will see if can get it to work :-) | 09:06 |
jrosser | ChosSimbaOne: there are two parts | 09:09 |
jrosser | there is an API part and the agent | 09:09 |
jrosser | the agent is making your routers and dhcp servers for project networks | 09:10 |
ChosSimbaOne | jrosser: Okay. But without a dedicated network host it all runs in the neutron_server_container? | 09:17 |
ChosSimbaOne | CeeMac: the opendalight is only relevant if I have the opendaylight SDN right? | 09:19 |
jrosser | no, the l3 agent is on bare metal, that used to be containerised but at some point around P/Q that was changed | 09:19 |
jrosser | and the neutron API runs in the container | 09:19 |
ChosSimbaOne | okay, now I follow :) Thank you for the clarificaton. | 09:21 |
CeeMac | ChosSimbaOne: where were you seeing the opendaylight? | 09:27 |
CeeMac | actually, i see it now | 09:27 |
CeeMac | that was in the original skel so i just left it there to keep things straight | 09:27 |
CeeMac | you could try leaving it out if you wanted | 09:28 |
ChosSimbaOne | OKay, super :-) | 09:28 |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_cinder master: Add global override for service bind address https://review.opendev.org/683074 | 09:28 |
noonedeadpunk | mornings | 09:29 |
ChosSimbaOne | CeeMac: Super, will try and see if my world explodes :-) | 09:30 |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_blazar master: Add global override for service bind address https://review.opendev.org/683075 | 09:30 |
noonedeadpunk | its high time upgrade jobs got reviewed:) https://review.opendev.org/#/c/671105/ | 09:30 |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_congress master: Add global override for service bind address https://review.opendev.org/683076 | 09:32 |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_horizon master: Allow to configure horizon bind address https://review.opendev.org/681718 | 09:32 |
noonedeadpunk | jrosser: ^ updated horizon patch to use openstack_service_bind_address | 09:33 |
jrosser | noonedeadpunk: hi | 09:33 |
noonedeadpunk | but I'm not sure right now what's the idea - will veriable with container address work properly with lxc? | 09:33 |
jrosser | my plan is to put openstack_service_bind_address in group_vars/all.yml | 09:34 |
jrosser | openstack_service_bind_address: "{{ management_address }}" | 09:34 |
noonedeadpunk | yeah, ok, I guess this should work then | 09:34 |
noonedeadpunk | I think we'll need to heavily update our docs during next cycle:) | 09:36 |
jrosser | yes | 09:36 |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_glance master: Add global override for service bind address https://review.opendev.org/683080 | 09:41 |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_gnocchi master: Add global override for service bind address https://review.opendev.org/683082 | 09:43 |
*** jawad_axd has joined #openstack-ansible | 09:44 | |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_heat master: Add global override for service bind address https://review.opendev.org/683084 | 09:45 |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_ironic master: Add global override for service bind address https://review.opendev.org/683086 | 09:47 |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_keystone master: Add global override for service bind address https://review.opendev.org/683089 | 09:52 |
*** udesale has quit IRC | 10:01 | |
*** cshen has quit IRC | 10:09 | |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_swift master: Revert "Update fallocate_reserve to work with py3" https://review.opendev.org/683102 | 10:13 |
*** cshen has joined #openstack-ansible | 10:25 | |
*** markvoelker has joined #openstack-ansible | 10:29 | |
*** cshen has quit IRC | 10:31 | |
*** udesale has joined #openstack-ansible | 10:32 | |
*** cshen has joined #openstack-ansible | 10:33 | |
*** markvoelker has quit IRC | 10:34 | |
*** trident has quit IRC | 10:35 | |
*** udesale has quit IRC | 10:37 | |
*** udesale has joined #openstack-ansible | 10:37 | |
*** cshen has quit IRC | 10:37 | |
*** cshen has joined #openstack-ansible | 10:38 | |
*** trident has joined #openstack-ansible | 10:48 | |
*** pcaruana has quit IRC | 10:54 | |
*** rpittau is now known as rpittau|afk | 10:56 | |
*** udesale has quit IRC | 11:01 | |
*** nicolasbock has joined #openstack-ansible | 11:06 | |
jrosser | noonedeadpunk: we need this https://review.opendev.org/#/c/682830/ | 11:10 |
