| *** gyee has quit IRC | 00:04 | |
| *** boxrick_ has joined #openstack-ansible | 00:05 | |
| *** kysse_ has joined #openstack-ansible | 00:07 | |
| *** chris_hultin|AWA has joined #openstack-ansible | 00:10 | |
| *** mgagne_ has joined #openstack-ansible | 00:10 | |
| *** adreznec_ has joined #openstack-ansible | 00:11 | |
| *** KeithMnemonic has quit IRC | 00:11 | |
| *** boxrick has quit IRC | 00:12 | |
| *** adreznec has quit IRC | 00:12 | |
| *** kysse has quit IRC | 00:12 | |
| *** chris_hultin has quit IRC | 00:12 | |
| *** mgagne has quit IRC | 00:12 | |
| *** yolanda has quit IRC | 00:12 | |
| *** dhellmann has quit IRC | 00:12 | |
| *** paulfantom has quit IRC | 00:12 | |
| *** logan- has quit IRC | 00:12 | |
| *** sdake has quit IRC | 00:12 | |
| *** pylot has quit IRC | 00:12 | |
| *** d34dh0r53 has quit IRC | 00:12 | |
| *** rickflare has quit IRC | 00:12 | |
| *** coolj has quit IRC | 00:12 | |
| *** aludwar has quit IRC | 00:12 | |
| *** sc68cal- has quit IRC | 00:12 | |
| *** neillc_ has quit IRC | 00:12 | |
| *** pabelanger has quit IRC | 00:12 | |
| *** raginbajin has quit IRC | 00:12 | |
| *** prometheanfire has quit IRC | 00:12 | |
| *** EmilienM has quit IRC | 00:12 | |
| *** portante has quit IRC | 00:12 | |
| *** SamYaple has quit IRC | 00:12 | |
| *** antonym has quit IRC | 00:12 | |
| *** bauruine has quit IRC | 00:12 | |
| *** lkoranda has quit IRC | 00:12 | |
| *** ChanServ has quit IRC | 00:12 | |
| *** chris_hultin|AWA is now known as chris_hultin | 00:12 | |
| *** adreznec_ is now known as adreznec | 00:12 | |
| *** odyssey4me has quit IRC | 00:12 | |
| *** odyssey4me has joined #openstack-ansible | 00:12 | |
| *** prometheanfire has joined #openstack-ansible | 00:13 | |
| *** aludwar has joined #openstack-ansible | 00:13 | |
| *** EmilienM has joined #openstack-ansible | 00:13 | |
| *** pabelanger has joined #openstack-ansible | 00:13 | |
| *** portante has joined #openstack-ansible | 00:13 | |
| *** SamYaple has joined #openstack-ansible | 00:13 | |
| *** sc68cal- has joined #openstack-ansible | 00:13 | |
| *** ChanServ has joined #openstack-ansible | 00:13 | |
| *** raginbajin has joined #openstack-ansible | 00:13 | |
| *** lkoranda has joined #openstack-ansible | 00:13 | |
| *** bauruine has joined #openstack-ansible | 00:13 | |
| *** antonym has joined #openstack-ansible | 00:13 | |
| *** neillc_ has joined #openstack-ansible | 00:13 | |
| *** barjavel.freenode.net sets mode: +o ChanServ | 00:13 | |
| *** flemingo has joined #openstack-ansible | 00:15 | |
| *** jra has joined #openstack-ansible | 00:19 | |
| *** flemingo has quit IRC | 00:20 | |
| jra | Could somebody help me understand how to configure multiple ceph auth profiles with OSA? I need a virsh secret registered for each (one for a replicated pool, another for an erasure-coded pool), but it seems like the role only handles a single keyring and secret? | 00:21 |
|---|---|---|
| *** throwsb1 has quit IRC | 00:23 | |
| *** keithmnemonic[m] has quit IRC | 00:23 | |
| *** csmart has quit IRC | 00:23 | |
| *** evrardjp[m] has quit IRC | 00:23 | |
| *** nurdie_ has quit IRC | 00:28 | |
| *** csmart has joined #openstack-ansible | 00:32 | |
| *** esberglu has joined #openstack-ansible | 00:35 | |
| *** esberglu has quit IRC | 00:40 | |
| *** osnaya has joined #openstack-ansible | 00:40 | |
| *** flemingo has joined #openstack-ansible | 00:41 | |
| *** vnogin has joined #openstack-ansible | 00:45 | |
| *** flemingo has quit IRC | 00:45 | |
| *** vnogin has quit IRC | 00:49 | |
| osnaya | does setup-hosts.yml install linux container related packages and start service on the target host? Or is it part of pre-requisite to be done manually? (couldn't find any manual steps as part of pre-requisites)... | 00:57 |
| *** flemingo has joined #openstack-ansible | 00:57 | |
| *** keithmnemonic[m] has joined #openstack-ansible | 01:00 | |
| *** evrardjp[m] has joined #openstack-ansible | 01:00 | |
| *** flemingo has quit IRC | 01:02 | |
| *** flemingo has joined #openstack-ansible | 01:07 | |
| *** flemingo has quit IRC | 01:11 | |
| osnaya | @evrardjp does setup-hosts.yml install linux container related packages and start service on the target host? Or is it part of pre-requisite to be done manually? (couldn't find any manual steps as part of pre-requisites)... | 01:12 |
| idlemind | osnaya the osa environment will take care of that | 01:15 |
| osnaya | @evrardjp setup-hosts.yml run successfully, console log shows 20+ containers created but on target host lxc list errors (saying conn refused, is Lxc service running?) so trying to figure out.... | 01:16 |
| osnaya | will be online again in few hours... and continue debug... thanks | 01:19 |
| *** osnaya has quit IRC | 01:19 | |
| *** nurdie has joined #openstack-ansible | 01:26 | |
| *** nurdie has quit IRC | 01:27 | |
| *** yolanda has joined #openstack-ansible | 01:28 | |
| *** dhellmann has joined #openstack-ansible | 01:28 | |
| *** paulfantom has joined #openstack-ansible | 01:28 | |
| *** logan- has joined #openstack-ansible | 01:28 | |
| *** sdake has joined #openstack-ansible | 01:28 | |
| *** pylot has joined #openstack-ansible | 01:28 | |
| *** d34dh0r53 has joined #openstack-ansible | 01:28 | |
| *** rickflare has joined #openstack-ansible | 01:28 | |
| *** coolj has joined #openstack-ansible | 01:28 | |
| *** hamzy has joined #openstack-ansible | 01:28 | |
| *** nurdie has joined #openstack-ansible | 01:28 | |
| *** adreznec has quit IRC | 01:30 | |
| *** keithmnemonic[m] has quit IRC | 01:32 | |
| *** csmart has quit IRC | 01:32 | |
| *** evrardjp[m] has quit IRC | 01:33 | |
| *** adreznec has joined #openstack-ansible | 01:33 | |
| *** fatdragon1 has joined #openstack-ansible | 01:43 | |
| *** flemingo has joined #openstack-ansible | 01:44 | |
| *** armaan has quit IRC | 01:45 | |
| *** armaan has joined #openstack-ansible | 01:46 | |
| *** csmart has joined #openstack-ansible | 01:47 | |
| *** csmart has quit IRC | 01:47 | |
| *** csmart has joined #openstack-ansible | 01:53 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_nova stable/queens: change python-libguestfs to python-guestfs for ubuntu https://review.openstack.org/551971 | 02:15 |
| *** flemingo_ has joined #openstack-ansible | 02:17 | |
| *** flemingo has quit IRC | 02:20 | |
| *** evrardjp[m] has joined #openstack-ansible | 02:20 | |
| *** keithmnemonic[m] has joined #openstack-ansible | 02:20 | |
| *** yifei1 has joined #openstack-ansible | 02:22 | |
| yifei1 | Hi all I have encountered an issue recently that the deployment fails while building gnocchi package (playing repo-install.yml) | 02:28 |
| yifei1 | ERROR like this: command 'build_py' has no such option 'pre_hook.build_config' | 02:29 |
| yifei1 | both pike and ocata branches have this issue for now | 02:29 |
| yifei1 | is there any one can help me about this? | 02:30 |
| yifei1 | evrardjp, hwoarang, cloudnull, odyssey4me: ping | 02:31 |
| openstackgerrit | Merged openstack/ansible-role-python_venv_build master: Revert the use of wheel synchronizing https://review.openstack.org/557744 | 02:37 |
| openstackgerrit | Merged openstack/ansible-role-python_venv_build master: Allow a venv to be downloaded from a URL https://review.openstack.org/557779 | 02:38 |
| *** sawblade_ has joined #openstack-ansible | 02:44 | |
| *** sawblade6 has quit IRC | 02:45 | |
| *** chhagarw has joined #openstack-ansible | 02:55 | |
| *** yifei1 has quit IRC | 02:57 | |
| *** chhagarw has quit IRC | 03:00 | |
| cloudnull | he yifei yes this is a known issue | 03:12 |
| cloudnull | **hi | 03:12 |
| cloudnull | https://review.openstack.org/#/c/557754/ | 03:12 |
| cloudnull | the problem is caused by a new version of PBR being released | 03:13 |
| cloudnull | that patch will fix it | 03:13 |
| cloudnull | however you could also just install it on the repo server to fix it | 03:13 |
| cloudnull | ansible -m shell -a 'pip install pbr==3.1.1' repo_all | 03:13 |
| *** savvas has joined #openstack-ansible | 03:17 | |
| *** fatdragon1 has quit IRC | 03:19 | |
