| *** cjloader has joined #openstack-ansible | 00:02 | |
| xdfil | glance container isn't mounting NFS at start from fstab. But if I do "mount -a" it mounts fine. it seems like the container config is simply ignoring fstab on boot | 00:07 |
|---|---|---|
| *** lbragstad has joined #openstack-ansible | 00:14 | |
| *** cjloader has quit IRC | 00:40 | |
| *** cjloader has joined #openstack-ansible | 00:40 | |
| *** cjloader has quit IRC | 00:45 | |
| *** cjloader has joined #openstack-ansible | 00:56 | |
| *** cjloader has quit IRC | 01:01 | |
| *** rstarmer has quit IRC | 01:01 | |
| mnaser | hmm | 01:03 |
| mnaser | osa doesn't do any active/passive setup for rbd, does it? | 01:04 |
| mnaser | it seems to deploy 3 different and independent hosts which means that while the cluster is still up, if a host goes down, all volumes linked to it wont be able to perform their work | 01:05 |
| mnaser | vs active/passive where they all share the same host name and then when one goes down, any other one can go up and take over | 01:06 |
| logan- | mnaser: yeah you can do that | 01:57 |
| logan- | it is not in the docs afaik | 01:57 |
| logan- | you're talkign about setting the cinder-volume hostname right? | 01:57 |
| mnaser | logan-: what coordinates the active/passive stuff though? and yes | 01:57 |
| mnaser | cause i think if you run the same host across all 3, only 1 instance should run at time | 01:58 |
| logan- | ive always run them active/active with the same hostname | 01:58 |
| mnaser | i guess i'll have to do some reading to check (though your feedback helps) | 01:59 |
| mnaser | i know tripleo uses pacemaker to do active/passive | 01:59 |
| logan- | i could see needing to do that if you actually have cinder-volume in the data path for iscsi or something | 01:59 |
| logan- | but for ceph it is basically just a metadata service | 01:59 |
| mnaser | logan-: im thinking in some cases of potentially having races | 02:00 |
| mnaser | but yeah i see your logic | 02:00 |
| logan- | im curious to hear what you find out. i was a concerned about the MQ side when I saw that advised somewhere in the juno/kilo timeframe, but ive run with it for that long and have not seen any problems | 02:02 |
| mnaser | logan-: if anything this would help the scale side of things | 02:03 |
| mnaser | i'm seeing some really old discussions | 02:04 |
| xdfil | hi | 02:04 |
| mnaser | logan-: ill check with the cinder team and see what they have to say about it | 02:11 |
| logan- | cool | 02:11 |
| *** cjloader has joined #openstack-ansible | 02:15 | |
| *** cjloader has quit IRC | 02:40 | |
| *** hw_wutianwei_ has joined #openstack-ansible | 02:43 | |
| *** dave-mccowan has quit IRC | 03:20 | |
| *** gkadam_ has quit IRC | 03:58 | |
| *** pmannidi has quit IRC | 04:47 | |
| openstackgerrit | James E. Blair proposed openstack/openstack-ansible-os_horizon master: Zuul: Remove project name https://review.openstack.org/538749 | 05:07 |
| openstackgerrit | James E. Blair proposed openstack/openstack-ansible-os_horizon stable/newton: Zuul: Remove project name https://review.openstack.org/538754 | 05:27 |
| *** rstarmer has joined #openstack-ansible | 05:31 | |
| *** gkadam has joined #openstack-ansible | 05:47 | |
| openstackgerrit | James E. Blair proposed openstack/openstack-ansible-os_horizon stable/ocata: Zuul: Remove project name https://review.openstack.org/538757 | 05:47 |
| *** hybridpollo has quit IRC | 05:48 | |
| *** rstarmer has quit IRC | 05:50 | |
| *** ivveh has joined #openstack-ansible | 05:59 | |
| openstackgerrit | James E. Blair proposed openstack/openstack-ansible-os_horizon stable/pike: Zuul: Remove project name https://review.openstack.org/538763 | 06:07 |
| *** rstarmer has joined #openstack-ansible | 06:07 | |
| *** hybridpollo has joined #openstack-ansible | 06:08 | |
| *** ANKITA has joined #openstack-ansible | 06:14 | |
| *** hybridpollo has quit IRC | 06:27 | |
| openstackgerrit | Tuan Luong-Anh proposed openstack/ansible-hardening master: Zuul: Remove project name https://review.openstack.org/538775 | 06:28 |
| *** ANKITA has quit IRC | 06:35 | |
| *** yolanda_ has quit IRC | 06:44 | |
| evrardjp | hello everyone -- FYI I am freezing the a-r-r https://review.openstack.org/#/c/538680/ for master, to prepare for milestone 3 | 06:51 |
| evrardjp | xdfil: bugs can be reported on https://bugs.launchpad.net/openstack-ansible | 06:52 |
| ivveh | good morning | 06:52 |
| evrardjp | xdfil: for mounting on boot, I am not sure which systemd service take care of that. But if there is a bug, please report it :) | 06:53 |
| evrardjp | ivveh: good morning! | 06:53 |
| *** ivveh has quit IRC | 07:14 | |
| *** rstarmer has quit IRC | 07:14 | |
| *** armaan has joined #openstack-ansible | 07:26 | |
| *** pcaruana has joined #openstack-ansible | 07:44 | |
| *** mbuil has joined #openstack-ansible | 08:00 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible master: Update all SHAs for 17.0.0.0b3 https://review.openstack.org/538680 | 08:03 |
| evrardjp | jmccrory: when you're available, I've added you to the review of https://review.openstack.org/#/c/538315/ | 08:09 |
| *** hamza21 has joined #openstack-ansible | 08:09 | |
| evrardjp | this should unblock O to P upgrades if I am not mistaken. | 08:09 |
| * prometheanfire hates the integrated gate | 08:16 | |
| evrardjp | why? | 08:17 |
| evrardjp | they didn't do anything wrong :p | 08:17 |
| evrardjp | yes they are slow | 08:17 |
| evrardjp | but not to the point I'd hate them ! | 08:17 |
| prometheanfire | it keeps on randomly failing my patches | 08:18 |
| prometheanfire | (and I do mean random) | 08:18 |
| prometheanfire | I'm thinking an extra 5 min would 'fix' it | 08:18 |
| evrardjp | in which branch? | 08:18 |
| prometheanfire | wonder if it could be blamed on meltdown... | 08:18 |
| evrardjp | Pike? | 08:18 |
| prometheanfire | not-osa | 08:18 |
| prometheanfire | requirements | 08:18 |
| evrardjp | oh ok | 08:18 |
| evrardjp | you can change the timeout | 08:19 |
| prometheanfire | just me venting is all | 08:19 |
| evrardjp | how long is your job rn? | 08:19 |
| prometheanfire | back up to 15 hours :| | 08:19 |
| evrardjp | woa | 08:19 |
| evrardjp | oh it's not your job that last 15h | 08:19 |
| evrardjp | it's the integrated queue | 08:20 |
| evrardjp | of fun, you have post merge issues too. | 08:20 |
| prometheanfire | doesn't look like we define it in our .zuul.d | 08:20 |
| evrardjp | http://logs.openstack.org/c5/c5053646aa1bbbd0b2f2a5b269ddb42a9f29d49e/post/publish-openstack-python-branch-tarball/8205ccb/job-output.txt.gz | 08:21 |
