| *** DanyC has quit IRC | 00:00 | |
| *** markvoelker has joined #openstack-ansible | 00:05 | |
| *** markvoelker has quit IRC | 00:10 | |
| *** ram5391 has joined #openstack-ansible | 00:16 | |
| ram5391 | odyssey4me are you still among us? | 00:16 |
|---|---|---|
| ram5391 | has anyone had issues with a deployment failing at nova services? Specifically on the add service users task? I'm getting a message: the output has been hidden due to the fact that 'no_log: true' was specified for this result | 00:29 |
| ram5391 | fresh deployment of stable/rocky | 00:29 |
| *** DanyC has joined #openstack-ansible | 00:56 | |
| *** DanyC has quit IRC | 01:00 | |
| *** markvoelker has joined #openstack-ansible | 01:05 | |
| *** markvoelker has quit IRC | 01:09 | |
| *** Emine has quit IRC | 01:22 | |
| *** Emine has joined #openstack-ansible | 01:23 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Add kolide cluster tests https://review.openstack.org/611472 | 01:31 |
| openstackgerrit | Merged openstack/openstack-ansible-ops master: Add osquery testing https://review.openstack.org/611226 | 01:32 |
| cloudnull | ram5391 not that ive seen recently | 01:33 |
| cloudnull | is that using the stable/rocky branch ? | 01:33 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_horizon master: remove basic distro packages from the repo server https://review.openstack.org/611475 | 01:50 |
| *** DanyC has joined #openstack-ansible | 01:57 | |
| *** DanyC has quit IRC | 02:01 | |
| *** Emine has quit IRC | 02:23 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Add kolide cluster tests https://review.openstack.org/611472 | 02:24 |
| ram5391 | cloudnull yes indeed. Just pulled a hour or two ago. | 02:28 |
| ram5391 | running another deployment now, will post logs if it fails again | 02:30 |
| openstackgerrit | Antony Messerli proposed openstack/openstack-ansible stable/pike: Stop MariaDB nodes to reset quorum during upgrade https://review.openstack.org/611438 | 02:37 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Add kolide cluster tests https://review.openstack.org/611472 | 02:40 |
| ram5391 | cloudnull the nova thing solved itself, but now I'm getting a weird ioctl error with neutron: http://paste.openstack.org/show/732392/ Is that familiar? | 02:52 |
| *** dave-mccowan has quit IRC | 02:56 | |
| *** DanyC has joined #openstack-ansible | 02:58 | |
| *** DanyC has quit IRC | 03:03 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Add kolide cluster tests https://review.openstack.org/611472 | 03:05 |
| ram5391 | I double-checked my config, it _seems_ okay but due to the vague error, I'm not sure even what to look for here | 03:06 |
| cloudnull | ram5391 nope. however i'd suspect something is up in the openstack_user_config | 03:06 |
| cloudnull | under the "provide_networks" section | 03:06 |
| cloudnull | could be a simple indentation error | 03:07 |
| ram5391 | found an IRC log with someone that had the "net_name" key as netname, but that wasn't the case. This is mine: http://paste.openstack.org/show/732394/ | 03:08 |
| ram5391 | I did have some crazy indentations, so I'll try with some of the whitespace cleaned up | 03:08 |
| cloudnull | looks good compared to my config https://gist.github.com/cloudnull/eab22ba500843343470a7fd1e3f55d1c#file-openstack_user_config-yml-L20-L74 | 03:12 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Add kolide cluster tests https://review.openstack.org/611472 | 03:12 |
| ram5391 | cloudnull they look so similar ;) | 03:15 |
| cloudnull | that's the only thing I cna think of that would generate a none type error like that | 03:16 |
| ram5391 | still no dice, is it maybe not able to reach something? | 03:20 |
| ram5391 | my interfaces file on the headnode is: http://paste.openstack.org/show/732395/ | 03:21 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Add kolide cluster tests https://review.openstack.org/611472 | 03:26 |
| ram5391 | cloudnull yep, still nothing after a complete hosts->infra->openstack. Any pointers on where to start looking? | 03:49 |
| *** DanyC has joined #openstack-ansible | 04:00 | |
| *** DanyC has quit IRC | 04:05 | |
| *** ram5391 has quit IRC | 04:44 | |
| *** udesale has joined #openstack-ansible | 04:45 | |
| *** DanyC has joined #openstack-ansible | 04:53 | |
| *** jonher has quit IRC | 04:57 | |
| *** jonher has joined #openstack-ansible | 04:57 | |
| *** DanyC has quit IRC | 04:58 | |
| *** chhagarw has joined #openstack-ansible | 05:00 | |
| *** lbragstad_503 has quit IRC | 05:27 | |
| *** lbragstad_503 has joined #openstack-ansible | 05:27 | |
| *** DanyC has joined #openstack-ansible | 05:54 | |
| *** DanyC has quit IRC | 05:58 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/ansible-role-python_venv_build master: Only install the appropriate distro packages https://review.openstack.org/611508 | 05:59 |
| *** cshen has joined #openstack-ansible | 06:14 | |
| *** Emine has joined #openstack-ansible | 06:37 | |
| *** Emine has quit IRC | 06:42 | |
| *** rodolof has joined #openstack-ansible | 06:44 | |
| *** fghaas has joined #openstack-ansible | 06:48 | |
| *** persia has quit IRC | 07:19 | |
| *** persia has joined #openstack-ansible | 07:21 | |
| *** jcosmao has quit IRC | 07:22 | |
| *** shardy has joined #openstack-ansible | 07:23 | |
| *** tosky has joined #openstack-ansible | 07:31 | |
| *** openstackgerrit has quit IRC | 07:35 | |
| *** openstackgerrit has joined #openstack-ansible | 07:40 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Fix rgw playbook pre_tasks https://review.openstack.org/611304 | 07:40 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Restore bionic/ceph voting jobs https://review.openstack.org/609965 | 07:40 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Refactor ceph server playbooks https://review.openstack.org/611146 | 07:40 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Ensure ceph server apt pinning is well defined https://review.openstack.org/611521 | 07:40 |
| openstackgerrit | Markos Chandras (hwoarang) proposed openstack/ansible-hardening master: Revert "file_perms: Skip rpm verification step for Leap 15" https://review.openstack.org/611524 | 07:51 |
| *** chhagarw has quit IRC | 07:53 | |
| *** thuydang has joined #openstack-ansible | 07:57 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_horizon stable/rocky: Remove the static when use include_tasks feature https://review.openstack.org/609574 | 08:09 |
| *** spatel has joined #openstack-ansible | 08:12 | |
| *** spatel has quit IRC | 08:17 | |
| *** shardy has quit IRC | 08:18 | |
| *** rodolof has quit IRC | 08:18 | |
| *** olivierb has joined #openstack-ansible | 08:20 | |
| *** shardy has joined #openstack-ansible | 08:26 | |
| *** faizy98 has joined #openstack-ansible | 08:26 | |
| *** DanyC has joined #openstack-ansible | 08:30 | |
| *** faizy_ has quit IRC | 08:30 | |
| *** hamzaachi has joined #openstack-ansible | 09:09 | |