| *** yifei1 has joined #openstack-ansible | 03:20 | |
| *** savvas has quit IRC | 03:21 | |
| *** dave-mccowan has quit IRC | 03:23 | |
| cloudnull | osnaya the command to list the containers would be lxc-ls | 03:25 |
| cloudnull | if you use the "lxc" command, that would interact with LXD. | 03:25 |
| cloudnull | it's super confusing but something that canonical has done. | 03:26 |
| * cloudnull has no idea why | 03:26 | |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-plugins master: Fix delegation to hosts not in inventory https://review.openstack.org/550665 | 03:29 |
| *** yifei1 has quit IRC | 03:51 | |
| *** poopcat has quit IRC | 04:04 | |
| *** flemingo_ has quit IRC | 04:13 | |
| *** chhagarw has joined #openstack-ansible | 04:17 | |
| openstackgerrit | Matthew Thode proposed openstack/openstack-ansible-os_tempest stable/queens: update for changes needed to pass refstack https://review.openstack.org/552905 | 04:26 |
| *** bhujay has joined #openstack-ansible | 04:27 | |
| openstackgerrit | Matthew Thode proposed openstack/openstack-ansible-os_tempest stable/pike: update for changes needed to pass refstack https://review.openstack.org/552906 | 04:28 |
| *** flemingo has joined #openstack-ansible | 04:39 | |
| *** gkadam has joined #openstack-ansible | 04:41 | |
| *** flemingo has quit IRC | 04:44 | |
| *** vnogin has joined #openstack-ansible | 04:45 | |
| *** vnogin has quit IRC | 04:50 | |
| *** markvoelker has quit IRC | 04:52 | |
| *** strobelight has joined #openstack-ansible | 05:09 | |
| *** strobelight has quit IRC | 05:13 | |
| *** strobelight has joined #openstack-ansible | 05:30 | |
| *** strobelight has quit IRC | 05:34 | |
| *** flemingo has joined #openstack-ansible | 05:42 | |
| *** gkadam has quit IRC | 05:43 | |
| *** strobelight has joined #openstack-ansible | 05:44 | |
| *** flemingo has quit IRC | 05:47 | |
| *** strobelight has quit IRC | 05:49 | |
| *** armaan has quit IRC | 05:52 | |
| *** armaan has joined #openstack-ansible | 05:53 | |
| *** markvoelker has joined #openstack-ansible | 05:53 | |
| *** bhujay has quit IRC | 05:57 | |
| *** flemingo has joined #openstack-ansible | 06:02 | |
| *** bhujay has joined #openstack-ansible | 06:03 | |
| *** flemingo has quit IRC | 06:07 | |
| *** indistylo has joined #openstack-ansible | 06:08 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-specs master: Add spec to refactor the dynamic inventory https://review.openstack.org/554453 | 06:16 |
| openstackgerrit | Merged openstack/openstack-ansible master: [Docs] Simplify docs configuration https://review.openstack.org/555880 | 06:16 |
| *** flemingo has joined #openstack-ansible | 06:18 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible master: Run the on metal job https://review.openstack.org/556521 | 06:20 |
| *** flemingo has quit IRC | 06:22 | |
| evrardjp | yifei: it's a day off for the uk | 06:25 |
| *** yifei1 has joined #openstack-ansible | 06:25 | |
| evrardjp | yifei1: ^ | 06:26 |
| yifei1 | evrardjp: Hi | 06:26 |
| yifei1 | evrardjp: Sorry I lost | 06:26 |
| evrardjp | I will make sure the issue is merged into the branches, and issue a release today | 06:26 |
| yifei1 | the connection before | 06:27 |
| evrardjp | yifei1: It's a day off in UK today | 06:27 |
| evrardjp | same for monday | 06:27 |
| yifei1 | evrardjp: OK thanks | 06:27 |
| evrardjp | but I will still try to issue a release today | 06:28 |
| evrardjp | which branch are you on? | 06:28 |
| *** flemingo has joined #openstack-ansible | 06:29 | |
| yifei1 | ocata and pike | 06:29 |
| yifei | cloudnull: Thanks to you | 06:32 |
| *** flemingo has quit IRC | 06:33 | |
| *** flemingo has joined #openstack-ansible | 06:38 | |
| *** aruns has joined #openstack-ansible | 06:38 | |
| *** indistylo has quit IRC | 06:40 | |
| *** Sha000000 has joined #openstack-ansible | 06:41 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-os_tempest master: [WIP] Overriding default timeouts https://review.openstack.org/557896 | 06:41 |
| *** flemingo has quit IRC | 06:42 | |
| *** flemingo has joined #openstack-ansible | 06:56 | |
| openstackgerrit | Merged openstack/ansible-config_template master: Utilise sorted to ensure no random changes https://review.openstack.org/554979 | 06:56 |
| *** flemingo has quit IRC | 07:01 | |
| *** arbrandes1 has joined #openstack-ansible | 07:01 | |
| prometheanfire | lol, how I know europe woke up | 07:01 |
| *** Taseer has quit IRC | 07:02 | |
| *** arbrandes has quit IRC | 07:03 | |
| *** DanyC has quit IRC | 07:12 | |
| *** DanyC has joined #openstack-ansible | 07:13 | |
| *** DanyC has quit IRC | 07:17 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-plugins master: Allow remote user to "become" when managing containers https://review.openstack.org/557901 | 07:19 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-plugins master: Allow remote user to "become" when managing containers https://review.openstack.org/557901 | 07:23 |
| *** flemingo has joined #openstack-ansible | 07:25 | |
| *** DanyC has joined #openstack-ansible | 07:28 | |
| *** flemingo has quit IRC | 07:30 | |
| *** DanyC has quit IRC | 07:30 | |
| *** DanyC has joined #openstack-ansible | 07:31 | |
| *** DanyC has quit IRC | 07:35 | |
| *** electrofelix has joined #openstack-ansible | 07:42 | |
| *** flemingo has joined #openstack-ansible | 07:46 | |
| *** flemingo has quit IRC | 07:51 | |
| *** Sha000000 has quit IRC | 07:57 | |
| *** DanyC has joined #openstack-ansible | 07:59 | |
| *** DanyC has quit IRC | 08:06 | |
| *** DanyC has joined #openstack-ansible | 08:06 | |
| *** bhujay has quit IRC | 08:16 | |
| *** DanyC has quit IRC | 08:28 | |
| *** DanyC has joined #openstack-ansible | 08:29 | |
| *** DanyC has quit IRC | 08:33 | |
| *** DanyC has joined #openstack-ansible | 08:35 | |
| *** DanyC has quit IRC | 08:36 | |
| *** DanyC has joined #openstack-ansible | 08:37 | |
| *** DanyC has quit IRC | 08:38 | |
| *** bhujay has joined #openstack-ansible | 08:39 | |
| *** gkadam has joined #openstack-ansible | 08:46 | |
| *** DanyC has joined #openstack-ansible | 08:47 | |
| *** ivve has joined #openstack-ansible | 08:54 | |
| *** Taseer has joined #openstack-ansible | 08:57 | |
| *** flemingo has joined #openstack-ansible | 08:59 | |
| *** vnogin has joined #openstack-ansible | 09:01 | |
| *** flemingo has quit IRC | 09:04 | |
| *** esberglu has joined #openstack-ansible | 09:24 | |
| *** flemingo has joined #openstack-ansible | 09:28 | |
| *** esberglu has quit IRC | 09:29 | |
| *** flemingo has quit IRC | 09:32 | |
| *** bhujay has quit IRC | 09:36 | |
| *** flemingo has joined #openstack-ansible | 09:46 | |
| *** flemingo has quit IRC | 09:50 | |
| *** fghaas has joined #openstack-ansible | 09:57 | |
| fghaas | evrardjp, odyssey4me: if one of you is around today, perhaps you'd like to take a look at https://gist.github.com/fghaas/3a66fae177a4a19652aa6372aa6a8947 and see if that summarizes the post-upgrade container removal tasks correctly. (In that gist, "deploy" is the node running OSA; "alice" is a controller node, "bob" a compute node, and "daisy" a cinder-volume node. | 09:58 |
| *** flemingo has joined #openstack-ansible | 10:08 | |
| *** flemingo has quit IRC | 10:13 | |
| odyssey4me | fghaas heh, I kicked of looking into it yesterday with https://etherpad.openstack.org/p/osa-pike-to-queens-cleanup-notes | 10:21 |
| odyssey4me | that looks pretty comprehensive, thanks fghaas ! | 10:25 |
| *** flemingo has joined #openstack-ansible | 10:27 | |
| *** flemingo has quit IRC | 10:32 | |
| *** armaan has quit IRC | 10:34 | |
| *** armaan has joined #openstack-ansible | 10:35 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-tests master: Do not generate the html ARA report any more https://review.openstack.org/557921 | 10:36 |
| *** flemingo has joined #openstack-ansible | 10:41 | |
| fghaas | odyssey4me: Given the fact that there's no "service remove" nor "service disable" command for Heat services, what's your suggestion for operators there? Just leave the service down and not worry about it? | 10:45 |
