| Bofu2MBP | Just did a re-bootstrap of galera and the compute nodes are all up, but not showing up within horizon and the nodes aren't running. Any ideas on if there's some part of the playbook to run again to just kind of ... uh ... check everything that already is running? lol. | 00:00 |
|---|---|---|
| odyssey4me | Bofu2MBP a multi-node setup? do you have some sort of time sync setup? also, check if the rabbitmq cluster is ok | 00:06 |
| Bofu2MBP | rabbit says it's good | 00:06 |
| *** fawadkhaliq has quit IRC | 00:07 | |
| Bofu2MBP | this is a cluster that was (physically) offline for 3-4 days | 00:07 |
| odyssey4me | Bofu2MBP then you're down to logs | 00:07 |
| *** fawadkhaliq has joined #openstack-ansible | 00:07 | |
| Bofu2MBP | sec | 00:07 |
| Bofu2MBP | Timed out waiting for nova-conductor. Is it running? Or did this service start before nova-conductor? Reattempting establishment of nova-conductor connection... | 00:07 |
| odyssey4me | ok, check the conductor service log in the conductor container | 00:08 |
| Bofu2MBP | odyssey4me I don't see that container. | 00:08 |
| Bofu2MBP | sec | 00:08 |
| Bofu2MBP | sql connection failed. doh | 00:09 |
| Bofu2MBP | guess I didn't fix galera in it's entirety. hm | 00:09 |
| odyssey4me | it might have gone offline for too long, so you may have to restart all services that use the db | 00:10 |
| Bofu2MBP | boo | 00:15 |
| Bofu2MBP | easy way to do that using the playbooks by chance? lol | 00:15 |
| odyssey4me | Bofu2MBP a cheat would be to change a common var, like set verbose to false, then run setup-openstack, then back to true and run it again | 00:17 |
| odyssey4me | you could probably limit it to specific tags and to all_containers to reduce the time taken | 00:17 |
| odyssey4me | but yeah, something like that would be pretty useful | 00:18 |
| *** KLevenstein has quit IRC | 00:26 | |
| Bofu2MBP | odyssey4me looks like I got it | 00:38 |
| Bofu2MBP | just had to reboot rabbit a few times to get them back in sync | 00:38 |
| Bofu2MBP | the rest cascaded back into place | 00:38 |
| odyssey4me | Bofu2MBP good to hear - sadly rabbitmq is often the cause of issues | 00:42 |
| Bofu2MBP | odyssey4me now just trying to figure out why it won't let me make a network through horizon. :( | 00:42 |
| *** thorst has joined #openstack-ansible | 00:44 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible: Set SHA's for Mitaka RC1 https://review.openstack.org/300238 | 00:45 |
| odyssey4me | jmccrory automagically cloudnull d34dh0r53 stevelle mattt hughsaunders andymccr please babysit https://review.openstack.org/300238 until it merges, our last day to apply for a release and have it processed is tomorrow | 00:48 |
| odyssey4me | it's almost 2am for me - time to zzz | 00:49 |
| *** thorst has quit IRC | 00:51 | |
| openstackgerrit | Steve Lewis proposed openstack/openstack-ansible-os_swift: Add ceilometer middleware when ceilometer enabled https://review.openstack.org/300246 | 00:51 |
| odyssey4me | stevelle I liked your first idea, but I like that one better. :) | 00:53 |
| stevelle | it worked, this is more obvious. obvious is better. now go zzz odyssey4me | 00:53 |
| *** markvoelker_ has quit IRC | 00:53 | |
| odyssey4me | thanks stevelle - agreed :) | 00:56 |
| *** Bofu2MBP has quit IRC | 01:02 | |
| *** thorst has joined #openstack-ansible | 01:17 | |
| *** asettle has quit IRC | 01:19 | |
