| *** tosky has quit IRC | 00:32 | |
| *** cshen has joined #openstack-ansible | 00:37 | |
| *** cshen has quit IRC | 00:42 | |
| mnaser | someone should ping spatel about sending a patch for this :) | 00:48 |
|---|---|---|
| *** cshen has joined #openstack-ansible | 00:53 | |
| *** cshen has quit IRC | 00:58 | |
| *** waxfire has quit IRC | 00:59 | |
| *** waxfire has joined #openstack-ansible | 00:59 | |
| *** macz_ has joined #openstack-ansible | 01:24 | |
| *** macz_ has quit IRC | 01:28 | |
| *** jamesden_ has quit IRC | 01:34 | |
| *** alvinstarr has joined #openstack-ansible | 01:35 | |
| *** spatel has joined #openstack-ansible | 01:56 | |
| *** macz_ has joined #openstack-ansible | 02:17 | |
| *** gyee has quit IRC | 02:21 | |
| *** macz_ has quit IRC | 02:21 | |
| *** cshen has joined #openstack-ansible | 02:54 | |
| *** cshen has quit IRC | 02:58 | |
| *** cshen has joined #openstack-ansible | 03:10 | |
| *** cshen has quit IRC | 03:14 | |
| *** dmsimard|off has joined #openstack-ansible | 04:18 | |
| *** udesale has joined #openstack-ansible | 04:51 | |
| *** spatel has quit IRC | 05:04 | |
| *** cshen has joined #openstack-ansible | 05:11 | |
| *** cshen has quit IRC | 05:15 | |
| *** cshen has joined #openstack-ansible | 05:26 | |
| *** cshen has quit IRC | 05:31 | |
| *** evrardjp has quit IRC | 05:36 | |
| *** evrardjp has joined #openstack-ansible | 05:36 | |
| *** shyamb has joined #openstack-ansible | 05:46 | |
| *** macz_ has joined #openstack-ansible | 05:53 | |
| *** macz_ has quit IRC | 05:58 | |
| *** cshen has joined #openstack-ansible | 06:02 | |
| *** cshen has quit IRC | 06:06 | |
| *** shyam89 has joined #openstack-ansible | 06:18 | |
| *** shyamb has quit IRC | 06:21 | |
| *** shyam89 has quit IRC | 06:35 | |
| openstackgerrit | Erik Berg proposed openstack/openstack-ansible-repo_server stable/train: Exclude repo_build_global_links_path from sync https://review.opendev.org/714841 | 06:37 |
| openstackgerrit | Erik Berg proposed openstack/openstack-ansible-repo_server stable/stein: Exclude repo_build_global_links_path from sync https://review.opendev.org/714842 | 06:37 |
| snadge | this is a novice question.. but if i've already deployed using openstack-ansible.. but I want to add a whole bunch of compute nodes.. can i just edit the configuration, add the nodes and re-run the 3 playbooks? | 06:50 |
| snadge | or should i reset the the controller and the compute node. and do them all from scratch | 06:51 |
| *** shyamb has joined #openstack-ansible | 06:59 | |
| *** shyam89 has joined #openstack-ansible | 07:07 | |
| *** shyamb has quit IRC | 07:09 | |
| *** cshen has joined #openstack-ansible | 07:11 | |
| *** vesper11 has quit IRC | 07:16 | |
| *** vesper has joined #openstack-ansible | 07:16 | |
| *** cshen has quit IRC | 07:16 | |
| *** miloa has joined #openstack-ansible | 07:20 | |
| *** shyam89 has quit IRC | 07:27 | |
| *** cshen has joined #openstack-ansible | 07:27 | |
| *** shyamb has joined #openstack-ansible | 07:27 | |
| openstackgerrit | Erik Berg proposed openstack/openstack-ansible-os_neutron master: Add systemd-devel to neutron_devel_distro_packages https://review.opendev.org/714403 | 07:39 |
| *** shyam89 has joined #openstack-ansible | 07:39 | |
| *** shyamb has quit IRC | 07:42 | |
| *** jbadiapa has joined #openstack-ansible | 07:43 | |
| *** shyam89 has quit IRC | 07:55 | |
| *** shyamb has joined #openstack-ansible | 07:56 | |
| openstackgerrit | Erik Berg proposed openstack/openstack-ansible-os_nova master: Align vars/redhat.yml with other distros https://review.opendev.org/714853 | 08:01 |
| *** DanyC has joined #openstack-ansible | 08:10 | |
| *** tosky has joined #openstack-ansible | 08:24 | |
| *** rpittau|afk is now known as rpittau | 08:31 | |
| *** DanyC has quit IRC | 08:33 | |
| *** DanyC has joined #openstack-ansible | 08:33 | |
| *** shyam89 has joined #openstack-ansible | 08:36 | |
| *** shyamb has quit IRC | 08:39 | |
| *** shyam89 has quit IRC | 09:11 | |
| *** shyamb has joined #openstack-ansible | 09:12 | |
| *** sshnaidm|afk is now known as sshnaidm | 09:15 | |
| *** shyamb has quit IRC | 09:18 | |
| *** jbadiapa has quit IRC | 09:27 | |
| *** shyamb has joined #openstack-ansible | 09:49 | |
| *** ivve has joined #openstack-ansible | 09:53 | |
| *** shyam89 has joined #openstack-ansible | 10:08 | |
| *** shyamb has quit IRC | 10:11 | |
| openstackgerrit | Merged openstack/openstack-ansible stable/rocky: Fix idempotency for os-lxc-container-setup.yml https://review.opendev.org/708668 | 10:12 |