jawad_axd | Hi! Can anyone suggest what is wrong with these settings http://paste.openstack.org/show/777691/ ? This is scenario when you have different network interfaces names on compute nodes and you need to define them for neutron_provider_networks.Would appreciate your help in this. | 11:10 |
noonedeadpunk | jrosser: sure, I saw it (and even thought that voted on it) | 11:11 |
noonedeadpunk | or decided to postpone reviews for later time:) | 11:12 |
noonedeadpunk | btw, jrosser, do you know how to get delegated host group_vars during delegate_to? | 11:12 |
jrosser | jawad_axd: i think the issue might be that you are trying to set the same var in more than one place, and user_variables.yml is the highest priority and will win over everything else | 11:13 |
jrosser | it is passed as a -e parameter to ansible-playbook which is highest precedence | 11:13 |
jrosser | noonedeadpunk: group_vars specifically? i think the only thing is hostvars[<thing>] ? | 11:14 |
jawad_axd | I need to define separate network_mappings variable for compute and network node. How can I do that? | 11:14 |
jrosser | jawad_axd: which release? | 11:15 |
jawad_axd | stein | 11:15 |
jawad_axd | centos7 | 11:15 |
noonedeadpunk | yep, I'm trying to achieve step 8 from https://docs.openstack.org/murano/latest/install/from-source.html#install-the-api-service-and-engine | 11:16 |
noonedeadpunk | by delegating to utility host, which obviously supposed top have python-muranoclient | 11:16 |
*** pcaruana has joined #openstack-ansible | 11:17 | |
noonedeadpunk | but, I need utility_venv_bin, which is defined here https://opendev.org/openstack/openstack-ansible/src/branch/master/inventory/group_vars/utility_all.yml | 11:17 |
noonedeadpunk | hostvars are empty for some reason:( | 11:18 |
jrosser | jawad_axd: there is excellent docuemtnation by jamesdenton here https://docs.openstack.org/openstack-ansible/latest/user/prod/provnet_groups.html | 11:18 |
jrosser | jawad_axd: you can do that in stein | 11:18 |
*** ansmith has quit IRC | 11:21 | |
jrosser | noonedeadpunk: i wonder if instead it is easier to do that from the murano_api[0] (or whatever it is called) host? | 11:22 |
noonedeadpunk | yeah, probably it's easier:) | 11:22 |
jrosser | then you use the murano venv, and you have (maybe?) all those files that need uploading already there? | 11:22 |
noonedeadpunk | but utiity is probably more porper place for such operations... | 11:22 |
jrosser | perhaps delegate_to: "{{ murano_service_bootstrap_host }}" | 11:22 |
jrosser | where do you make the zip file? | 11:23 |
noonedeadpunk | I was thinking about utility | 11:23 |
noonedeadpunk | so, I need to clone repo, do archive, and run murano client | 11:24 |
jrosser | does all the zip file content get installed with the service? | 11:24 |
noonedeadpunk | but yep, the easier thing to do this on murano host | 11:24 |
jrosser | what actually is it btw in the zip? | 11:25 |
noonedeadpunk | nope, they don't install it, which is pretty ridiculous | 11:25 |
jrosser | hah | 11:25 |
jawad_axd | jrosser: It helped. Thanks | 11:26 |
noonedeadpunk | so, it's some internal core packages without which murano can't init any environment | 11:26 |
noonedeadpunk | the most dissapointing thing, that they ship some of them http://paste.openstack.org/show/777696/ | 11:27 |
noonedeadpunk | but not the essential one | 11:27 |
noonedeadpunk | which is io.murano.Environment | 11:27 |
*** macz has joined #openstack-ansible | 11:28 | |
noonedeadpunk | so yep, I'll probably do this on murano_api[0] | 11:30 |
noonedeadpunk | if you don't mind:) | 11:30 |