| *** thorst has quit IRC | 01:23 | |
| *** ChrisBenson has quit IRC | 01:33 | |
| *** fawadkhaliq has quit IRC | 01:40 | |
| *** thorst has joined #openstack-ansible | 01:42 | |
| *** thorst has quit IRC | 01:43 | |
| *** shanec_ has quit IRC | 01:49 | |
| *** saneax is now known as saneax_AFK | 01:53 | |
| *** fawadkhaliq has joined #openstack-ansible | 01:54 | |
| *** asettle has joined #openstack-ansible | 01:55 | |
| *** asettle has quit IRC | 02:01 | |
| *** asettle has joined #openstack-ansible | 02:23 | |
| *** fawadkhaliq has quit IRC | 02:33 | |
| *** asettle has quit IRC | 02:39 | |
| *** fawadkhaliq has joined #openstack-ansible | 02:42 | |
| *** fawadkhaliq has quit IRC | 02:43 | |
| *** asettle has joined #openstack-ansible | 02:45 | |
| *** fawadkhaliq has joined #openstack-ansible | 02:45 | |
| *** shanec_ has joined #openstack-ansible | 02:47 | |
| *** asettle has quit IRC | 02:50 | |
| *** fawadkhaliq has quit IRC | 02:53 | |
| *** fawadkhaliq has joined #openstack-ansible | 02:53 | |
| *** fawadkhaliq has quit IRC | 03:08 | |
| *** fawadkhaliq has joined #openstack-ansible | 03:10 | |
| *** fawadkhaliq has quit IRC | 03:12 | |
| *** krotscheck is now known as krotscheck_vaca | 03:20 | |
| *** fawadkhaliq has joined #openstack-ansible | 03:32 | |
| *** ChrisBenson has joined #openstack-ansible | 03:46 | |
| *** markvoelker has joined #openstack-ansible | 03:47 | |
| *** markvoelker_ has joined #openstack-ansible | 03:56 | |
| *** markvoelker has quit IRC | 03:56 | |
| *** asettle has joined #openstack-ansible | 04:17 | |
| *** asettle has quit IRC | 04:21 | |
| *** ChrisBenson has quit IRC | 04:22 | |
| *** fawadkhaliq has quit IRC | 04:26 | |
| *** sdake has joined #openstack-ansible | 04:29 | |
| *** javeriak has joined #openstack-ansible | 04:29 | |
| *** markvoelker_ has quit IRC | 04:33 | |
| *** markvoelker has joined #openstack-ansible | 04:33 | |
| *** shausy has joined #openstack-ansible | 04:37 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_ceilometer: Switch ceilometer to use keystone v3 https://review.openstack.org/300239 | 04:37 |
| openstackgerrit | Merged openstack/openstack-ansible-os_swift: Add ceilometer middleware when ceilometer enabled https://review.openstack.org/300246 | 04:38 |
| *** markvoelker has quit IRC | 04:38 | |
| *** rohanp_ has quit IRC | 04:40 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_swift: Clean up deprecation warning in swift proxy https://review.openstack.org/300153 | 04:40 |
| openstackgerrit | Merged openstack/openstack-ansible-pip_install: Add pip_get_pip_options to fall back Install PIP https://review.openstack.org/296114 | 04:56 |
| *** fawadkhaliq has joined #openstack-ansible | 04:57 | |
| *** arslan has joined #openstack-ansible | 05:01 | |
| *** arslan has quit IRC | 05:06 | |
| *** saneax_AFK is now known as saneax | 05:09 | |
| *** daneyon has quit IRC | 05:10 | |
| *** arslan has joined #openstack-ansible | 05:19 | |
| *** markvoelker has joined #openstack-ansible | 05:29 | |
| *** sdake_ has joined #openstack-ansible | 05:30 | |
| *** sdake_ has quit IRC | 05:31 | |
| *** sdake has quit IRC | 05:32 | |
| *** javeriak has quit IRC | 05:34 | |
| *** shausy has quit IRC | 05:35 | |
| *** shausy has joined #openstack-ansible | 05:36 | |
| *** markvoelker has quit IRC | 05:37 | |
| *** arslan has quit IRC | 05:38 | |
| *** shanec_ has quit IRC | 05:41 | |