| fghaas | (I'm never a big fan of poking around in the database for things like that) | 10:45 |
| odyssey4me | fghaas is there a heat-manage perhaps? | 10:45 |
| odyssey4me | I think that cinder-manage may allow you to remove a service. | 10:46 |
| *** flemingo has quit IRC | 10:46 | |
| odyssey4me | You'd have to run that from the venv in the service container though. | 10:46 |
| fghaas | Yup, heat-manage service clean, indeed | 10:48 |
| fghaas | Also, agree that it's pretty tedious to have to shell into the container _and_ activate the venv for that. Meaning, this is a fine task to automate, but until it's automated, I wouldn't necessarily recommend that people do it manually | 10:49 |
| *** chhagarw has quit IRC | 10:51 | |
| odyssey4me | fghaas just use: ansible -m shell -a "/openstack/venvs/.../bin/...-manage ..." heat_all[0] | 10:52 |
| odyssey4me | or something like that | 10:52 |
| odyssey4me | the venv shouldn't require activation due to the python shebang referring to the venv python | 10:53 |
| fghaas | Right. One more question: after I drop the old containers from the inventory, running openstack-hosts-setup.yml doesn't remove their entries from the /etc/hosts file. Is that intentional? | 10:59 |
| *** hamza21 has joined #openstack-ansible | 10:59 | |
| fghaas | If so, and there's really no way to do this with a playbook, that's another post-upgrade cleanup task I'd need to add. | 10:59 |
| *** DanyC has quit IRC | 10:59 | |
| *** aruns has quit IRC | 11:01 | |
| odyssey4me | oh really? that sounds like a bug | 11:01 |
| fghaas | No, not really. I could easily see how it would be intentional, that is, just not touching any entries that OSA doesn't know about. | 11:02 |
| fghaas | Which it doesn't, after removal from the inventory. | 11:02 |
| *** flemingo has joined #openstack-ansible | 11:03 | |
| odyssey4me | ah, that kinda makes sense | 11:03 |
| odyssey4me | hmm, we should get rid of that etc/hosts thing and just recommend using unbound instead | 11:04 |
| odyssey4me | anyway | 11:04 |
| odyssey4me | I guess that means another clean-up task | 11:04 |
| *** flemingo has quit IRC | 11:08 | |
| *** niraj_singh has joined #openstack-ansible | 11:10 | |
| odyssey4me | fghaas oh, now that I think about it - the procedure should probably be to disable all those agents/services prior to deleting them - that way they get to finish up whatever tasks they currently have executing and the other agents will take over | 11:10 |
| fghaas | huh? They don't do that when we do a graceful container shutdown? (lxc-destroy -f does lxc-stop, not lxc-stop -k) | 11:12 |
| odyssey4me | that just shuts down the service in the normal way | 11:12 |
| odyssey4me | I guess it depends on how much you care about transactions that are in motion | 11:13 |
| odyssey4me | disabling the service, then waiting a little, effectively drains the service first | 11:13 |
| odyssey4me | versus a service shutdown which will try to train, then kill | 11:13 |
| odyssey4me | the default timeout is 2 mins | 11:14 |
| *** savvas has joined #openstack-ansible | 11:14 | |
| odyssey4me | I guess that's fine for most environments | 11:14 |
| odyssey4me | fghaas added a comment for the container cleanup - it can be done without knowing the container names | 11:17 |
| odyssey4me | afk for a bit | 11:18 |
| spotz | fghaas and odyssey4me - Things should finish their work prior to deleting, unless a force is used maybe? Anyways going riding back in a few hourse:) | 11:29 |
| *** flemingo has joined #openstack-ansible | 11:30 | |
| *** jwitko has quit IRC | 11:31 | |
| fghaas | the --force in lxc-destroy means that it allows you to destroy a running container but shutting it down _gracefully_ first. If you actually wanted do destroy a running container non-gracefully, you'd have to use lxc-stop -k followed by lxc-destroy without -f | 11:32 |
| fghaas | But enjoy the ride :) | 11:33 |
| *** flemingo has quit IRC | 11:34 | |
| *** admin0 has joined #openstack-ansible | 11:38 | |
| *** savvas has quit IRC | 11:41 | |
| *** savvas has joined #openstack-ansible | 11:42 | |
| admin0 | stable/queens -- greenfield ... python pw-token-gen.py --file /etc/openstack_deploy/user_secrets.yml | 11:42 |
| admin0 | => Missing Dependency, "PyYAML" | 11:42 |
| admin0 | 11:42 | |
| admin0 | shouldn't bootstrap install this ? | 11:45 |
| openstackgerrit | Merged openstack/ansible-config_template master: Retain the ordering when using with MultiStringOp https://review.openstack.org/556956 | 11:46 |
| *** mrhillsman has quit IRC | 11:48 | |
| *** hamza21 has quit IRC | 11:53 | |
| *** savvas_ has joined #openstack-ansible | 11:55 | |
| *** hamza21 has joined #openstack-ansible | 11:57 | |
| *** flemingo has joined #openstack-ansible | 11:58 | |
| *** savvas has quit IRC | 11:58 | |
| *** flemingo has quit IRC | 12:02 | |
| *** flemingo has joined #openstack-ansible | 12:07 | |
| admin0 | just ./ pw-token-gen.py --file /etc/openstack_deploy/user_secrets.yml worked fine | 12:07 |
| *** flemingo has quit IRC | 12:11 | |
| *** admin0 has quit IRC | 12:12 | |
| *** hamza21 has quit IRC | 12:15 | |
| *** hamza21 has joined #openstack-ansible | 12:15 | |
| *** odyssey4me has quit IRC | 12:19 | |
| *** odyssey4me has joined #openstack-ansible | 12:19 | |
| *** markvoelker has quit IRC | 12:20 | |
| *** markvoelker has joined #openstack-ansible | 12:20 | |
| *** markvoelker has quit IRC | 12:23 | |
| *** markvoelker has joined #openstack-ansible | 12:25 | |
| *** markvoelker has quit IRC | 12:29 | |
| *** DanyC has joined #openstack-ansible | 12:30 | |
| *** strobelight has joined #openstack-ansible | 12:33 | |
| *** flemingo has joined #openstack-ansible | 12:34 | |
| *** markvoelker has joined #openstack-ansible | 12:34 | |
| *** DanyC has quit IRC | 12:36 | |
| *** DanyC has joined #openstack-ansible | 12:36 | |
| *** flemingo has quit IRC | 12:38 | |
| *** DanyC has quit IRC | 12:40 | |
| *** hamza21 has quit IRC | 12:44 | |
| *** hamza21 has joined #openstack-ansible | 12:44 | |
| *** vnogin has quit IRC | 12:46 | |
| openstackgerrit | Merged openstack/openstack-ansible stable/pike: Update repo_build role for PBR fix https://review.openstack.org/557812 | 12:46 |
| *** vnogin has joined #openstack-ansible | 12:49 | |
| *** flemingo has joined #openstack-ansible | 12:55 | |
| *** flemingo has quit IRC | 13:00 | |
| *** strobelight_ has joined #openstack-ansible | 13:04 | |
| *** strobelight is now known as Guest71457 | 13:04 | |
| *** strobelight_ is now known as strobelight | 13:05 | |
| *** Guest71457 has quit IRC | 13:07 | |
| *** KeithMnemonic has joined #openstack-ansible | 13:16 | |
| *** electrofelix has quit IRC | 13:21 | |
| *** flemingo has joined #openstack-ansible | 13:22 | |
| *** flemingo has quit IRC | 13:26 | |
| *** fghaas has quit IRC | 13:31 | |
| *** dave-mccowan has joined #openstack-ansible | 13:33 | |
| *** esberglu has joined #openstack-ansible | 13:34 | |
| *** 3NAAAIX4G has joined #openstack-ansible | 13:34 | |
| *** dave-mccowan has quit IRC | 13:37 | |
| *** fghaas has joined #openstack-ansible | 13:39 | |
| *** fghaas has quit IRC | 13:39 | |
| *** fghaas has joined #openstack-ansible | 13:39 | |
| *** dave-mccowan has joined #openstack-ansible | 13:40 | |
| *** flemingo has joined #openstack-ansible | 13:42 | |
| cloudnull | mornings | 13:46 |
| *** goldenfr_ has joined #openstack-ansible | 13:47 | |
| *** flemingo has quit IRC | 13:47 | |
| *** goldenfr_ is now known as goldenfri | 13:48 | |
| *** throwsb1 has joined #openstack-ansible | 13:50 | |
| d34dh0r53 | o/ cloudnull | 13:51 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible master: Add playbook to ship journals from hosts https://review.openstack.org/555982 | 13:52 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible master: Converge distro interfaces to systemd-networkd https://review.openstack.org/552721 | 13:53 |