jawad_axd | jrosser: You guys are doing great work and must be busy with your priority work.I just want to know if some came up with fix for stein/centos7 heat-engine service issue http://paste.openstack.org/show/777695/ http://paste.openstack.org/show/777694/ | 11:32 |
*** macz has quit IRC | 11:32 | |
jrosser | noonedeadpunk: also we use the deployment host a lot for this | 11:39 |
jrosser | download to the deploy host | 11:39 |
jrosser | zip up / whatever | 11:40 |
jrosser | copy to target, delegate to murano[0] and use the cli | 11:40 |
jrosser | thats very similar to how we do things like HATop and is also friendly to environments where the compute nodes don't have much external connectivity, but the deploy host might | 11:41 |
noonedeadpunk | it will slow down things because of upload - no sense in copying it as this can be done remotely.... | 11:41 |
jrosser | you need to think of people who want near offline installs too | 11:41 |
noonedeadpunk | the still will need to set murano_git_repo | 11:42 |
noonedeadpunk | we don't build venvs on the deployment host, are we? | 11:43 |
noonedeadpunk | and these files are part of the default murano_git_repo | 11:47 |
openstackgerrit | Ralf Haferkamp proposed openstack/ansible-role-uwsgi master: Fix distro install on openSUSE https://review.opendev.org/683120 | 11:59 |
*** goldyfruit_ has quit IRC | 12:11 | |
openstackgerrit | Ralf Haferkamp proposed openstack/ansible-role-uwsgi master: Fix distro install on openSUSE https://review.opendev.org/683120 | 12:13 |
*** markvoelker has joined #openstack-ansible | 12:13 | |
openstackgerrit | Ralf Haferkamp proposed openstack/openstack-ansible-os_placement master: Fix distro install on openSUSE https://review.opendev.org/683123 | 12:16 |
*** vakuznet has joined #openstack-ansible | 12:16 | |
openstackgerrit | Andrew Gibb proposed openstack/openstack-ansible-os_trove master: Add global override for service bind address https://review.opendev.org/683129 | 12:25 |
*** mgariepy has quit IRC | 12:37 | |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_magnum master: Add global override for service bind address https://review.opendev.org/683133 | 12:43 |
*** ansmith has joined #openstack-ansible | 12:44 | |
openstackgerrit | Andrew Gibb proposed openstack/openstack-ansible-os_tacker master: Add global override for service bind address https://review.opendev.org/683134 | 12:44 |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_manila master: Add global override for service bind address https://review.opendev.org/683136 | 12:46 |
*** markvoelker has quit IRC | 12:46 | |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_masakari master: Add global override for service bind address https://review.opendev.org/683138 | 12:48 |
openstackgerrit | Andrew Gibb proposed openstack/openstack-ansible-os_sahara master: Add global override for service bind address https://review.opendev.org/683139 | 12:49 |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_mistral master: Add global override for service bind address https://review.opendev.org/683140 | 12:49 |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_neutron master: Add global override for service bind address https://review.opendev.org/683141 | 12:50 |
openstackgerrit | Andrew Gibb proposed openstack/openstack-ansible-os_placement master: Add global override for service bind address https://review.opendev.org/683142 | 12:53 |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_nova master: Add global override for service bind address https://review.opendev.org/683143 | 12:54 |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_octavia master: Add global override for service bind address https://review.opendev.org/683144 | 12:55 |
openstackgerrit | Andrew Gibb proposed openstack/openstack-ansible-os_panko master: Add global override for service bind address https://review.opendev.org/683145 | 12:56 |
jrosser | jawad_axd: how did you configure it to log to systemd? in stein the default is to log to files afaik | 13:07 |