| *** asettle has joined #openstack-ansible | 05:43 | |
| *** javeriak has joined #openstack-ansible | 05:43 | |
| *** shanec_ has joined #openstack-ansible | 05:46 | |
| *** shanec_ has quit IRC | 05:50 | |
| *** javeriak has quit IRC | 05:54 | |
| *** shanec_ has joined #openstack-ansible | 05:55 | |
| *** shanec_ has quit IRC | 05:59 | |
| *** arslan has joined #openstack-ansible | 06:01 | |
| *** shanec_ has joined #openstack-ansible | 06:04 | |
| *** ChrisBenson has joined #openstack-ansible | 06:06 | |
| Tebro | is the openstack AIO usable for real stuff? Other than just testing? | 06:06 |
| *** shanec_ has quit IRC | 06:08 | |
| *** mikelk has joined #openstack-ansible | 06:13 | |
| *** shanec_ has joined #openstack-ansible | 06:13 | |
| *** ChrisBenson has quit IRC | 06:16 | |
| *** shanec_ has quit IRC | 06:18 | |
| *** arslan has quit IRC | 06:19 | |
| *** javeriak has joined #openstack-ansible | 06:20 | |
| openstackgerrit | Merged openstack/openstack-ansible: Set SHA's for Mitaka RC1 https://review.openstack.org/300238 | 06:21 |
| openstackgerrit | Merged openstack/openstack-ansible: Add haproxy-install to setup-infrastructure.yml https://review.openstack.org/299507 | 06:21 |
| *** shanec_ has joined #openstack-ansible | 06:22 | |
| *** Mudpuppy has quit IRC | 06:23 | |
| *** Mudpuppy has joined #openstack-ansible | 06:24 | |
| *** shanec_ has quit IRC | 06:27 | |
| *** Mudpuppy has quit IRC | 06:28 | |
| *** mikelk has quit IRC | 06:29 | |
| *** shanec_ has joined #openstack-ansible | 06:32 | |
| *** asettle has quit IRC | 06:35 | |
| *** shanec_ has quit IRC | 06:36 | |
| *** sdake has joined #openstack-ansible | 06:36 | |
| *** fawadkhaliq has quit IRC | 06:38 | |
| *** fawadkhaliq has joined #openstack-ansible | 06:38 | |
| *** asettle has joined #openstack-ansible | 06:39 | |
| *** shanec_ has joined #openstack-ansible | 06:41 | |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-os_neutron: Remove iptables checksum rule script https://review.openstack.org/293067 | 06:43 |
| *** shanec_ has quit IRC | 06:45 | |
| *** shanec_ has joined #openstack-ansible | 06:50 | |
| *** asettle has quit IRC | 06:52 | |
| *** Mudpuppy has joined #openstack-ansible | 06:54 | |
| *** shanec_ has quit IRC | 06:54 | |
| *** asettle has joined #openstack-ansible | 06:57 | |
| *** shanec_ has joined #openstack-ansible | 06:59 | |
| *** Mudpuppy has quit IRC | 06:59 | |
| *** javeriak has quit IRC | 07:02 | |
| *** asettle has quit IRC | 07:02 | |
| *** shanec_ has quit IRC | 07:04 | |
| *** phiche1 has joined #openstack-ansible | 07:08 | |
| *** shanec_ has joined #openstack-ansible | 07:08 | |
| *** admin0 has joined #openstack-ansible | 07:10 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_swift: Include reseller admin role correctly https://review.openstack.org/300216 | 07:11 |
| *** phiche has quit IRC | 07:11 | |
| *** shanec_ has quit IRC | 07:13 | |
| *** javeriak has joined #openstack-ansible | 07:13 | |
| *** asettle has joined #openstack-ansible | 07:14 | |
| *** mgoddard has joined #openstack-ansible | 07:14 | |
| *** asettle has quit IRC | 07:14 | |
| *** jamielennox is now known as jamielennox|away | 07:17 | |
| *** shanec_ has joined #openstack-ansible | 07:17 | |
| *** unlaudable has joined #openstack-ansible | 07:18 | |
| *** markvoelker has joined #openstack-ansible | 07:19 | |
| *** admin0 has quit IRC | 07:21 | |