| jrosser | snadge: we ave documentation for adding a compute node here https://docs.openstack.org/openstack-ansible/latest/admin/scale-environment.html | 10:16 |
| *** sshnaidm has quit IRC | 10:18 | |
| *** sshnaidm has joined #openstack-ansible | 10:19 | |
| *** tremble has joined #openstack-ansible | 10:22 | |
| *** tremble has left #openstack-ansible | 10:23 | |
| *** crazzy has quit IRC | 10:29 | |
| *** trident has quit IRC | 10:29 | |
| *** crazzy has joined #openstack-ansible | 10:30 | |
| jrosser | noonedeadpunk: seems this is broken for all the things on centos/train https://zuul.opendev.org/t/openstack/build/fbbcff01aabc43c0a4b649be04b3e2e3/log/job-output.txt#3360 | 10:31 |
| *** trident has joined #openstack-ansible | 10:31 | |
| *** trident has quit IRC | 10:33 | |
| *** trident has joined #openstack-ansible | 10:37 | |
| *** gshippey has joined #openstack-ansible | 10:51 | |
| openstackgerrit | Chandan Kumar (raukadah) proposed openstack/openstack-ansible-os_tempest master: Added tempest_tempestconf_image var for tempestconf image https://review.opendev.org/714601 | 11:00 |
| *** tosky is now known as tosky_ | 11:24 | |
| *** yolanda has quit IRC | 11:27 | |
| *** DanyC has quit IRC | 11:30 | |
| *** DanyC has joined #openstack-ansible | 11:31 | |
| *** DanyC has joined #openstack-ansible | 11:31 | |
| *** yolanda has joined #openstack-ansible | 11:32 | |
| *** sshnaidm has quit IRC | 11:34 | |
| *** sshnaidm has joined #openstack-ansible | 11:34 | |
| *** tosky_ is now known as tosky | 11:40 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_keystone stable/train: [DNM] add debug to test constraints URL https://review.opendev.org/714921 | 11:41 |
| *** rpittau is now known as rpittau|bbl | 11:42 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-tests stable/train: Set requirements_git_install_branch during functional tests https://review.opendev.org/714486 | 11:42 |
| noonedeadpunk | jrosser: yeah I saw that yestarday and was going to investigate today | 11:47 |
| noonedeadpunk | ebbex: seems like all rocky stuff is merged? | 11:47 |
| jrosser | ok, i am still looking at this u-c stuff, didnt get much time yesterday to spend on that | 11:47 |
| CeeMac | morning | 11:51 |
| noonedeadpunk | o/ | 11:51 |
| *** spatel has joined #openstack-ansible | 11:53 | |
| *** shyam89 has quit IRC | 11:55 | |
| *** rh-jelabarre has joined #openstack-ansible | 11:56 | |
| *** spatel has quit IRC | 11:58 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible stable/rocky: Bump OSA stable/rocky https://review.opendev.org/714926 | 12:08 |
| *** sshnaidm is now known as sshnaidm|afk | 12:08 | |
| ebbex | noonedeadpunk: Yup. looks good to me. | 12:17 |
| *** jamesden_ has joined #openstack-ansible | 12:18 | |
| *** shyamb has joined #openstack-ansible | 12:23 | |
| *** sshnaidm|afk is now known as sshnaidm | 12:24 | |
| *** shyamb has quit IRC | 12:29 | |
| *** ansmith_ has quit IRC | 12:31 | |
| *** ansmith has joined #openstack-ansible | 12:31 | |
| *** mathlin has quit IRC | 12:33 | |
| *** mathlin_ has joined #openstack-ansible | 12:33 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_horizon master: Do not become `horizon` user for distro installs https://review.opendev.org/714931 | 12:37 |
| *** udesale_ has joined #openstack-ansible | 12:42 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_horizon master: Do not become `horizon` user for distro installs https://review.opendev.org/714931 | 12:44 |
| *** udesale has quit IRC | 12:45 | |
| *** spatel has joined #openstack-ansible | 12:59 | |
| *** rpittau|bbl is now known as rpittau | 13:00 | |
| chandankumar | jrosser, noonedeadpunk please have a look when free https://review.opendev.org/714601 thanks! | 13:06 |
| *** ryneq has joined #openstack-ansible | 13:25 | |
| noonedeadpunk | chandankumar: not really sure I got what problem do we solve with that variable (or what feature we implement). | 13:25 |
| noonedeadpunk | Doesn't providing url will cause image download? | 13:25 |
| *** jbadiapa has joined #openstack-ansible | 13:25 | |
| noonedeadpunk | btw, do can you recall why we need images doubled in tempest_images_map? | 13:27 |
| chandankumar | noonedeadpunk, we wanted to use custom image https://review.opendev.org/#/c/714609/ | 13:28 |
| chandankumar | but current tempest_images has a list of images | 13:28 |
