| *** tosky has quit IRC | 00:07 | |
| xdfil | cloudnull: disregard, it was an inventory syntax issue because I moved cinder_backend definition from o_u_c to u_v | 00:09 |
|---|---|---|
| openstackgerrit | caoyuan proposed openstack/openstack-ansible-os_zaqar master: use include_tasks instead of include https://review.openstack.org/593428 | 00:10 |
| xdfil | that being said, this is a perfect example of how my time is consumed by OSA over trivial issues | 00:12 |
| openstackgerrit | caoyuan proposed openstack/openstack-ansible-os_zaqar master: use include_tasks instead of include https://review.openstack.org/593428 | 00:12 |
| xdfil | I know a lot of emphasis has been put on speeding up OSA run time | 00:12 |
| xdfil | I could deal with longer run times if I didn't have to deal with as many "trivial" issues... thats a more difficult problem to solve though | 00:13 |
| openstackgerrit | caoyuan proposed openstack/openstack-ansible-os_zaqar master: use include_tasks instead of include https://review.openstack.org/593428 | 00:14 |
| openstackgerrit | caoyuan proposed openstack/openstack-ansible-os_zaqar master: use include_tasks instead of include https://review.openstack.org/593428 | 00:16 |
| openstackgerrit | caoyuan proposed openstack/openstack-ansible-os_aodh master: Remove the unnecessary = https://review.openstack.org/583995 | 00:19 |
| openstackgerrit | caoyuan proposed openstack/openstack-ansible-os_glance master: use include_tasks instead of include https://review.openstack.org/593439 | 00:21 |
| *** spartakos has quit IRC | 00:30 | |
| openstackgerrit | caoyuan proposed openstack/openstack-ansible-openstack_openrc stable/queens: Fix the home-page https://review.openstack.org/600282 | 00:34 |
| openstackgerrit | caoyuan proposed openstack/openstack-ansible-os_searchlight stable/queens: Fix the home-page link https://review.openstack.org/600283 | 00:34 |
| *** gregoryo has joined #openstack-ansible | 00:34 | |
| openstackgerrit | caoyuan proposed openstack/openstack-ansible-galera_client stable/queens: Fix the home-page link https://review.openstack.org/600284 | 00:34 |
| *** cshen has joined #openstack-ansible | 00:46 | |
| *** cshen has quit IRC | 00:50 | |
| *** gyee has quit IRC | 00:53 | |
| *** blinkiz has quit IRC | 01:19 | |
| *** mmercer has quit IRC | 01:22 | |
| *** blinkiz has joined #openstack-ansible | 01:31 | |
| *** lbragstad has quit IRC | 01:43 | |
| *** lbragstad has joined #openstack-ansible | 01:44 | |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-repo_build master: Update pypi url to new url Pypi url changed from [1] to [2] [1] https://pypi.python.org/{var} [2] https://pypi.org/{var} https://review.openstack.org/600293 | 02:06 |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-repo_build master: Update pypi url to new url https://review.openstack.org/600293 | 02:08 |
| *** lbragstad has quit IRC | 02:09 | |
| *** spartakos has joined #openstack-ansible | 02:10 | |
| *** nicolasbock has quit IRC | 02:18 | |
| *** aicilamd has quit IRC | 02:20 | |
| *** aicilamd has joined #openstack-ansible | 02:23 | |
| *** aicilamd has quit IRC | 02:24 | |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-os_zun master: Replace Chinese punctuation with English punctuation https://review.openstack.org/600296 | 02:34 |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-os_zun master: [Trivial Fix] Replace Chinese punctuation with English punctuation https://review.openstack.org/600296 | 02:35 |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-os_keystone master: [Trivial Fix] Replace Chinese punctuation with English punctuation https://review.openstack.org/600297 | 02:37 |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-os_glance master: [Trivial Fix] Replace Chinese punctuation with English punctuation https://review.openstack.org/600298 | 02:39 |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-memcached_server master: [Trivial Fix] Replace Chinese punctuation with English punctuation https://review.openstack.org/600299 | 02:40 |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-os_monasca master: [Trivial Fix] Replace Chinese punctuation with English punctuation https://review.openstack.org/600300 | 02:41 |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-os_masakari master: [Trivial Fix] Replace Chinese punctuation with English punctuation https://review.openstack.org/600301 | 02:41 |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-os_magnum master: [Trivial Fix] Replace Chinese punctuation with English punctuation https://review.openstack.org/600302 | 02:42 |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-os_designate master: [Trivial Fix] Replace Chinese punctuation with English punctuation https://review.openstack.org/600303 | 02:43 |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-os_barbican master: [Trivial Fix] Replace Chinese punctuation with English punctuation https://review.openstack.org/600304 | 02:44 |
| openstackgerrit | wangqi proposed openstack/openstack-ansible-os_blazar master: format rst code and add bug link https://review.openstack.org/600305 | 02:44 |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-openstack_hosts master: [Trivial Fix] Replace Chinese punctuation with English punctuation https://review.openstack.org/600306 | 02:45 |
| openstackgerrit | wangqi proposed openstack/openstack-ansible-os_magnum master: Add bugs url link to README https://review.openstack.org/600307 | 02:47 |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-os_swift master: [Trivial Fix] Replace Chinese punctuation with English punctuation https://review.openstack.org/600308 | 02:48 |
| openstackgerrit | wangqi proposed openstack/openstack-ansible-os_cinder master: Add bugs url link to README https://review.openstack.org/600309 | 02:49 |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-os_sahara master: [Trivial Fix] Replace Chinese punctuation with English punctuation https://review.openstack.org/600310 | 02:51 |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-os_nova master: [Trivial Fix] Replace Chinese punctuation with English punctuation https://review.openstack.org/600311 | 02:52 |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-os_octavia master: [Trivial Fix] Replace Chinese punctuation with English punctuation https://review.openstack.org/600312 | 02:53 |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-nspawn_container_create master: [Trivial Fix] Replace Chinese punctuation with English punctuation https://review.openstack.org/600313 | 02:54 |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-galera_server master: [Trivial Fix] Replace Chinese punctuation with English punctuation https://review.openstack.org/600314 | 02:55 |
| openstackgerrit | wangqi proposed openstack/openstack-ansible-os_nova master: format rst code and add bug link https://review.openstack.org/600315 | 02:57 |
| openstackgerrit | wangqi proposed openstack/openstack-ansible-haproxy_server master: add bug link https://review.openstack.org/600317 | 03:11 |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-repo_build master: [Trivial Fix] Replace Chinese punctuation with English punctuation https://review.openstack.org/600318 | 03:24 |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-os_watcher master: [Trivial Fix] Replace Chinese punctuation with English punctuation https://review.openstack.org/600319 | 03:25 |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-os_searchlight master: [Trivial Fix] Replace Chinese punctuation with English punctuation https://review.openstack.org/600320 | 03:26 |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-os_rally master: [Trivial Fix] Replace Chinese punctuation with English punctuation https://review.openstack.org/600321 | 03:27 |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-os_monasca-agent master: [Trivial Fix] Replace Chinese punctuation with English punctuation https://review.openstack.org/600322 | 03:27 |