| cloudnull | o/ d34dh0r53 | 13:54 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_glance master: Convert role to use a common systemd mount role https://review.openstack.org/552185 | 13:55 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_neutron master: Convert role to use a common systemd service role https://review.openstack.org/556952 | 13:57 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_nova master: Convert role to use a common systemd service role https://review.openstack.org/466179 | 13:58 |
| d34dh0r53 | cloudnull: Why all the dummy interfaces? | 13:58 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible master: Use production like interfaces in the gate https://review.openstack.org/553636 | 13:59 |
| cloudnull | d34dh0r53: the systemd-networkd services needs an interface to plug the birdges into | 14:02 |
| *** flemingo has joined #openstack-ansible | 14:02 | |
| cloudnull | in the aio we leave the bridges empty | 14:02 |
| d34dh0r53 | cloudnull: yeah, was figuring that out. Do we need a br_vxlan in systemd_networks? | 14:02 |
| d34dh0r53 | nevermind, I'm blind this morning | 14:03 |
| d34dh0r53 | it's there | 14:03 |
| cloudnull | yes, we should have "br-vxlan" | 14:03 |
| d34dh0r53 | id-10-t error | 14:04 |
| cloudnull | ha! delay in my bouncer :) | 14:04 |
| * cloudnull things he needs to go tune the bouncer up its been acting a fool. | 14:04 | |
| d34dh0r53 | yeah, I've noticed it's been slow | 14:05 |
| odyssey4me | cloudnull FYI nothing in the integrated gate will pass until https://review.openstack.org/556994 merges | 14:05 |
| cloudnull | doh! | 14:05 |
| odyssey4me | you can rebase on top of that if you like to validate whether something's working | 14:05 |
| * cloudnull should have asked about the integrated gate status before rebasing some of those PRs | 14:06 | |
| cloudnull | looks like its in the merge queue | 14:07 |
| *** flemingo has quit IRC | 14:07 | |
| cloudnull | while waiting on that, anyone want to review https://review.openstack.org/#/c/557901 | 14:08 |
| odyssey4me | busy working on the pike->queens cleanup, with thanks to fghaas' great work in https://gist.github.com/fghaas/3a66fae177a4a19652aa6372aa6a8947 | 14:12 |
| cloudnull | neat! | 14:21 |
| cloudnull | that looks nice. | 14:21 |
| odyssey4me | yep, busy automating it into run-upgrade now and will add docs explaining what it does | 14:21 |
| odyssey4me | thanks to fghaas I don't need to think too much :) | 14:21 |
| cloudnull | which is always a good thing on a firday | 14:22 |
| *** flemingo has joined #openstack-ansible | 14:22 | |
| cloudnull | ** Friday | 14:22 |
| odyssey4me | yep, easter friday on top of that :p | 14:22 |
| cloudnull | On the eve of Zombie Jesus. | 14:22 |
| cloudnull | on that note, Wife and I were watching "Santa Clarita Diet" last night, its a pretty good show. | 14:23 |
| shananigans | I can't stop laughing at Nathan Fillion's talking head. | 14:25 |
| cloudnull | its a funny show. | 14:26 |
| *** flemingo has quit IRC | 14:26 | |
| shananigans | My wife refuses to watch though. | 14:29 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-plugins master: Allow remote user to "become" when managing containers https://review.openstack.org/557901 | 14:31 |
| cloudnull | its horror-comedy, not everyone is into that. Buts its pretty damn funny. | 14:33 |
| cloudnull | my wife was the one that got me into it | 14:33 |
| cloudnull | we're still on Season 1 but we binged most of it last night | 14:33 |
| fghaas | cloudnull: oh hell yeah awesome on that last change | 14:45 |
| *** flemingo has joined #openstack-ansible | 14:45 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible master: Convert rsyslog to an include_task with group_vars https://review.openstack.org/556002 | 14:46 |
| fghaas | I've been wanting to say PermitRootLogin no on my OSA-managed boxes for a loooong time | 14:46 |
| *** flemingo has quit IRC | 14:50 | |
| spotz | anyone know what day dmsimard's talk is gonna be? Was thinking we could do our outing and incorporate that | 14:51 |
| dmsimard | spotz: what talk? | 14:51 |
| spotz | dmsimard: At the meetup | 14:51 |
| dmsimard | which one? lol | 14:51 |
| * dmsimard talks too much | 14:51 | |
| dmsimard | the ansible vancouver meetup ? | 14:51 |
| spotz | dmsimard: you're doing more then one meetup during summit? | 14:52 |
| cloudnull | fghaas: thanks! | 14:52 |
| dmsimard | spotz: no but I did a talk this week and I'm doing another soon.. | 14:52 |
| cloudnull | I was trying to use my the plugin with my home lab which looks like http://paste.openstack.org/show/718025/ | 14:52 |
| cloudnull | now i can :) | 14:52 |
| spotz | dmsimard: Summit one:) | 14:52 |
| dmsimard | spotz: I haven't had time to formalize a date with the organizer yet | 14:52 |
| cloudnull | that's very much the hope, Having no root login an option would be wonderful | 14:53 |
| spotz | dmsimard: Well let us know. If you want us there that is:) | 14:54 |
| dmsimard | spotz: I also kinda need to talk with Robyn, see if we can do something at the summit or if we're going to do something external (which likely needs to be cleared by foundation staff) | 14:54 |
| odyssey4me | fghaas IIRC we now have permitrootlogin set to only work with a key now | 14:55 |
| odyssey4me | https://github.com/openstack/openstack-ansible/blob/3c74b99ea4b035b8ccfc53ca2a4169941b777316/releasenotes/notes/sshd-permit-root-login-without-password-948ec79c6508c19b.yaml | 14:55 |
| spotz | dmsimard: Well let us know, we try to get together a few times and would love to support you | 14:55 |
| dmsimard | spotz: sounds good to me, cloudnull might have wanted a spot too :) | 14:56 |
| fghaas | yeah and I'm being a zealot and I don't want PermitRootLogin without-password either, I just want sudo to be the only way anyone becomes root. :) | 14:56 |
| odyssey4me | not perfect, but we should actually switch to assuming non-root by default - baby steps ;) | 14:56 |
| fghaas | so using become on target: yay, doubleplusgood | 14:56 |
| cloudnull | ++ | 14:56 |
| *** KeithMnemonic has quit IRC | 14:57 | |
| cloudnull | dmsimard: spotz: IDK if im going quite yet, but if i do get to go I'd love a spot. | 14:57 |
| spotz | cloudnull dmsimard I'll cross my fingers | 15:04 |
| evrardjp | odyssey4me: is cloudnull helping you on the upgrade part ? | 15:11 |
| evrardjp | I remember he promised :p | 15:11 |
| odyssey4me | evrardjp no need, fghaas has done all the heavy lifting ;) | 15:11 |
| evrardjp | ok | 15:11 |
| evrardjp | well it was just to know | 15:11 |
| odyssey4me | cloudnull shoulda woulda coulda | 15:12 |
| odyssey4me | :p | 15:12 |
| evrardjp | promises only bind those who believe in them, I agree. | 15:12 |
| spotz | evrardjp: You're on vacation... go vacay!:) | 15:12 |
| odyssey4me | dmsimard so https://review.openstack.org/557921 didn't work when trying to look at the report for our results... does the DB need a special name or something? | 15:13 |
| * odyssey4me is on vacation too, technically... and thus there is a margarita in-hand. ;) | 15:17 | |
| evrardjp | odyssey4me: lucky basterd. | 15:18 |
| spotz | evrardjp: odyssey4me Can I just have a margarita? | 15:18 |
| odyssey4me | even luckier if you had spelled it right ;) | 15:19 |
| evrardjp | odyssey4me: I just came back of the supermarket with everything that would go fine with your margarita though. | 15:19 |
| evrardjp | odyssey4me: haha I am not sure it's the netiquette with the right spelling :p | 15:19 |
| odyssey4me | evrardjp with a margarita in-hand, everything is perfectly fine | 15:20 |
| evrardjp | the bloody mary is only for the morning? | 15:20 |
| odyssey4me | mornings are relative | 15:21 |
| *** vnogin has quit IRC | 15:22 | |
| evrardjp | :) | 15:23 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-openstack_hosts master: Add retries to requirement package install https://review.openstack.org/557943 | 15:25 |
| odyssey4me | apparently it's the time where mirrors are updated and everything fails | 15:26 |
| odyssey4me | at least in the pub cloud I'm doing things | 15:27 |
| odyssey4me | oh, this isn't good - I'm using archive.ubuntu.com | 15:29 |
| *** RossSchulman[m] has joined #openstack-ansible | 15:33 | |