| *** gkadam has joined #openstack-ansible | 09:21 | |
| *** gkadam has quit IRC | 09:21 | |
| *** fghaas has quit IRC | 09:39 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_swift master: releasenotes: oslo-messaging-separate-backends add project name https://review.openstack.org/610614 | 09:46 |
| openstackgerrit | Merged openstack/openstack-ansible-ops master: Add vars to configure prometheus module https://review.openstack.org/611394 | 09:47 |
| openstackgerrit | Merged openstack/openstack-ansible-ops master: Add ability to deploy Octavia load balancer service https://review.openstack.org/600155 | 09:47 |
| hwoarang | odyssey4me: re https://review.openstack.org/#/c/610902/ rechecks wont help. something regressed on stable/rocky and i can't reproduce locally | 09:48 |
| openstackgerrit | Merged openstack/openstack-ansible-os_magnum master: releasenotes: oslo-messaging-separate-backends add project name https://review.openstack.org/610612 | 09:48 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Upgrade ceph to mimic release https://review.openstack.org/610165 | 09:48 |
| openstackgerrit | Merged openstack/openstack-ansible-os_octavia master: releasenotes: oslo-messaging-separate-backends add project name https://review.openstack.org/610615 | 09:50 |
| *** priteau has joined #openstack-ansible | 09:52 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_keystone master: releasenotes: oslo-messaging-separate-backends add project name https://review.openstack.org/610613 | 09:53 |
| *** jbadiapa has quit IRC | 09:58 | |
| *** jonher_ has joined #openstack-ansible | 10:03 | |
| *** jonher has quit IRC | 10:06 | |
| *** jonher_ is now known as jonher | 10:06 | |
| jrosser | hwoarang: nice to see the opensuse ceph job pass there | 10:10 |
| hwoarang | :) | 10:12 |
| jrosser | we should make that vote asap to keep it passing | 10:12 |
| odyssey4me | hwoarang: https://review.openstack.org/610363 then? | 10:13 |
| hwoarang | odyssey4me: what? | 10:14 |
| hwoarang | you said that Xenial jobs are replaced but i can still see them | 10:14 |
| hwoarang | so wanted to check if everything is right like jrosser said | 10:14 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/rocky: Use the compute kit + horizon for all distros https://review.openstack.org/610363 | 10:15 |
| odyssey4me | sorry, missed that in the commit msg - that was from the master job changes | 10:16 |
| jrosser | it's important because the xenial/rocky situation is confusing, partiuclary wrt UCA | 10:16 |
| odyssey4me | distro xenial is removed, but xenial source remains for rocky | 10:17 |
| hwoarang | odyssey4me: what mirror issues wrt opensuse btw? | 10:18 |
| hwoarang | all of them are passing now? | 10:19 |
| odyssey4me | hwoarang: this was submitted some time ago when there were mirror issues persistently for more than a week - I still have https://review.openstack.org/#/c/609353/5 up for master to restore them, but that's not been passing | 10:19 |
| hwoarang | perhaps I can recheck | 10:20 |
| hwoarang | sorry was away last week and i am not aware of mirror issues right now | 10:20 |
| odyssey4me | I don't know if there's some way that we can get the opensuse mirrors tested in a 'proposed' repo before promoting them because opensuse is failing every week due to package changes. | 10:20 |
| odyssey4me | Another way to resolve this may be to switch opensuse testing to external CI, if perhaps XCI can do external CI? | 10:21 |
| hwoarang | all of the failures are on tempest | 10:21 |
| hwoarang | can't see packaging issues there. It's normally the hardware that's broken in the mirrors not the packages themselves :) | 10:22 |
| odyssey4me | I know that https://review.openstack.org/611508 may fix some things - but that failed the integrated build, so I think we may have to resolve some of those first. | 10:22 |
| * hwoarang is trying to catch up | 10:24 | |
| odyssey4me | :/ no useful logs for that glance error | 10:24 |
| odyssey4me | I'm working on an octavia issue - then I'll switch back to that one and do a build to see what's going on. | 10:25 |
| odyssey4me | Could do with another review for https://review.openstack.org/607522 and https://goo.gl/k9cCfF | 10:26 |
| noonedeadpunk | Folks, I know, that my deployment differs from the guide, but update of the https://github.com/openstack/openstack-ansible/blob/stable/rocky/playbooks/common-tasks/dynamic-address-fact.yml in rocky seems not to work for my deployment... | 10:35 |
| noonedeadpunk | The point is, that in Q it was possible to configure tunnel without br-vlan bridge, but to set it to the regular interface. And here there is no ability to set non_metal_query, if neutron is placed on metal. | 10:36 |
| noonedeadpunk | Probably it's worth to offer the deployer ability to provide his own version of query? | 10:38 |
| noonedeadpunk | or I'm wrong, as I've missed "|| {{ non_metal_query }}" part of metal_query, sorry. | 10:41 |
| openstackgerrit | Merged openstack/openstack-ansible-os_neutron stable/rocky: zuul: Switch Xenial jobs to Bionic https://review.openstack.org/606389 | 10:46 |
| *** hamzaachi has quit IRC | 10:53 | |
| *** chhagarw has joined #openstack-ansible | 10:57 | |
| nicolasbock | hwoarang (IRC): The mirror issues were resolved as far as I know. It was due to an out of sync OpenStack mirror with mismatching package lists. And since https://review.openstack.org/#/c/609395/ is merged the sync is more frequent now, hopefully reducing this kind of issue. | 11:03 |
| *** udesale has quit IRC | 11:04 | |
| odyssey4me | hope so - I'd like to see success for that for a few weeks though | 11:05 |
| odyssey4me | thanks for that nicolasbock | 11:05 |
| nicolasbock | No problem odyssey4me | 11:06 |
| nicolasbock | But the packaging issue that hwoarang (IRC) was mentioning is pretty mysterious | 11:06 |
| *** jbadiapa has joined #openstack-ansible | 11:14 | |
| *** hamzaachi has joined #openstack-ansible | 11:16 | |
| hwoarang | the one in stable/rocky? yeah ... | 11:16 |
| hwoarang | running AIO locally know to see if I can reproduce it | 11:16 |
| nicolasbock | Thanks hwoarang (IRC) | 11:25 |
| openstackgerrit | Merged openstack/openstack-ansible-os_keystone master: Only do sshd enablement task on the first play host https://review.openstack.org/607522 | 11:26 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_octavia master: Ensure that the flavor/network facts are available to all targets https://review.openstack.org/611389 | 11:31 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_keystone stable/rocky: Only do sshd enablement task on the first play host https://review.openstack.org/611575 | 11:31 |
| openstackgerrit | Merged openstack/openstack-ansible-os_trove stable/rocky: use include_tasks instead of include https://review.openstack.org/608107 | 11:31 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_octavia master: Ensure that the flavor/network facts are available to all targets https://review.openstack.org/611389 | 11:35 |