| openstackgerrit | inspurericzhang proposed openstack/openstack-ansible-os_cloudkitty master: [Trivial Fix] Replace Chinese punctuation with English punctuation https://review.openstack.org/600323 | 03:28 |
| *** gkadam has joined #openstack-ansible | 03:31 | |
| *** spartakos has quit IRC | 03:35 | |
| *** nwonknu has quit IRC | 03:48 | |
| *** radeks has joined #openstack-ansible | 03:49 | |
| *** radeks_ has joined #openstack-ansible | 03:51 | |
| *** radeks has quit IRC | 03:54 | |
| *** nwonknu has joined #openstack-ansible | 04:00 | |
| *** radeks_ has quit IRC | 04:03 | |
| *** poopcat has quit IRC | 04:03 | |
| *** markvoelker has joined #openstack-ansible | 04:06 | |
| *** markvoelker has quit IRC | 04:11 | |
| *** radeks_ has joined #openstack-ansible | 04:30 | |
| *** marst has joined #openstack-ansible | 04:40 | |
| *** shyamb has joined #openstack-ansible | 04:45 | |
| *** Bhujay has joined #openstack-ansible | 04:45 | |
| *** radeks_ has quit IRC | 05:06 | |
| *** markvoelker has joined #openstack-ansible | 05:07 | |
| *** cshen has joined #openstack-ansible | 05:15 | |
| *** cshen has quit IRC | 05:20 | |
| *** shyamb has quit IRC | 05:25 | |
| *** cshen has joined #openstack-ansible | 05:30 | |
| *** cshen has quit IRC | 05:35 | |
| *** shyamb has joined #openstack-ansible | 05:36 | |
| *** udesale has joined #openstack-ansible | 05:38 | |
| *** gkadam_ has joined #openstack-ansible | 05:41 | |
| *** gkadam has quit IRC | 05:43 | |
| *** gkadam__ has joined #openstack-ansible | 05:45 | |
| *** gkadam_ has quit IRC | 05:47 | |
| *** DanyC has joined #openstack-ansible | 05:58 | |
| *** chandankumar is now known as chkumar|ruck | 05:58 | |
| *** rgogunskiy has joined #openstack-ansible | 05:59 | |
| *** Bhujay has quit IRC | 06:07 | |
| *** vnogin has quit IRC | 06:12 | |
| *** vnogin has joined #openstack-ansible | 06:12 | |
| *** Bhujay has joined #openstack-ansible | 06:18 | |
| *** pcaruana has joined #openstack-ansible | 06:25 | |
| *** DanyC has quit IRC | 06:27 | |
| *** udesale has quit IRC | 06:29 | |
| *** cshen has joined #openstack-ansible | 06:32 | |
| ivve | anyone running 17.1.0 with ceph backend? | 06:34 |
| ivve | gnocchi related quesion... | 06:35 |
| *** shyam89 has joined #openstack-ansible | 06:37 | |
| *** shyamb has quit IRC | 06:41 | |
| *** markvoelker has quit IRC | 06:45 | |
| *** fghaas has joined #openstack-ansible | 06:50 | |
| ivve | nevermind, it had to do with syntax from old -> new config | 06:55 |
| ivve | pike -> queens | 06:55 |
| *** Miouge is now known as miouge | 07:00 | |
| *** aludwar has quit IRC | 07:01 | |
| evrardjp | good morning | 07:02 |
| *** shardy has joined #openstack-ansible | 07:04 | |
| *** aludwar has joined #openstack-ansible | 07:05 | |
| *** Emine has joined #openstack-ansible | 07:06 | |
| *** priteau has joined #openstack-ansible | 07:09 | |
| jrosser | Morning | 07:14 |
| *** Taseer has joined #openstack-ansible | 07:14 | |
| jrosser | evrardjp: could we get this on the way https://review.openstack.org/#/c/594926/ ? | 07:16 |
| evrardjp | jrosser: wow | 07:19 |
| evrardjp | great | 07:19 |
| *** vnogin has quit IRC | 07:20 | |
| evrardjp | jrosser: I -1-ed for the documentation | 07:20 |
| evrardjp | I think if we add this as voting, we make it kinda official, so we could add the little changes in the docs : ) | 07:21 |
| jrosser | Could you be a bit more specific on “other documentations”? :) | 07:22 |
| evrardjp | I am listing right now | 07:22 |
| jrosser | Excellent, I’ll sort it | 07:22 |
| hwoarang | good morning | 07:23 |
| hwoarang | fyi, CI problems due to failing opensuse jobs are being discussed on #infra. infra mirrors are heavily outdated | 07:24 |
| evrardjp | jrosser: updated my comment | 07:25 |
| evrardjp | hwoarang: oh | 07:25 |
| evrardjp | is that since there was a change of the mirrors used in infra for suse? | 07:26 |
| evrardjp | hwoarang: ^ | 07:26 |
| hwoarang | was there such a change? | 07:26 |
| jrosser | evrardjp: thanks - there’s also a bit of a checking to do that all the right bionic bits are backported to rocky | 07:27 |
| jrosser | Like role tests and meta info | 07:27 |
| hwoarang | hmm the mirror changed for regular repos but for OBS is the same. and it's the OBS sync that's failing now | 07:27 |
| *** udesale has joined #openstack-ansible | 07:27 | |
| evrardjp | hwoarang: I see, but let's discuss this matter in #openstack-infra | 07:30 |
| *** pcaruana has quit IRC | 07:36 | |
| *** shyam89 has quit IRC | 07:38 | |
| Taseer | evrardjp: can you please review this before the PTG? https://review.openstack.org/#/c/571388/ | 07:39 |
| *** tosky has joined #openstack-ansible | 07:41 | |
| *** luksky has joined #openstack-ansible | 07:41 | |
| evrardjp | Taseer: lgtm | 07:42 |
| *** markvoelker has joined #openstack-ansible | 07:43 | |
| openstackgerrit | lvxianguo proposed openstack/openstack-ansible-os_blazar master: modify the format of release notes link https://review.openstack.org/600361 | 07:47 |
| *** pcaruana has joined #openstack-ansible | 07:48 | |
| ivve | o hey guys, gnocchi playbooks should "pip install uwsgi" into gnocchi containers as it is missing | 07:51 |
| ivve | gnocchi-api complains about: ERROR gnocchi.cli.api: Unable to find `uwsgi'. | 07:51 |
| ivve | 17.1.0 (ubuntu 16.04) | 07:52 |
| *** hamzaachi_ has joined #openstack-ansible | 07:52 | |
| ivve | its looking in $PATH, perhaps it should be looking elsewhere? | 07:52 |
| *** radeks has joined #openstack-ansible | 07:52 | |
| *** priteau has quit IRC | 07:52 | |
| ivve | then its a configuration thingie for gnocchi-api | 07:52 |
| Taseer | evrardjp: Thanks! | 07:54 |
| *** hamzaachi_ has quit IRC | 07:57 | |
| *** cshen has quit IRC | 07:57 | |
| *** hamzaachi_ has joined #openstack-ansible | 07:58 | |
| *** shyamb has joined #openstack-ansible | 07:58 | |
| evrardjp | ivve: check it's doing it in the right venv location ? | 08:06 |
| evrardjp | jrosser: could you ping me when you have the new patch handy? :D so that I can review quickly :) | 08:07 |
| ivve | gnocchi_pip_packages: not containing uwsgi, should it? | 08:07 |
| ioni | hi guys | 08:09 |
| ioni | evrardjp, can i pm you regarding a video that i saw earlier about rocky release? | 08:09 |
| *** shyamb has quit IRC | 08:14 | |
| noonedeadpunk | mornings everyone | 08:15 |
| evrardjp | ioni: you can try, but I am not sure I will have time to watch it? :) | 08:16 |
| noonedeadpunk | evrardjp: if you have some time for reviews, maybe you'll take a look at https://review.openstack.org/#/c/600072/ ? | 08:16 |
| evrardjp | ioni: shoot :) | 08:16 |
| ioni | evrardjp, it seems that i found the information | 08:17 |
| ioni | evrardjp, i was asking about mohamed(i hope that i wrote his name correctly) plugged his company and i didn't understand his company name | 08:17 |
| ioni | :D | 08:17 |
| ioni | and wondering if you know | 08:17 |
| ivve | evrardjp: im no expert in python venvs but i cant see any playbook installing the uwsgi pip package at all.. anywhere in conjunction with gnocchi atleast | 08:18 |
| *** Taseer has quit IRC | 08:19 | |
| ivve | meaning, it doesn't do it at all | 08:19 |
| evrardjp | noonedeadpunk: I just commented | 08:20 |