*** SmearedBeard has joined #openstack-ansible | 13:08 | |
jawad_axd | I did set use_journal = True in /etc/heat/heat.conf. | 13:08 |
jawad_axd | I have two stein setup both are using 19.0.2 tag. Both setups have same issue with heat engine service. | 13:12 |
jrosser | this will be becasue there is nowhere in stein that the systemd-python library is installed for heat, i expect | 13:13 |
jrosser | in master / train we switched all the logging to systemd so these things would have been fixed on that branch | 13:14 |
jrosser | jawad_axd: here is the patch in master https://github.com/openstack/openstack-ansible-os_heat/commit/88c29bb66547d791c0035f688d5097a7a7632fd7 | 13:14 |
jrosser | noonedeadpunk: i am tracking the bind-to-mgmt stuff a bit here https://etherpad.openstack.org/p/osa-bind-to-mgmt | 13:16 |
jrosser | the roles with missing patches don't have an obvious place in defaults/main.yml that the bind address is set | 13:16 |
jrosser | they need a bit more investigation | 13:16 |
*** goldyfruit_ has joined #openstack-ansible | 13:16 | |
noonedeadpunk | ok, will have a look at them | 13:18 |
jawad_axd | jrosser: Thanks. I am looking into it. I am improving my python skills, would love to contribute in coming time. | 13:18 |
jrosser | noonedeadpunk: i'm going to -W the existing patches for now and make a DNM to test them all in the integrated repo | 13:19 |
jrosser | jawad_axd: contributions are always welcome :) | 13:19 |
*** macz has joined #openstack-ansible | 13:21 | |
*** udesale has joined #openstack-ansible | 13:23 | |
*** udesale has quit IRC | 13:24 | |
*** udesale has joined #openstack-ansible | 13:25 | |
*** macz has quit IRC | 13:26 | |
*** pcaruana has quit IRC | 13:28 | |
jawad_axd | jrosser: That patch worked !!! Thumbs up man. Heat-engine service got started properly after patch. | 13:40 |
*** jawad_axd has quit IRC | 13:51 | |
*** jawad_axd has joined #openstack-ansible | 13:53 | |
*** pcaruana has joined #openstack-ansible | 13:54 | |
*** jawad_axd has quit IRC | 13:57 | |
*** BjoernT has joined #openstack-ansible | 13:59 | |
*** BjoernT_ has joined #openstack-ansible | 14:02 | |
*** macz has joined #openstack-ansible | 14:03 | |
*** BjoernT has quit IRC | 14:04 | |
*** macz has quit IRC | 14:07 | |
*** ivve has quit IRC | 14:09 | |
*** goldyfruit___ has joined #openstack-ansible | 14:15 | |
*** goldyfruit_ has quit IRC | 14:18 | |
*** macz has joined #openstack-ansible | 14:19 | |
*** macz has quit IRC | 14:23 | |
*** pcaruana has quit IRC | 14:33 | |
*** SmearedBeard has quit IRC | 14:33 | |
*** spatel has joined #openstack-ansible | 14:33 | |
*** jawad_axd has joined #openstack-ansible | 14:40 | |
*** jawad_axd has quit IRC | 14:44 | |
*** macz has joined #openstack-ansible | 14:45 | |
mnaser | jrosser: did you use my tool to generate those patches by any chance :D | 14:48 |
jrosser | ctrl-r | 14:49 |
mnaser | aww | 14:49 |
jrosser | type fast :) | 14:49 |
mnaser | sorry i cant keep up | 14:49 |
jrosser | no worries | 14:50 |
*** macz has quit IRC | 14:50 | |
jrosser | we’re still churning stuff out to finish various bits | 14:50 |
jrosser | but that’s soon going to turn into needing to merge it all which might need a few more bodies | 14:51 |
*** gyee has joined #openstack-ansible | 14:56 | |
*** tosky has quit IRC | 14:57 | |
*** tosky has joined #openstack-ansible | 14:59 | |
mnaser | ive personally been pretty backed up with a lot of stuff going on until ansiblefest | 15:00 |
*** gillesMo has joined #openstack-ansible | 15:00 | |
mnaser | i think i'll be able to breathe a bit more when it comes time for that | 15:00 |
*** gillesMo has quit IRC | 15:00 | |
*** cshen has quit IRC | 15:11 | |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: Add support to deploy murano https://review.opendev.org/679496 | 15:14 |
*** gillesMo has joined #openstack-ansible | 15:20 | |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_murano master: Initial commit to os_murano role https://review.opendev.org/677830 | 15:32 |
*** kopecmartin is now known as kopecmartin|off | 15:35 | |
*** jawad_axd has joined #openstack-ansible | 15:42 | |
*** jawad_axd has quit IRC | 15:47 | |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_designate master: Add global override for service bind address https://review.opendev.org/683192 | 15:50 |
cjloader | morning | 15:50 |