| *** shanec_ has quit IRC | 07:22 | |
| *** admin0 has joined #openstack-ansible | 07:23 | |
| *** javeriak has quit IRC | 07:24 | |
| *** pcaruana has joined #openstack-ansible | 07:25 | |
| *** markvoelker has quit IRC | 07:25 | |
| *** shanec_ has joined #openstack-ansible | 07:27 | |
| *** shanec_ has quit IRC | 07:31 | |
| *** shanec_ has joined #openstack-ansible | 07:36 | |
| odyssey4me | Tebro If you'd like an April's fool's joke, then yes. | 07:36 |
| Tebro | :D | 07:37 |
| admin0 | morning | 07:37 |
| odyssey4me | The truth is that while everything used to execute an AIO, barring the AIO bootstrap, is usable for production. The fact is that a production environment requires at least 3 controllers to ensure that Galera & Rabbit clusters have tie-breakers for the HA setup which we implement. | 07:38 |
| Tebro | Alright. | 07:39 |
| *** shanec_ has quit IRC | 07:40 | |
| *** shanec_ has joined #openstack-ansible | 07:45 | |
| *** shanec_ has quit IRC | 07:49 | |
| openstackgerrit | Merged openstack/openstack-ansible: Ensure the OpenStack gate has access to the logs https://review.openstack.org/299571 | 07:50 |
| *** karimb has joined #openstack-ansible | 07:52 | |
| *** shanec_ has joined #openstack-ansible | 07:54 | |
| *** fawadkhaliq has quit IRC | 07:56 | |
| *** shanec_ has quit IRC | 07:58 | |
| *** shanec_ has joined #openstack-ansible | 08:03 | |
| *** mikelk has joined #openstack-ansible | 08:08 | |
| *** shanec_ has quit IRC | 08:08 | |
| *** shanec_ has joined #openstack-ansible | 08:13 | |
| admin0 | logan-: here ? | 08:15 |
| *** shanec_ has quit IRC | 08:17 | |
| *** admin0_ has joined #openstack-ansible | 08:21 | |
| *** admin0 has quit IRC | 08:22 | |
| *** admin0_ is now known as admin0 | 08:22 | |
| *** shanec_ has joined #openstack-ansible | 08:22 | |
| *** phiche has joined #openstack-ansible | 08:24 | |
| *** iceyao has joined #openstack-ansible | 08:26 | |
| *** phiche1 has quit IRC | 08:26 | |
| *** shanec_ has quit IRC | 08:26 | |
| *** shanec_ has joined #openstack-ansible | 08:31 | |
| *** javeriak has joined #openstack-ansible | 08:35 | |
| *** shanec_ has quit IRC | 08:35 | |
| *** javeriak has quit IRC | 08:40 | |
| *** shanec_ has joined #openstack-ansible | 08:40 | |
| openstackgerrit | git-harry proposed openstack/openstack-ansible: Remove clustering config from rabbitmq.config https://review.openstack.org/299836 | 08:41 |
| *** shanec_ has quit IRC | 08:44 | |
| *** shanec_ has joined #openstack-ansible | 08:49 | |
| *** kmARC has quit IRC | 08:50 | |
| *** kmARC has joined #openstack-ansible | 08:50 | |
| *** phiche1 has joined #openstack-ansible | 08:53 | |
| *** shanec_ has quit IRC | 08:54 | |
| odyssey4me | jmccrory automagically cloudnull d34dh0r53 stevelle mattt hughsaunders andymccr FYI as we are tagged as an independent release project, today is the last day for us to tag a release before the summit, and we don't get to do RC's... so hold by beer and watch this: https://review.openstack.org/300355 | 08:55 |
| * hughsaunders holds beer | 08:55 | |
| *** phiche has quit IRC | 08:55 | |
| odyssey4me | is gerrit being flaky, or do I have a bad connection? | 08:57 |
| mattt | releasing today eh? nice! | 08:58 |
| *** shanec_ has joined #openstack-ansible | 08:58 | |
| odyssey4me | yeah, well - what that does mean is that we'll have to do a bit of backporting and such in the next few weeks as we iron out any more testing and issues we find so we can tag 13.0.1 before the summit | 09:00 |