| evrardjp | if you update the commit message I will balance my vote for a +1/-1. :) | 08:20 |
| ivve | there is not a single playbook mentioning uwsgi in the os-gnocchi | 08:21 |
| noonedeadpunk | evrardjp: I've tested it in my production, but on queens | 08:21 |
| evrardjp | ioni: mnaser's company is vexxhost: https://vexxhost.com/ | 08:21 |
| noonedeadpunk | I'll update commit message with reference | 08:21 |
| ioni | evrardjp, thanks | 08:21 |
| evrardjp | ivve: we might have a gap there and something requiring fixes | 08:22 |
| evrardjp | ivve: which branch? | 08:22 |
| *** olivierb has joined #openstack-ansible | 08:22 | |
| evrardjp | ivve: https://github.com/openstack/openstack-ansible-os_keystone/blob/master/defaults/main.yml#L448 does not appear in https://github.com/openstack/openstack-ansible-os_gnocchi/blob/master/defaults/main.yml#L177-L184 | 08:24 |
| evrardjp | ivve: which probably means that gnocchi was not moved to uwsgi | 08:24 |
| evrardjp | ivve: if you're willing to do it, it would be helpful | 08:24 |
| evrardjp | ivve: it seems to rely on apache setup, which is something we'd like to change... patches welcome | 08:25 |
| openstackgerrit | Dmitriy Rabotjagov (noonedeadpunk) proposed openstack/openstack-ansible-os_ceilometer master: Changed messaging_urls for notification agent https://review.openstack.org/600072 | 08:25 |
| *** DanyC has joined #openstack-ansible | 08:26 | |
| ivve | evrardjp: 17.1.0 | 08:26 |
| ivve | however gnocchi venv is called 17.1.1 | 08:27 |
| ivve | the openstack-ansible tag is: 17.1.0-3-g0ceb08d | 08:28 |
| jrosser | evrardjp: just so i get this right first time - what should be the supported ubuntu releases for master? depending on the answer to that adding these docs changes means that patch won't be a clean backport to rocky btw | 08:29 |
| evrardjp | noonedeadpunk: I am curious what's the difference in generated transport_url? Maybe I am blind... | 08:31 |
| evrardjp | noonedeadpunk: oh I understand | 08:32 |
| noonedeadpunk | evrardjp: difference only in the way of generation. I decided to move to one construction. Also this line is now not so long) | 08:32 |
| evrardjp | could you clarify in the commit message that the input of ceilometer_oslomsg_rpc_servers is a list instead of a comma separated string ? | 08:32 |
| *** DanyC has quit IRC | 08:32 | |
| *** shyamb has joined #openstack-ansible | 08:34 | |
| ivve | evrardjp: this is my conclusion, adding uwsgi to the gnocchi_pip_packages solves the problem.. however i don't think im qualified to say its an accepted solution due to software infrastructure requirements such as you mentioned :) | 08:35 |
| evrardjp | ivve: I don't think it's enough :) It would work, but I think we should work on getting uwsgi done for gnocchi instead. | 08:39 |
| noonedeadpunk | yep, give me a moment, I'm not sure about list anymore | 08:40 |
| ivve | evrardjp: you're the boss! im just saying the deployment fails with current playbook | 08:41 |
| evrardjp | noonedeadpunk: I checked, and reviewed | 08:42 |
| ivve | and thats why im not qualified ;) | 08:42 |
| evrardjp | the macros is a plus not a necessary thing | 08:42 |
| evrardjp | ivve: I've never been the _boss_ :) | 08:42 |
| *** gregoryo has quit IRC | 08:42 | |
| evrardjp | ivve: makes sense if gnocchi now requires uwsgi | 08:42 |
| evrardjp | which concerns me even mroe | 08:43 |
| ivve | you're qualified to make the decisions, im not :) | 08:43 |
| ivve | measures doesn't end up in ceph and can't be queried for without the package | 08:44 |
| ivve | (or whatever backen you're using) | 08:44 |
| ivve | backend | 08:44 |
| noonedeadpunk | evrardjp: now I a bit confused about join of rabbitmq_host_group with comas and then splitting it for every service... | 08:44 |
| ivve | its wierd because that wasn't a problem in pike, the playbooks are very similar. ut i guess there are changes in gnocchi that breaks it | 08:45 |
| ivve | but* | 08:45 |
| noonedeadpunk | looks like smth that worth changing | 08:45 |
| noonedeadpunk | and macros is really good idea. I've completelly forgotten about them | 08:47 |
| *** udesale has quit IRC | 08:47 | |
| *** udesale has joined #openstack-ansible | 08:47 | |
| *** udesale has quit IRC | 08:48 | |
| *** udesale has joined #openstack-ansible | 08:48 | |
| *** electrofelix has joined #openstack-ansible | 08:48 | |
| *** hamzaachi has quit IRC | 08:50 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Move bionic jobs to voting, add release note. https://review.openstack.org/594926 | 08:50 |
| jrosser | evrardjp: ^ how about that? | 08:51 |
| evrardjp | noonedeadpunk: I would love seeing these inputs as lists instead of comma separated strings | 08:51 |
| evrardjp | noonedeadpunk: it makes so much more sense to give proper data | 08:51 |
| evrardjp | noonedeadpunk: however, you're there for a big ride, I tell you :p | 08:51 |
| evrardjp | release notes and all the stuff :) | 08:51 |
| evrardjp | but it's a good change, and I am happy to discuss this. | 08:52 |
| evrardjp | we should discuss it in a meeting with a larger audience maybe? | 08:52 |
| evrardjp | jrosser: you got my vote :) | 08:52 |
| noonedeadpunk | evrardjp: yep, maybe. And its really tons of work which should be done in pressed time because of dependencies... | 08:55 |
| evrardjp | noonedeadpunk: yeah, so I think for your patch, the questions are "what failure do you see" and "have you filed a bug with that?" | 09:07 |
| evrardjp | because with those two answered, we can help on backporting, should backport happen | 09:07 |
| *** DanyC has joined #openstack-ansible | 09:07 | |
| *** hwoarang_ has joined #openstack-ansible | 09:11 | |
| evrardjp | hey, following a discussion with jrosser I think it would be nice to talk about tc elections that are coming up | 09:11 |
| evrardjp | I just wanted to raise the fact that there is something in the community going on, and it's easy to miss it if you're not involved | 09:13 |
| *** hwoarang has quit IRC | 09:15 | |
| evrardjp | I don't want this channel to become a campaining platform, but I just think that if you're a contributor of OpenStack-Ansible, and you're interested in OpenStack, you can help shaping it by talking to the TC candidates about their vision for OpenStack, and then vote for the ones you feel are in line with your requirements | 09:15 |
| *** udesale has quit IRC | 09:16 | |
| *** vnogin has joined #openstack-ansible | 09:16 | |
| evrardjp | FYI information about the next elections are here: https://governance.openstack.org/election/ | 09:16 |
| evrardjp | So please, make sure your emails are properly configured in gerrit (and openstack website!) so that you can receive your ballot at the good email adress :) | 09:18 |
| odyssey4me | morning folks, any update on the suse mirror story? | 09:19 |
| noonedeadpunk | evrardjp: yep, got it, will vote) I'll double check my account at openstack website, but on gerrit emails are ok) | 09:19 |
| jrosser | thanks for the pointers evrardjp - if people are not subscribed to the ML (which i wasn't for a looooong time) or don't grok the acronyms used then this stuff happens invisibly | 09:20 |
| *** vnogin has quit IRC | 09:21 | |
| evrardjp | odyssey4me: yeah, there is a change up | 09:21 |
| evrardjp | odyssey4me: so there is a hardware failure on one of the things to mirror | 09:21 |
| *** shyamb has quit IRC | 09:22 | |
| evrardjp | and the script basically stops there, so no sync happens. | 09:22 |
| evrardjp | dirk has proposed a patch to allow || true | 09:22 |
| evrardjp | this will need to merge in infra before a sync can happen | 09:22 |