| *** electrofelix has joined #openstack-ansible | 11:38 | |
| *** strobelight has quit IRC | 11:39 | |
| *** chhagarw has quit IRC | 11:39 | |
| noonedeadpunk | I've located problem with dynamic-address-fact.yml, and it has roots in my interface name - ib1.87d2. And json_query('ansible_ib1.87d2') returns error, as interpret '.' as a key in dict... | 11:42 |
| noonedeadpunk | So "{{ hostvars[inventory_hostname]['ansible_ib1.87d2']['ipv4']['address'] }}" returns IP address while "{{ hostvars[inventory_hostname] | json_query('ansible_ib1.87d2.ipv4.address') }}" returns error | 11:46 |
| openstackgerrit | Merged openstack/openstack-ansible stable/rocky: fix typo in releasenotes https://review.openstack.org/610519 | 11:48 |
| odyssey4me | nicolasbock: mind if I tweak up https://review.openstack.org/#/c/611393/1 with some more fixes I found? | 11:53 |
| *** hamzaachi has quit IRC | 11:54 | |
| openstackgerrit | Merged openstack/openstack-ansible master: fix typo in releasenotes https://review.openstack.org/610505 | 11:54 |
| nicolasbock | Not at all odyssey4me | 11:55 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_octavia master: Fix updating amphora images https://review.openstack.org/611393 | 11:58 |
| jrosser | odyssey4me: this looks wrong http://logs.openstack.org/04/611304/2/check/openstack-ansible-deploy-aio_metal-ubuntu-bionic/3e2f407/job-output.txt.gz#_2018-10-18_08_03_50_805909 | 11:58 |
| odyssey4me | nicolasbock: I'd appreciate it if you could test that - unfortunately my env has a broken glance, so the image parts aren't working. :/ | 11:59 |
| nicolasbock | Yes, I'll give it a spin | 11:59 |
| odyssey4me | jrosser: oh dear, I don't know if infra has more mirrors of the ceph repo | 11:59 |
| jrosser | odyssey4me: it looks like there there http://mirror.iad.rax.openstack.org/ | 12:00 |
| odyssey4me | perhaps it's better to use a proxy instead | 12:00 |
| *** faizy98 has quit IRC | 12:00 | |
| *** faizy98 has joined #openstack-ansible | 12:01 | |
| jrosser | is that data that is pushed into the test instance with the path set like that? | 12:01 |
| jrosser | by the look of it we don't consume that var | 12:04 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ops master: MNAIO: Remove legacy os-infra_hosts group https://review.openstack.org/611582 | 12:06 |
| odyssey4me | jrosser: ah, then we should probably remove that var setting to avoid the confusion | 12:07 |
| odyssey4me | If no-one else is using it, then we should also ditch the mirror. | 12:07 |
| jrosser | it'd be useful if it wasnt hardwired to hammer release | 12:07 |
| openstackgerrit | Dmitriy Rabotjagov (noonedeadpunk) proposed openstack/openstack-ansible master: Fixed ability of usage interfaces/bridges with dots (aliases) https://review.openstack.org/611585 | 12:09 |
| *** hamzaachi has joined #openstack-ansible | 12:09 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ops master: MNAIO: Increase default infra_vm_server_ram to 16384 https://review.openstack.org/611586 | 12:13 |
| odyssey4me | jrosser: yeah, perhaps more useful if there was a proxy pointing at the ceph repositories so that they could be used for multiple releases... I think that mirror is one of the originals, before we had proxies | 12:13 |
| *** spatel has joined #openstack-ansible | 12:14 | |
| *** ansmith has joined #openstack-ansible | 12:15 | |
| *** thuydang has quit IRC | 12:18 | |
| *** thuydang has joined #openstack-ansible | 12:19 | |
| *** spatel has quit IRC | 12:19 | |
| *** strattao has joined #openstack-ansible | 12:20 | |
| jrosser | hwoarang: the opensuse ceph job is tricking us http://logs.openstack.org/65/610165/3/check/openstack-ansible-deploy-distro_ceph-opensuse-423/dbd8c73/job-output.txt.gz#_2018-10-18_10_02_33_715911 | 12:22 |
| openstackgerrit | Taseer Ahmed proposed openstack/openstack-ansible master: Integrate Blazar with OpenStack Ansible https://review.openstack.org/549956 | 12:33 |
| odyssey4me | jrosser hwoarang mind giving this backport a review? https://review.openstack.org/611575 | 12:34 |
| *** fghaas has joined #openstack-ansible | 12:34 | |
| *** gkadam has joined #openstack-ansible | 12:34 | |
| jrosser | odyssey4me: its a very crafty construct used there - i had to stare hard at the master version of that earlier! | 12:35 |
| odyssey4me | jrosser: yeah, it's another way of doing in-line serial tasks - rather than using the 'serial' argument in a play | 12:36 |
| odyssey4me | useful for some things | 12:37 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ops master: MNAIO: Remove glance_default_store override https://review.openstack.org/611590 | 12:46 |
| *** kaiokmo has joined #openstack-ansible | 12:48 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ops master: MNAIO: Remove lxc_container_backing_store override https://review.openstack.org/611591 | 12:49 |
| openstackgerrit | Nicolas Bock proposed openstack/openstack-ansible master: Include script and line numbers in checks https://review.openstack.org/611592 | 12:49 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ops master: MNAIO: Move the ELK-related comment inside the conditional https://review.openstack.org/611593 | 12:51 |
| openstackgerrit | Dmitriy Rabotjagov (noonedeadpunk) proposed openstack/openstack-ansible master: Fixed ability of usage interfaces/bridges with dots (aliases) https://review.openstack.org/611585 | 12:57 |
| *** strobelight has joined #openstack-ansible | 12:57 | |
| hwoarang | jrosser: odyssey4me i am in a meeting will reply back soon | 12:58 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_octavia master: Fix updating amphora images https://review.openstack.org/611393 | 13:06 |
| hwoarang | jrosser: you are right sorry :( are you going to submit a patch or should I? | 13:08 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_octavia stable/rocky: Only run OpenStack tasks once https://review.openstack.org/608757 | 13:08 |
| mgariepy | hwoarang, not sure what is going on for this but I do have a really hard time with suse and my nova patch: http://logs.openstack.org/89/605789/6/gate/openstack-ansible-functional-distro_install-opensuse-423/3ee4157/logs/openstack/openstack1/neutron/neutron-dhcp-agent.log.txt.gz | 13:16 |
| hwoarang | hmm spells like apparmor | 13:18 |
| hwoarang | *smells | 13:18 |
| hwoarang | what patch is this? | 13:18 |
| mgariepy | https://review.openstack.org/#/c/605789/ | 13:19 |
| mgariepy | why is the check passes and not the gate ? | 13:19 |
| hwoarang | that's good question | 13:21 |
| hwoarang | the logs suggest that apparmor manipulation does not happen at all | 13:21 |
| hwoarang | makes no sense | 13:21 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ops master: MNAIO: Make cinder/compute/swift hosts conditional https://review.openstack.org/611603 | 13:23 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ops master: MNAIO: Make cinder/compute/swift hosts conditional https://review.openstack.org/611603 | 13:25 |