| prometheanfire | that did get to github at least... | 08:22 |
| prometheanfire | https://github.com/openstack/requirements/commit/c5053646aa1bbbd0b2f2a5b269ddb42a9f29d49e | 08:22 |
| evrardjp | they all inherit from openstack-tox | 08:22 |
| evrardjp | but you shouldn't probably change it, your job timeouts are fine I think | 08:22 |
| evrardjp | well to what I see rn | 08:23 |
| evrardjp | it's just that the queue is big | 08:23 |
| evrardjp | post-merge | 08:23 |
| evrardjp | what I meant | 08:23 |
| evrardjp | it's on the tarballs I guess | 08:23 |
| evrardjp | that's not recheckable | 08:24 |
| evrardjp | maybe worth discussing it with infra | 08:24 |
| evrardjp | but yeah, if you think you have jobs timeouts that are too painful, just change the parent, create a new parent for tox-requirements, and change the timeout there | 08:25 |
| evrardjp | just need to add this parent into jobs.yaml if I am not mistaken, with just this override. | 08:25 |
| evrardjp | should be a 3 line change, and a sed. | 08:25 |
| evrardjp | well, more a sed and then a 3 line addition :) | 08:26 |
| evrardjp | 4 line* | 08:26 |
| prometheanfire | stuff like https://review.openstack.org/536793 | 08:26 |
| evrardjp | yu had many issues due to legacy-tempest-dsvm-neutron-full and legacy-requirements-integration-dsvm. on top of that, they take a lot of time. but yeah you're hitting more and more the 30+ minutes mark in your cross-jobs. | 08:28 |
| evrardjp | neutron and nova being the ones that take a little longer, why not only changing those 2? | 08:28 |
| evrardjp | let me check the stats quick | 08:29 |
| evrardjp | yeah cross-neutron-py35 seem to be a more regular offender, but it's not that bad | 08:31 |
| evrardjp | look at http://zuul.openstack.org/builds.html , pipeline gate (or check), job: cross-neutron-py35, project: openstack/requirements | 08:32 |
| evrardjp | it's not too bad :D | 08:32 |
| *** peri has joined #openstack-ansible | 08:36 | |
| *** sxc731 has joined #openstack-ansible | 08:41 | |
| *** masber has quit IRC | 08:41 | |
| *** hamza21 has quit IRC | 08:42 | |
| *** masber has joined #openstack-ansible | 08:42 | |
| *** hamza21 has joined #openstack-ansible | 08:42 | |
| *** cjloader has joined #openstack-ansible | 08:48 | |
| openstackgerrit | Taseer Ahmed proposed openstack/openstack-ansible-os_congress master: Introduce os_congress role in gerrit https://review.openstack.org/522491 | 08:51 |
| *** cjloader has quit IRC | 08:53 | |
| *** cxc99 has quit IRC | 08:57 | |
| *** cxc99 has joined #openstack-ansible | 09:00 | |
| *** jafeha has joined #openstack-ansible | 09:17 | |
| jafeha | good morning | 09:19 |
| prometheanfire | moin | 09:19 |
| hwoarang | good morning | 09:20 |
| evrardjp | good morning jafeha hwoarang | 09:28 |
| openstackgerrit | Manuel Buil proposed openstack/openstack-ansible-os_tacker master: Modify the etc directory in default/main https://review.openstack.org/538855 | 09:32 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-os_keystone master: Fix memcache caching w/ multiple cache servers https://review.openstack.org/533314 | 09:40 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-os_keystone master: [WIP] Add ansible modules idempotency testing https://review.openstack.org/538856 | 09:42 |
| *** shardy has joined #openstack-ansible | 09:43 | |
| *** vnogin has joined #openstack-ansible | 09:46 | |
| *** vnogin has quit IRC | 09:46 | |
| *** cjloader has joined #openstack-ansible | 09:48 | |
| *** cjloader has quit IRC | 09:52 | |
| odyssey4me | o/ | 10:02 |
| *** bauruine has quit IRC | 10:07 | |
| *** bauruine has joined #openstack-ansible | 10:08 | |
| evrardjp | hello odyssey4me | 10:13 |
| bndzor_ | hmm im not having the best performance with ceph and i cant figure out why.. is there anything i should tune ? | 10:28 |
| evrardjp | bndzor_: we are simply using ceph ansible, maybe there are setting in ceph ansible you can use | 10:32 |
| evrardjp | else I'd say, check your network stack and all. | 10:33 |
| *** threestrands_ has joined #openstack-ansible | 10:33 | |
| *** threestrands has quit IRC | 10:36 | |
| *** vnogin has joined #openstack-ansible | 10:37 | |
| *** jappleii__ has joined #openstack-ansible | 10:37 | |
| *** jappleii__ has quit IRC | 10:38 | |
| *** jappleii__ has joined #openstack-ansible | 10:39 | |
| *** Neptu_ has joined #openstack-ansible | 10:39 | |
| *** threestrands_ has quit IRC | 10:41 | |
| *** vnogin has quit IRC | 10:42 | |
| evrardjp | odyssey4me: hey, maybe rcarrillocruz got a present for you | 10:42 |
| *** rcarrillocruz has joined #openstack-ansible | 10:44 | |
| * rcarrillocruz waves | 10:44 | |
| evrardjp | hello rcarrillocruz ! | 10:45 |
| evrardjp | I just pinged odyssey4me , telling him you have presents | 10:46 |
| evrardjp | because he is a nodepool user he will like it :p | 10:46 |
| evrardjp | At least, I think. | 10:46 |
| evrardjp | oh whie you are here rcarrillocruz could you check this for me please? https://github.com/ansible/ansible/pull/35053 | 10:47 |
| evrardjp | between many pr, it just needs a shipit :) | 10:48 |
| *** cjloader has joined #openstack-ansible | 10:48 | |
| openstackgerrit | James E. Blair proposed openstack/openstack-ansible-galera_server master: Zuul: Remove project name https://review.openstack.org/538868 | 10:49 |
| odyssey4me | evrardjp yep - discussing it in #zuul | 10:50 |
| evrardjp | oh I see... too many channels, I am lost! | 10:51 |
| evrardjp | let me join #zuul | 10:51 |
| *** cjloader has quit IRC | 10:53 | |
| -openstackstatus- NOTICE: Zuul is currently under heavy load. Do not *recheck* or *approve* any changes. | 11:02 | |
| *** electrofelix has joined #openstack-ansible | 11:06 | |
| openstackgerrit | James E. Blair proposed openstack/openstack-ansible-galera_server stable/newton: Zuul: Remove project name https://review.openstack.org/538876 | 11:09 |
| *** vnogin has joined #openstack-ansible | 11:10 | |
| openstackgerrit | James E. Blair proposed openstack/openstack-ansible-galera_server stable/ocata: Zuul: Remove project name https://review.openstack.org/538883 | 11:29 |
| *** stuartgr has joined #openstack-ansible | 11:33 | |
| *** cjloader has joined #openstack-ansible | 11:48 | |
| openstackgerrit | James E. Blair proposed openstack/openstack-ansible-galera_server stable/pike: Zuul: Remove project name https://review.openstack.org/538889 | 11:49 |
| *** cjloader has quit IRC | 11:52 | |
| *** zenirc369 has joined #openstack-ansible | 11:58 | |