| evrardjp | jrosser: yes I think that's why it's important to share it -- we can all work on our sides of the pond, but at the end, it will make our life better to work together :) | 09:23 |
| *** hwoarang_ is now known as hwoarang | 09:23 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-tests stable/queens: Revert "Revert "Run with KVM if we can in tests"" https://review.openstack.org/600392 | 09:29 |
| odyssey4me | evrardjp: heh, perhaps the mirror tooling should be configured to allow multiple sources - so if one fails, perhaps one of the others will work | 09:32 |
| odyssey4me | FYI I've proposed https://review.openstack.org/600392 because I'm really, really tired of rechecking https://review.openstack.org/599621 | 09:33 |
| evrardjp | odyssey4me: only for q? | 09:34 |
| odyssey4me | evrardjp: it's already done some time ago in r & s | 09:35 |
| odyssey4me | it's also already done in the integrated gate for q - just not role tests | 09:35 |
| evrardjp | tired after 4? Imagine my 20! | 09:35 |
| evrardjp | hahahaha | 09:35 |
| odyssey4me | well, that's only one of the patches I've had to recheck multiple times | 09:36 |
| evrardjp | odyssey4me: did you see jrosser 's patch here: https://review.openstack.org/#/c/594926/6 : ) a milestone! | 09:36 |
| odyssey4me | evrardjp: I did, as evidenced by my vote ;) | 09:36 |
| evrardjp | wierd, as I don't see the gerrit notification! | 09:36 |
| evrardjp | a refresh shows up your vote indeed! | 09:37 |
| *** vnogin has joined #openstack-ansible | 09:37 | |
| evrardjp | woot | 09:37 |
| evrardjp | cool! | 09:37 |
| odyssey4me | It'd be nice to get that ported back to R asap. I guess after the PTG we can start pulling Xenial out from S. | 09:37 |
| odyssey4me | jrosser: have you found that nova-compute comes up just fine after a reboot of the host? | 09:38 |
| *** shyamb has joined #openstack-ansible | 09:38 | |
| odyssey4me | I'm finding that even with https://github.com/openstack/openstack-ansible-os_nova/commit/da2b8f44440e92a9a25af17dd9737f7177d1e5bc it seems to come up (the service is running), but instances will not build on it | 09:39 |
| odyssey4me | I've tried restarting libvirt and nova-compute, but that doesn't work. I've tried different orders too. However, running the os-nova-install.yml playbook against it makes it work again... | 09:40 |
| odyssey4me | I'm busy inspecting exactly what happens and what changes when the playbook runs against it to try and figure out what the thing is that matters, but am wondering whether you've encountered this? | 09:40 |
| *** dabukalam has joined #openstack-ansible | 09:41 | |
| odyssey4me | evrardjp: mind taking a peek at https://review.openstack.org/598562 ? | 09:43 |
| *** shyam89 has joined #openstack-ansible | 09:46 | |
| jrosser | odyssey4me: tbh I've not rebooted a host like that yet, you're seeing this on mnaio i'm guessing? | 09:46 |
| evrardjp | odyssey4me: shouldn't this be in the topic create-rocky? | 09:46 |
| odyssey4me | evrardjp: perhaps, it gets the topic changed automatically on submission due to the bug | 09:47 |
| evrardjp | odyssey4me: I am updating the topic so we can track it for next time | 09:47 |
| odyssey4me | happy to have it changed | 09:47 |
| odyssey4me | jrosser: yep, seeing it on MNAIO when restoring an environment from images - basically I have everything coming up right, computes are running, the services are running, but creating instances fails | 09:48 |
| odyssey4me | I initially suspected rabbit, but that's fine - then suspected the control plane services - but restarting them makes no difference. Then I thought it might be start order, but that doesn't seem to make a difference. | 09:49 |
| odyssey4me | Now I'm starting to wonder whether it's something like missing /var/run directories or something - busy digging into it. | 09:49 |
| shyam89 | Hi | 09:49 |
| *** shyamb has quit IRC | 09:50 | |
| shyam89 | jaosorior: | 09:50 |
| shyam89 | On compute node of overcloud, I see ownership of "/var/lib/nova/instances/snapshots" as root:root | 09:51 |
| odyssey4me | jrosser: I figure that anything that stops a service from coming up properly after a reboot is something we want to. solve. | 09:51 |
| shyam89 | all other directories in "/var/lib/nova" has nova:nova ownership | 09:51 |
| shyam89 | jaosorior: Is it expected or some issue? | 09:52 |
| *** udesale has joined #openstack-ansible | 09:53 | |
| shyam89 | Tengu: shardy: owalsh: Any thoughts here? | 09:56 |
| *** shyam89 has quit IRC | 09:56 | |
| *** shyamb has joined #openstack-ansible | 09:56 | |
| odyssey4me | shyamb: none of the people you're talking to are in this channel | 09:57 |
| odyssey4me | Are you perhaps meaning to talk in #tripleo ? | 09:57 |
| shyamb | ohh, sorry | 09:57 |
| shyamb | yeah | 09:57 |
| odyssey4me | :) | 09:57 |
| shyamb | :D | 09:57 |
| openstackgerrit | Merged openstack/openstack-ansible-os_blazar master: Fix functional tests for Blazar https://review.openstack.org/571388 | 09:58 |
| *** shyamb has quit IRC | 10:03 | |
| *** shyamb has joined #openstack-ansible | 10:03 | |
| *** nicolasbock has joined #openstack-ansible | 10:31 | |
| *** luksky has quit IRC | 10:40 | |
| *** luksky has joined #openstack-ansible | 10:42 | |
| *** shyamb has quit IRC | 10:43 | |
| *** dave-mccowan has joined #openstack-ansible | 10:47 | |
| *** shyamb has joined #openstack-ansible | 10:54 | |
| openstackgerrit | Dmitriy Rabotjagov (noonedeadpunk) proposed openstack/openstack-ansible-os_ceilometer master: Changed messaging_urls for notification agent https://review.openstack.org/600072 | 10:57 |
| odyssey4me | interesting, the re-run of os-nova-install changes nova.conf - switching virt_type from 'kvm' to 'qemu' and adding 'cpu_mode = none' | 10:59 |
| *** dcdamien has quit IRC | 11:00 | |
| *** dcdamien has joined #openstack-ansible | 11:03 | |
| noonedeadpunk | evrardjp: macros were made https://review.openstack.org/600072 | 11:04 |
| evrardjp | odyssey4me: wow that sounds bad | 11:05 |
| odyssey4me | evrardjp: hmm, looks like nested virt isn't enabled on this host - digging into it, this may just be a change in config between xenial and bionic | 11:08 |
| odyssey4me | the images were built on xenial, but are being used on bionic | 11:09 |
| odyssey4me | yep, the xenial host has it enabled | 11:09 |
| odyssey4me | hmm, both have the option enabled.... hmm | 11:10 |
| odyssey4me | one is just not active | 11:10 |
| evrardjp | noonedeadpunk: I gave a suggestion on how to simplify this macro further, and make it easier to read (change function name to be explicit that this works for non_rpc, etc.) | 11:17 |
| odyssey4me | jrosser or hwoarang - mind adding your review to https://review.openstack.org/600392 ? | 11:21 |
| hwoarang | on it | 11:22 |
| odyssey4me | tyvm | 11:22 |
| hwoarang | again? ( | 11:22 |
| hwoarang | so be it | 11:22 |
| odyssey4me | yeah :/ | 11:22 |
| odyssey4me | looks like OVH may be having the same issues I am with my testing :p | 11:22 |
| *** Emine has quit IRC | 11:23 | |
| noonedeadpunk | evrardjp: is rewrite of hosts outside of the loop ok? | 11:23 |
| *** Emine has joined #openstack-ansible | 11:25 | |
| noonedeadpunk | ah, you mean to simplify it by removing... actually maybe it's worth to leave in place if we'll ever decide to move from joining the list in order to split it in another role? | 11:25 |
| noonedeadpunk | and this will be smth that will make transition easier? | 11:26 |
| noonedeadpunk | but yeah, this needs discussion | 11:26 |
| blinkiz | hello. On infrastructure nodes, is there any reason to have suggested br-vlan bridge? I Only see need for Internet/management, br-mgmt and br-vxlan on infrastructure nodes? | 11:26 |