| hwoarang | mgariepy: i think there is a race condition there one moment | 13:28 |
| hwoarang | will look into it in ~30min i am still in a meeting | 13:30 |
| mgariepy | thanks a lot hwoarang | 13:30 |
| *** lbragstad_503 is now known as lbragstad | 13:32 | |
| *** thuydang has quit IRC | 13:35 | |
| jrosser | odyssey4me: you mentioned problems with glance earlier - was it like this? http://logs.openstack.org/65/610165/3/check/openstack-ansible-deploy-aio_lxc-ubuntu-bionic/ad6ba11/job-output.txt.gz#_2018-10-18_12_02_04_667327 | 13:37 |
| openstackgerrit | Merged openstack/openstack-ansible master: Fix rgw playbook pre_tasks https://review.openstack.org/611304 | 13:38 |
| odyssey4me | jrosser: no, it was my own config issue in an MNAIO - hence https://review.openstack.org/#/c/611590/1 | 13:38 |
| jrosser | hrrm, ok, i'll recheck that | 13:38 |
| odyssey4me | was also getting fails because of https://review.openstack.org/#/c/611586/1 - I must really spend a little time working on getting the uwsgi cheaper algo going for autoscaling workers | 13:38 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Add missing ceph scenario to opensuse distro ceph job https://review.openstack.org/611610 | 13:42 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible stable/rocky: Fix rgw playbook pre_tasks https://review.openstack.org/611611 | 13:43 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ops master: MNAIO: Make cinder/compute/swift hosts conditional https://review.openstack.org/611603 | 13:44 |
| openstackgerrit | Merged openstack/openstack-ansible-os_nova stable/rocky: Correct ansible conditional jinja2 warning https://review.openstack.org/609701 | 13:46 |
| *** jcosmao has joined #openstack-ansible | 13:47 | |
| spotz | Any reason barbican can't be installed in an AIO? | 13:48 |
| odyssey4me | spotz: some works was done for master & rocky to make that work - not sure about queens/pike | 13:50 |
| openstackgerrit | Markos Chandras (hwoarang) proposed openstack/openstack-ansible-os_neutron master: tasks: neutron_install: Fix apparmor delegation to physical hosts. https://review.openstack.org/611613 | 13:50 |
| spotz | odyssey4me: I'll let jungleboyj know, he's trying an install now but might be queens | 13:51 |
| odyssey4me | spotz: https://review.openstack.org/#/q/I10a07d5e4b2b3fbf122a79f15e27a9ca34799221 | 13:51 |
| odyssey4me | thta doesn't mean it won't work for queens, but there's no AIO test for it | 13:51 |
| odyssey4me | the fixes that required seem to have been only related to rocky/master anyway and only related to AIO testing. I think we have some folks working on barbican for queens and things seem to be going right, mostly | 13:52 |
| openstackgerrit | Merged openstack/openstack-ansible-os_horizon stable/rocky: Add ability for HTTP access to horizon https://review.openstack.org/611322 | 13:58 |
| *** gkadam has quit IRC | 13:59 | |
| *** gkadam has joined #openstack-ansible | 13:59 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_keystone stable/rocky: Only do sshd enablement task on the first play host https://review.openstack.org/611575 | 14:01 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Add kolide cluster tests https://review.openstack.org/611472 | 14:06 |
| odyssey4me | spotz: did you want https://review.openstack.org/611322 any further back than rocky? | 14:11 |
| spotz | odyssey4me: I've got my hack on Queens, but I think it's more of how far back do we think people might need it? | 14:12 |
| odyssey4me | perhaps queens would be good too, let me port it back | 14:12 |
| spotz | ok | 14:13 |
| *** hamzy has quit IRC | 14:15 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_horizon stable/queens: Add ability for HTTP access to horizon https://review.openstack.org/611621 | 14:15 |
| *** cshen has quit IRC | 14:19 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_swift stable/rocky: releasenotes: oslo-messaging-separate-backends add project name https://review.openstack.org/611623 | 14:25 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_magnum stable/rocky: releasenotes: oslo-messaging-separate-backends add project name https://review.openstack.org/611624 | 14:26 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_octavia stable/rocky: releasenotes: oslo-messaging-separate-backends add project name https://review.openstack.org/611625 | 14:26 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_keystone stable/rocky: releasenotes: oslo-messaging-separate-backends add project name https://review.openstack.org/611626 | 14:26 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_heat stable/rocky: releasenotes: oslo-messaging-separate-backends add project name https://review.openstack.org/611628 | 14:28 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_heat stable/rocky: Remove python-keystoneclient package install https://review.openstack.org/603378 | 14:29 |
| *** thuydang has joined #openstack-ansible | 14:31 | |
| *** gkadam_ has joined #openstack-ansible | 14:33 | |
| *** gkadam has quit IRC | 14:35 | |
| *** hamzy has joined #openstack-ansible | 14:42 | |
| *** tobias-urdin has joined #openstack-ansible | 14:42 | |
| *** Taseer has joined #openstack-ansible | 14:43 | |
| *** weezS has joined #openstack-ansible | 14:50 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Cleanup the osquery role https://review.openstack.org/611641 | 14:54 |
| *** thuydang has quit IRC | 15:02 | |
| *** strattao has quit IRC | 15:07 | |
| spotz | FInishing up my course but ping if needed for reviews | 15:09 |
| *** strattao has joined #openstack-ansible | 15:12 | |
| *** Taseer has quit IRC | 15:25 | |
| noonedeadpunk | folks, it seems that I did some wrong cherry-pick (or missed one) for ceilometer role. And 2 lines differs between master and rocky - https://github.com/openstack/openstack-ansible-os_ceilometer/blob/master/templates/ceilometer.conf.j2#L39 Is it okey, or I should fix it somehow? | 15:28 |
| *** gkadam_ has quit IRC | 15:38 | |
| *** suggestable has joined #openstack-ansible | 15:39 | |
| suggestable | Hey guys! | 15:39 |
| spotz | noonedeadpunk: if it's wrong you may be able to patch it, or abandon and fix | 15:40 |
| spotz | hey suggestable | 15:40 |
| suggestable | Deploying stable/rocky from distro source on Bionic. Setup-infrastructure.yml completes successfully, but healthcheck-infrastructure.yml fails during the line 71-91 repo check. | 15:41 |
| suggestable | Is this a bug? | 15:41 |
| spotz | suggestable: I'm not sure we're bionic for rocky | 15:41 |
| suggestable | spotz: documentation suggests OSA supports Bionic for Rocky (but not Queens or earlier). | 15:42 |
| spotz | suggestable: Then it could be a bug:) | 15:42 |
| *** hamzy has quit IRC | 15:42 | |
| *** hamzy has joined #openstack-ansible | 15:43 | |
| noonedeadpunk | spotz: it's not wrong, but absent in rocky:) | 15:43 |
| suggestable | The earlier check (lines 22-36) has a "when: install_method == 'source'" line, which the later repo check misses. | 15:43 |