| zenirc369 | How to use grouped/combined register variable output values [values can be repeated/duplicate] of multiple hosts? | 11:59 |
| zenirc369 | https://stackoverflow.com/questions/48499649/how-to-use-grouped-combined-register-variable-output-values-values-can-be-repea | 11:59 |
| *** vnogin has quit IRC | 12:02 | |
| *** vnogin has joined #openstack-ansible | 12:03 | |
| evrardjp | zenirc369: combine filter combines dicts | 12:17 |
| evrardjp | zenirc369: + operator concats lists | 12:17 |
| evrardjp | math operators should do the trick too | 12:17 |
| evrardjp | math filters* | 12:17 |
| evrardjp | like | unique | 12:17 |
| evrardjp | of difference | 12:17 |
| evrardjp | I don't know what you are looking for but grouping seems that | 12:18 |
| evrardjp | also if you're expecting always the same answer etc. maybe run_once can help. | 12:18 |
| evrardjp | not sure what the issue is, so giving you a generic feedback zenirc369 | 12:19 |
| evrardjp | http://docs.ansible.com/ansible/latest/playbooks_filters.html#set-theory-filters for the set theory filters | 12:19 |
| zenirc369 | filtering is fine | 12:20 |
| zenirc369 | I have idea about filtering the list | 12:21 |
| zenirc369 | but at first I need all the cidr values to be grouped to form a list | 12:21 |
| *** sxc731 has quit IRC | 12:23 | |
| evrardjp | I am not sure what the problem is. | 12:24 |
| evrardjp | If your variable is a host var, it's meant for being scoped as a host var | 12:24 |
| evrardjp | if it's a group_var, then any host in the group will have the same content | 12:25 |
| evrardjp | (unless usual ansible overrides) | 12:25 |
| evrardjp | it all depends on your case | 12:25 |
| evrardjp | think about set fact, delegation, etc. | 12:25 |
| evrardjp | and think about your structure of data | 12:25 |
| evrardjp | (and where your overrides are) | 12:26 |
| evrardjp | but this isn't openstack-ansible specific, is it? | 12:26 |
| *** sxc731 has joined #openstack-ansible | 12:27 | |
| *** vnogin has quit IRC | 12:34 | |
| *** vnogin has joined #openstack-ansible | 12:41 | |
| *** vnogin has quit IRC | 12:42 | |
| zenirc369 | the environment I'm trying to implment is on openstack | 12:44 |
| *** cjloader has joined #openstack-ansible | 12:48 | |
| zenirc369 | evrardjp: thnks for your inputs :) | 12:49 |
| *** armaan_ has joined #openstack-ansible | 12:49 | |
| *** vnogin has joined #openstack-ansible | 12:51 | |
| *** armaan has quit IRC | 12:52 | |
| *** cjloader has quit IRC | 12:52 | |
| openstackgerrit | Markos Chandras (hwoarang) proposed openstack/openstack-ansible master: Vagrantfile: Use rsync to copy working directory to VM https://review.openstack.org/538903 | 12:56 |
| armaan_ | evrardjp: hello, Does this https://review.openstack.org/#/c/538407/ mean that i can start testing O->P upgrade? :) | 13:02 |
| evrardjp | armaan_: nope that was for P-Q role test upgrade | 13:02 |
| evrardjp | armaan_: https://review.openstack.org/#/c/538315/ is the one for O-P | 13:03 |
| evrardjp | I have yet to test it manually too though, so I'd wait. | 13:03 |
| evrardjp | It should work. | 13:03 |
| armaan_ | evrardjp: gotcha! thanks for letting me know :) | 13:03 |
| evrardjp | and I think i tested it last week, but it was long ago. | 13:03 |
| evrardjp | :D | 13:03 |
| armaan_ | ohh okk :) | 13:04 |
| evrardjp | armaan_: oh I tested it this morning to make sure. I got 91% api downtime for glance | 13:05 |
| evrardjp | during the process | 13:05 |
| armaan_ | eeehhh ... 91% | 13:05 |
| *** zenirc369 has quit IRC | 13:06 | |
| evrardjp | armaan_: http://paste.openstack.org/show/656884/ | 13:06 |
| evrardjp | but that's a major upgrade too. | 13:06 |
| evrardjp | so it succeeded | 13:06 |
| evrardjp | and that's api downtime, maybe not that important to you. | 13:06 |
| evrardjp | But I will have a look in the future to reduce that downtime value | 13:07 |
| evrardjp | anyway, with that patch in, it should be better. | 13:07 |
| armaan_ | evrardjp: we will inform our customers about the service window and after that i think we can live with any downtime :) | 13:07 |
| evrardjp | what a luxury :) | 13:08 |
| evrardjp | I think the most important/worrying part on your case is magnum... | 13:08 |
| armaan_ | evrardjp: hehe ... yes. | 13:08 |
| armaan_ | evrardjp: How can i remove the magnum service properly before upgrading? | 13:09 |
| armaan_ | evrardjp: Destroy containers, remove from inventory? | 13:09 |
| *** hamza21 has quit IRC | 13:15 | |
| *** vnogin has quit IRC | 13:16 | |
| evrardjp | yes | 13:19 |
| evrardjp | probably remove the env.d or conf.d overrides if any | 13:19 |
| odyssey4me | remove the endpoint too, I expect? | 13:21 |
| odyssey4me | endpoint(s) | 13:21 |
| armaan_ | evrardjp odyssey4me: Roger that. thanks! | 13:22 |
| *** woodard has quit IRC | 13:26 | |
| *** woodard has joined #openstack-ansible | 13:27 | |
| evrardjp | odyssey4me: armaan_ yeah probably worth cleaning up tenants, dbs, and stuff like that | 13:28 |
| evrardjp | cloudnull: are you there? | 13:28 |
| armaan_ | evrardjp: oh, you mean removing service user from service tenant etc | 13:29 |
| evrardjp | yes if need be. But that's maybe extra cleanup | 13:30 |
| evrardjp | as usual, take a backup first :p | 13:30 |
| -openstackstatus- NOTICE: Zuul is currently under heavy load. Do not *recheck* or *approve* any changes until we give the go ahead. | 13:30 | |
| *** ChanServ changes topic to "Zuul is currently under heavy load. Do not *recheck* or *approve* any changes until we give the go ahead." | 13:30 | |
| evrardjp | but... my patches! | 13:31 |
| evrardjp | :D | 13:31 |
| *** vnogin has joined #openstack-ansible | 13:34 | |
| mnaser | so, RBD specific, OSA doesn't have anything that does active/passive cinder-volume processes, multiple processes with the same "host" is not recommended by the cinder team | 13:46 |
| mnaser | i know tripleo uses pacemaker to do active/passive. it might be a bit overkill to run pacemaker just for this (or is it not?) | 13:47 |
| mnaser | logan-: ^ | 13:47 |
| odyssey4me | nope, we don't do anything like that | 13:47 |
| odyssey4me | each cinder-volume host is its own | 13:47 |
| mnaser | odyssey4me: the thing i have with that is that if a specific cinder-volume service goes down for some reason, technically the service is still operational but all volumes become un-managable | 13:47 |
| *** cjloader has joined #openstack-ansible | 13:48 | |
| odyssey4me | yes, that is true | 13:48 |
| odyssey4me | IIRC cinder-manage now has a command to migrate the management over to another cinder-volume instance | 13:48 |