| *** hamzaachi_ has quit IRC | 11:28 | |
| blinkiz | Am guessing all my provider networks will be on br-vlan? | 11:29 |
| *** hamzaachi_ has joined #openstack-ansible | 11:29 | |
| openstackgerrit | Dmitriy Rabotjagov (noonedeadpunk) proposed openstack/openstack-ansible-os_ceilometer master: Changed messaging_urls for notification agent https://review.openstack.org/600072 | 11:34 |
| evrardjp | noonedeadpunk: I think I pasted an example code... If we discuss about changing that "interface" from comma separated strings to a list, then the function will not change with my example code, merely the call. | 11:34 |
| noonedeadpunk | evrardjp: sorry, i meant not change of the function, but the change of the template, as then we'll need to pass another var into the macros | 11:35 |
| evrardjp | noonedeadpunk: let me review it once again :) | 11:35 |
| jrosser | blinkiz: your answer is in here https://docs.openstack.org/openstack-ansible/latest/user/prod/example.html | 11:35 |
| jrosser | blinkiz: look in particular there at how eth12 is wired up to br-vlan. This allows untagged traffic on br-vlan to be the external network. Neutron requires an interface rather than a bridge in it's external network config, hence eth12 | 11:36 |
| noonedeadpunk | evrardjp: if you think this this simplification is really worth it - no problem in implementing it. But I think macros in this way is more flexible | 11:37 |
| jrosser | blinkiz: remember that this is just an example, you don't have to do it like that if you don't want to. It's possible to bring in external networks pretty much how you wish | 11:37 |
| blinkiz | jrosser, Thanks for the explanation. | 11:39 |
| blinkiz | noonedeadpunk, What do you mean with macros | 11:39 |
| noonedeadpunk | blinkiz: we were talking about this patch https://review.openstack.org/600072 today I was a bit surprised with the way of making rabbitmq_servers, where we firstly join list into string, separated with comas and then splitting it back inside configuration templates. | 11:42 |
| blinkiz | noonedeadpunk, aha, got it :) | 11:42 |
| noonedeadpunk | So my concern was to write this macros in a way, that it won't mind about what is passed into it - list or string and would make it's work in both cases And if we move away from such join/split we won't need to change at least ceilometer role) | 11:44 |
| evrardjp | well the macro was just to simplify and DRY | 11:45 |
| evrardjp | so I think we are good | 11:45 |
| evrardjp | it's now cosmetic :) | 11:45 |
| noonedeadpunk | actually yes) | 11:45 |
| evrardjp | but cosmetic is sometimes important, as it helps being maintainable. If we have a very big jinja that's unreadable, then it becomes harder. | 11:45 |
| evrardjp | to maintain* | 11:46 |
| noonedeadpunk | the problem is, that transport_url and messaging_urls are really long, so this split won't be really in sight | 11:47 |
| evrardjp | noonedeadpunk: so for your info, the host_groups parameter could as well have been replaced with a map function on the call :) | 11:47 |
| evrardjp | so that you pass about the same thing on each of them, but generate the right string | 11:48 |
| evrardjp | I can show you that later when I have finished with what I am doing | 11:48 |
| evrardjp | which may take days! :p | 11:48 |
| noonedeadpunk | evrardjp: and there's no common example in existing roles? | 11:51 |
| evrardjp | you are setting all the good precedents noonedeadpunk :D | 11:53 |
| *** shyamb has quit IRC | 11:57 | |
| *** hamzaachi_ has quit IRC | 12:01 | |
| *** shyamb has joined #openstack-ansible | 12:05 | |
| blinkiz | I would like to ask about br-vxlan interface. Why is that a bridge on infrastructure nodes? I have currently no containers attached to br-vxlan bridge. In /etc/openstack_deploy/openstack_user_config.yml I can see that br-vxlan should be assigned to group_binds neutron_linuxbridge_agent. On my infrastructure nodes I do not have L2 bridges. Maybe I will in the future? | 12:06 |
| jrosser | blinkiz: do you have dedicated network nodes? | 12:07 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ops master: MNAIO: Ensure that virt-networks are properly setup https://review.openstack.org/600424 | 12:07 |
| blinkiz | maybe am missing something because on my infrastructure nodes I have four brqxxxxxx bridges that has vxlan-xx interfaces | 12:07 |
| blinkiz | jrosser, no, I do not currently have that. This functionality is on infrastructure nodes at the moment. | 12:08 |
| jrosser | blinkiz: if you don't have dedicated network nodes then your infra nodes are running the neutron agents | 12:08 |
| jrosser | and as such they need to create vxlan interfaces for the router and dhcp namespaces and plumb that all up on the infra nodes | 12:08 |
| jrosser | which i think is what you are seeing there with those brq..... bridges | 12:09 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ops master: MNAIO: Ensure that virt-networks are properly setup https://review.openstack.org/600424 | 12:11 |
| openstackgerrit | Merged openstack/openstack-ansible-os_nova stable/queens: Ensure nova-compute service start after libvirtd https://review.openstack.org/599621 | 12:19 |
| *** vnogin has quit IRC | 12:20 | |
| odyssey4me | yay! it finally merged :) | 12:20 |
| guilhermesp | that's nice odyssey4me ! :D | 12:21 |
| odyssey4me | guilhermesp: yep, now that's in all branches - so the start order is right when our next sha bump happens | 12:22 |
| openstackgerrit | Nicolas Bock proposed openstack/openstack-ansible stable/pike: scripts: bootstrap-ansible.sh: Do not error on 'alternatives' command https://review.openstack.org/600427 | 12:22 |
| *** markvoelker has quit IRC | 12:22 | |
| guilhermesp | odyssey4me: nice nice nice | 12:22 |
| odyssey4me | thanks nicolasbock - I appreciate your petience :) | 12:23 |
| blinkiz | In recommendation for network setup there are four interfaces for infrastructure nodes. eth0, eth1, eth2, eth3 which creates bond0 and bond1. bond0 is br-mgmt and upstream Internet. bond1 is br-vxlan and br-vlan. Does this mean bond0 will be happy with 10x2 Gbit and bond1 is at 40x2 Gbit? Or do I miss something on bond0 that will require high bandwidth from a dual 40 Gbit? | 12:23 |
| odyssey4me | *patience | 12:23 |
| nicolasbock | It didn't take that long odyssey4me ;) | 12:23 |
| *** shyamb has quit IRC | 12:24 | |
| *** shyamb has joined #openstack-ansible | 12:24 | |
| nicolasbock | Could you review https://review.openstack.org/#/c/600272/ odyssey4me ? | 12:24 |
| openstackgerrit | Dmitriy Rabotjagov (noonedeadpunk) proposed openstack/openstack-ansible-os_ceilometer master: Changed messaging_urls for notification agent https://review.openstack.org/600072 | 12:24 |
| odyssey4me | nicolasbock: I'll have to trust you on that one. Perhaps evrardjp or hwoarang should also look at it. | 12:25 |
| jamesdenton | blinkiz that configuration should be OK for the majority of uses | 12:25 |
| nicolasbock | Good ides | 12:25 |
| nicolasbock | idea | 12:25 |
| noonedeadpunk | evrardjp: I've placed changes, but left split in macros as for now. So if you think, that it should be outside of macros - I'll change it now, but my opinion is that it's ok to be inside macros | 12:27 |
| blinkiz | jamesdenton, Thanks for the response. Do I even need 10x2 Gbit for bond0? Even 1 Gbit is enough here. br-mgmt does not need bandwidth, right? | 12:29 |
| jamesdenton | It depends. We typically run 10G there to better support Glance-related traffic. We standardize on 10G interfaces all around, usually | 12:30 |