| *** gyee has joined #openstack-ansible | 15:43 | |
| suggestable | Is the health check error safe for me to ignore during my deployment run, or is it indicative of a problem? I'd expect the health check to skip checking the repo when deploying from distro packages. | 15:48 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Additional playbook cleanup and use stable release https://review.openstack.org/611661 | 15:51 |
| spotz | noonedeadpunk: I meant the patch:) | 15:52 |
| bgmccollum | because it's barbican't? | 15:52 |
| spotz | bgmccollum: Be nice! | 15:55 |
| bgmccollum | summoning the dad joke spirit of major | 15:55 |
| spotz | heheh | 15:55 |
| *** macza has joined #openstack-ansible | 15:56 | |
| openstackgerrit | Markos Chandras (hwoarang) proposed openstack/openstack-ansible-os_neutron master: tasks: neutron_install: Fix group for neutron hosts https://review.openstack.org/611613 | 15:59 |
| hwoarang | mgariepy: wanna make your nova patch depend on https://review.openstack.org/611613 so we can test it? | 15:59 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: clean up readme https://review.openstack.org/611668 | 16:01 |
| *** nsmeds__ has joined #openstack-ansible | 16:01 | |
| openstackgerrit | Marc Gariépy (mgariepy) proposed openstack/openstack-ansible-os_nova master: Update the pci config for nova. https://review.openstack.org/605789 | 16:06 |
| mgariepy | done hwoarang | 16:07 |
| *** hamzaachi has quit IRC | 16:08 | |
| *** olivierb has quit IRC | 16:08 | |
| *** jungleboyj has joined #openstack-ansible | 16:11 | |
| jungleboyj | Howdy all. | 16:11 |
| jungleboyj | spotz: I have tried to run the setup-hosts and it keeps failing: Temporary failure resolving 'us.archive.ubuntu.com' | 16:12 |
| jungleboyj | I am able to ping the host from the command line though. | 16:12 |
| jungleboyj | DNS seems to be working. | 16:12 |
| spotz | jungleboyj: weird | 16:13 |
| spotz | what happens if you do a apt update on the host? | 16:13 |
| *** suggestable has quit IRC | 16:15 | |
| jungleboyj | spotz: Let me see. | 16:15 |
| jungleboyj | All good. | 16:16 |
| *** cshen has joined #openstack-ansible | 16:16 | |
| spotz | and if you re-run the role now? | 16:16 |
| jungleboyj | Trying. | 16:17 |
| jungleboyj | You are going to wish you didn't recommend this. I break everything. | 16:17 |
| spotz | And I break cinder all the tme:) | 16:17 |
| jungleboyj | spotz: Fair enough and we love you for it. | 16:22 |
| jungleboyj | Failed a third time. | 16:23 |
| jungleboyj | It is in [openstack_hosts : Remove the blacklisted packages] | 16:23 |
| spotz | jungleboyj: booting up a VM | 16:24 |
| jungleboyj | Like the container has a network issue? | 16:24 |
| spotz | jungleboyj: Nono gonna install queens | 16:26 |
| jungleboyj | Ok. | 16:26 |
| odyssey4me | jungleboyj: how did you do the config prep - is this an AIO and did you use bootstrap-aio, or did you prep config and host config by hand? | 16:26 |
| jungleboyj | odyssey4me: This is AIO and I used bootstrap-aio | 16:27 |
| *** vnogin has joined #openstack-ansible | 16:27 | |
| *** cshen has quit IRC | 16:28 | |
| odyssey4me | ok, bootstrap-aio does some funky business because it's a test tool - I've seen breakage like that before sometimes, but have never tried to work it out | 16:28 |
| odyssey4me | which branch is this? | 16:28 |
| jungleboyj | 17.1.2 | 16:29 |
| spotz | odyssey4me: He should be doing queens off of https://docs.openstack.org/openstack-ansible/latest/user/aio/quickstart.html | 16:29 |
| jungleboyj | spotz: I am in one of the containers and it looks like DNS isn't working. | 16:30 |
| odyssey4me | hmm, queens - that's pre systemd-networkd, so the networking should be pretty straightforward | 16:30 |
| odyssey4me | ok, so dns works on the host, but not inside the container? | 16:30 |
| jungleboyj | odyssey4me: Correct. | 16:30 |
| odyssey4me | dns would use an outbound link via a nat on the host through the lxc bridge | 16:30 |
| * jungleboyj is proud that I figured out how to get into the container that quick. | 16:31 | |
| spotz | odyssey4me: jungleboyj - I just started a build with stable/queens vs a tag | 16:31 |
| jungleboyj | Can't ping my nameserver either. | 16:31 |
| spotz | jungleboyj: ssh or lxc-attach?:) | 16:31 |
| odyssey4me | are you running ubuntu, centos or opensuse? | 16:31 |
| jungleboyj | spotz: ssh. | 16:31 |
| jungleboyj | ubuntu 16.04.5 | 16:31 |
| openstackgerrit | Merged openstack/openstack-ansible-ops master: Add kolide cluster tests https://review.openstack.org/611472 | 16:31 |
| *** fghaas has quit IRC | 16:32 | |
| jungleboyj | Ironically, this is why I got off packstack. Networking issues from the VMs. | 16:32 |
| *** vnogin has quit IRC | 16:32 | |
| odyssey4me | is the lxcbr0 up? | 16:32 |
| *** macza has quit IRC | 16:33 | |
| odyssey4me | does your container have addresses for all its interfaces? | 16:33 |
| odyssey4me | 'lxc-ls -f' from the host will give you all containers and their addresses | 16:33 |
| odyssey4me | some will have two, some will have more | 16:33 |
| jungleboyj | lxcbr0 is 10.255.255.1 | 16:33 |
| jungleboyj | No, only the second interface has an IP. The 172. interface. | 16:34 |
| odyssey4me | nothing on the first one? | 16:34 |
| jungleboyj | odyssey4me: No. | 16:34 |
| odyssey4me | right, sounds like dnsmasq isn't running on the host | 16:34 |
| jungleboyj | Ah. That would be a problem. | 16:34 |
| jungleboyj | Hmm, there are two dnsmasq processes. | 16:35 |
| odyssey4me | lxc-system-manage is a script we have which fixes things magically for you | 16:35 |
| jungleboyj | One owned by libvirt+ and one by root. | 16:35 |
| jungleboyj | odyssey4me: Want me to give that a shot? | 16:37 |
| *** cshen has joined #openstack-ansible | 16:37 | |
| odyssey4me | jungleboyj: Yeah, I think so. I suck at networking, and queens is aging so I can't remember all the right things to do - but that script has options to restart dnsmasq specifically for lxc. | 16:38 |
| jungleboyj | odyssey4me: OMG! That is what is killing me too. I have been banging my head on networking issues for a while here. I just want it to work. | 16:39 |
| spotz | Thanks for stepping in to help odyssey4me:) I pester jungleboyj for broken stuff all the time:) | 16:39 |
| jungleboyj | spotz: Is my sister from another mister. | 16:40 |
| jungleboyj | BTW, I panic'ed a bit when the automation disabled my ssh password access. | 16:40 |
| jungleboyj | odyssey4me: Looks like I want to do a system-rebuild? | 16:41 |
| spotz | jungleboyj: I did kinda warn you about that, I mentioned hacking ssh for my students:) | 16:42 |
| jungleboyj | spotz: Oh, yeah, you did. I was fighting so many issues at the time that Iost that. | 16:42 |
| *** shardy has quit IRC | 16:43 | |
| odyssey4me | jungleboyj: dnsmasq-restart ? | 16:46 |