| *** DanyC has joined #openstack-ansible | 15:33 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-os_tempest master: [WIP] Overriding default timeouts https://review.openstack.org/557896 | 15:34 |
| RossSchulman[m] | Could anyone help me out by telling me how to tear down a failed AIO attempt back to nothing so I can try again? | 15:36 |
| d34dh0r53 | bloody mary is acceptable all day, every day | 15:36 |
| odyssey4me | ^ ++ | 15:36 |
| odyssey4me | RossSchulman[m] delete the instance/VM and start again | 15:36 |
| *** admin0 has joined #openstack-ansible | 15:37 | |
| *** flemingo has joined #openstack-ansible | 15:37 | |
| RossSchulman[m] | Well, its installing on top of Ubuntu on bare metal, so you mean just reinstall ubuntu? | 15:38 |
| evrardjp | odyssey4me: hey you said to me that the instances that fail for disk space from infra are on rax public cloud, and their flavor is performance1-8 ? | 15:38 |
| odyssey4me | RossSchulman[m] AIO's are often better done on VM's or instances for exactly that reason | 15:38 |
| odyssey4me | evrardjp yes | 15:38 |
| evrardjp | odyssey4me: I can deny that. | 15:39 |
| evrardjp | I mean it's something else | 15:39 |
| odyssey4me | evrardjp ORLY? | 15:39 |
| evrardjp | that is flavor list: | 15:39 |
| evrardjp | | performance1-8 | 8 GB Performance | 8192 | 40 | 80 | 8 | N/A | | 15:39 |
| *** savvas_ has quit IRC | 15:39 | |
| evrardjp | the diskspace is 80 | 15:39 |
| evrardjp | where what I see on the last failed instance, is that the xvde disk is 50 GB. | 15:39 |
| RossSchulman[m] | odyssey4me: Got it, makes sense. Thanks. | 15:40 |
| evrardjp | with 24 GB assigned to /openstack it reduces the size for the btrfs | 15:40 |
| evrardjp | 60GB* | 15:40 |
| evrardjp | however, weird thing, we have something that gathers the disk usage on the node info | 15:40 |
| evrardjp | and that doesn't show 100% | 15:40 |
| odyssey4me | RossSchulman[m] A good test environment is one that can be reset to nothing quickly and easily. | 15:41 |
| *** flemingo has quit IRC | 15:41 | |
| odyssey4me | evrardjp yep, that's one o fthe reasons why that's been so hard to figure out | 15:42 |
| *** 3NAAAIX4G has quit IRC | 15:42 | |
| RossSchulman[m] | odyssey4me: For sure. I was hoping to grow out from the AIO into a cluster which is why I didn't use containers. | 15:42 |
| RossSchulman[m] | But that's probably taking the wrong approach. | 15:42 |
| idlemind | hey team; new one for the day. running stable/pike on centos7. updated my git repo to latest (stable/pike still) and ran setup-everything. it's choking on os-keystone-install.yml at "Create DB for service" saying invalid credentials. debug looks like it's using root username but i can't confirm the root password it's excluded from logs. deleted the container and recreated it to no avail. logged into the container and | 15:43 |
| idlemind | mysql command line is able to connect. not sure were to go | 15:43 |
| odyssey4me | RossSchulman[m] it can be done, although the default AIO setup needs some pretty serious adjustment for networking to get there | 15:43 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible master: Test the main repo with next version of ansible https://review.openstack.org/557820 | 15:44 |
| RossSchulman[m] | odyssey4me: Understood. So if I only have the one server right now but want to get started, am I better served by just doing the control plane ansible stuff? This server is going to be the controller. | 15:45 |
| evrardjp | odyssey4me: I think it's a spike of disk usage that exceed the pool of disk and get discarded | 15:46 |
| evrardjp | But I need one example to test | 15:46 |
| evrardjp | also I lost the one that was failing with what I said, to show you :( | 15:47 |
| admin0 | RossSchulman[m], what you can do is create a few KVM images interally .. say 1 controller , 1 log and 1 storage .. and then use the base as nova compute | 15:48 |
| admin0 | that way, you can move/migrate compute to another machine | 15:48 |
| *** savvas has joined #openstack-ansible | 15:48 | |
| *** savvas has quit IRC | 15:48 | |
| *** savvas has joined #openstack-ansible | 15:48 | |
| odyssey4me | RossSchulman[m] one option is what admin0 suggested, another is to do the AIO and include it as a compute host, but make sure you're setting up the networking yourself and not using the developer AIO which uses a dummy bridge | 15:49 |
| RossSchulman[m] | Would I need to take controller/log/storage out of KVM later or can they just live there? | 15:50 |
| *** flemingo has joined #openstack-ansible | 15:51 | |
| admin0 | RossSchulman[m], when you got a real controller later, add the 2nd controller, run the playbooks, then delete the 1st one | 15:52 |
| odyssey4me | RossSchulman[m] if you go the KVM route, then each VM is a host which you manage on your own, so it's up to you | 15:52 |
| admin0 | or just rsync the whole thing :) | 15:52 |
| admin0 | dd | 15:52 |
| admin0 | ssh rsync nc | 15:52 |
| admin0 | lots of ways to go | 15:52 |
| *** KeithMnemonic has joined #openstack-ansible | 15:53 | |
| odyssey4me | if you transition a VM to a real host, then as admin0 suggests there are many options - we try to treat hosts as disposable, within reason | 15:53 |
| odyssey4me | one caution though is that a VM as a compute host doesn't perform all that well unless you can do nested virt | 15:54 |
| odyssey4me | it's fine for a test system though | 15:54 |
| *** flemingo has quit IRC | 15:55 | |
| *** Sha000000 has joined #openstack-ansible | 15:57 | |
| *** gkadam_ has joined #openstack-ansible | 16:03 | |
| admin0 | RossSchulman[m], what is your system specs ? | 16:04 |
| *** Sha000000 has quit IRC | 16:04 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-plugins master: Allow remote user to "become" when managing containers https://review.openstack.org/557901 | 16:05 |
| *** gkadam__ has joined #openstack-ansible | 16:05 | |
| *** gkadam has quit IRC | 16:07 | |
| *** gkadam_ has quit IRC | 16:09 | |
| *** flemingo has joined #openstack-ansible | 16:13 | |
| RossSchulman[m] | admin0: This one is 12 cpus @ 3.3 Ghz, 62GB RAM, and 250GB SSD | 16:16 |
| *** flemingo has quit IRC | 16:18 | |
| *** fatdragon has joined #openstack-ansible | 16:22 | |
| *** flemingo has joined #openstack-ansible | 16:37 | |
| *** savvas has quit IRC | 16:39 | |
| openstackgerrit | Merged openstack/openstack-ansible-openstack_hosts master: Add retries to requirement package install https://review.openstack.org/557943 | 16:41 |
| *** flemingo has quit IRC | 16:42 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible master: [WIP] Use dstat freespace plugin https://review.openstack.org/557965 | 16:43 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-openstack_hosts stable/queens: Add retries to requirement package install https://review.openstack.org/557966 | 16:43 |
| *** DanyC_ has joined #openstack-ansible | 16:44 | |
| *** DanyC has quit IRC | 16:44 | |
| odyssey4me | @cloudnull @evrardjp let's make https://review.openstack.org/557966 go too :) | 16:45 |
| evrardjp | already reviewed | 16:45 |
| evrardjp | if the dstat thing works and is not a problem, we should backport it to Q | 16:46 |
| odyssey4me | 'cos you're jedi | 16:46 |
| evrardjp | see what's going ther | 16:46 |
| evrardjp | e | 16:46 |
| odyssey4me | agreed | 16:46 |
| *** savvas has joined #openstack-ansible | 16:47 | |
| odyssey4me | cores, this has been waiting for some time and improves a greenfield deploy time: https://review.openstack.org/553746 | 16:50 |
| *** flemingo has joined #openstack-ansible | 16:50 | |
| RossSchulman[m] | Am I correct that the deployment workstation can't be a Mac? | 16:50 |
| evrardjp | RossSchulman[m]: yes | 16:51 |
| odyssey4me | RossSchulman[m] it possibly could, but you'd have to do your own ansible & config bootstrapping | 16:51 |
| evrardjp | in the docs the deploymenent workstation requirements are listed | 16:51 |
| evrardjp | just run a vm | 16:51 |
| evrardjp | :p | 16:51 |
| evrardjp | take a certain amount of ram if you're doing a large deploy | 16:51 |