| *** shanec_ has quit IRC | 09:03 | |
| *** shanec_ has joined #openstack-ansible | 09:07 | |
| *** markvoelker has joined #openstack-ansible | 09:11 | |
| *** shanec_ has quit IRC | 09:12 | |
| *** sdake has quit IRC | 09:13 | |
| *** markvoelker has quit IRC | 09:16 | |
| *** shanec_ has joined #openstack-ansible | 09:17 | |
| *** shanec_ has quit IRC | 09:21 | |
| *** shanec_ has joined #openstack-ansible | 09:26 | |
| *** shanec_ has quit IRC | 09:30 | |
| *** Oku_OS has joined #openstack-ansible | 09:31 | |
| *** Oku_OS_ has joined #openstack-ansible | 09:32 | |
| *** shanec_ has joined #openstack-ansible | 09:35 | |
| *** Oku_OS has quit IRC | 09:36 | |
| *** shanec_ has quit IRC | 09:39 | |
| *** shanec_ has joined #openstack-ansible | 09:44 | |
| *** shanec_ has quit IRC | 09:49 | |
| *** shanec_ has joined #openstack-ansible | 09:53 | |
| *** admin0 has quit IRC | 09:56 | |
| *** shanec_ has quit IRC | 09:58 | |
| *** admin0 has joined #openstack-ansible | 09:59 | |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-os_nova: Update test dependencies https://review.openstack.org/300393 | 09:59 |
| *** shausy has quit IRC | 10:01 | |
| *** shanec_ has joined #openstack-ansible | 10:03 | |
| *** iceyao has quit IRC | 10:05 | |
| *** markvoelker has joined #openstack-ansible | 10:06 | |
| *** shanec_ has quit IRC | 10:07 | |
| *** shanec_ has joined #openstack-ansible | 10:12 | |
| *** markvoelker has quit IRC | 10:12 | |
| *** admin0 has quit IRC | 10:12 | |
| *** shanec_ has quit IRC | 10:16 | |
| *** admin0 has joined #openstack-ansible | 10:17 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_keystone: removed duplicate key https://review.openstack.org/299760 | 10:18 |
| *** shanec_ has joined #openstack-ansible | 10:21 | |
| *** shanec_ has quit IRC | 10:25 | |
| openstackgerrit | Merged openstack/openstack-ansible-galera_server: removed duplicate key https://review.openstack.org/299759 | 10:27 |
| *** shanec_ has joined #openstack-ansible | 10:30 | |
| *** shanec_ has quit IRC | 10:35 | |
| *** kong has quit IRC | 10:36 | |
| *** phiche1 has quit IRC | 10:39 | |
| *** shanec_ has joined #openstack-ansible | 10:39 | |
| *** iceyao has joined #openstack-ansible | 10:41 | |
| *** shanec_ has quit IRC | 10:43 | |
| *** shanec_ has joined #openstack-ansible | 10:48 | |
| *** javeriak has joined #openstack-ansible | 10:49 | |
| *** shanec_ has quit IRC | 10:53 | |
| *** admin0 has quit IRC | 10:57 | |
| *** shanec_ has joined #openstack-ansible | 10:57 | |
| openstackgerrit | Merged openstack/openstack-ansible: Remove hard-coded pip indexes from repo-build playbook https://review.openstack.org/299656 | 10:58 |
| openstackgerrit | Merged openstack/openstack-ansible: removed duplicate key https://review.openstack.org/299762 | 10:58 |
| *** admin0 has joined #openstack-ansible | 11:02 | |
| *** markvoelker has joined #openstack-ansible | 11:02 | |
| *** shanec_ has quit IRC | 11:02 | |
| *** markvoelker has quit IRC | 11:06 | |
| *** iceyao has quit IRC | 11:06 | |
| *** shanec_ has joined #openstack-ansible | 11:07 | |
| *** shanec_ has quit IRC | 11:11 | |
| *** shanec_ has joined #openstack-ansible | 11:16 | |
| openstackgerrit | Hugh Saunders proposed openstack/openstack-ansible-plugins: Add allocation pool support to neutron plugin https://review.openstack.org/300421 | 11:16 |