| mnaser | so any maintenance on controllers can cause a control plane outage (unless you start messing around the db) | 13:48 |
| mnaser | oh TIL, let me read about that | 13:48 |
| *** vnogin has quit IRC | 13:49 | |
| mnaser | cinder-manage volume update_host --currenthost <current host> --newhost <new host> | 13:49 |
| mnaser | Updates the host name of all volumes currently associated with a specified host. | 13:49 |
| mnaser | indeed | 13:49 |
| *** vnogin has joined #openstack-ansible | 13:49 | |
| openstackgerrit | James E. Blair proposed openstack/openstack-ansible-repo_build master: Zuul: Remove project name https://review.openstack.org/538926 | 13:50 |
| odyssey4me | yep, so instead of adding more moving parts we prefer to think that if an admin can't get the service back up the admin can use that command instead | 13:50 |
| mnaser | odyssey4me: fair thinking | 13:50 |
| mnaser | that's a pretty clean mitigation, didnt know about that command | 13:51 |
| *** cjloader has quit IRC | 13:53 | |
| evrardjp | maybe that should be part of ops docs. | 13:57 |
| mgariepy | morning everyone | 14:00 |
| evrardjp | morning mgariepy | 14:00 |
| mgariepy | how are things going this morning ? | 14:01 |
| evrardjp | zuul is in a bad shape, see chan title | 14:02 |
| evrardjp | I hope my patch to freeze stuff for m3 will pass, I am already late | 14:03 |
| evrardjp | then unfreeze and here we go again! | 14:03 |
| *** RandomTech has joined #openstack-ansible | 14:03 | |
| openstackgerrit | James E. Blair proposed openstack/openstack-ansible-repo_build stable/newton: Zuul: Remove project name https://review.openstack.org/538928 | 14:10 |
| evrardjp | xdfil: I asked on openstack-charms, no frozen uid/gid for their ubuntu packages, so we can probably implement Red Hat/IBM's uids/gids. | 14:14 |
| *** dmsimard is now known as dmsimard|afk | 14:17 | |
| *** pabelanger has joined #openstack-ansible | 14:24 | |
| pabelanger | morning! | 14:24 |
| pabelanger | can you please top approving patches | 14:24 |
| *** cmart has joined #openstack-ansible | 14:24 | |
| pabelanger | specifically status:open+topic:zuulv3-projects | 14:24 |
| pabelanger | we are running into memory issues and likely need to schdule those reviews for approval | 14:25 |
| evrardjp | top approving or stop approving? | 14:26 |
| evrardjp | pabelanger: I haven't rechecked/+W patches since the message, not sure about others. | 14:27 |
| evrardjp | pabelanger: how can we help? | 14:27 |
| -openstackstatus- NOTICE: we've been able to restart zuul, and re-enqueue changes for gate. Please hold off on recheck or approves, we are still recovering. More info shortly. | 14:28 | |
| pabelanger | evrardjp: if you don't mind reaching out to team mates using the following topic: https://review.openstack.org/#/q/status:open+topic:zuulv3-projects | 14:29 |
| pabelanger | can see some openstack ansible patches approved | 14:29 |
| pabelanger | sadly, this isn't the best way to handle it, but until we clear patches already approved we run the risk of running out of memory again | 14:30 |
| openstackgerrit | James E. Blair proposed openstack/openstack-ansible-repo_build stable/ocata: Zuul: Remove project name https://review.openstack.org/538931 | 14:30 |
| odyssey4me | it might be by someone who's not yet looking at IRC - maybe just browsing on their phone | 14:30 |
| odyssey4me | it also seems like jeblair has new patch submissions being done at some sort of interval | 14:30 |
| mnaser | odyssey4me: every 20 minutes | 14:30 |
| pabelanger | yes | 14:31 |
| pabelanger | working on stopping that too | 14:31 |
| pabelanger | mnaser: if you also don't mind approving puppet changes, that would be helpful | 14:32 |
| evrardjp | not sure I understood, you mean they shouldn't go in, and we should -2 right now? | 14:32 |
| evrardjp | or it's the other way around? | 14:32 |
| pabelanger | no, things approved is fine for now. Already loaded into memory | 14:32 |
| evrardjp | because the patches approved on this topic were approved at 9:40 | 14:32 |
| evrardjp | ok | 14:32 |
| pabelanger | but +W is what we need to hold off on for the moment | 14:33 |
| evrardjp | well, we'll just say to stay calm until the storm is over. | 14:33 |
| pabelanger | yah | 14:33 |
| evrardjp | yeah ok understood. | 14:33 |
| pabelanger | thanks | 14:33 |
| *** gkadam has quit IRC | 14:33 | |
| evrardjp | hwoarang: mhayden mgariepy andymccr odyssey4me cloudnull see the message just here: do not recheck or +w please, until further infra notice! Thanks! | 14:34 |
| *** dave-mccowan has joined #openstack-ansible | 14:36 | |
| *** zenirc369 has joined #openstack-ansible | 14:39 | |
| openstackgerrit | Markos Chandras (hwoarang) proposed openstack/openstack-ansible-os_neutron master: tasks: Ensure Open vSwitch is started for all providers that need it https://review.openstack.org/538933 | 14:40 |
| *** santacloud has joined #openstack-ansible | 14:41 | |
| *** zenirc369 has quit IRC | 14:44 | |
| *** SerenaFeng has joined #openstack-ansible | 14:46 | |
| *** zenirc369 has joined #openstack-ansible | 14:47 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_horizon stable/ocata: Zuul: Remove project name https://review.openstack.org/538757 | 14:47 |
| dcdamien | https://review.openstack.org/#/c/538056/4 | 14:48 |
| dcdamien | mhayden: can you elaborate? | 14:48 |
| *** indistylo has joined #openstack-ansible | 14:48 | |
| openstackgerrit | James E. Blair proposed openstack/openstack-ansible-repo_build stable/pike: Zuul: Remove project name https://review.openstack.org/538935 | 14:50 |
| *** vnogin has quit IRC | 14:50 | |
| *** esberglu has joined #openstack-ansible | 14:51 | |
| *** vnogin has joined #openstack-ansible | 14:51 | |
| *** cjloader has joined #openstack-ansible | 14:53 | |
| RandomTech | Hey, can someone explain to me how external networking is determined in terms of the playbooks? | 14:53 |
| evrardjp | what do you mean? | 14:54 |
| RandomTech | one moment while i think of how to word this please, sorry | 14:56 |
| *** cjloader has quit IRC | 14:57 | |
| *** cjloader has joined #openstack-ansible | 14:57 | |
| RandomTech | in openstack_user_config, which network is normally used for the creation of floating ips. or how do i make an extra network for that purpose | 14:58 |
| *** hw_wutianwei has joined #openstack-ansible | 14:58 | |
| evrardjp | dcdamien: answered for him | 14:58 |
| evrardjp | RandomTech: understood the question | 14:59 |
| evrardjp | so | 14:59 |
| evrardjp | the provider_networks can arrive to your neutron agent containers. | 14:59 |