*** mgoddard has quit IRC | 15:58 | |
openstackgerrit | Ralf Haferkamp proposed openstack/openstack-ansible-os_cinder master: Fix distro install on openSUSE https://review.opendev.org/683195 | 16:00 |
openstackgerrit | Ralf Haferkamp proposed openstack/openstack-ansible-os_tempest master: Fix distro install on openSUSE https://review.opendev.org/683197 | 16:05 |
*** openstackgerrit has quit IRC | 16:06 | |
*** macz has joined #openstack-ansible | 16:10 | |
*** macz has quit IRC | 16:15 | |
*** markvoelker has joined #openstack-ansible | 16:15 | |
cjloader | so noonedeadpunk with your venv_build_vars comment, do I add each of the vars? | 16:20 |
noonedeadpunk | yeah, it was kind of the idea. SO the thing is, that we want to use import in main.yml for tags to be working | 16:22 |
noonedeadpunk | So we can't use with_items there | 16:22 |
cjloader | thats why I had include | 16:22 |
cjloader | but I understand now | 16:23 |
*** openstackgerrit has joined #openstack-ansible | 16:25 | |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_nova master: Only clone console repos on source based installs https://review.opendev.org/666463 | 16:25 |
*** macz has joined #openstack-ansible | 16:25 | |
*** udesale has quit IRC | 16:27 | |
*** udesale has joined #openstack-ansible | 16:28 | |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_nova master: [redhat] Update variables to be version agnostic https://review.opendev.org/658259 | 16:37 |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_neutron master: [redhat] Update variables to be version agnostic https://review.opendev.org/658261 | 16:40 |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_glance master: Turn the filesystem directory to be configurable https://review.opendev.org/613924 | 16:45 |
*** udesale has quit IRC | 16:52 | |
*** mbuil has quit IRC | 16:59 | |
*** mbuil has joined #openstack-ansible | 16:59 | |
*** pcaruana has joined #openstack-ansible | 17:01 | |
*** markvoelker has quit IRC | 17:01 | |
*** markvoelker has joined #openstack-ansible | 17:04 | |
openstackgerrit | Cam J. Loader (cjloader) proposed openstack/openstack-ansible-os_ironic master: Adds configuration for the Ironic Inspector https://review.opendev.org/680284 | 17:05 |
cjloader | noonedeadpunk: like that? | 17:05 |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_neutron master: Implement uWSGI for neutron-api https://review.opendev.org/486156 | 17:06 |
*** zbr|ruck is now known as zbr | 17:07 | |
noonedeadpunk | cjloader: I'd say you can do this with single update | 17:08 |
odyssey4me | noonedeadpunk: apologies for jumping into your conversation, but I noticed you mention import vs include, and I've done some stuff around that recently | 17:08 |
noonedeadpunk | odyssey4me: you're always welcome:) | 17:08 |
odyssey4me | it turns out that include_tasks/include_role will work with tags, but only if the include_tasks/include_role task is tagged with the always tag | 17:09 |
cjloader | hi odyssey4me | 17:09 |
odyssey4me | o/ cjloader noonedeadpunk :) | 17:09 |
noonedeadpunk | yeah, you're right, always trigger it. | 17:09 |
noonedeadpunk | But I guess specific tags are valuable as well. | 17:09 |
odyssey4me | there is some detail here: https://docs.ansible.com/ansible/latest/porting_guides/porting_guide_2.5.html#dynamic-includes-and-attribute-inheritance | 17:10 |
cjloader | and noonedeadpunk I updated all the things you asked me to | 17:10 |
odyssey4me | it's not well described though | 17:10 |
*** newbie has joined #openstack-ansible | 17:10 | |
odyssey4me | you can also apply tags to the included tasks using the 'apply' key word: https://docs.ansible.com/ansible/latest/modules/include_tasks_module.html | 17:11 |