| *** iceyao has joined #openstack-ansible | 11:17 | |
| openstackgerrit | Hugh Saunders proposed openstack/openstack-ansible-plugins: Add allocation pool support to neutron module https://review.openstack.org/300421 | 11:19 |
| *** shanec_ has quit IRC | 11:20 | |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-os_neutron: Neutron test cleanup https://review.openstack.org/300423 | 11:20 |
| *** phiche has joined #openstack-ansible | 11:22 | |
| *** shanec_ has joined #openstack-ansible | 11:25 | |
| *** shanec_ has quit IRC | 11:29 | |
| *** weshay has quit IRC | 11:30 | |
| *** Bofu2U has joined #openstack-ansible | 11:31 | |
| *** javeriak has quit IRC | 11:31 | |
| *** shanec_ has joined #openstack-ansible | 11:34 | |
| *** thorst has joined #openstack-ansible | 11:35 | |
| *** shanec_ has quit IRC | 11:39 | |
| winggundamth | hi. I'm trying to deploy swift on top of working OSA infra. I'm using openstack-ansible --forks 20 os-swift-install.yml but it didn't deploy swift proxy. | 11:43 |
| *** shanec_ has joined #openstack-ansible | 11:43 | |
| Bofu2U | did you define the nodes for swift in your config | 11:44 |
| odyssey4me | winggundamth yeah, you need to define these groups in your openstack_user_config or in a file in conf.d: https://github.com/openstack/openstack-ansible/blob/master/etc/openstack_deploy/conf.d/swift.yml.aio | 11:45 |
| odyssey4me | the groups swift-procy_hosts and swift_hosts are both required required and need hosts allocated | 11:45 |
| *** neilus has quit IRC | 11:46 | |
| winggundamth | swift-proxy_hosts or swift_proxy-hosts? | 11:47 |
| Bofu2U | odyssey4me thanks for your assistance last night btw | 11:47 |
| odyssey4me | it's covered quite extensively in http://docs.openstack.org/developer/openstack-ansible/install-guide/configure-swift.html | 11:47 |
| winggundamth | from swift.yml.example it used # Level: swift_proxy-hosts (required) | 11:47 |
| winggundamth | so I follow that | 11:48 |
| odyssey4me | winggundamth https://github.com/openstack/openstack-ansible/blob/master/etc/openstack_deploy/conf.d/swift.yml.aio#L17 is what's used for gate testing, so I would believe that - have you found it documented differently elsewhere? | 11:48 |
| *** shanec_ has quit IRC | 11:48 | |
| odyssey4me | Bofu2U sure, no problem :) | 11:48 |
| winggundamth | odyssey4me: https://github.com/openstack/openstack-ansible/blob/master/etc/openstack_deploy/conf.d/swift.yml.example#L313 | 11:49 |
| *** retreved has joined #openstack-ansible | 11:50 | |
| mhayden | mornin' | 11:52 |
| winggundamth | odyseey4me: after I change it to what you suggested so it works! | 11:52 |
| * winggundamth reading http://docs.openstack.org/contributor-guide/ while run OSA | 11:52 | |
| *** shanec_ has joined #openstack-ansible | 11:53 | |
| odyssey4me | winggundamth ah, that'd be a typo - would you be able to submit a patch to fix that? | 11:53 |
| * odyssey4me rages as duplicated documentation | 11:53 | |
| odyssey4me | perhaps rage is a little strong | 11:54 |
| * odyssey4me shakes a fist at duplicated documentation | 11:54 | |
| winggundamth | odyseey4me: I just got mentor from Rackspace Australia today and she gave me contributor guide so I'll start reading it and try to submit a patch :) | 11:55 |
| *** pjm6 has joined #openstack-ansible | 11:55 | |
| odyssey4me | winggundamth awesome :) we look forward to many patches then :D | 11:55 |