| blinkiz | jamesdenton, glance traffic.. okay.. Do you think there will be a need for 40 Gbit on br-mgmt in the future? | 12:34 |
| jamesdenton | highly doubtful. But, one might use bond0 for other traffic types (e.g. ceph replication), or sometimes collapse all traffic on a single bond and might make a case for 40G | 12:35 |
| openstackgerrit | Merged openstack/openstack-ansible-tests stable/queens: Revert "Revert "Run with KVM if we can in tests"" https://review.openstack.org/600392 | 12:35 |
| jamesdenton | but i haven't really seen anyone max out a 10G interface consistently, let alone a 20G bond | 12:36 |
| blinkiz | jamesdenton, thanks for all the responses. I now feel more confided to run at 10 Gbit for br-mgmt and have 40x2 Gbit for br-vlan and br-vxlan. :) | 12:37 |
| jamesdenton | should be pretty beefy. good luck | 12:37 |
| jrosser | jamesdenton: quick question - do you run snmpd on network nodes? on mine is is going bezerk over a long period | 12:38 |
| openstackgerrit | Dmitriy Rabotjagov (noonedeadpunk) proposed openstack/openstack-ansible-os_ceilometer master: Changed messaging_urls for notification agent https://review.openstack.org/600072 | 12:38 |
| jamesdenton | jrosser not that i know of. what's happening? | 12:38 |
| jrosser | snmpd is consuming 100% of one cpu core and 50% of the RAM on the box | 12:39 |
| jamesdenton | yeah that sounds like a problem :D | 12:39 |
| noonedeadpunk | evrardjp: you won, explicit is better than implicit :) | 12:39 |
| jrosser | my hunch is that it sees all the neutron interfaces coming/going and just grows and grows | 12:39 |
| jamesdenton | maybe? But on a network node you might have some l3 and dhcp interfaces. i would think a compute would be more problematic | 12:40 |
| jamesdenton | unless you're polling everything from those nodes. not sure what your setup looks like | 12:40 |
| openstackgerrit | Dmitriy Rabotjagov (noonedeadpunk) proposed openstack/openstack-ansible-os_ceilometer master: Changed messaging_urls for notification agent https://review.openstack.org/600072 | 12:41 |
| jrosser | it doesnt seem to grow in the same way on a compute node | 12:41 |
| jamesdenton | maybe it doesn't like veth interfaces? | 12:42 |
| jamesdenton | i'm at a loss | 12:42 |
| *** ansmith has joined #openstack-ansible | 12:43 | |
| jrosser | no worries - i'll dig a bit. Ideally i just make it ignore anything thats not a physical interface | 12:43 |
| *** hamzaachi_ has joined #openstack-ansible | 12:43 | |
| *** aicilamd has joined #openstack-ansible | 12:46 | |
| jamesdenton | is cloudnull at PTG all week? | 12:47 |
| jamesdenton | or thats next week, isnt it | 12:47 |
| *** lbragstad has joined #openstack-ansible | 12:54 | |
| evrardjp | noonedeadpunk: you know your zen: ) | 13:08 |
| *** lbragstad has quit IRC | 13:09 | |
| guilhermesp | ops guys, how viable is change the i ranges of container, tunnel and storage networks in a production env? I assume that is a risky operation.. | 13:14 |
| noonedeadpunk | evrardjp: I have a question - why do we need ?ssl=0 here https://github.com/openstack/openstack-ansible-os_ceilometer/blob/6e0502655ad9dcdf45c47d9c35196865e10d8d50/templates/ceilometer.conf.j2#L12 and if we need it - why it's absent for messaging_urls? | 13:14 |
| *** lbragstad has joined #openstack-ansible | 13:15 | |
| noonedeadpunk | will it be ok to add ?ssl=0 for all messaging_urls or may I remove it from else section in transport_url? | 13:16 |
| evrardjp | noonedeadpunk: that's a very good question, that I'd like to pass to amsmith, which currently isn't in our channel. odyssey4me maybe you can answer that one? :) | 13:17 |
| openstackgerrit | Michael Vollman proposed openstack/ansible-config_template master: Compare dict vars to determine changed https://review.openstack.org/582532 | 13:18 |
| evrardjp | I think it would be okay to add ?ssl=0 | 13:18 |
| jamesdenton | guilhermesp that's not really something i'd want to attempt. those IPs end up in all sorts of config files. | 13:19 |
| *** aicilamd has quit IRC | 13:19 | |
| jamesdenton | not to mention configured in the containers themselves | 13:19 |
| evrardjp | guilhermesp: yeah I agree with jamesdenton . It's also on all the nodes /etc/hosts files | 13:20 |
| guilhermesp | yep jamesdenton ... the scenario is a cloud that was deployed with wrong ip ranges... I think is almost impossible to avoid downtimes since we need first to change the container ips and then reconfigure all the service to look at the new ips | 13:20 |
| evrardjp | you will probably break rabbitmq which can be fun. | 13:20 |
| jamesdenton | how deployed is deployed? :D | 13:20 |
| noonedeadpunk | Now I don't have any really working ceilometer on rocky to test it, and on queens there's no oslo_messaging, so ?ssl=0 breaks ceilometer | 13:20 |
| evrardjp | noonedeadpunk: I see | 13:21 |
| guilhermesp | yep evrardjp could be more pain than joy :P | 13:21 |
| evrardjp | guilhermesp: you got it ;) | 13:21 |
| jamesdenton | i mean, realistically, every cloud could have the same ranges. None of them have L3 gateways - the only IP you might hit externally would be the VIP. | 13:21 |
| evrardjp | or vips depending on you do things, but yeah, internal is very self contained | 13:21 |
| jamesdenton | right | 13:22 |
| guilhermesp | I see jamesdenton and evrardjp so I see.. probably I'm the first one that is trying to attempt such thing | 13:23 |
| guilhermesp | hahah | 13:23 |
| openstackgerrit | Michael Vollman proposed openstack/ansible-config_template master: Compare dict vars to determine changed https://review.openstack.org/582532 | 13:23 |
| jamesdenton | probably :) but it may be avoidable altogether unless you require non-overlapping subnets as some sort of policy. | 13:23 |
| *** rgogunskiy has quit IRC | 13:24 | |
| jrosser | i remember mnaser talking about doing something like this? | 13:24 |
| guilhermesp | I believe that this was the scenario :P | 13:25 |
| openstackgerrit | Dmitriy Rabotjagov (noonedeadpunk) proposed openstack/openstack-ansible-os_ceilometer master: Changed messaging_urls for notification agent https://review.openstack.org/600072 | 13:25 |
| *** aicilamd has joined #openstack-ansible | 13:27 | |
| noonedeadpunk | actually there's oslo_messaging on queens, but the a concept seems to be changed with hybrid messaging. But I'm still not sure, that ?ssl=0 won't break ceilometer with rabbit notifications | 13:33 |
| *** rgogunskiy has joined #openstack-ansible | 13:35 | |
| *** shyamb has quit IRC | 13:35 | |
| *** kevintynfron has joined #openstack-ansible | 13:35 | |
| hwoarang | nicolasbock: in all honesty, i dont understand why the backport is needed to be fair. we do not install pip anymore so dropping the line is probably more appropriate than || true | 13:37 |
| hwoarang | but it's ok like this i suppose. it does not do any harm | 13:37 |
| nicolasbock | Maybe I am not understanding this correctly, but isn't the issue that if pip is not installed then the test will fail hwoarang ? | 13:41 |
| nicolasbock | And yes, there is get_pip but that's not until later in the script | 13:41 |
| hwoarang | probably not, because we use virtualenv, to set pip in virtualenv | 13:41 |
| nicolasbock | Let me double check this then | 13:42 |
| hwoarang | i dont see anything calling 'pip' before get_pip | 13:42 |
| hwoarang | but the line you modified does not install pip. it merely selects the python2 version. so if python-pip is not installed, this line is pointless | 13:43 |
| hwoarang | actually all it does it so not break the script if python-pip is not installed. it's almost equivalent to dropping the line entirely :) | 13:44 |