| RossSchulman[m] | Yeah, will just VM. :) | 16:52 |
| evrardjp | RossSchulman[m]: alternatively use the first node of your infra | 16:52 |
| evrardjp | probably has more ram | 16:52 |
| *** savvas has quit IRC | 16:52 | |
| *** flemingo has quit IRC | 16:54 | |
| *** savvas has joined #openstack-ansible | 16:56 | |
| *** savvas has quit IRC | 17:01 | |
| *** gkadam__ has quit IRC | 17:03 | |
| idlemind | hey team; new one for the day. running stable/pike on centos7. updated my git repo to latest (stable/pike still) and ran setup-everything. it's choking on os-keystone-install.yml at "Create DB for service" saying invalid credentials. debug looks like it's using root username but i can't confirm the root password it's excluded from logs. deleted the container and recreated it to no avail. logged into the container and | 17:04 |
| idlemind | mysql command line is able to connect. not sure were to go | 17:04 |
| *** savvas has joined #openstack-ansible | 17:05 | |
| *** savvas_ has joined #openstack-ansible | 17:08 | |
| openstackgerrit | Merged openstack/openstack-ansible-openstack_hosts stable/queens: Add retries to requirement package install https://review.openstack.org/557966 | 17:08 |
| *** savvas has quit IRC | 17:10 | |
| *** flemingo has joined #openstack-ansible | 17:12 | |
| admin0 | evrardjp, i have the same issue again | 17:12 |
| odyssey4me | idlemind confirm that what's in the .my.cnf file which you used to connect to the DB matches the credentials supplied in your user_secrets.yml file | 17:13 |
| idlemind | no .my.cnf on the container in question | 17:13 |
| idlemind | should i create one | 17:13 |
| idlemind | or should it be the .my.cnf in the deployment host? | 17:13 |
| odyssey4me | no, the .my.cnf used is the one on the galera container - the db create tasks are delegated there | 17:14 |
| idlemind | ah k | 17:14 |
| idlemind | i'll check that | 17:14 |
| *** flemingo_ has joined #openstack-ansible | 17:14 | |
| admin0 | odyssey4me, https://gist.github.com/a1git/e8a4e53213394ceb0a59d95e132ed7a1 -- have you seen such issue before ? | 17:15 |
| odyssey4me | admin0 I think that might be what https://review.openstack.org/557805 is trying to solve | 17:17 |
| *** flemingo has quit IRC | 17:17 | |
| *** gkadam__ has joined #openstack-ansible | 17:18 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-lxc_hosts master: Add mount options for better machinectl performance https://review.openstack.org/557969 | 17:20 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible master: Add mount options for better machinectl performance https://review.openstack.org/557970 | 17:23 |
| admin0 | odyssey4me, if i just copy that file, paste it .. it will help me move ahead ? | 17:26 |
| odyssey4me | admin0 I dunno, my suggestion was that perhaps it would help | 17:27 |
| admin0 | do i need to rerun from setup-hosts | 17:28 |
| *** DanyC_ has quit IRC | 17:28 | |
| odyssey4me | admin0 that should only require a re-run of repo-install.yml | 17:28 |
| odyssey4me | although you may have to use the flags to force rebuilding wheels/venvs | 17:28 |
| admin0 | odyssey4me, never done that . how :) | 17:31 |
| odyssey4me | in https://github.com/openstack/openstack-ansible-repo_build/blob/master/defaults/main.yml you'll find notes about it: openstack-ansible repo-build.yml -e repo_build_git_reclone=yes -e repo_build_wheel_rebuild=yes -e repo_build_venv_rebuild | 17:33 |
| odyssey4me | =yes | 17:33 |
| odyssey4me | with those you can force each stage of the repo build to be done, regardless of what it thinks is ok | 17:34 |
| *** jayant has joined #openstack-ansible | 17:34 | |
| *** osnaya has joined #openstack-ansible | 17:35 | |
| *** Sha0000 has joined #openstack-ansible | 17:36 | |
| *** dave-mccowan has quit IRC | 17:39 | |
| *** dave-mccowan has joined #openstack-ansible | 17:39 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/ansible-role-systemd_mount master: Change mount states to ensure uptime https://review.openstack.org/557973 | 17:42 |
| openstackgerrit | Merged openstack/openstack-ansible-os_tempest stable/queens: update for changes needed to pass refstack https://review.openstack.org/552905 | 17:43 |
| *** savvas has joined #openstack-ansible | 17:43 | |
| admin0 | the only change is the version .. doing that and will rerun | 17:43 |
| *** savvas_ has quit IRC | 17:46 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/ansible-role-systemd_mount master: Change mount states to ensure uptime https://review.openstack.org/557973 | 17:49 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/ansible-role-systemd_mount master: Change mount states to ensure uptime https://review.openstack.org/557973 | 17:50 |
| openstackgerrit | Merged openstack/openstack-ansible-os_tempest stable/pike: update for changes needed to pass refstack https://review.openstack.org/552906 | 17:52 |
| admin0 | will this also fix this ? | 17:53 |
| *** savvas_ has joined #openstack-ansible | 17:53 | |
| *** savvas has quit IRC | 17:57 | |
| *** EmilienM has quit IRC | 17:58 | |
| *** jayant has quit IRC | 17:59 | |
| *** Sha0000 has quit IRC | 18:01 | |
| *** sawblade6 has joined #openstack-ansible | 18:08 | |
| *** sawblade_ has quit IRC | 18:11 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-lxc_hosts master: Add mount options for better machinectl performance https://review.openstack.org/557969 | 18:12 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/ansible-role-systemd_mount master: Change mount states to ensure uptime https://review.openstack.org/557973 | 18:18 |
| idlemind | odyssey4me the .my.cnf on my galera container does have the correct root password per my secrets file and running mysql -u root while attached to it seems to work w/o issue (show databases shows my db's) | 18:24 |
| odyssey4me | idlemind odd, nothing else comes to mind right now | 18:24 |
| idlemind | k thx for at least giving me something to check | 18:25 |
| jrosser | cloudnull: are there similar uptime issues with the networkd role? we had an inadvertent "service networking restart" across all our hosts which was fairly interesting to unpick the mess afterwards | 18:25 |
| cloudnull | jrosser: there could be. | 18:26 |
| osnaya | @cloudnull @admin0 @evrardjp @odyssey4me @idlemind - clarification needed on physical network/vlan please. please see http://paste.openstack.org/show/718034/. for details | 18:26 |
| cloudnull | are you using the networkd role for your host networks? | 18:26 |
| jrosser | the end result of that was adding one-shot behavoiur to our network role so that you couldnt have that kind of accident | 18:26 |
| jrosser | no we have something that manages the standard ubuntu networking | 18:27 |
| cloudnull | oh ok. | 18:27 |
| cloudnull | with networkd "service networking restart" would be "service systemd-networkd restart" | 18:27 |
| jrosser | but very similar inadvertant restart -> disaster potential i guess | 18:27 |
| cloudnull | a restart of networkd would not take down your existing networks | 18:27 |
| cloudnull | it would add things to your network but not modify them | 18:28 |
| cloudnull | s/not modify/harm/ | 18:28 |
| idlemind | osnaya https://docs.openstack.org/project-deploy-guide/openstack-ansible/newton/app-networking.html#network-appendix | 18:28 |
| jrosser | cloudnull: https://github.com/systemd/systemd/issues/3031 ? | 18:28 |
| idlemind | the network diagrams section may be helpful ... it's recommended to have the deployment host on your equivalent of br-mgmt but as long as access between it and br-mgmt doesn't block anything and is routable you shouldn't have problems | 18:29 |
| idlemind | for me; i target the br-mgmt ip in my definition's which in your case appear to be 200.x.x.x | 18:30 |
| cloudnull | jrosser: there's no "networkctl $VERB #ETH" yet. | 18:31 |
| cloudnull | though that would be nice to have | 18:31 |
| jrosser | yeah, so currently do you always have to restart networkd on any change, and does that bounce everything you've got configured? | 18:32 |