| pjm6 | good morning guys :) | 11:56 |
| winggundamth | lol | 11:56 |
| winggundamth | evening to you all too :P | 11:56 |
| logan- | admin0: I am now | 11:56 |
| odyssey4me | winggundamth what you can possibly do as a starting point is file a bug for anything you find wrong, and if you discover how to fix it then add a comment about how you fixed it | 11:56 |
| odyssey4me | then at least someone can pick that up and patch it | 11:57 |
| admin0 | you mentioned something about rwx authenticaiton on ceph .. i want to update the blog .. https://www.openstackfaq.com/openstack-ansible-ceph/ — can you tell me the change :) | 11:57 |
| admin0 | so that its compatible with future :) | 11:57 |
| odyssey4me | it also acts as a reminder list for you when you get to the point of submitting your own patches | 11:57 |
| *** shanec_ has quit IRC | 11:57 | |
| logan- | yeah. CFA: Basement MMR Cage 27 Rack: 103.07 Shelf: MP75 Ports: 37-38 | 11:59 |
| logan- | er.. clipboard :( | 11:59 |
| admin0 | :D | 11:59 |
| winggundamth | odyseey4me: so I should start to submit a bug from here right? https://bugs.launchpad.net/openstack-ansible | 11:59 |
| admin0 | now we know where you mine bitcoins :) | 11:59 |
| logan- | https://www.openstackfaq.com/ceph-setup-ubuntu-three-node/ on there where you add teh user accounts | 11:59 |
| admin0 | ‘allow rwx’ ? | 12:00 |
| logan- | set the auth caps on client.cinder to rwx for the glance-images pool | 12:00 |
| admin0 | ok | 12:00 |
| logan- | at the very end of the line | 12:00 |
| odyssey4me | admin0 fyi, all the IRC discussions in this channel are available for reference in http://eavesdrop.openstack.org/irclogs/%23openstack-ansible/ | 12:00 |
| odyssey4me | winggundamth yep, that's correct | 12:01 |
| * winggundamth report a bug | 12:01 | |
| logan- | nova will be using those permissions to write snapshots to the glance pool in Mitaka | 12:02 |
| *** shanec_ has joined #openstack-ansible | 12:02 | |
| winggundamth | https://bugs.launchpad.net/openstack-ansible/+bug/1564873 | 12:05 |
| openstack | Launchpad bug 1564873 in openstack-ansible "typo on swift.yml.example" [Undecided,New] | 12:05 |
| winggundamth | oh so it automatically shows in here | 12:05 |
| *** shanec_ has quit IRC | 12:06 | |
| odyssey4me | winggundamth yep :) spotz_zzz may pick that up, but if you're ready to submit a patch later today or this coming week and it's still not yet patches then you're welcome to go ahead and do so | 12:06 |
| winggundamth | odyssey4me: to submit a patch. I should follow here right? http://docs.openstack.org/contributor-guide/quickstart/first-timers.html#set-up-git-and-git-review | 12:09 |
| *** javeriak has joined #openstack-ansible | 12:10 | |
| *** ChanServ changes topic to "Launchpad: https://launchpad.net/openstack-ansible || Weekly Meetings: https://wiki.openstack.org/wiki/Meetings/openstack-ansible || Review Dashboard: https://goo.gl/tTmdgs" | 12:10 | |
| *** shanec_ has joined #openstack-ansible | 12:11 | |
| odyssey4me | winggundamth have you done the CLA mentioned in http://docs.openstack.org/contributor-guide/quickstart/first-timers.html#setting-up-for-contribution ? | 12:12 |
| winggundamth | yes. I have ssh, git and clone the code to my laptop | 12:12 |
| winggundamth | which branch I should checkout before doing change? | 12:12 |
| *** Mudpuppy has joined #openstack-ansible | 12:13 | |