| evrardjp | RandomTech: for example: https://github.com/openstack/openstack-ansible/blob/master/etc/openstack_deploy/openstack_user_config.yml.aio#L93 | 15:00 |
| evrardjp | or https://github.com/openstack/openstack-ansible/blob/master/etc/openstack_deploy/openstack_user_config.yml.aio#L84 | 15:00 |
| evrardjp | in this case: https://github.com/openstack/openstack-ansible/blob/master/etc/openstack_deploy/openstack_user_config.yml.aio#L76-L84 | 15:00 |
| evrardjp | you are passing a host bind override (a host nic directly) named eth12, to the container that can then use that eth12 nic. | 15:01 |
| evrardjp | because the type is flat, we are generating all you need to use it as a flat network (which means no .1q tagging) | 15:01 |
| evrardjp | this: https://github.com/openstack/openstack-ansible/blob/master/etc/openstack_deploy/openstack_user_config.yml.aio#L85-L93 is different | 15:01 |
| openstackgerrit | Merged openstack/openstack-ansible-os_tacker stable/pike: Zuul: Remove project name https://review.openstack.org/538404 | 15:02 |
| openstackgerrit | Merged openstack/openstack-ansible-os_tacker master: Zuul: Remove project name https://review.openstack.org/538401 | 15:02 |
| evrardjp | it allows the container to have a nic that's allowing .1q traffic (i.e vlans) | 15:02 |
| evrardjp | whose VLAN ID is matching the "range" stanza. | 15:03 |
| *** sxc731 has quit IRC | 15:03 | |
| evrardjp | to be more specific, there is a nic that's passed to the container (eth11), and on that nic you can use .1q for neutron. | 15:03 |
| RandomTech | so groupbind: neutron_linuxbridge_agent allows for the container to attach to that network. host_bind_override passes a nic from the host directly to the container. net_type tells it if its flat or not? | 15:05 |
| *** SerenaFeng has quit IRC | 15:06 | |
| RandomTech | would i use ip_from_q if i was wanting to assign floating evrardjp ? | 15:07 |
| evrardjp | nope. | 15:09 |
| evrardjp | after the deploy of openstack, you should use the openstack cli to create your networks | 15:09 |
| RandomTech | okay. would you mind if i show you what i typed after i try configuring one? | 15:10 |
| RandomTech | im pretty new to openstack in general | 15:10 |
| jrosser | RandomTech: here is a little thing we made to turn a fresh openstack "useful" https://github.com/openstack/openstack-ansible-ops/blob/master/multi-node-aio/playbooks/openstack-service-setup.yml | 15:10 |
| evrardjp | I am not sure if I can help, but I am pretty sure someone might help you here. | 15:10 |
| cloudnull | mornings | 15:11 |
| evrardjp | cloudnull: good morning! | 15:11 |
| evrardjp | please don't do patches, recheck patches or approve patches until infra is feeling better :) | 15:11 |
| cloudnull | ha ok | 15:11 |
| evrardjp | cloudnull: do you mind, in the meantime, help me on a small challenge? | 15:11 |
| * cloudnull goes to recheck everything :) | 15:12 | |
| evrardjp | bypassing stupid pylint is the challenge | 15:12 |
| cloudnull | sure what's up ? | 15:12 |
| RandomTech | evrardjp: this is what i came up with: http://paste.openstack.org/show/657041/ | 15:12 |
| *** gus has quit IRC | 15:13 | |
| RandomTech | assuming i make a br-external of course | 15:13 |
| evrardjp | cloudnull: let's talk on pm, easier to not pollute this chan | 15:13 |
| *** gus has joined #openstack-ansible | 15:14 | |
| *** vnogin has quit IRC | 15:20 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_magnum stable/ocata: Zuul: Remove project name https://review.openstack.org/538634 | 15:23 |
| *** indistylo has quit IRC | 15:27 | |
| *** bauruine has quit IRC | 15:28 | |
| *** bauruine has joined #openstack-ansible | 15:34 | |
| *** sxc731 has joined #openstack-ansible | 15:44 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_sahara stable/pike: Zuul: Remove project name https://review.openstack.org/538386 | 15:45 |
| openstackgerrit | Merged openstack/openstack-ansible-os_sahara stable/ocata: Zuul: Remove project name https://review.openstack.org/538385 | 15:45 |
| openstackgerrit | Merged openstack/openstack-ansible-os_sahara stable/newton: Zuul: Remove project name https://review.openstack.org/538384 | 15:45 |
| openstackgerrit | Merged openstack/openstack-ansible-os_horizon stable/newton: Zuul: Remove project name https://review.openstack.org/538754 | 15:45 |
| openstackgerrit | Merged openstack/openstack-ansible-os_horizon master: Zuul: Remove project name https://review.openstack.org/538749 | 15:45 |
| evrardjp | it's raining merges! | 15:45 |
| *** esberglu has quit IRC | 15:48 | |
| *** esberglu has joined #openstack-ansible | 15:49 | |
| *** phalmos has joined #openstack-ansible | 15:57 | |
| RandomTech | evrardjp: did you get a chance to look at the paste i did? sorry i had to try to fix my boss's mac | 15:59 |
| *** vnogin has joined #openstack-ansible | 16:00 | |
| *** hw_wutianwei has quit IRC | 16:01 | |
| *** vnogin has quit IRC | 16:04 | |
| *** vnogin has joined #openstack-ansible | 16:04 | |
| *** santacloud has quit IRC | 16:05 | |
| *** jappleii__ has quit IRC | 16:05 | |
| RandomTech | or could you take a look cloudnull? | 16:05 |
| *** phalmos has quit IRC | 16:11 | |
| *** woodard has quit IRC | 16:18 | |
| *** ChanServ changes topic to "Launchpad: https://launchpad.net/openstack-ansible || Weekly Meetings: https://wiki.openstack.org/wiki/Meetings/openstack-ansible || Review Dashboard: https://goo.gl/tTmdgs" | 16:19 | |
| -openstackstatus- NOTICE: zuul.o.o is back online, feel free to recheck / approve patches. | 16:19 | |
| cloudnull | RandomTech: I'd set the config to something like so: http://paste.openstack.org/show/657143 | 16:19 |
| cloudnull | RandomTech: and have the interface file to be something like http://paste.openstack.org/show/657141 | 16:19 |
| *** vnogin has quit IRC | 16:19 | |
| cloudnull | that would create a "br-external" bridge and hang an eth5 interface off of the bridge allowing neutron to use a flat network | 16:21 |
| jrosser | cloudnull: does that need an IP on the bridge? | 16:21 |
| cloudnull | it does not need an IP. | 16:21 |
| jrosser | ok, what i really meant was there was one in the paste and i was puzzled | 16:21 |
| cloudnull | the IP would only be useful if you wanted to have access to the instances from the physical host. | 16:21 |
| jrosser | right, i see | 16:22 |
| cloudnull | if you have multiple subnets that you were going to use for flat networks then it wouldn't be all that usefule . | 16:22 |
| cloudnull | **useful | 16:22 |
| *** chyka has joined #openstack-ansible | 16:22 | |