| odyssey4me | gotta love that cloudnull wit: https://github.com/openstack/openstack-ansible-lxc_hosts/blob/stable/queens/templates/lxc-system-manage.j2#L297-L312 | 16:46 |
| noonedeadpunk | One question: I've completed upgrade from Q to R, and it seems, that I can not launch aany playbook anymore, as clouds.yaml is required for it. But I can not find it, so am I right, that it wasn't created during upgrade? | 16:46 |
| odyssey4me | noonedeadpunk: clouds.yaml for R is now only on the deploy node, and in the utility container - it is in one or two more places, but we're actively trying to cut down on the presence to improve security | 16:47 |
| * jungleboyj laughs. | 16:47 | |
| jungleboyj | Fire, Fire! | 16:47 |
| cloudnull | gotta use moar fire | 16:48 |
| jungleboyj | Hmmm, stopping dnsmasq doesn't actually stop it. | 16:49 |
| jungleboyj | So when I try to restart it says the address is already in use. | 16:49 |
| cloudnull | there | 16:49 |
| cloudnull | **there's the lxc-dnsmasq process | 16:49 |
| cloudnull | which has its own service unit | 16:49 |
| noonedeadpunk | odyssey4me: ah, thanks, found it. it is placed inside /root/ directory, and I was using "sudo -s", which didn't set root ENV for my session. | 16:50 |
| jungleboyj | But this should stop it. Right? lxc-system-manage dnsmasq-stop | 16:50 |
| cloudnull | it should, however to be sure `systemctl stop lxc-dnsmasq` | 16:52 |
| jungleboyj | Ok. Will give that a shot. | 16:52 |
| cloudnull | if the veth pairs are broken to the containers you can run `lxc-veth-check` | 16:53 |
| * cloudnull doesn't really know what's broken or we're talking about, just saw odyssey4me's ping :) | 16:54 | |
| odyssey4me | cloudnull: it's an AIO, and DNS from inside the containers isn't working | 16:55 |
| odyssey4me | oh, and it's queens | 16:55 |
| jungleboyj | cloudnull: For some reason my containers can't get to the outside world. | 16:55 |
| cloudnull | lxc-dnsmaq would be my first guess. | 16:57 |
| cloudnull | veth pair brokenness my second | 16:57 |
| spotz | cloudnull: We expect you to just KNOW:) | 17:01 |
| spotz | He had issues with networking on packstack which is why I suggested he try OSA | 17:01 |
| cloudnull | if the bridges are up/down'd the veth pairs will break and that can cause the issue. | 17:02 |
| cloudnull | you can check that using the `lxc-veth-check` script | 17:02 |
| cloudnull | also if you lxc-attach to the contianer you can ping the gateway | 17:02 |
| cloudnull | if its reachable then we should be ok from a wiring point of view | 17:02 |
| jungleboyj | Hmm, restarted my containers and now they don't get any ipaddresses. | 17:05 |
| jungleboyj | So, dnsmasq is all messed up. | 17:05 |
| openstackgerrit | David Wilde proposed openstack/openstack-ansible-ops master: MNAIO: Increase default logging_vm_server_ram to 16384 https://review.openstack.org/611679 | 17:06 |
| jungleboyj | odyssey4me: When I try to start with lxc-system-manage it just complains that 10.255.255.1 is already in use. | 17:09 |
| jungleboyj | lxc-veth-check just returns 2.0.8 . | 17:10 |
| jungleboyj | Not sure what that means. | 17:10 |
| cloudnull | jungleboyj `killall dnsmasq` and restart the lxc-dnsmasq process? | 17:11 |
| jungleboyj | I did that too. So I don't know what is using the port already. Do you know the port number>? | 17:11 |
| *** fghaas has joined #openstack-ansible | 17:15 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_octavia master: Fix updating amphora images https://review.openstack.org/611393 | 17:16 |
| antonym | anyone know if there's a cache to remove after removing a destroyed container from inventory using inventory-manage.py -r ? it doesn't show up in the list initially until i run a playbook and then it magically appears again in the list, it looks like the entries get removed from openstack_inventory.json but there's still an entry in openstack_hostnames_ips. I'm wondering if it's readding an entry | 17:17 |
| antonym | from that | 17:17 |
| odyssey4me | antonym: you have to remove it from openstack_user_config, then use inventory-manage to remove it from the inventory | 17:17 |
| odyssey4me | if it's still in openstack_user_config, then next inventory execution will rebuild it | 17:18 |
| jungleboyj | odyssey4me: Should I try using Rocky instead? | 17:19 |
| antonym | ah... ok, i'll give that a whirl... i'm guessing we'll need to do some more cleanup in the upgrade plays that queens runs at the end then | 17:19 |
| jungleboyj | Was doing Queens as there is a bug in Horizon that I was trying to avoid but if this goes off of stable/rocky for deployment I might be ok. | 17:20 |
| jungleboyj | If it is using Git instead of what is in Pip. | 17:20 |
| odyssey4me | jungleboyj: rocky's alright I think - there're still some things trickling in... I'm busy doing multi-node tests right now with it and things seem functional | 17:21 |
| jungleboyj | odyssey4me: Do I need to wipe the system and start over or is there a better way to get there? | 17:21 |
| odyssey4me | jungleboyj: I use VM's for AIO's - easier to kill and rebuild. And yeah, because it does all sort of crap it's probably better to rekick. | 17:22 |
| jungleboyj | Boo. Ok. | 17:22 |
| *** nsmeds__ has quit IRC | 17:23 | |
| jungleboyj | odyssey4me: I am trying to get my box to the point where I can have VMs. :-) | 17:23 |
| * jungleboyj this is why I can't have nice things | 17:24 | |
| *** nsmeds__ has joined #openstack-ansible | 17:25 | |
| odyssey4me | jungleboyj: soo... how much ram have you got? | 17:29 |
| odyssey4me | and how fast is your disk? | 17:29 |
| jungleboyj | It is an x3650 M3 with 48 GB of ram and 15k disks | 17:29 |
| odyssey4me | ok, and nested virt is supported? | 17:29 |
| jungleboyj | 6gb SAS | 17:29 |
| jungleboyj | odyssey4me: Should be. Xenon 2650 processors or something like that. | 17:30 |
| odyssey4me | did you take a peek at https://github.com/openstack/openstack-ansible-ops/tree/master/multi-node-aio ? you could use it to prep a cluster, and rekick it very simply if it all goes wrong | 17:30 |
| odyssey4me | I have it on my TODO to actually have that tooling support doing AIO's - I think it can, but it needs some docs to tell you how. | 17:31 |
| odyssey4me | that tooling keeps the host pretty lean - only the VM's get dirty | 17:31 |
| *** mmercer has joined #openstack-ansible | 17:31 | |
| jungleboyj | odyssey4me: So I just read through that and it is a bit over my head. | 17:34 |
| * jungleboyj is just a lowly storage guy. | 17:34 | |
| odyssey4me | :) | 17:35 |
| jungleboyj | odyssey4me: Oh wait, so you are saying use VirtManager to create instances on the Host and then deploy OSA in a VM there. | 17:36 |
| odyssey4me | yep, pretty much | 17:36 |
| odyssey4me | you don't even need any of that tooling | 17:36 |
| jungleboyj | Ah! Gotcha. | 17:36 |
| odyssey4me | for now, until you're more familiar with the AIO, just install libvirt and virt-manager on a fresh host - then build your AIO in there | 17:36 |