| *** chkumar|ruck is now known as chkumar|out | 13:44 | |
| nicolasbock | Yes that's true :) | 13:48 |
| nicolasbock | Also what if we have python3-pip installed but not the python2 package? | 13:48 |
| nicolasbock | Then this will fail as well and cause issues later on? | 13:48 |
| nicolasbock | I mean the alternative call will fail | 13:49 |
| nicolasbock | And it can't select python2-pip | 13:49 |
| nicolasbock | Which as I understand is an issue | 13:49 |
| nicolasbock | I don't know whether that question is purely academic though :) | 13:49 |
| *** goldenfri has joined #openstack-ansible | 13:56 | |
| hwoarang | yes the alternative could fail in that case too | 14:03 |
| hwoarang | bottom point is that we dont really care whether pip is installed on the host or not, so we should not try to configure it | 14:03 |
| odyssey4me | hwoarang: so should that be removed for both queens and pike? | 14:04 |
| odyssey4me | the alternative could be to move that line until after pip is installed | 14:05 |
| hwoarang | not sure. i would simply drop it because suse has a new enough virtualenv so we dont install pip on the host | 14:06 |
| hwoarang | but i fail to remember if that's true for 42.3 and if that applies to pike and/or queens | 14:06 |
| hwoarang | i need to investigate | 14:06 |
| *** kevintynfron has quit IRC | 14:22 | |
| odyssey4me | hwoarang: for queens and pike, we need to install pip on the host because we do pip installs everywhere... trying to backport all the changes to prevent that is just not practical or advisable | 14:28 |
| hwoarang | odyssey4me: ok i just don't see python-pip being installed in bootstrap-ansible.sh at all | 14:28 |
| hwoarang | so we are not using the distro package, so the 'alternatives' line has no effect | 14:29 |
| odyssey4me | hwoarang: yep, because in pike we used get-pip.py to install pip | 14:29 |
| odyssey4me | same in queens | 14:29 |
| hwoarang | "Skip host pip installs for ansible bootstrap" b95eafb0eeaaa8700b1293aec9a68b252602af00 | 14:29 |
| hwoarang | this removed python-pip from installed packages on the host. so the alternatives line is useless i think. but i could be missing something | 14:30 |
| hwoarang | anyway it does not matter much | 14:30 |
| odyssey4me | yep, true | 14:30 |
| *** skiedude has joined #openstack-ansible | 14:31 | |
| *** luksky has quit IRC | 14:31 | |
| *** goldenfri has quit IRC | 14:33 | |
| *** Bhujay has quit IRC | 14:34 | |
| openstackgerrit | Taseer Ahmed proposed openstack/openstack-ansible master: Integrate Blazar with OpenStack Ansible https://review.openstack.org/549956 | 14:35 |
| nicolasbock | If the host has python-pip already installed though hwoarang ? Not by OSA but by the user for some reason? | 14:41 |
| openstackgerrit | Taseer Ahmed proposed openstack/openstack-ansible master: Integrate Blazar with OpenStack Ansible https://review.openstack.org/549956 | 14:42 |
| hwoarang | nicolasbock: i am not sure what happens when the pip_install runs. i need to bring this memory back to my frontal lobe | 14:44 |
| nicolasbock | Good luck with that ;) | 14:44 |
| openstackgerrit | weezer su proposed openstack/openstack-ansible master: Add one test case to the TestMergeDictUnit for same key dict merge https://review.openstack.org/600160 | 14:47 |
| *** hamzy has quit IRC | 15:04 | |
| *** gyee has joined #openstack-ansible | 15:11 | |
| openstackgerrit | Merged openstack/openstack-ansible stable/ocata: Make sure the branch to track is in YAML https://review.openstack.org/599602 | 15:15 |
| *** rgogunskiy has quit IRC | 15:19 | |
| *** spartakos has joined #openstack-ansible | 15:24 | |
| *** fghaas has quit IRC | 15:24 | |
| jamesdenton | mnaser are you using haproxy for your prod cloud(s)? and if so, is it on the infras or a dedicated set of nodes? | 15:25 |
| *** hamzy has joined #openstack-ansible | 15:27 | |
| jrosser | who knows about config_template, and what to do when it blows up? | 15:33 |
| *** udesale has quit IRC | 15:35 | |
| *** tosky has quit IRC | 15:36 | |
| jrosser | oh wow, the config_template in the plugins repo and ansible-config_template arent the same :/ | 15:39 |
| odyssey4me | they are not, and I've been asking what the inhibitors are to moving to the independent repo for weeks now with no answer :/ | 15:40 |
| *** macza has joined #openstack-ansible | 15:42 | |
| *** pcaruana has quit IRC | 15:42 | |
| olivierb | hello everyone, this may be a very dummy question but I noticed a potentially missed reference to pike in https://github.com/openstack/openstack-ansible/blob/stable/queens/tests/roles/bootstrap-host/vars/redhat.yml#L16 (after seeing that the others which are left seem to be valid as they are used to test upgrade from pike to queens). Does this makes sense ? | 15:43 |
| olivierb | or won't it matter when AIO is deployed ? | 15:44 |
| jrosser | odyssey4me: those config_template differences cause actual brok for the ELK deploy, i'm test driving this https://review.openstack.org/#/c/597329/ | 15:46 |
| jrosser | i think cloudnull must be developing against config_template in the plugins repo, and the elk_metrics_6x embedded ansible wires up the independant repo version | 15:48 |
| *** luksky has joined #openstack-ansible | 15:55 | |
| *** Leo_m has joined #openstack-ansible | 15:58 | |
| *** radeks has quit IRC | 15:58 | |
| *** radeks has joined #openstack-ansible | 15:58 | |
| openstackgerrit | Merged openstack/openstack-ansible stable/pike: Make sure the branch to track is in YAML https://review.openstack.org/599599 | 16:00 |
| *** pcaruana has joined #openstack-ansible | 16:01 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ops master: MNAIO: Ensure that nested virt is enabled on host https://review.openstack.org/600470 | 16:06 |
| *** goldenfri has joined #openstack-ansible | 16:10 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ops master: MNAIO: Ensure that nested virt is enabled on host https://review.openstack.org/600470 | 16:13 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ops master: MNAIO: Ensure that nested virt is enabled on host https://review.openstack.org/600470 | 16:14 |
| *** luksky has quit IRC | 16:15 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ops master: MNAIO: Ensure that nested virt is enabled on host https://review.openstack.org/600470 | 16:16 |
| *** spatel has joined #openstack-ansible | 16:24 | |
| *** hamzaachi_ has quit IRC | 16:30 | |
| *** noonedeadpunk has quit IRC | 16:30 | |
| *** noonedeadpunk has joined #openstack-ansible | 16:30 | |
| *** aicilamd has quit IRC | 16:33 | |
| *** aicilamd has joined #openstack-ansible | 16:35 | |
| *** aicilamd has quit IRC | 16:35 | |
| odyssey4me | jrosser: I vaguely remember seeing a patch put in somewhere which changes 'kvm' to 'qemu' - and I think it might have related to bionic, but am not sure. Does this ring a bell? | 16:54 |
| openstackgerrit | Merged openstack/openstack-ansible-ops master: MNAIO: Ensure that virt-networks are properly setup https://review.openstack.org/600424 | 16:54 |
| jrosser | odyssey4me: theres nothing specific to bionic in that regard, but likwise i remember discussion possibly with logan- about that? | 16:55 |
| odyssey4me | well, after much spelunking over the last few days, I discovered that nested virt was not enabled on my bionic host | 16:56 |
| odyssey4me | I tried to track down why, because any new host I build has it enabled. | 16:56 |
| mnaser | jamesdenton: haproxy on infras | 16:56 |
| spotz | odyssey4me: weird | 16:56 |