| *** bauruine has quit IRC | 16:22 | |
| *** chyka has quit IRC | 16:22 | |
| *** bauruine has joined #openstack-ansible | 16:23 | |
| *** chyka has joined #openstack-ansible | 16:23 | |
| RandomTech | cloudnull: why net_name: flat? and were just wanting this as a network to assign floating ips on | 16:24 |
| *** pcaruana has quit IRC | 16:25 | |
| RandomTech | and what interface file are you talking about? the hosts? | 16:26 |
| cloudnull | the netname is used within neutron | 16:27 |
| RandomTech | so that would allow me to see it in the interface when creating networks? | 16:28 |
| *** phalmos has joined #openstack-ansible | 16:30 | |
| RandomTech | cloudnull: why is it bridging to br-vlan-veth? | 16:31 |
| cloudnull | opps line 17 was a typo | 16:35 |
| cloudnull | that should have been br-external-veth | 16:35 |
| *** armaan has joined #openstack-ansible | 16:36 | |
| *** sxc731 has quit IRC | 16:38 | |
| *** armaan_ has quit IRC | 16:39 | |
| RandomTech | this seems alot more complicated than the other bridges | 16:39 |
| cloudnull | it is a bit more complex but the setup has been working for us for quite a while now. | 16:40 |
| cloudnull | it's effectively just create an available interface so neutron can bind ot | 16:41 |
| cloudnull | **tp | 16:41 |
| cloudnull | hahahah. | 16:41 |
| cloudnull | I can't type | 16:41 |
| cloudnull | mind sharing the br-external interface config you have now ? | 16:41 |
| RandomTech | i was just setting this up one moment while i type something up | 16:42 |
| cloudnull | ok. | 16:42 |
| *** phalmos has quit IRC | 16:42 | |
| *** sxc731 has joined #openstack-ansible | 16:47 | |
| RandomTech | this is my template file for the interface files cloudnull: http://paste.openstack.org/show/657192/ | 16:48 |
| RandomTech | so far at least | 16:49 |
| cloudnull | I'd add "offload-sg off" to all of the bridges | 16:50 |
| cloudnull | and change the br-external to look like http://paste.openstack.org/show/657197/ | 16:50 |
| cloudnull | otherwise the rest looks good | 16:51 |
| *** Leo_m has joined #openstack-ansible | 16:51 | |
| RandomTech | why line 5-7? | 16:51 |
| cloudnull | that's what creates the "eth5" interface | 16:52 |
| cloudnull | which neutron will bind to for the flat network type | 16:52 |
| cloudnull | technically that would only need to be done on compute nodes | 16:52 |
| RandomTech | could i just have it bind to the existing public interface? | 16:52 |
| cloudnull | ** compute and network nodes | 16:53 |
| RandomTech | the enp7s0f1 | 16:53 |
| cloudnull | yes you could have it just bind to that, if you set the hostbind override to that address and that address is the same across the fleet. | 16:53 |
| cloudnull | s/address/interface/ | 16:54 |
| RandomTech | it should be | 16:54 |
| *** ndusek has joined #openstack-ansible | 16:55 | |
| ndusek | hey all | 16:55 |
| cloudnull | when using a flat network, neutron will attempt to create bridges on a given interface, so if nothing else is bridging that interface it'll work great. | 16:56 |
| cloudnull | hanging the veth off a br-external interface is just a way to ensure an open interface is available but there are a lot of ways to get that done. | 16:56 |
| RandomTech | so like this? http://paste.openstack.org/show/657205/ | 16:56 |
| cloudnull | o/ ndusek | 16:56 |
| ndusek | I just deployed openstack ansible all-in-one last week. Had to do a reboot today and none of my containers came back and I can't restart them | 16:57 |
| cloudnull | RandomTech: yes | 16:57 |
| RandomTech | the only thing bridging off of that interface is the br-external i just created in the template | 16:57 |
| ndusek | any clues as to what might be going on here? | 16:57 |
| ndusek | I can provide lxc log info | 16:57 |
| cloudnull | on the compute nodes, the flat network type will run on "enp7s0f1" and within the containers neutron will run on "eth5" which will all be part of the same l2. | 16:58 |
| cloudnull | RandomTech: will that template be run on compute nodes too? if so then you might need to make the br-external interface conditional. | 16:59 |
| RandomTech | the template is run on all of the host | 16:59 |
| RandomTech | currently anyway | 16:59 |
| RandomTech | i used it to help do preliminary setup before running openstack playbooks | 17:00 |
| *** kstev has joined #openstack-ansible | 17:01 | |
| cloudnull | RandomTech: http://paste.openstack.org/show/657210 | 17:01 |
| cloudnull | so if the node is not in the compute group it'll create the bridge | 17:02 |
| cloudnull | otherwise it'll just rely on the interface which is defined earlier in your template. | 17:02 |
| RandomTech | that works with the template command in ansible? | 17:03 |
| *** rstarmer has joined #openstack-ansible | 17:03 | |
| cloudnull | yes | 17:03 |
| cloudnull | the template module is effectively the copy module with jinja2 templating | 17:04 |
| RandomTech | im not sure how to integrate this with the openstack-ansible inventory honestly | 17:04 |
| RandomTech | we were just running a ansible script sepearately on the bare metal | 17:06 |
| cloudnull | i'd probably continue running that with an external inventory. | 17:08 |
| cloudnull | you could however add the needed options to the openstack_user_config.yml | 17:08 |
| RandomTech | the external inventory doesnt have a group['nova_compute'] | 17:09 |
| cloudnull | looks like you added a hash for "openstack_storage_ip" | 17:09 |
| RandomTech | heres an example of one of the computes /etc/network/interfaces file after we ran it (pre adding external stuff): http://paste.openstack.org/show/657221/ | 17:12 |
| *** akasurde has joined #openstack-ansible | 17:14 | |
| cloudnull | RandomTech: http://paste.openstack.org/show/657228/ | 17:15 |
| cloudnull | if you wanted to extend the osa inventory with the options needed for your template ^ | 17:15 |
| *** phalmos has joined #openstack-ansible | 17:15 | |
| cloudnull | RandomTech: to make that work in your template all you'd need to change is "{{ openstack_storage_ip[bare_metal[inventory_hostname]] }}" to "{{ openstack_storage_ip['bare_metal'] }}" | 17:17 |
| *** woodard has joined #openstack-ansible | 17:17 | |
| *** akasurde has quit IRC | 17:17 | |
| RandomTech | would is cause a problem if the computes have a br-external? | 17:19 |
| RandomTech | sorry for all the questions, my boss basically told me set up openstack with no prior knowledge of it | 17:25 |
| RandomTech | all i knew was that it was cluster computing | 17:25 |
| *** guhcampos has joined #openstack-ansible | 17:25 | |