noonedeadpunk | odyssey4me: oh, you mean that we should just place tag inside every task of the included file? | 17:11 |
odyssey4me | well, it depends on what you want | 17:11 |
odyssey4me | if you want all tasks inside an included task file to be tagged the same, then use the apply key word on the include task | 17:12 |
odyssey4me | but make sure to also tag the include task 'always' so that the tags are useful | 17:12 |
odyssey4me | if you can point me at a review and explain what you want, I can comment in review how to achieve it :) | 17:13 |
cjloader | https://review.opendev.org/#/c/680284/ | 17:13 |
noonedeadpunk | yeah, i see. Thanks for sharing that | 17:13 |
cjloader | tasks/main.yml | 17:13 |
noonedeadpunk | odyssey4me: actually cjloader already made import intead:) | 17:13 |
*** pcaruana has quit IRC | 17:14 | |
odyssey4me | if there's no conditional, or no need for dynamic - then use import | 17:14 |
noonedeadpunk | I think imports are cleaner anyway... | 17:14 |
cjloader | there was a conditional | 17:14 |
cjloader | but I removed/ used import | 17:14 |
newbie | hey guys, I'm running linuxbridge agent and recently upgraded from Rocky to stein. I'm finding that the linuxbridgeagent running on my neutron nodes (the ones running on copute nodes are fine) are dying and getting privsep errors. | 17:14 |
noonedeadpunk | odyssey4me: I think conditionals might be aplied. There's restrictions for iterations and variables inside task path | 17:15 |
newbie | d getting privsep errors. | 17:15 |
newbie | Specifically getting an error that says it could't drop privileges and one Value Error where max_bin_len was exceeded | 17:15 |
odyssey4me | ok, so what happens with a conditional on an import is that the conditional gets applied to every task in the task file or role | 17:15 |
odyssey4me | that means slower processing | 17:15 |
odyssey4me | if you want it faster, then use include - if you need to use tags inside the task file or role, then apply the always tag to the include task | 17:16 |
noonedeadpunk | than we need to move everything back to includes:( | 17:16 |
odyssey4me | well, do what works for now - but perhaps it's worth an experiment to validate... no-one needs slow processing ;) | 17:16 |
cjloader | i originally had include | 17:17 |
odyssey4me | I only sat to figure it all out recently, and in our situation we *had* to use include because there was a loop for the set of tasks. | 17:17 |
noonedeadpunk | Yeah, than you don't have another choice | 17:17 |
cjloader | but just switched to import because of the looping | 17:18 |
noonedeadpunk | cjloader: in my opinion - let's leave things as is for now | 17:18 |
odyssey4me | cjloader: import won't allow a loop | 17:18 |
cjloader | yea | 17:18 |
cjloader | thats why I had incclude | 17:18 |
odyssey4me | you'd have to loop inside the imported task set, rather than loop from the outside | 17:19 |
noonedeadpunk | If we will decide to speedup things - we'll do migrations across all roles | 17:19 |
*** newbie has quit IRC | 17:19 | |
cjloader | noonedeadpunk: ok | 17:19 |
noonedeadpunk | just to be consisten for now in used things | 17:19 |
*** newbie has joined #openstack-ansible | 17:19 | |
cjloader | thx odyssey4me for your advice | 17:20 |
noonedeadpunk | but yeah, thanks odyssey4me for sharing this | 17:20 |
odyssey4me | I think I may just write up a blog post for this, because it's a bit obscure. | 17:20 |
noonedeadpunk | your input as always very valuable | 17:20 |
noonedeadpunk | btw, since you're here, what do you think about placement migration ? I guess it works now, but now sure how good things I'm doing for this https://review.opendev.org/#/c/664867/ | 17:21 |
cjloader | noonedeadpunk: do my recent changes address your comments? | 17:21 |
noonedeadpunk | actually it's your's patch:) | 17:21 |
noonedeadpunk | cjloader: yess | 17:21 |
odyssey4me | noonedeadpunk: oh yay, thanks for resurrecting it | 17:22 |