| cloudnull | yes we have to restart it, but no it does not bounce existing interfaces. | 18:33 |
| idlemind | odyssey4me login_host is point to haproxy ... me wonders if haproxy thinks mysql is down ... i seem to remember some issues w/that in the past | 18:33 |
| osnaya | @idlemind I am referring to the diagram. Need clarification if infra1's and compute1's physical 10.xx IP or their corresponding VLAN10 IP's to be given. and if deployment host needs to be on the same VLAN as br-mgmt? The network diag doesn't show a separate deployment host. So need clarification | 18:33 |
| odyssey4me | idlemind that's a possibility, and easy enough to discover with hatop - I forget how to use it though | 18:33 |
| idlemind | ya i turned on the web ui lol | 18:33 |
| jrosser | cloudnull: have you experimented much with using networkd to go from config A -> B -> C... where B might involve rearranging the interfaces created in A, and so on | 18:34 |
| *** mrhillsman has joined #openstack-ansible | 18:34 | |
| cloudnull | its a bit more work on the administrators end, but if i had an interface, like br-thing, and wanted to change the config for it I have two options: one use `ip` commands to add the config i want and then write my changes to the appropriate network file, or two write the changes I want to the network file, delete the interface and restart networkd | 18:34 |
| cloudnull | when Ive had to make changes to a network interface I tend to add the changes I want with `ip` commands and then update the network files so the changes I'm making are persistent. | 18:37 |
| odyssey4me | sounds painful | 18:37 |
| odyssey4me | yet not much different to using an interfaces file | 18:37 |
| cloudnull | yup | 18:38 |
| jrosser | so on ubuntu systems we've had to modify the network restart script a bit | 18:38 |
| jrosser | so that it does ifdown -a with the old config, slips in the new config then does ifup -a | 18:38 |
| jrosser | otherwise it tries to take down non-existant interfaces in the new config when you restart the service | 18:38 |
| cloudnull | odyssey4me: a lot of the networkd interactions are the same as before. its just the configs are now common between distros. | 18:38 |
| cloudnull | id say we're no better or worse off in terms of administrative overhead | 18:39 |
| cloudnull | just a lot less code to carry :) | 18:39 |
| idlemind | osnaya in your environment you've defined br-mgmt as the 200.x.x.x/24 network ... you didn't indicate anything for this 10.140.x.x network so based solely on your pastebin ... the vip ip's should be the 200.x.x.x ip and the ip's in the other areas shoud be the 200.x.x.x ip's ... that said it doesn't mean it won't work as long as the deployment host can reach the 10.x.140.x ip's it just might not be as "logically clear" | 18:39 |
| cloudnull | jrosser: in your case you wouldnt have to worry much about that | 18:40 |
| idlemind | osnaya do you intend for 200.x.x.x to be the equivalent of br-mgmt in the architecture drawings? | 18:40 |
| cloudnull | with networkd | 18:40 |
| *** strobelight has quit IRC | 18:40 | |
| *** DanyC has joined #openstack-ansible | 18:40 | |
| cloudnull | because a service restart wouldnt take down the existing interfaces. | 18:42 |
| osnaya | @idlemind yes br-mgmt is with VLAN10 (200.xx IPs). Currently my deployment host is not VLAN 10, so had a doubt, if it needs to be. Currently deployment and target hosts can communicate over 10.xx network and only target hosts are on VLAN10 (not deployment host... so wondering how it will communicate with any containers, since containers addresses from inventory are all on 200.xx range) | 18:43 |
| jrosser | cloudnull: yes adding new interfaces certainly sounds less disruptive with networkd, but changing an existing one maybe less easy | 18:44 |
| cloudnull | yea, you'd have to set the link down (or delete it) then restart networkd | 18:45 |
| idlemind | segway into this other convo about networkd usage ... seems like a good idea to standardize something as critical as network configuration across platforms especially as the project adds more supported platforms a lot less weight to carry for each distro added | 18:45 |
| cloudnull | idlemind: https://review.openstack.org/#/c/552721/ :) | 18:45 |
| cloudnull | its only for test. but once that pattern is worked out we can get a better overall example in place for prod | 18:46 |
| * cloudnull just needs the integrated gate to be happy again | 18:46 | |
| idlemind | cloudnull cool i'll look at it. i know i hit some issues w/veth pairs on centos | 18:46 |
| jrosser | i think that expectations need to be managed a bit though - expecting to update/change config after you've deployed is full of traps | 18:46 |
| odyssey4me | idlemind yep, that's the plan - less tech debt, more consistency - it makes it much easier for the community to help each other regardless of platform when the underlying tech is the same | 18:47 |
| idlemind | osnaya ya that's why it's suggested to put the deployment host on br-mgmt or ensure you have connectivity from the deployment host to br-mgmt | 18:47 |
| jrosser | ^ with whatever tool manages it | 18:47 |
| cloudnull | jrosser: totally agree. | 18:47 |
| cloudnull | we may be able to do some more fencing in the networkd role, and create something that allows for smart reloads? | 18:48 |
| cloudnull | I've not done much on that yet. | 18:48 |
| jrosser | i would certinaly thing about it as we effectively blew away an entire cloud with a networking restart | 18:48 |
| logan- | jrosser: yeah networkd is not very idempotent right now, thats my biggest issue | 18:49 |
| cloudnull | I am however using the networkd role to manage a couple smaller pieces of infra, so I can say it works as-is even if it may have some sharp edges atm. | 18:50 |
| *** savvas_ has quit IRC | 18:50 | |
| jrosser | oh yes it's all good :) i was just interested in seeing if you'd tackled some of the issues we've faced | 18:50 |
| cloudnull | not yet, but hopefully soon :) | 18:51 |
| *** savvas has joined #openstack-ansible | 18:52 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-tests master: Do not generate the html ARA report any more https://review.openstack.org/557921 | 18:55 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-lxc_hosts master: Add mount options for better machinectl performance https://review.openstack.org/557969 | 18:57 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/ansible-role-systemd_mount master: Change mount states to ensure uptime https://review.openstack.org/557973 | 18:59 |
| jrosser | osnaya: you can indeed have your deploy host on the same untagged network as the rest of your physical hosts, and the mgmt network can be a seperate vlan. This is exactly what i do. | 18:59 |
| jrosser | The deploy host does not have to be on the mgmt network, the docs just suggest this for simplicity (i guess) | 19:00 |
| *** savvas has quit IRC | 19:03 | |
| *** savvas has joined #openstack-ansible | 19:04 | |
| *** osnaya has quit IRC | 19:08 | |
| *** savvas has quit IRC | 19:08 | |
| *** throwsb1 has quit IRC | 19:11 | |
| *** savvas has joined #openstack-ansible | 19:11 | |
| openstackgerrit | Merged openstack/openstack-ansible-tests master: Disable developer mode for tempest role tests https://review.openstack.org/557720 | 19:13 |
| *** savvas has quit IRC | 19:15 | |
| idlemind | grr ... on galera container ... clustercheck gives an everything good http 200 but telnet localhost 9200 says percona xtradb cluster node is not synced | 19:20 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-tests stable/queens: Disable developer mode for tempest role tests https://review.openstack.org/557988 | 19:22 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-tests stable/pike: Disable developer mode for tempest role tests https://review.openstack.org/557989 | 19:23 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-tests stable/ocata: Disable developer mode for tempest role tests https://review.openstack.org/557990 | 19:24 |