| chandankumar | noonedeadpunk, I think we added the tempest images map so that we can consume different images based on different arch | 13:30 |
| noonedeadpunk | I guess yes, but why there's 2 same images https://opendev.org/openstack/openstack-ansible-os_tempest/src/branch/master/vars/main.yml#L24-L31 ? | 13:30 |
| noonedeadpunk | (the same for arm) | 13:31 |
| chandankumar | no idea why we added it twice | 13:32 |
| chandankumar | noonedeadpunk, let me send a patch to remove duplicates | 13:34 |
| noonedeadpunk | chandankumar: regarding your patch... I don't really like what we have now... Actually profile can contain lots of other things except of the image.... | 13:37 |
| chandankumar | noonedeadpunk, do we want to create a custom profile on ourside? | 13:38 |
| *** Adri2000 has quit IRC | 13:39 | |
| jrosser | there was somthing in tempest about needing an image and an alterative image - i don't remember why | 13:40 |
| noonedeadpunk | yes, but these ones have the same naming.... | 13:40 |
| noonedeadpunk | but maybe we really need them doubled | 13:41 |
| noonedeadpunk | I've jsut asked:) | 13:41 |
| noonedeadpunk | chandankumar: the thing is that currently we'll need another variable to set image_disk_format and so on... | 13:42 |
| noonedeadpunk | I see 3 options: mergning your commit with some edits, setting fact `tempest_tempestconf_profile` on your side or adding same override variable and merging 2 dicts together later | 13:44 |
| noonedeadpunk | regarding edits I've placed a comment on patch | 13:44 |
| chandankumar | noonedeadpunk, let me fix current patch | 13:45 |
| chandankumar | noonedeadpunk, then coming to next one | 13:45 |
| noonedeadpunk | hm, ok, now not 100% sure that omit will work there.... as tempest_tempest_conf_overrides which have one is always defined | 13:46 |
| noonedeadpunk | so wait editing) | 13:47 |
| noonedeadpunk | let me think about it for some time | 13:47 |
| chandankumar | noonedeadpunk, ok | 13:48 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_keystone stable/train: [DNM] add debug to test constraints URL https://review.opendev.org/714921 | 13:53 |
| *** dwilde has joined #openstack-ansible | 13:55 | |
| *** andrewbonney has joined #openstack-ansible | 14:04 | |
| *** aedc has joined #openstack-ansible | 14:13 | |
| *** aedc has quit IRC | 14:13 | |
| openstackgerrit | Chandan Kumar (raukadah) proposed openstack/openstack-ansible-os_tempest master: Added tempest_tempestconf_image var for tempestconf image https://review.opendev.org/714601 | 14:19 |
| *** prometheanfire has quit IRC | 14:19 | |
| *** prometheanfire has joined #openstack-ansible | 14:23 | |
| *** Adri2000 has joined #openstack-ansible | 14:29 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-tests stable/train: Set requirements_git_install_branch during functional tests https://review.opendev.org/714486 | 14:32 |
| *** cshen has quit IRC | 14:42 | |
| spotz | hey all, poke if anything holding up stuff needs a review | 14:49 |
| *** macz_ has joined #openstack-ansible | 14:52 | |
| *** cshen has joined #openstack-ansible | 14:53 | |
| *** udesale_ has quit IRC | 14:57 | |
| *** cshen has quit IRC | 14:58 | |
| noonedeadpunk | I'd appreciate votes on https://review.opendev.org/#/c/710776/ and https://review.opendev.org/#/c/713222/ :) it's not blocking anything, but still) | 14:58 |
| *** cshen has joined #openstack-ansible | 15:09 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-tests stable/train: Set requirements_git_install_branch during functional tests https://review.opendev.org/714486 | 15:10 |
| *** prometheanfire has quit IRC | 15:19 | |
| *** macz_ has quit IRC | 15:20 | |
| *** macz_ has joined #openstack-ansible | 15:21 | |
| *** prometheanfire has joined #openstack-ansible | 15:23 | |
| spotz | noonedeadpunk: I've pulled them up:) | 15:24 |
| spotz | I just normally kinda squirrel on reviews so if I'm not paying attention there's nothing to squirrel on!:) | 15:24 |
| spotz | done | 15:25 |
| noonedeadpunk | thanks) | 15:25 |
| noonedeadpunk | chandankumar: does it actually work for you? As I have failure here https://opendev.org/openstack/openstack-ansible-os_tempest/src/branch/master/tasks/tempestconf.yml#L31-L43 | 15:27 |
| noonedeadpunk | as there's no venv under directory http://paste.openstack.org/show/791142/ So does the rest part of it works for you? | 15:28 |
| *** arxcruz|rover is now known as arxcruz | 15:30 | |
| *** gyee has joined #openstack-ansible | 15:40 | |