| *** guhcampos_ has joined #openstack-ansible | 17:25 | |
| *** phalmos has quit IRC | 17:26 | |
| *** armaan_ has joined #openstack-ansible | 17:29 | |
| xdfil | evrardjp: cool, i've been thinking about the optimal way to handle it. currently if you set a uid value OSA will try to change it but if the existing uid has any running processes it will fail and leave you with a small mess to clean up | 17:29 |
| xdfil | well it could be a big mess if it is in production | 17:30 |
| *** sxc731 has quit IRC | 17:30 | |
| *** armaan_ has quit IRC | 17:31 | |
| *** mbuil has quit IRC | 17:31 | |
| *** armaan_ has joined #openstack-ansible | 17:32 | |
| *** armaan has quit IRC | 17:32 | |
| *** cjloader has quit IRC | 17:33 | |
| *** cjloader has joined #openstack-ansible | 17:33 | |
| *** cjloader has quit IRC | 17:39 | |
| *** cjloader has joined #openstack-ansible | 17:40 | |
| cloudnull | RandomTech: if the compute nodes have "br-external" then you'll need to hang the veth off that bridge | 17:40 |
| *** cjloader has quit IRC | 17:41 | |
| *** cjloader has joined #openstack-ansible | 17:42 | |
| *** cjloader has quit IRC | 17:45 | |
| *** cjloader has joined #openstack-ansible | 17:46 | |
| RandomTech | do the infrastucture or storage nodes need the bridge cloudnull ? | 17:46 |
| cloudnull | yes. | 17:48 |
| *** cjloader has quit IRC | 17:48 | |
| *** cjloader has joined #openstack-ansible | 17:48 | |
| cloudnull | the exception would be network and compute | 17:48 |
| RandomTech | how will the routing work if we hange the veth? | 17:48 |
| cloudnull | everything will pass through the bridege | 17:48 |
| RandomTech | it looks like well have to do that as our storage and compute are combined | 17:49 |
| cloudnull | so eth5 > br-external > physical | 17:49 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-os_glance master: Set glance images dir owners each time https://review.openstack.org/533189 | 17:49 |
| RandomTech | so just use this for the bridge in my template? : http://paste.openstack.org/show/657197/ | 17:49 |
| *** peri has quit IRC | 18:02 | |
| *** phalmos has joined #openstack-ansible | 18:03 | |
| *** RandomTech has quit IRC | 18:05 | |
| *** phalmos has quit IRC | 18:09 | |
| *** stuartgr has quit IRC | 18:29 | |
| xdfil | where is OSA saving iptables stuff at? | 18:40 |
| xdfil | there are firewall rules, but I don't see where they are getting loaded from | 18:46 |
| *** kstev has quit IRC | 18:46 | |
| cloudnull | xdfil: if it's an openstack firewall rule it'd come from nova / neutron | 18:56 |
| cloudnull | we don't do much in osa for firewall rules, outside of the aio | 18:56 |
| *** shardy has quit IRC | 18:56 | |
| *** ivve has quit IRC | 18:57 | |
| xdfil | thanks | 19:00 |
| *** poopcat has joined #openstack-ansible | 19:02 | |
| *** ndusek has left #openstack-ansible | 19:04 | |
| openstackgerrit | Merged openstack/openstack-ansible master: Update all SHAs for 17.0.0.0b3 https://review.openstack.org/538680 | 19:05 |
| *** poopcat1 has joined #openstack-ansible | 19:06 | |
| *** poopcat has quit IRC | 19:08 | |
| *** cjloader has quit IRC | 19:08 | |
| *** cjloader has joined #openstack-ansible | 19:09 | |
| *** ivve has joined #openstack-ansible | 19:11 | |
| *** phalmos has joined #openstack-ansible | 19:12 | |
| *** hamza21 has joined #openstack-ansible | 19:14 | |
| *** mgariepy has quit IRC | 19:15 | |
| *** rstarmer has quit IRC | 19:17 | |
| *** rstarmer has joined #openstack-ansible | 19:19 | |
| *** mgariepy has joined #openstack-ansible | 19:24 | |
| *** john51 has quit IRC | 19:30 | |
| *** john51 has joined #openstack-ansible | 19:31 | |
| *** phalmos has quit IRC | 19:37 | |
| *** rstarmer has quit IRC | 19:39 | |
| *** phalmos has joined #openstack-ansible | 19:39 | |
| *** rstarmer has joined #openstack-ansible | 19:39 | |
| *** hamza21 has quit IRC | 19:47 | |
| *** phalmos has quit IRC | 19:53 | |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-os_nova master: Fix SELinux file contexts for nova's ssh keys https://review.openstack.org/534891 | 19:55 |
| *** phalmos has joined #openstack-ansible | 19:57 | |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-os_trove master: Updated from global requirements https://review.openstack.org/535074 | 19:59 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-os_nova master: Adjust deprecated nova console variables https://review.openstack.org/532836 | 20:01 |
| *** ivve has quit IRC | 20:03 | |
| *** pramodrj07 has joined #openstack-ansible | 20:04 | |
| *** hybridpollo has joined #openstack-ansible | 20:05 | |
| *** aludwar has joined #openstack-ansible | 20:05 | |
| *** aludwar has quit IRC | 20:06 | |
| *** aludwar has joined #openstack-ansible | 20:07 | |
| *** aludwar has quit IRC | 20:11 | |
| *** hybridpollo has quit IRC | 20:11 | |
| *** hybridpollo has joined #openstack-ansible | 20:12 | |
| *** ivve has joined #openstack-ansible | 20:15 | |
| *** rstarmer has quit IRC | 20:17 | |
| mgariepy | any cores to push this one over ? https://review.openstack.org/#/c/538324/ | 20:17 |
| openstackgerrit | Merged openstack/openstack-ansible-rsyslog_server master: Zuul: Remove project name https://review.openstack.org/538620 | 20:31 |
| *** kstev has joined #openstack-ansible | 20:54 | |
| *** pbandark has joined #openstack-ansible | 21:05 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_glance master: Set glance images dir owners each time https://review.openstack.org/533189 | 21:15 |
| cloudnull | mgariepy: done. | 21:17 |
| mgariepy | thanks cloudnull | 21:18 |
| cloudnull | anytime ! | 21:19 |
| *** kstev1 has joined #openstack-ansible | 21:27 | |
| openstackgerrit | Marc Gariépy (mgariepy) proposed openstack/openstack-ansible-lxc_hosts master: Change default for sshd to only all keybased auth https://review.openstack.org/539020 | 21:27 |
| mgariepy | mhayden, cloudnull, ^^ | 21:27 |
| *** kstev has quit IRC | 21:28 | |
| cloudnull | i'm good with that | 21:29 |
| mgariepy | on this cya tomorrow. | 21:30 |
| cloudnull | take care mgariepy | 21:30 |
| *** poopcat1 is now known as poopcat | 21:37 | |
| mhayden | mgariepy: ol' evrardjp makes a valid point on a reno | 21:41 |
| mhayden | even though people shouldn't be using passwords by default, it's worth mentioning | 21:42 |
| *** woodard has quit IRC | 21:43 | |
| idlemind | running stable/pike on a freshly installed centos 7 host and i failed on lxc_hosts get init.scope pids.max value. the log says "file not found: /sys/fs/cgroup/pids/init.scope/pids.max" | 21:56 |
| idlemind | i verified it doesn't exist, any thoughts on how i should correct it or is it a false positive? | 21:56 |