| odyssey4me | Unfortunately I couldn't find the root cause. The qemu-system-x86 package puts a modprobe.d file into place to enable it, and has a postinst script to ensure it's loaded. | 16:57 |
| odyssey4me | Anyway, that's why I decided to push up https://review.openstack.org/600470 - just to make sure it doesn't hit anyone else, and to make sure I don't waste my time again. :p | 16:57 |
| odyssey4me | hwoarang evrardjp nicolasbock any news on fixing up the suse mirror? | 16:59 |
| *** olivierb has quit IRC | 16:59 | |
| *** shardy has quit IRC | 17:00 | |
| spotz | sounds like the postinst didn't kick or errored but my guess is you've used it often enough this was a one off? | 17:01 |
| jamesdenton | thanks mnaser. are you bonding NICs? single or dual? just curious | 17:02 |
| mnaser | jamesdenton: nope, no bonded nics, but we allow connecting directly to public network on vms so | 17:02 |
| odyssey4me | spotz: yeah, it may have errored - the script does '|| true', which means the error wouldn't have shown up... but that task will, so it provides a safety net in case | 17:03 |
| jamesdenton | are you running everything on one nic? or split control/data across two? | 17:03 |
| mnaser | jamesdenton: all on one but 40g nics have plenty | 17:04 |
| jamesdenton | ahh gotcha. yeah you've got room to grow :D | 17:04 |
| mnaser | yep plenty of space there :) | 17:04 |
| spotz | odyssey4me: Yeah which is a good thing:) | 17:05 |
| openstackgerrit | Jacob Wagner proposed openstack/openstack-ansible-ops master: Added playbooks to snapshot and revert snapshot MNAIO deployments https://review.openstack.org/600482 | 17:06 |
| *** gkadam__ has quit IRC | 17:09 | |
| *** spartakos has quit IRC | 17:11 | |
| *** ChanServ sets mode: -rf | 17:18 | |
| *** rickflare has joined #openstack-ansible | 17:18 | |
| *** MrWatson has joined #openstack-ansible | 17:18 | |
| *** DanyC has quit IRC | 17:22 | |
| *** pcaruana has quit IRC | 17:27 | |
| openstackgerrit | Dmitriy Rabotjagov (noonedeadpunk) proposed openstack/openstack-ansible-os_ceilometer master: Changed messaging_urls for notification agent https://review.openstack.org/600072 | 17:31 |
| mgariepy | anyone seen qemu process spinning 100% and allocating memory a few MB / sec ? | 17:35 |
| odyssey4me | I'm out for the night - cheers folks! | 17:35 |
| mgariepy | cya odyssey4me | 17:35 |
| *** spartakos has joined #openstack-ansible | 17:35 | |
| kaiokmo | have a good one odyssey4me | 17:39 |
| mgariepy | skiedude, did you progress on your setup ? | 17:40 |
| skiedude | yes! finally got all the playbooks to finish | 17:41 |
| skiedude | everything appears to be up | 17:41 |
| skiedude | THe semi unfortunate part now is trying to figure out if I can convert our existing project setup ansible scripts to work with this kind fo setup, or if I have to start from scratch | 17:41 |
| skiedude | Our current/old setup is everyting just running as services on the infra/compute hosts, now that its in containers, and behind haproxy, I need to figure out the best way to automate creations of affinity groups, projects, vms etc | 17:42 |
| mnaser | mgariepy: spinning a really big instance? | 17:44 |
| mnaser | i've see nthat if its a really big server | 17:45 |
| mgariepy | only 120g of ram | 17:48 |
| mgariepy | the first 70ish G fills fast. | 17:49 |
| mgariepy | but after that 1 core and slowly over a lot of minutes | 17:49 |
| *** olivierb has joined #openstack-ansible | 17:58 | |
| *** fghaas has joined #openstack-ansible | 17:58 | |
| *** electrofelix has quit IRC | 17:59 | |
| *** pcaruana has joined #openstack-ansible | 18:02 | |
| *** poopcat has joined #openstack-ansible | 18:02 | |
| mgariepy | mnaser, seems to be the kernel. | 18:03 |
| mgariepy | ish. | 18:04 |
| *** openstackgerrit has quit IRC | 18:05 | |
| *** olivierb has quit IRC | 18:06 | |
| *** hamzaachi has joined #openstack-ansible | 18:12 | |
| *** spartakos has quit IRC | 18:15 | |
| *** aicilamd has joined #openstack-ansible | 18:23 | |
| *** poopcat has quit IRC | 18:38 | |
| *** poopcat has joined #openstack-ansible | 18:52 | |
| *** flaviosr_ has joined #openstack-ansible | 18:59 | |
| hwoarang | odyssey4me: the mirror is fixed now | 19:00 |
| hwoarang | at least the infra i checked is synced again | 19:01 |
| *** flaviosr has quit IRC | 19:02 | |
| *** mbuil has quit IRC | 19:04 | |
| *** mbuil has joined #openstack-ansible | 19:07 | |
| *** nyloc has joined #openstack-ansible | 19:08 | |
| *** olivierb has joined #openstack-ansible | 19:09 | |
| *** fghaas has quit IRC | 19:11 | |
| *** spatel has quit IRC | 19:17 | |
| *** spatel has joined #openstack-ansible | 19:20 | |
| *** spartakos has joined #openstack-ansible | 19:30 | |
| *** openstackgerrit has joined #openstack-ansible | 19:31 | |
| openstackgerrit | Mohammed Naser proposed openstack/openstack-ansible-os_nova master: Revert region name option to os_region_name for Cinder https://review.openstack.org/596576 | 19:31 |
| *** hamzy has quit IRC | 19:34 | |
| *** pcaruana has quit IRC | 19:35 | |
| *** olivierb has quit IRC | 19:35 | |
| *** Emine has quit IRC | 19:37 | |
| *** Emine has joined #openstack-ansible | 19:38 | |
| *** goldenfri has quit IRC | 19:53 | |
| spatel | ostack-compute-40 : Sep 6 15:46:40 : neutron : user NOT in sudoers ; TTY=unknown ; PWD=/ ; USER=root ; COMMAND=/openstack/venvs/neutron-17.0.8/bin/neutron-rootwrap /etc/neutron/rootwrap.conf iptables-save -t raw | 19:55 |
| spatel | I do have /etc/sudoers.d/neutron_sudoers file | 19:56 |
| *** johanssone has joined #openstack-ansible | 19:59 | |
| mgariepy | latest hwe kernel (4.15.0-33-generic) do have some weirdness. | 20:03 |
| mgariepy | allocating memory do take quite a long time on this one. testing proposed one that seems to fix the issue so far. | 20:03 |
| *** goldenfri has joined #openstack-ansible | 20:05 | |
| *** hamzaachi has quit IRC | 20:11 | |
| *** hamzaachi has joined #openstack-ansible | 20:11 | |
| *** tosky has joined #openstack-ansible | 20:19 | |
| *** radeks has quit IRC | 20:31 | |
| *** nicolasbock has quit IRC | 20:34 | |
| *** goldenfri has quit IRC | 20:45 | |
| *** spatel has quit IRC | 20:46 | |
| *** ansmith has quit IRC | 20:47 | |
| *** savvas_ has joined #openstack-ansible | 20:48 | |
| *** savvas_ has quit IRC | 20:48 | |
| *** savvas has joined #openstack-ansible | 20:49 | |
| *** nicolasbock has joined #openstack-ansible | 20:54 | |
| openstackgerrit | Merged openstack/openstack-ansible-ops master: MNAIO: Ensure that nested virt is enabled on host https://review.openstack.org/600470 | 20:57 |
| *** DanyC has joined #openstack-ansible | 21:05 | |
| *** luksky has joined #openstack-ansible | 21:06 | |
| *** bgmccollum has joined #openstack-ansible | 21:16 | |
| *** bgmccollum has left #openstack-ansible | 21:17 | |
| *** bgmccollum has joined #openstack-ansible | 21:17 | |
| *** hamzaachi has quit IRC | 21:56 | |
| *** dave-mccowan has quit IRC | 22:00 | |
| *** luksky has quit IRC | 22:01 | |
| *** skiedude has quit IRC | 22:06 | |
| *** dave-mccowan has joined #openstack-ansible | 22:24 | |
| *** DanyC has quit IRC | 22:25 | |
| *** jamesdenton has quit IRC | 22:34 | |
| *** Leo_m has quit IRC | 22:40 | |
| *** tosky has quit IRC | 22:59 | |
| *** dave-mccowan has quit IRC | 23:01 | |
| *** macza has quit IRC | 23:08 | |
| *** spatel has joined #openstack-ansible | 23:10 | |
| *** spatel has quit IRC | 23:14 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_horizon master: Ensure dashboard plugins are found by py_pkgs https://review.openstack.org/600025 | 23:44 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!