| noonedeadpunk | ebbex: can we finish https://review.opendev.org/#/c/710803/ so that to unblock rocky EM? | 15:46 |
| chandankumar | noonedeadpunk, let me check | 15:59 |
| chandankumar | noonedeadpunk, yes tempest conf got failed | 16:02 |
| noonedeadpunk | ok, so we have bigger problem I gues :( | 16:03 |
| noonedeadpunk | btw omit do not work there:( | 16:04 |
| noonedeadpunk | ok I need to finish smth and will return back to that investigation | 16:04 |
| chandankumar | yes sure | 16:06 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_horizon master: Refactor memcached_servers https://review.opendev.org/713236 | 16:06 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-tests stable/train: Set requirements_git_install_branch during functional tests https://review.opendev.org/714486 | 16:09 |
| *** andrea15 has quit IRC | 16:18 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_aodh master: Refactor memcached_servers https://review.opendev.org/714990 | 16:38 |
| *** jbadiapa has quit IRC | 16:40 | |
| openstackgerrit | Merged openstack/openstack-ansible-repo_server stable/stein: Exclude repo_build_global_links_path from sync https://review.opendev.org/714842 | 16:44 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_mistral master: Refactor memcached_servers https://review.opendev.org/714996 | 16:52 |
| openstackgerrit | Merged openstack/openstack-ansible-os_neutron master: Add systemd-devel to neutron_devel_distro_packages https://review.opendev.org/714403 | 16:56 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_murano master: Refactor memcached_servers https://review.opendev.org/715000 | 16:56 |
| *** jamesden_ has quit IRC | 16:59 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_zun master: Refactor memcached_servers https://review.opendev.org/715001 | 16:59 |
| *** jamesden_ has joined #openstack-ansible | 16:59 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_nova master: Align vars/redhat.yml with other distros https://review.opendev.org/714853 | 17:04 |
| openstackgerrit | James Denton proposed openstack/openstack-ansible-ops master: Implements support for OVS+DVR in an MNAIO https://review.opendev.org/715002 | 17:06 |
| jrosser | noonedeadpunk: can you have a look at this https://github.com/openstack/openstack-ansible-os_keystone/blob/master/defaults/main.yml#L43 | 17:09 |
| jrosser | what would requirements_git_install_branch be on the different branches? | 17:09 |
| noonedeadpunk | I guess we tke it from https://opendev.org/openstack/openstack-ansible/src/branch/master/playbooks/defaults/repo_packages/openstack_services.yml#L34 | 17:11 |
| jrosser | so the end of this line https://zuul.opendev.org/t/openstack/build/3502cf06d498414b825d464e0a28d9ee/log/job-output.txt#5616 | 17:12 |
| mnaser | jrosser, noonedeadpunk, guilhermesp: i think i just had something click to a reason behind why our ansible runs may have kinda slowed down when we added python_venv_build | 17:13 |
| jrosser | that is indeed 404 athttps://releases.openstack.org/constraints/upper/stable/train | 17:13 |
| noonedeadpunk | oh | 17:13 |
| jrosser | but it is ok at https://releases.openstack.org/constraints/upper/train | 17:13 |
| mnaser | we started including the openstack services vars for all of the roles/inventory as extra vars | 17:13 |
| *** rpittau is now known as rpittau|afk | 17:13 | |
| jrosser | so either we had this broken for ever or the redirects have moved or something | 17:14 |
| mnaser | more specifically _all_ those vars are being copied into the inventory of every single host -- https://github.com/openstack/openstack-ansible/blob/master/playbooks/defaults/repo_packages/openstack_services.yml | 17:14 |
| noonedeadpunk | oh.... | 17:14 |
| noonedeadpunk | that's interesting.... | 17:15 |
| mnaser | so MAYBE we can move it to group_vars for every single service/role ? | 17:16 |
| jrosser | that would be a bit ugly for the tool that bumps the releases | 17:17 |
| noonedeadpunk | mnaser: do we have some env to test it out? | 17:17 |
| openstackgerrit | Merged openstack/openstack-ansible-os_gnocchi master: Refactor memcached_servers https://review.opendev.org/713233 | 17:18 |
| noonedeadpunk | I wonder if ansible actually should cache all role defaults? | 17:18 |
| spotz | jrosser: didn't things get renamed/moved recently? | 17:20 |
| mnaser | jrosser: i mean for the bumps, we already use a tool right now, right? | 17:21 |
| jrosser | we do | 17:22 |