| odyssey4me | if it breaks, you can wpe and rebuild again and again until you have figured out how to do exactly what you want | 17:37 |
| odyssey4me | of course virt-manager allows you to bridge the the host and all those nice things if you need it | 17:37 |
| jungleboyj | Man, been a while since I used VirtManager. But then I could create one Beefy VM on there and do the AIO in there but not have to rebuild the whole box. | 17:37 |
| odyssey4me | yep - saves you tons of time and heartache, and gives you the freedom to experiment | 17:37 |
| jungleboyj | I forgot that we used to do that on boxes like this in our lab. | 17:38 |
| jungleboyj | And once I get a cloud working it doesn't really matter that it is within a VM. :-) | 17:39 |
| odyssey4me | with nested virt, nope - not at all | 17:39 |
| jungleboyj | Wheeee. | 17:39 |
| jungleboyj | odyssey4me: I don't suppose the networking is more complicated as the VM thinks it is baremetal. | 17:43 |
| odyssey4me | jungleboyj: the AIO pretty much just needs eth0, it does all the rest - so a standard ubuntu image for the vm will work just fine | 17:44 |
| *** hamzy has quit IRC | 17:44 | |
| jungleboyj | odyssey4me: And that will make things less complicated given this box has like 4 network adapters. I am only using one. Wondering if that is also complicating things. | 17:45 |
| *** hamzy has joined #openstack-ansible | 17:45 | |
| *** electrofelix has quit IRC | 17:48 | |
| *** priteau has quit IRC | 17:52 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_octavia master: Fix updating amphora images https://review.openstack.org/611393 | 18:06 |
| odyssey4me | nicolasbock: so, as it turns out - if there are multiple matches for the image, os_image_facts returns an error saying so... it does not return a list | 18:08 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_horizon stable/queens: Add ability for HTTP access to horizon https://review.openstack.org/611621 | 18:12 |
| odyssey4me | spotz jrosser I had a slight indentation error in https://review.openstack.org/611621 - hopefully all fixed now | 18:12 |
| spotz | odyssey4me: looking now | 18:13 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Cleanup the osquery role https://review.openstack.org/611641 | 18:14 |
| openstackgerrit | Antony Messerli proposed openstack/openstack-ansible master: Force overwrite of env.d files when upgrading https://review.openstack.org/611689 | 18:15 |
| nicolasbock | odyssey4me: well that is a bit inconvenient :( | 18:15 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Additional playbook cleanup and use stable release https://review.openstack.org/611661 | 18:15 |
| antonym | odyssey4me: thanks for the hint, tracked down the issue, env.d files weren't getting replaced | 18:15 |
| odyssey4me | antonym: I think those tasks were meant to identify if the /etc/openstack_deploy/env.d files in place were the same as the stock ones, and remove them if they were or leave them if they weren't. | 18:17 |
| *** lbragstad has quit IRC | 18:17 | |
| antonym | yeah, just as is, it wasn't replacing any of them, that's why all the containers that were supposed to be removed were still popping up | 18:18 |
| antonym | i guess if force is off it doesn't touch it if something is already there | 18:18 |
| odyssey4me | well, it'd be ideal if the user space files were removed instead - maybe backed up somewhere | 18:18 |
| *** DanyC_ has joined #openstack-ansible | 18:18 | |
| antonym | that's handled with the openstack_deploy backup right? | 18:18 |
| odyssey4me | in the old days, we had no in-tree env.d files - it was all user space | 18:19 |
| antonym | that's the first thing it does | 18:19 |
| odyssey4me | well, yeah - makes sense - and the groups will still exist in the inventory afterwards anyway | 18:19 |
| odyssey4me | so if there were any special customisations, they can be restored from backup after the upgrade and carried | 18:20 |
| antonym | yeah, that's what i was thinking, seems relatively safe | 18:20 |
| odyssey4me | but this transition happened at a very specific point in time - we actually shouldn't even have all that in master | 18:20 |
| odyssey4me | I think the transition was m->n ? | 18:20 |
| *** hamzaachi has joined #openstack-ansible | 18:21 | |
| odyssey4me | yep, m->n: https://github.com/openstack/openstack-ansible/tree/mitaka-eol/etc/openstack_deploy/env.d | 18:21 |
| odyssey4me | so on newton, those files should be removed - then from newton onwards, they should be left alone, because they're definitely customised | 18:21 |
| *** DanyC has quit IRC | 18:22 | |
| *** DanyC_ has quit IRC | 18:22 | |
| openstackgerrit | Merged openstack/openstack-ansible-ops master: MNAIO: Remove legacy os-infra_hosts group https://review.openstack.org/611582 | 18:23 |
| openstackgerrit | Merged openstack/openstack-ansible-ops master: MNAIO: Increase default infra_vm_server_ram to 16384 https://review.openstack.org/611586 | 18:23 |
| openstackgerrit | Merged openstack/openstack-ansible-ops master: MNAIO: Remove glance_default_store override https://review.openstack.org/611590 | 18:23 |
| openstackgerrit | Merged openstack/openstack-ansible-ops master: MNAIO: Remove lxc_container_backing_store override https://review.openstack.org/611591 | 18:23 |
| openstackgerrit | Merged openstack/openstack-ansible-ops master: MNAIO: Move the ELK-related comment inside the conditional https://review.openstack.org/611593 | 18:23 |
| openstackgerrit | Merged openstack/openstack-ansible-ops master: MNAIO: Make cinder/compute/swift hosts conditional https://review.openstack.org/611603 | 18:23 |
| antonym | so the env.d should be cleaned out when starting the newton upgrade and then the set provided in inventory/env.d should be what's always there? | 18:24 |
| odyssey4me | antonym: no, from newton onwards - nothing is needed in /etc/openstack_deploy/env.d at all | 18:25 |
| odyssey4me | so on the newton upgrade, remove them - from then on, ignore the folder entirely | 18:25 |
| odyssey4me | I guess if starting at newton, clean up the folder before upgrading from there | 18:26 |
| odyssey4me | time for me to head on out - cheers folks! | 18:28 |
| antonym | odyssey4me: ah, ok, got it, thanks! have a good night | 18:28 |
| jungleboyj | odyssey4me: Oh nice, I can dedicate one of the other network adapters to the VM. Need to go plug it in but this could work really well. | 18:39 |
| *** hamzaachi has quit IRC | 18:46 | |
| *** lbragstad has joined #openstack-ansible | 18:48 | |
| jungleboyj | Sweet! | 18:59 |
| *** hamzy has quit IRC | 19:07 | |
| *** hamzy has joined #openstack-ansible | 19:08 | |
| *** lbragstad has quit IRC | 19:21 | |
| jrosser | this seems to have worked if any opensuse folks want to double check the log https://review.openstack.org/#/c/611610/ | 19:23 |
| *** lbragstad has joined #openstack-ansible | 19:24 | |
| *** fghaas has quit IRC | 19:35 | |
| spotz | jungleboyj: FYI, my plain stab;e/queens AIO install finished and I do have DNS in the containers. Hope odyssey4me got you sorted out | 19:38 |