| cloudnull | idlemind: looks like we use a block / rescue there | 22:01 |
| idlemind | ^^ ahh fix is in upstream openstack-ansible-lxc_hosts (master) but not in their stable/pike ... seems older versions of systemd don't support init.scope so the patch to fix it basically skips the task that failed. i'll ignore for now and watch for the patch to move into stable/pike | 22:01 |
| idlemind | https://github.com/openstack/openstack-ansible-lxc_hosts/commit/165bd142b1d40467284f6b13457888b599b2bebd | 22:01 |
| cloudnull | idlemind: https://github.com/openstack/openstack-ansible-lxc_hosts/blob/stable/pike/tasks/lxc_post_install.yml#L74-L117 | 22:02 |
| cloudnull | it should have been caught by the rescue and printed out the debug info | 22:02 |
| cloudnull | however there's a really poor user experience there | 22:03 |
| cloudnull | it'll show that the task failed | 22:03 |
| cloudnull | however the playbook succeeds | 22:03 |
| cloudnull | we recently added https://review.openstack.org/#/c/530959/ | 22:03 |
| cloudnull | which will notify the user if the playbook failed by looking at the return code | 22:04 |
| cloudnull | because of this issue. | 22:04 |
| cloudnull | the UX however should imporve in ansible 2.5, but we're not there yet. | 22:04 |
| idlemind | cloudnull yup i see the additional debug in green below the failed as well good to see additional pretty print information as well | 22:04 |
| idlemind | cloudnull also seems the "fix" is upstream in the sub repo so it just has to get cherry picked into stable/pike for centos 7 | 22:05 |
| cloudnull | ** This is all assuming the task failed, was caught by the rescue and everything else succeeded. | 22:05 |
| idlemind | or the right sha adjusted in openstack-ansible for osa-lxc_hosts however it is linked | 22:05 |
| *** threestrands has joined #openstack-ansible | 22:05 | |
| *** threestrands has quit IRC | 22:05 | |
| *** threestrands has joined #openstack-ansible | 22:05 | |
| cloudnull | lets go see maybe, maybe it's in the branch but not in a stable tag yet | 22:05 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-lxc_hosts stable/pike: tasks: lxc_post_install: Handle missing init.scope scope https://review.openstack.org/539026 | 22:07 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible stable/pike: Add a return code check to the osa wrapper https://review.openstack.org/539027 | 22:07 |
| cloudnull | yup, both needed to be cherry-picked | 22:07 |
| idlemind | speedy speedy | 22:07 |
| *** threestrands_ has joined #openstack-ansible | 22:07 | |
| cloudnull | so you can cherry-pick those into your checkout for now | 22:08 |
| cloudnull | and as soon as they're merged they'll be included in the next stable tag | 22:08 |
| *** threestrands has quit IRC | 22:08 | |
| *** threestrands_ has quit IRC | 22:08 | |
| cloudnull | if you have time to review either of those pr's it'd be greatly appreciated. | 22:08 |
| *** threestrands has joined #openstack-ansible | 22:08 | |
| *** threestrands has quit IRC | 22:08 | |
| *** threestrands has joined #openstack-ansible | 22:08 | |
| idlemind | cloudnull any tips on how i'd do that? just cherry-pick them on my side and post feedback? | 22:10 |
| cloudnull | cd to the dir w/ the role "/etc/ansible/roles/lxc_hosts" and run `git fetch https://git.openstack.org/openstack/openstack-ansible-lxc_hosts refs/changes/26/539026/1 && git checkout FETCH_HEAD` | 22:10 |
| idlemind | sweet thx | 22:11 |
| cloudnull | in gerrit in the top right corner there's the "download" button | 22:11 |
| cloudnull | you can see a couple options for cherry-pick | 22:11 |
| idlemind | running another playbook atm so when that finishes i'll grab the change and re-run to see if it works | 22:11 |
| cloudnull | checkout, etc | 22:11 |
| cloudnull | cool! | 22:11 |
| idlemind | nice didn't know the download thing had the cheater commands | 22:12 |
| *** rstarmer has joined #openstack-ansible | 22:16 | |
| *** threestrands_ has joined #openstack-ansible | 22:21 | |
| *** threestrands has quit IRC | 22:23 | |
| *** ndusek has joined #openstack-ansible | 22:26 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_cloudkitty master: Zuul: Remove project name https://review.openstack.org/538610 | 22:27 |
| *** woodard has joined #openstack-ansible | 22:29 | |
| *** dave-mccowan has quit IRC | 22:31 | |
| *** Leo_m_ has joined #openstack-ansible | 22:32 | |
| *** Leo_m has quit IRC | 22:35 | |
| *** jappleii__ has joined #openstack-ansible | 22:35 | |
| *** jappleii__ has quit IRC | 22:36 | |
| *** jappleii__ has joined #openstack-ansible | 22:37 | |
| *** Leo_m_ has quit IRC | 22:37 | |
| *** threestrands_ has quit IRC | 22:37 | |
| cloudnull | idlemind: yup. | 22:40 |
| * cloudnull loves me some cheater commands | 22:40 | |
| cloudnull | I'm headed off , take care all | 22:41 |
| idlemind | nn | 22:41 |
| openstackgerrit | Merged openstack/openstack-ansible-os_neutron master: Zuul: Remove project name https://review.openstack.org/538524 | 22:41 |
| *** rstarmer has quit IRC | 22:47 | |
| *** ndusek has quit IRC | 22:49 | |
| *** phalmos_ has joined #openstack-ansible | 22:56 | |
| *** kstev has joined #openstack-ansible | 22:57 | |
| *** kstev1 has quit IRC | 22:57 | |
| *** rstarmer has joined #openstack-ansible | 22:58 | |
| *** phalmos has quit IRC | 22:59 | |
| openstackgerrit | Merged openstack/openstack-ansible-plugins master: Zuul: Remove project name https://review.openstack.org/538613 | 23:01 |
| *** pmannidi has joined #openstack-ansible | 23:01 | |
| *** kstev has quit IRC | 23:08 | |
| *** rstarmer has quit IRC | 23:26 | |
| *** zenirc369 has quit IRC | 23:29 | |
| *** zenirc369 has joined #openstack-ansible | 23:33 | |
| *** cjloader has quit IRC | 23:38 | |
| *** pbandark has quit IRC | 23:42 | |
| *** logan- has quit IRC | 23:45 | |
| *** errr has quit IRC | 23:45 | |
| *** Oku_OS-away has quit IRC | 23:45 | |
| *** xar- has quit IRC | 23:45 | |
| *** jillr has quit IRC | 23:45 | |
| *** logan- has joined #openstack-ansible | 23:51 | |
| *** errr has joined #openstack-ansible | 23:51 | |
| *** Oku_OS-away has joined #openstack-ansible | 23:51 | |
| *** xar- has joined #openstack-ansible | 23:51 | |
| *** jillr has joined #openstack-ansible | 23:51 | |
| openstackgerrit | Merged openstack/openstack-ansible stable/pike: Remove default_bind_mount_logs from config changes. https://review.openstack.org/538324 | 23:52 |
| *** dave-mccowan has joined #openstack-ansible | 23:53 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!