| jrosser | also bootstrap-ansible parses that file iirc | 17:22 |
| mnaser | hmm, for what exactly? | 17:22 |
| mnaser | maybe we can move group_vars in folders | 17:22 |
| mnaser | so doing things like group_vars/nova_all/services.yml | 17:22 |
| mnaser | and then globb-ing that instead | 17:23 |
| noonedeadpunk | won't that be overwriten bu user group_vars? | 17:23 |
| noonedeadpunk | (not sure - jsut asking) | 17:23 |
| mnaser | noonedeadpunk: if a user wants to override them, they'll still be able to | 17:23 |
| mnaser | using user_variables | 17:23 |
| jrosser | we did just add this https://github.com/openstack/openstack-ansible/commit/b3932f0e8241344ef9ece8c028ebb70bd548defc | 17:24 |
| noonedeadpunk | no, I mean if user wants to define their host_vars for example | 17:24 |
| mnaser | it should still work as the inventory we use takes higher prio, i think | 17:24 |
| noonedeadpunk | s/host_vars/group_vars | 17:24 |
| mnaser | noonedeadpunk: yeah i got you, i think it's okay, we can test that | 17:25 |
| noonedeadpunk | or you mean to create that into openstack_deploy folder which users will be copying? | 17:25 |
| mnaser | jrosser: yeah, we'll have to look into that as well | 17:25 |
| mnaser | no we add it into osa's group_vars and the user can override it in their own group_vars if they want (or user_variables) | 17:26 |
| mnaser | or we could split the openstack services into many small files for example in the same folder | 17:26 |
| mnaser | in my experience, ansible variable copying is really slow, so it could have a huge impact on performance | 17:26 |
| jrosser | maybe now should be the time we deal with the number of variables | 17:27 |
| noonedeadpunk | but I'm not really sure I got why it get cahced for each host... | 17:27 |
| jrosser | talked about it a lot | 17:27 |
| mnaser | i think the more we get things out of "all" the better we can help things | 17:28 |
| mnaser | noonedeadpunk: because all our roles do this - https://github.com/openstack/openstack-ansible/blob/master/playbooks/os-aodh-install.yml#L22 | 17:28 |
| noonedeadpunk | maybe, maybe.... we should install memcaced on deploy host and use it as a cache storage? | 17:29 |
| noonedeadpunk | It should be the way faster than storing facts in yaml/json files? | 17:29 |
| jrosser | but vars != facts ? | 17:30 |
| noonedeadpunk | we can store there as well as inventory cache as fact cache | 17:31 |
| *** DanyC has quit IRC | 17:32 | |
| noonedeadpunk | but I think inventory cache is just disabled now? | 17:32 |
| mnaser | we already use fact cache | 17:32 |
| mnaser | and when you have a big fact cache it's really slow | 17:32 |
| *** DanyC has joined #openstack-ansible | 17:32 | |
| noonedeadpunk | but we do store it in simple file? which is actually slow by design... | 17:33 |
| noonedeadpunk | especially when big | 17:33 |
| mnaser | it's a json file, and yeah, it can get really slow because it loads it every single task im pretty sure | 17:33 |
| noonedeadpunk | but my point is that in case of storing it in some nosql db which is in memory should make it the way faster? | 17:34 |
| mnaser | i think i tried that once at some point | 17:34 |
| mnaser | and it was just as bad | 17:34 |
| noonedeadpunk | but yes, removing duplication seems like a good idea anyway | 17:34 |
| mnaser | i think the issue was that ansible stupid-ly loads teh json string every task.. | 17:35 |
| noonedeadpunk | just we need to be sure that moving to group_vars will really solve that issue | 17:35 |
| *** evrardjp has quit IRC | 17:36 | |
| *** evrardjp has joined #openstack-ansible | 17:36 | |
| *** DanyC has quit IRC | 17:37 | |
| jrosser | i am pretty much stuck with understanding how this requirements url stuff is supposed to work | 17:37 |
| noonedeadpunk | so it would be kinda great having such slow env and try out different concepts wil running some noop role | 17:37 |
| noonedeadpunk | just moving everything to group_vars will really make us to update bump script, and it would be sad if we won't have speedup in the end... | 17:42 |
| *** DanyC has joined #openstack-ansible | 17:42 | |
| mnaser | jrosser: whats the problem? | 17:46 |
| mnaser | noonedeadpunk: i mean we can try moving it manually for once and se how much faster it runs | 17:46 |
| *** cshen has quit IRC | 17:47 | |
| noonedeadpunk | mnaser: but we will see no difference in AIO | 17:47 |
| noonedeadpunk | (probably) | 17:47 |
| mnaser | noonedeadpunk: we would for containerized deploys | 17:47 |