cjloader | as for OSA depends... I'm going to iterate later on that... i'm just trying to get my role into OSA noonedeadpunk | 17:22 |
noonedeadpunk | odyssey4me: related change actually make use of it and tests upgrade path | 17:23 |
odyssey4me | noonedeadpunk: whoah! https://review.opendev.org/#/c/671105/33 :) | 17:24 |
noonedeadpunk | yeah, there was lots of rebases:) | 17:24 |
cjloader | odyssey4me: ty for that catch, was remnant from include | 17:24 |
*** pcaruana has joined #openstack-ansible | 17:25 | |
openstackgerrit | Cam J. Loader (cjloader) proposed openstack/openstack-ansible-os_ironic master: Adds configuration for the Ironic Inspector https://review.opendev.org/680284 | 17:25 |
odyssey4me | noonedeadpunk: well, I can't really comment on the patch since I haven't really followed all the changes in the past 6-9 months - but it seems to work at least for the metal jobs, which is a great start! | 17:26 |
odyssey4me | better than nothing, that's for sure | 17:26 |
noonedeadpunk | for lxc there's a problem with timeouts:( | 17:26 |
cjloader | odyssey4me: my patch is fairly hefty lol | 17:27 |
noonedeadpunk | and for lxc kinda 30% of things might be dropped out... | 17:27 |
noonedeadpunk | cjloader: the second one is optional | 17:29 |
noonedeadpunk | but qwuotes around dict keys are really missing | 17:30 |
cjloader | ok | 17:34 |
*** jawad_axd has joined #openstack-ansible | 17:37 | |
newbie | ok, so I was able to sort my linuxbridge agent issue. I had to downgrade msgpack...not sure if this was a requirements issue or if did something worng at some point | 17:40 |
openstackgerrit | Cam J. Loader (cjloader) proposed openstack/openstack-ansible-os_ironic master: Adds configuration for the Ironic Inspector https://review.opendev.org/680284 | 17:45 |
cjloader | noonedeadpunk: addressed | 17:45 |
*** pcaruana has quit IRC | 17:55 | |
*** ansmith_ has joined #openstack-ansible | 17:58 | |
*** ansmith has quit IRC | 18:00 | |
openstackgerrit | Merged openstack/openstack-ansible-os_nova stable/stein: Restore nova-key tag https://review.opendev.org/681253 | 18:03 |
*** jawad_axd has quit IRC | 18:21 | |
*** trident has quit IRC | 18:45 | |
*** macz has quit IRC | 18:47 | |
*** newbie has quit IRC | 18:51 | |
*** radeks has quit IRC | 18:51 | |
*** trident has joined #openstack-ansible | 18:55 | |
*** trident has quit IRC | 19:01 | |
*** cshen has joined #openstack-ansible | 19:05 | |
*** trident has joined #openstack-ansible | 19:10 | |
*** cshen has quit IRC | 19:19 | |
*** ansmith_ has quit IRC | 19:25 | |
jrosser | can we get another vote on this? https://review.opendev.org/#/c/682830/ | 19:26 |
jrosser | and another here https://review.opendev.org/#/c/681610/ | 19:30 |
*** cshen has joined #openstack-ansible | 19:30 | |
jrosser | and another here https://review.opendev.org/#/c/681724/ | 19:31 |
*** ansmith_ has joined #openstack-ansible | 19:31 | |
cjloader | simple patches jrosser | 19:32 |
jrosser | i'm looking for +2 +W :) | 19:32 |
cjloader | make me core lol | 19:32 |
cjloader | jk | 19:32 |
*** cshen has quit IRC | 19:36 | |
openstackgerrit | Cam J. Loader (cjloader) proposed openstack/openstack-ansible-os_ironic master: Adds configuration for the Ironic Inspector https://review.opendev.org/680284 | 19:37 |
cjloader | jrosser: can you take another glance at my patch? | 19:41 |
*** macz has joined #openstack-ansible | 19:49 | |
openstackgerrit | Merged openstack/openstack-ansible-os_octavia master: Create mgmt subnet with gateway if provided https://review.opendev.org/681642 | 19:50 |
*** pcaruana has joined #openstack-ansible | 19:51 | |
*** pcaruana has quit IRC | 20:31 | |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: [WIP] Bind services to mgmt network addresses https://review.opendev.org/670051 | 20:34 |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: [WIP] Bind services to mgmt network addresses https://review.opendev.org/670051 | 20:34 |
*** ansmith_ has quit IRC | 20:34 | |
guilhermesp | jrosser: I guees we will need to update all roles? ^ | 20:37 |