| *** savvas has joined #openstack-ansible | 19:29 | |
| *** gkadam__ has quit IRC | 19:31 | |
| *** savvas has quit IRC | 19:34 | |
| *** DanyC has quit IRC | 19:35 | |
| *** fatdragon has quit IRC | 19:37 | |
| *** savvas has joined #openstack-ansible | 19:39 | |
| *** savvas has quit IRC | 19:43 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/ansible-role-systemd_mount master: Change mount states to ensure uptime https://review.openstack.org/557973 | 19:47 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-lxc_hosts master: Add mount options for better machinectl performance https://review.openstack.org/557969 | 19:50 |
| *** savvas has joined #openstack-ansible | 19:57 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_keystone stable/queens: Only try to stop services that exist https://review.openstack.org/553746 | 19:58 |
| idlemind | hey there champ ... how do i force my osa to pull down new copies of repo's based on a sha bump? the bug i'm hitting with mysql is fixed in the sha listed in ansible-role-requirements but it doesn't seem to be applying | 19:58 |
| odyssey4me | idlemind if you change a SHA/tag for the /optnopenstack-ansible git repo, then you should always do the following after: ./scripts/bootstrap-ansible.sh | 19:59 |
| odyssey4me | more here: https://docs.openstack.org/openstack-ansible/pike/user/minor-upgrade.html | 19:59 |
| idlemind | odyssey4me thx so it's safe to run bootstrap-ansible.sh to pick up the new code from the various sha bumps | 20:00 |
| idlemind | it won't mess up user secrets and stuff | 20:00 |
| odyssey4me | nope, it only affects the ansible runtime and the ansible roles - no user settings | 20:00 |
| idlemind | cool cool; guess i never thought of it as a minor upgrade but that's what it is | 20:02 |
| *** savvas has quit IRC | 20:02 | |
| idlemind | i'm betting that'll solve my issue lickity split | 20:02 |
| *** savvas has joined #openstack-ansible | 20:03 | |
| odyssey4me | I'm out for the w/end - cheerio all! | 20:08 |
| idlemind | lter thx | 20:09 |
| *** hamza21 has quit IRC | 20:12 | |
| idlemind | bootstrap-ansible is now failing to clone rsyslog_client, ansible-bird and ansible-ceph-mon (still in stable/pike) | 20:18 |
| cloudnull | idlemind: maybe github is having issues ? | 20:19 |
| *** savvas has quit IRC | 20:19 | |
| *** savvas has joined #openstack-ansible | 20:20 | |
| idlemind | the rysyslog_client one is looking at git.openstack.org and the commit it looks for seems to exist but it's complaining about a bunch of refs/tags not being valid | 20:20 |
| idlemind | other 2 are github ones | 20:20 |
| idlemind | the commits exist for sure in ansible-bird but is version in ansible for git cloning equal to a branch or tag? | 20:21 |
| idlemind | ok so version can be branch, tag or hash | 20:23 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible master: Add mount options for better machinectl performance https://review.openstack.org/557970 | 20:27 |
| cloudnull | idlemind: is the task failing? | 20:27 |
| idlemind | http://paste.openstack.org/show/718039/ | 20:28 |
| idlemind | Ya | 20:28 |
| idlemind | hmm seems a separate playbook for rsyslog_client went through (just a random one i wrote up from the logs) ... time to run bootstrap again | 20:31 |
| cloudnull | it kinda looks like its failing to fetch all of the REFs | 20:31 |
| cloudnull | though Im not sure why that would b e | 20:31 |
| *** armaan has quit IRC | 20:33 | |
| *** armaan has joined #openstack-ansible | 20:33 | |
| idlemind | hmm failed again in bootstrap ... | 20:35 |
| idlemind | does the bootstrap process pull in more of the repo than the git module in ansible? | 20:38 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-lxc_hosts master: Add mount options for better machinectl performance https://review.openstack.org/557969 | 20:38 |
| cloudnull | it will pull in the plugins | 20:39 |
| cloudnull | if they're already on the disk | 20:39 |
| cloudnull | you could flush the roles (rm -rf /etc/ansible/roles/*) | 20:39 |
| cloudnull | then rebootstrap | 20:39 |
| *** savvas has quit IRC | 20:40 | |
| cloudnull | if you think there's some conflict | 20:40 |
| *** savvas has joined #openstack-ansible | 20:40 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible master: Add mount options for better machinectl performance https://review.openstack.org/557970 | 20:41 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible stable/queens: Update repo_build role for PBR fix https://review.openstack.org/557805 | 20:42 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible master: Update run_tests.sh and remove tests-repo-clone.sh https://review.openstack.org/556994 | 20:42 |
| *** marst has quit IRC | 20:46 | |
| idlemind | cloudnull thx i'll try that | 20:55 |
| idlemind | cloudnull the wizard almighty | 20:58 |
| idlemind | where do i send the beer? | 20:58 |
| openstackgerrit | Shannon Mitchell proposed openstack/openstack-ansible-ops master: Issues with osa-ops mnaio service config with existing entries https://review.openstack.org/557999 | 21:15 |
| *** marst has joined #openstack-ansible | 21:15 | |
| cloudnull | idlemind: will you be at the summit? | 21:16 |
| * cloudnull forgets if we talked about this already | 21:17 | |
| *** armaan has quit IRC | 21:17 | |
| *** armaan has joined #openstack-ansible | 21:18 | |
| *** savvas has quit IRC | 21:21 | |
| bgmccollum | working on an upgrade from newton to ocata, following the upgrade guide. seeing lots of "No matching distribution found for..." during the bootstrap-ansible.sh run. should this not be the case? | 21:28 |
| cloudnull | bgmccollum: you seeing that when running the script ? | 21:29 |
| bgmccollum | yes... | 21:29 |
| bgmccollum | cloudnull -- http://paste.openstack.org/show/718042/ | 21:30 |
| cloudnull | ah - "No matching distribution found for Jinja2===2.8.1 (from -c https://git.openstack.org/cgit/openstack/requirements/plain/upper-constraints.txt?id=ca6ba8a28325681a87f7ca16954d2c2e72b6d61c (line 8))" ? | 21:31 |
| cloudnull | that should be ok | 21:31 |
| cloudnull | it doesn't look like the script failed | 21:32 |
| cloudnull | ending in "System is bootstrapped and ready for use.' | 21:32 |
| bgmccollum | the script didnt fail, but the warning were concerning... | 21:32 |
| bgmccollum | setuptools too | 21:32 |
| bgmccollum | OK to proceed? | 21:33 |
| cloudnull | yea, should be fine. it's just preping the venv that ansible is run within. if you want to be REALLY sure it's OK, you can nuke that directory (/opt/ansible-runtime) and rerun the bootstrap script . | 21:33 |
| bgmccollum | ill archive it and try that | 21:34 |
| admin0 | odyssey4me, i can confirm that, that change fixed my issue | 21:47 |
| *** admin0 has quit IRC | 22:02 | |
| *** admin0 has joined #openstack-ansible | 22:02 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible master: Add mount options for better machinectl performance https://review.openstack.org/557970 | 22:04 |
| admin0 | => ImportError: No module named prettytable | 22:05 |
| admin0 | -- how do I solve this ? | 22:05 |
| *** admin0 has quit IRC | 22:15 | |
| *** admin0 has joined #openstack-ansible | 22:16 | |
| *** vnogin has joined #openstack-ansible | 22:30 | |
| *** vnogin has quit IRC | 22:34 | |
| bgmccollum | upgrade successful...! | 22:39 |
| admin0 | bgmccollum, pike -> queens ? | 22:41 |
| bgmccollum | admin0 i wish...newton to ocata...baby steps. | 22:48 |
| *** admin0 has quit IRC | 22:51 | |
| logan- | cloudnull: if you dont get around to it i should be able to hack on the tests for your connection module stuff this weekend | 23:14 |
| *** strobelight has joined #openstack-ansible | 23:23 | |
| *** strobelight_ has joined #openstack-ansible | 23:24 | |
| *** strobelight is now known as Guest27941 | 23:24 | |
| *** strobelight_ is now known as strobelight | 23:25 | |
| *** Guest27941 has quit IRC | 23:28 | |
| *** Sha0000 has joined #openstack-ansible | 23:28 | |
| *** lbragstad has quit IRC | 23:41 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!