| mnaser | because now the inventory for the utility would have a lot less things, etc | 17:47 |
| *** cshen has joined #openstack-ansible | 17:48 | |
| noonedeadpunk | yeah, maybe. At least we can start from that | 17:50 |
| noonedeadpunk | jrosser: redirects in releases seems to be the same, so not really sure how it was working all this time... | 17:51 |
| noonedeadpunk | maybe let's ask releases if it ever worked... | 17:51 |
| noonedeadpunk | mnaser: just time spent was increasing dramatically with 5+ hosts iirc. | 17:52 |
| mnaser | noonedeadpunk: yeah i've seen that behaviour and so has spotz | 17:52 |
| mnaser | spatel: ** | 17:52 |
| spatel | yes | 17:52 |
| mnaser | but its one of those magical unicorns i have no clue why it happens | 17:52 |
| *** cshen has quit IRC | 17:52 | |
| noonedeadpunk | agree | 17:53 |
| spatel | whats up | 17:53 |
| *** miloa has quit IRC | 17:53 | |
| noonedeadpunk | have someone asked ansible folks about it? | 17:53 |
| spatel | mnaser: sorry trying to catchup all previous messages | 17:55 |
| noonedeadpunk | spatel: we're talking about the reasons why ansible slow downs with encreasing amount of hosts | 17:58 |
| spatel | noonedeadpunk: we have 330 compute nodes and its damn slow | 17:58 |
| spatel | soon i am going to add 200 more compute in that cluster so not sure what will happen after that... :( | 17:59 |
| noonedeadpunk | ok, so I'll try to test things in aio with lxc, and if have some positive result will update you | 18:00 |
| noonedeadpunk | so maybe you can try solution on bigger scale before updating everything | 18:01 |
| spatel | are we going to use nosql stuff | 18:01 |
| noonedeadpunk | jrosser: wait, we don't have backported that releases stuff yet? | 18:05 |
| noonedeadpunk | So it's just broken on master only? | 18:05 |
| *** andrewbonney has quit IRC | 18:06 | |
| jrosser | noonedeadpunk: i am trying to debug it on train | 18:09 |
| noonedeadpunk | yeah, but we have different url there https://github.com/openstack/openstack-ansible-os_keystone/blob/stable/train/defaults/main.yml#L43 | 18:10 |
| jrosser | i am not able to explain how this is happening https://zuul.opendev.org/t/openstack/build/3502cf06d498414b825d464e0a28d9ee/log/job-output.txt#5616 | 18:10 |
| noonedeadpunk | so we probably need to rollback change on master... | 18:10 |
| jrosser | and it's not keystone, its actually tempest | 18:10 |
| noonedeadpunk | but for tempest it is changed | 18:12 |
| jrosser | :/ i am going round in circles on this all day - now i have no clue what is wrong | 18:12 |
| jrosser | i started fixing some bits here https://review.opendev.org/#/c/714486/ | 18:13 |
| noonedeadpunk | so seems that you've fixed smth at least:) | 18:17 |
| noonedeadpunk | I mean that it wasn't working previously I guess | 18:17 |
| jrosser | yeah :) | 18:19 |
| noonedeadpunk | evrardjp: we're already discussing that but your input is pretty valuable:) | 18:19 |
| evrardjp | so apparently you're using head of the branch for tempest? | 18:19 |
| noonedeadpunk | actually jrosser get this https://zuul.opendev.org/t/openstack/build/3502cf06d498414b825d464e0a28d9ee/log/job-output.txt#5616 | 18:20 |
| evrardjp | sorry | 18:20 |
| evrardjp | head of master* | 18:20 |
| noonedeadpunk | on train | 18:20 |
| jrosser | hmm? isnt that what i linked earlier? | 18:21 |
| noonedeadpunk | yeah, just posted the same one | 18:21 |
| evrardjp | I think now that the roles have been moved back to use the integrated repo, couldn't you use the information there? | 18:22 |
| noonedeadpunk | jrosser: oh..... we do have SHA as for `requirements_git_install_branch` | 18:22 |
| jrosser | this is not integrated repo tests though | 18:22 |
| jrosser | now i am very confused | 18:23 |
| noonedeadpunk | and I guees you've used requirements_git_track_branch for requirements_git_install_branch | 18:23 |
| *** cshen has joined #openstack-ansible | 18:23 | |
| evrardjp | anyway I will leave you to your debugging, I am sure you have the situation at hand :) | 18:23 |
| noonedeadpunk | so when you provide sha to the rewuirements it redirects correctly | 18:23 |
| noonedeadpunk | evrardjp: yeah, sorry for disturbing you:) | 18:23 |
| jrosser | this isnt the integrated repo | 18:24 |
| evrardjp | no problem :) | 18:24 |
| noonedeadpunk | jrosser: so here https://review.opendev.org/#/c/714486/7/test-vars.yml you set `requirements_git_install_branch` to `stable/train` | 18:24 |