guilhermesp | all right I can see a burst of patches here https://review.opendev.org/#/q/topic:bind-to-mgmt+(status:open+OR+status:merged) | 20:37 |
jrosser | yeah i did them all today, execpt those where there isnt a var in defaults | 20:38 |
jrosser | heres where we are https://etherpad.openstack.org/p/osa-bind-to-mgmt | 20:38 |
guilhermesp | oh nice jrosser | 20:39 |
jrosser | guilhermesp: if you are able to review those 3 patches i mentioned up there ^^ just earlier it would unblock some other stuff | 20:40 |
guilhermesp | /me looks | 20:40 |
guilhermesp | ups | 20:40 |
jrosser | guilhermesp: so my plan was that we have openstack_service_bind_address for openstack-ansible-os_* | 20:40 |
guilhermesp | anyways, lookign | 20:40 |
jrosser | and then we deal with the infrastructure things as needed in group_vars | 20:41 |
jrosser | which lets you do the thing for galera however is needed eventually | 20:41 |
*** sshnaidm is now known as sshnaidm|off | 20:42 | |
guilhermesp | all right reviews done jrosser | 20:42 |
jrosser | thankyou | 20:42 |
guilhermesp | thanks for the work with mgmt stuff, I will come back to galera changes asap | 20:42 |
jrosser | cool - i think this is a really nice thing to get merged | 20:43 |
openstackgerrit | Merged openstack/openstack-ansible-os_octavia master: Octavia use barbican internalURL https://review.opendev.org/678871 | 20:45 |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_placement master: Add global override for service bind address https://review.opendev.org/683142 | 20:45 |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_neutron master: Move provider_networks into os_neutron https://review.opendev.org/658126 | 20:49 |
openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_neutron master: Move provider_networks into os_neutron https://review.opendev.org/658126 | 20:49 |
openstackgerrit | Cam J. Loader (cjloader) proposed openstack/openstack-ansible-os_ironic master: Adds configuration for the Ironic Inspector https://review.opendev.org/680284 | 21:16 |
openstackgerrit | Merged openstack/openstack-ansible-os_nova stable/stein: tasks: add conditionals for ironic https://review.opendev.org/673879 | 21:19 |
*** markvoelker has quit IRC | 21:20 | |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: Add support to deploy murano https://review.opendev.org/679496 | 21:24 |
*** ansmith_ has joined #openstack-ansible | 21:34 | |
*** luksky has quit IRC | 21:41 | |
*** markvoelker has joined #openstack-ansible | 21:48 | |
*** BjoernT_ has quit IRC | 22:00 | |
*** weifan has joined #openstack-ansible | 22:09 | |
*** weifan has quit IRC | 22:09 | |
*** macz has quit IRC | 22:25 | |
*** markvoelker has quit IRC | 22:32 | |
openstackgerrit | Merged openstack/openstack-ansible-os_neutron master: Mark calico jobs as non voting due to python3 bug https://review.opendev.org/682830 | 22:39 |
*** spatel has quit IRC | 22:41 | |
openstackgerrit | Cam J. Loader (cjloader) proposed openstack/openstack-ansible-os_ironic master: Adds configuration for the Ironic Inspector https://review.opendev.org/680284 | 22:42 |
openstackgerrit | Merged openstack/openstack-ansible-os_nova master: Only clone console repos on source based installs https://review.opendev.org/666463 | 22:46 |
openstackgerrit | Merged openstack/openstack-ansible-os_masakari master: Add uwsgi role to role requirements https://review.opendev.org/681724 | 22:48 |
openstackgerrit | Merged openstack/openstack-ansible-os_nova master: [redhat] Update variables to be version agnostic https://review.opendev.org/658259 | 22:49 |
*** macz has joined #openstack-ansible | 22:52 | |
*** macz has quit IRC | 22:54 | |
*** macz has joined #openstack-ansible | 22:54 | |
*** goldyfruit___ has quit IRC | 23:00 | |
*** tosky has quit IRC | 23:12 | |
*** ondrejburian has quit IRC | 23:23 | |
*** kaiokmo has quit IRC | 23:23 | |
*** ondrejburian has joined #openstack-ansible | 23:24 | |
*** macz has quit IRC | 23:54 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!