| openstackgerrit | Markos Chandras (hwoarang) proposed openstack/openstack-ansible-os_swift master: zuul: Switch to distro package installation template https://review.openstack.org/606056 | 19:39 |
| *** hamzy has quit IRC | 19:41 | |
| openstackgerrit | Chandan Kumar proposed openstack/openstack-ansible-os_tempest master: Added support for installing tempest from distro https://review.openstack.org/591424 | 19:41 |
| *** hamzaachi has joined #openstack-ansible | 19:45 | |
| jungleboyj | spotz: I am simplifying things by moving into a VM on the system to make it simpler and easier to recover. | 19:45 |
| jungleboyj | Off to enjoy nice weather with my boys for a bit. Will try the OSA deployment again later in the VM I spun up. | 19:46 |
| openstackgerrit | Merged openstack/openstack-ansible-os_watcher stable/rocky: Replace Chinese punctuation with English punctuation https://review.openstack.org/609446 | 19:49 |
| openstackgerrit | Merged openstack/openstack-ansible-os_trove stable/rocky: Replace Chinese punctuation with English punctuation https://review.openstack.org/609448 | 19:51 |
| *** cshen has quit IRC | 19:55 | |
| *** DanyC has joined #openstack-ansible | 19:58 | |
| *** cshen has joined #openstack-ansible | 20:00 | |
| *** hamzy has joined #openstack-ansible | 20:10 | |
| *** strattao has quit IRC | 20:13 | |
| *** bgmccollum has quit IRC | 20:15 | |
| *** strattao has joined #openstack-ansible | 20:18 | |
| *** nsmeds__ has quit IRC | 20:19 | |
| *** bgmccollum has joined #openstack-ansible | 20:19 | |
| *** strattao has quit IRC | 20:25 | |
| *** nsmeds__ has joined #openstack-ansible | 20:25 | |
| *** bgmccollum has quit IRC | 20:28 | |
| *** ThiagoCMC has joined #openstack-ansible | 20:28 | |
| ThiagoCMC | Hey guys, can I set "container_tech: spawn" globally? If yes, where to put it? | 20:29 |
| ThiagoCMC | *nspawn | 20:29 |
| ThiagoCMC | Under "global_overrides:"? | 20:29 |
| *** bgmccollum has joined #openstack-ansible | 20:30 | |
| *** openstackgerrit has quit IRC | 20:36 | |
| *** pcaruana has quit IRC | 20:44 | |
| *** ThiagoCMC has quit IRC | 21:05 | |
| *** DanyC has quit IRC | 21:14 | |
| *** cshen has quit IRC | 21:17 | |
| cloudnull | ThiagoCMC I typically set it as a host variable | 21:19 |
| cloudnull | though you can put that under the global overrides as well | 21:19 |
| *** openstackgerrit has joined #openstack-ansible | 21:20 | |
| openstackgerrit | Merged openstack/openstack-ansible-ops master: MNAIO: Increase default logging_vm_server_ram to 16384 https://review.openstack.org/611679 | 21:20 |
| openstackgerrit | Merged openstack/openstack-ansible stable/rocky: releasenotes: oslo-messaging-separate-backends add project name https://review.openstack.org/610518 | 21:23 |
| openstackgerrit | Merged openstack/openstack-ansible-os_blazar stable/rocky: Default MQ RPC/Notify credentials/vhosts to match https://review.openstack.org/609627 | 21:32 |
| *** strobelight has quit IRC | 21:39 | |
| *** ThiagoCMC has joined #openstack-ansible | 21:41 | |
| ThiagoCMC | How to set "container_tech: spawn" globally, under "global_overrides:"? | 21:42 |
| cloudnull | ThiagoCMC yes. That should work | 21:52 |
| ThiagoCMC | Ok, thanks! :-) | 21:53 |
| cloudnull | https://gist.github.com/cloudnull/eab22ba500843343470a7fd1e3f55d1c#file-openstack_user_config-yml-L97 - | 21:54 |
| cloudnull | I typically set it up as a host variable | 21:55 |
| cloudnull | however a "global_variable" should work too | 21:56 |
| ThiagoCMC | Sounds awesome! BTW, currently, for lab tests, I'm using KVM as a controller, network node and compute nodes. However, it consumes too much resources. Do you know if it is possible to deploy a lab on LXD, instead of KVM? | 21:58 |
| cloudnull | lxd is largely just lxc | 22:00 |
| cloudnull | if you were going to do lxd I'd recommend just running with a base deployment on the hardware | 22:00 |
| cloudnull | as for KVM, are the VMs taking too many resources or something else? | 22:01 |
| cloudnull | maybe you can turndown the workers ? | 22:01 |
| cloudnull | to save resources within the VMs. | 22:01 |
| cloudnull | these configs are all from my home lab however I too needed to save cycles to maximize my hardware. | 22:02 |
| cloudnull | https://gist.github.com/cloudnull/eab22ba500843343470a7fd1e3f55d1c#file-user_variables-yml-L183-L220 | 22:02 |
| cloudnull | my infra node is on a 4core (8thread) dell work station while my compute nodes are 16 core opterons | 22:02 |
| cloudnull | so i tuned down the infra worker threads so I didn't overwhelm my infra box | 22:03 |
| cloudnull | you could probably do to the same within your environment so that you don't need hefty VMs within KVM | 22:03 |
| ThiagoCMC | Hmm... Interesting! | 22:12 |
| ThiagoCMC | Thanks for sharing this! :-D | 22:12 |
| *** cshen has joined #openstack-ansible | 22:13 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_heat stable/rocky: Remove python-keystoneclient package install https://review.openstack.org/603378 | 22:13 |
| *** cshen has quit IRC | 22:17 | |
| *** hamzaachi has quit IRC | 22:22 | |
| *** hamzaachi has joined #openstack-ansible | 22:23 | |
| *** hamzaachi has quit IRC | 22:24 | |
| *** ThiagoCMC has quit IRC | 22:27 | |
| *** weezS has quit IRC | 22:30 | |
| nsmeds__ | cloudnull shares random user_config files, and TIL about YAML anchors and aliases | 22:34 |
| nsmeds__ | :partyparrot: | 22:34 |
| cloudnull | ha! | 22:34 |
| cloudnull | ThiagoCMC anytime ! | 22:35 |
| cloudnull | all of my home lab configs are in gists it makes it easy to share how i do things. | 22:36 |
| * cloudnull might just put all those bits into a proper github repo | 22:36 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops master: Cleanup the osquery role https://review.openstack.org/611641 | 22:56 |
| *** vnogin has joined #openstack-ansible | 23:00 | |
| *** tosky has quit IRC | 23:02 | |
| *** vnogin has quit IRC | 23:05 | |
| *** mmercer has quit IRC | 23:09 | |
| *** mmercer has joined #openstack-ansible | 23:09 | |
| *** greatgatsby has joined #openstack-ansible | 23:12 | |
| greatgatsby | Hello. If I use the os_server module with userdata that defines cloud-config, and that cloud-config specifies network configuration, am I correct in assuming that I should see the correct config written to /etc/network/interfaces.d/ when an ubuntu 16.04 machine is provisioned? | 23:14 |
| openstackgerrit | Merged openstack/openstack-ansible stable/pike: Stop MariaDB nodes to reset quorum during upgrade https://review.openstack.org/611438 | 23:24 |
| openstackgerrit | Merged openstack/openstack-ansible stable/pike: Update galera_server SHA to include cluster fixes https://review.openstack.org/611419 | 23:24 |
| *** nsmeds__ has quit IRC | 23:25 | |
| *** markvoelker has joined #openstack-ansible | 23:57 | |
| *** gyee has quit IRC | 23:59 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!