| evrardjp | You might want to consider something else, if you have extra cycles: produce venvs into a release, and publish them... The roles would be by far simplified. | 18:24 |
| jrosser | noonedeadpunk: correct | 18:25 |
| noonedeadpunk | jrosser: but generally `requirements_git_install_branch` is just SHA, not branch | 18:26 |
| noonedeadpunk | so probasbly we need to do some spilt by `/` and take last element | 18:27 |
| jrosser | https://review.opendev.org/#/c/714486/5/test-vars.yml | 18:27 |
| noonedeadpunk | evrardjp: not really sure I've catched the idea | 18:27 |
| jrosser | that failed in surprising ways too | 18:27 |
| noonedeadpunk | hum... | 18:27 |
| *** cshen has quit IRC | 18:27 | |
| noonedeadpunk | "404 Client Error: Not Found for url: https://opendev.org/openstack/requirements/raw/branch/master/train/upper-constraints.txt" | 18:28 |
| noonedeadpunk | lol | 18:28 |
| noonedeadpunk | but it was failing on keystone | 18:29 |
| jrosser | doesnt it fail in exactly the same line though - temepst again? | 18:29 |
| noonedeadpunk | I guess this might be because we use different url building for roles... | 18:30 |
| noonedeadpunk | nope - it's failing on keystone... | 18:30 |
| noonedeadpunk | https://zuul.opendev.org/t/openstack/build/cc8b05c3a58e4038b338905df6aaed22/log/job-output.txt#5067 | 18:30 |
| jrosser | are you sure? scroll back and it's the tempest role running against keystone1 (i know!!!!) | 18:30 |
| noonedeadpunk | but it's failing while building kystone venv | 18:31 |
| jrosser | oh you are right, i'm looking at the other PS | 18:31 |
| noonedeadpunk | and on patch 7 it's failing while building tempest bvenv | 18:31 |
| jrosser | yes that got further | 18:31 |
| jrosser | and WTF at https://opendev.org/openstack/requirements/raw/branch/master/train/upper-constraints.txt | 18:32 |
| jrosser | i cannot see how that came to be either | 18:32 |
| noonedeadpunk | that is nice question:))) | 18:32 |
| noonedeadpunk | while you have `"keystone_upper_constraints_url": "https://opendev.org/openstack/requirements/raw/train/upper-constraints.txt"` | 18:33 |
| jrosser | yeah | 18:34 |
| noonedeadpunk | but you should have stable/train for that url | 18:34 |
| noonedeadpunk | so opendev get you redirected to https://opendev.org/openstack/requirements/raw/branch/master/train/upper-constraints.txt | 18:34 |
| jrosser | yes so for that PS introducing the split on the branch name was wrong | 18:35 |
| noonedeadpunk | so we should either backport releases url to keystone and stuff | 18:35 |
| jrosser | so i took that out, and then it got further to tempets | 18:35 |
| jrosser | *tempest | 18:35 |
| noonedeadpunk | jrosser: it was wrong for keystone, but was ok for tempest :p | 18:35 |
| * jrosser head->desk | 18:35 | |
| noonedeadpunk | and the latest patch if ok for keystone but wrong for tempest | 18:36 |
| jrosser | what am i missing in the code? | 18:36 |
| jrosser | i've been trying to find a difference between os_tempest and os_keystone to explain this | 18:36 |
| *** irclogbot_1 has quit IRC | 18:37 | |
| noonedeadpunk | so keystone_upper_constraints_url has `https://opendev.org/openstack/requirements/raw/` url and tempest has https://releases.openstack.org/constraints/upper | 18:37 |
| noonedeadpunk | So I think you need to set in https://review.opendev.org/#/c/714486/5/test-vars.yml requirements_git_url | 18:37 |
| noonedeadpunk | instead of requirements_git_install_branch | 18:38 |
| noonedeadpunk | like requirements_git_url: "https://releases.openstack.org/constraints/upper/{{ test_branch.split('/')[-1] }}" | 18:39 |
| noonedeadpunk | or requirements_git_url: "{{ 'https://opendev.org/openstack/requirements/raw/' ~ requirements_git_install_branch | default('master') ~ '/upper-constraints.txt' }}" | 18:39 |
| jrosser | hmm | 18:40 |
| noonedeadpunk | * requirements_git_url: "{{ 'https://opendev.org/openstack/requirements/raw/' ~ test_branch ~ '/upper-constraints.txt' }}" | 18:40 |
| noonedeadpunk | or we need to backport/rollback stuff | 18:40 |
| jrosser | or should we patch os_tempest | 18:40 |
| jrosser | becasue that seems to be the odd one out on train | 18:40 |
| jrosser | i just compare keystone and glance and they are the same | 18:41 |
| noonedeadpunk | yeah, that's the thing as well | 18:41 |
| noonedeadpunk | but not really sure it's the only one... | 18:41 |
| noonedeadpunk | and not sure that all master is updated at the moment | 18:41 |
| jrosser | we will probably be in a situation of circular deps again | 18:41 |
| jrosser | i think your idea to use requirements_git_url in test-vars is good because it will allow to fix these other things | 18:42 |
| jrosser | right i will have a go at that tomorrow | 18:43 |
| noonedeadpunk | we can actually do it now, then update tempest and sync master, and change test-vars to use requirements_git_install_branch again | 18:43 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-tests stable/train: Set requirements_git_install_branch during functional tests https://review.opendev.org/714486 | 18:47 |
| jrosser | ^ lets try that | 18:47 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-tests stable/train: Set requirements_git_install_branch during functional tests https://review.opendev.org/714486 | 18:49 |
| openstackgerrit | Merged openstack/openstack-ansible master: Set bionic upgrade job to voting https://review.opendev.org/713222 | 18:49 |
| jrosser | or even that, which avoids a redirect | 18:49 |
| *** rh-jelabarre has quit IRC | 18:52 | |
| *** rh-jelabarre has joined #openstack-ansible | 18:52 | |
| *** irclogbot_0 has joined #openstack-ansible | 19:01 | |
| *** DanyC has quit IRC | 19:06 | |
| *** DanyC has joined #openstack-ansible | 19:06 | |
| *** DanyC has quit IRC | 19:10 | |
| *** cshen has joined #openstack-ansible | 19:11 | |
| *** cshen has quit IRC | 19:16 | |
| *** macz_ has quit IRC | 19:17 | |
| *** cshen has joined #openstack-ansible | 19:27 | |
| *** cshen has quit IRC | 19:31 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-tests stable/train: Set requirements_git_install_branch during functional tests https://review.opendev.org/714486 | 19:34 |
| *** irclogbot_0 has quit IRC | 19:37 | |
| *** irclogbot_1 has joined #openstack-ansible | 19:39 | |
| *** Openk10s has quit IRC | 19:41 | |
| *** irclogbot_1 has quit IRC | 19:42 | |
| *** irclogbot_1 has joined #openstack-ansible | 19:45 | |
| *** macz_ has joined #openstack-ansible | 19:50 | |
| *** irclogbot_1 has quit IRC | 20:00 | |
| *** irclogbot_3 has joined #openstack-ansible | 20:03 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_horizon master: Add murano-dashboard https://review.opendev.org/710776 | 20:06 |
| *** irclogbot_3 has quit IRC | 20:12 | |
| *** jamesden_ has quit IRC | 20:12 | |
| *** irclogbot_2 has joined #openstack-ansible | 20:15 | |
| *** irclogbot_2 has quit IRC | 20:16 | |
| *** irclogbot_0 has joined #openstack-ansible | 20:21 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_zun master: Refactor memcached_servers https://review.opendev.org/715001 | 20:26 |
| *** tosky has quit IRC | 20:40 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-tests stable/train: Set requirements_git_url during functional tests https://review.opendev.org/714486 | 20:51 |
| *** tosky has joined #openstack-ansible | 20:54 | |
| *** cshen has joined #openstack-ansible | 21:01 | |
| *** cshen has quit IRC | 21:06 | |
| *** fghaas has joined #openstack-ansible | 21:07 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_mistral master: Refactor memcached_servers https://review.opendev.org/714996 | 21:14 |
| *** cshen has joined #openstack-ansible | 21:17 | |
| *** cshen has quit IRC | 21:22 | |
| *** DanyC has joined #openstack-ansible | 21:36 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_murano master: Refactor memcached_servers https://review.opendev.org/715000 | 21:38 |
| *** gshippey has quit IRC | 21:39 | |
| *** fghaas has quit IRC | 21:41 | |
| *** spatel has quit IRC | 21:52 | |
| *** jonmasters has joined #openstack-ansible | 22:01 | |
| *** oligau has quit IRC | 22:03 | |
| *** rh-jelabarre has quit IRC | 22:16 | |
| *** Jeffrey4l has quit IRC | 22:27 | |
| *** Jeffrey4l has joined #openstack-ansible | 22:29 | |
| bjoernt | Hey @mnaser any plans when 20.0.3 get cut? | 22:53 |
| openstackgerrit | Merged openstack/openstack-ansible-os_horizon master: Do not become `horizon` user for distro installs https://review.opendev.org/714931 | 23:16 |
| *** cshen has joined #openstack-ansible | 23:18 | |
| *** cshen has quit IRC | 23:23 | |
| *** macz_ has quit IRC | 23:24 | |
| *** cshen has joined #openstack-ansible | 23:34 | |
| *** cshen has quit IRC | 23:39 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_horizon master: Refactor memcached_servers https://review.opendev.org/713236 | 23:58 |
| *** kplant has joined #openstack-ansible | 23:58 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!