| odyssey4me | winggundamth all changes start in the master branches, as the other named branches are considered 'stable' and changes to them need to fit into the backport rules | 12:13 |
| winggundamth | sound complicated. but I should focus on master branch first right? | 12:14 |
| odyssey4me | winggundamth yes, always | 12:14 |
| javeriak | odyssey4me is there any service of script in OSA that keeps track of the liveness of the target nodes? Something we can query? | 12:14 |
| odyssey4me | javeriak 'ansible all -m ping' ? | 12:15 |
| odyssey4me | other than that, no - monitoring is something that we don't specifically cover because everyone likes their own flavors of things | 12:15 |
| *** shanec_ has quit IRC | 12:15 | |
| odyssey4me | and because no-one's really come forward with any concrete proposals either | 12:16 |
| winggundamth | odyssey4me: I confuse here. Update the repository and create a new topic branch as described in the Starting a Change section of the Infrastructure manual. | 12:16 |
| odyssey4me | ah don't worry about that | 12:17 |
| winggundamth | it means that I have to create new branch right? how about the naming of a branch? | 12:17 |
| odyssey4me | in your git repo, use 'git checkout -b <topic>' | 12:17 |
| odyssey4me | the topic is typically something like bug/xxxxx (the x'es are numbers) for a bug | 12:17 |
| odyssey4me | in your case you should go to https://bugs.launchpad.net/openstack-ansible/+bug/1564873 and assign the bug to yourself, so that everyone else knows you'll be submitting a patch for it | 12:18 |
| openstack | Launchpad bug 1564873 in openstack-ansible "typo on swift.yml.example" [Undecided,New] | 12:18 |
| *** Mudpuppy has quit IRC | 12:18 | |
| winggundamth | ok | 12:18 |
| odyssey4me | then in your git clone, do 'git checkout -b bug/1564873' | 12:18 |
| odyssey4me | then edit whatever needs to be edited | 12:18 |
| odyssey4me | then commit it according to the commit guidelines | 12:19 |
| odyssey4me | then 'git review' to push up the patch | 12:19 |
| * winggundamth trying | 12:20 | |
| openstackgerrit | Merged openstack/openstack-ansible: Include security role in setup-hosts.yml https://review.openstack.org/290526 | 12:20 |
| *** shanec_ has joined #openstack-ansible | 12:20 | |
| javeriak | odyssey4me yea i know about the ansible command, was wondering if there as any service implemented; if not i we can come up with something of ours; no issue in running that on the deploy is there? would the rpc folks have any objection to extra services running on the deploy node? As in my experience some distro's are very particular about what they allow | 12:21 |
| odyssey4me | javeriak well, RPC does have monitoring services that are put on top of OSA - it uses RAX Monitoring as a Service | 12:22 |
| odyssey4me | but as it's RAX specific, and proprietary to RAX, it's not included in OSA | 12:22 |
| javeriak | odyssey4me yes ive been looking at their maas, actually looking into integrating with it | 12:23 |
| javeriak | what i was looking for was a fuel cli type command that gives the liveness status of all the nodes; redhat had one too, where the foreman kept a check of the nodes i believe | 12:24 |
| *** shanec_ has quit IRC | 12:25 | |
| odyssey4me | javeriak hmm, at this time we don't have anything like that - it would be nice if we could start to work in some operational things like that | 12:26 |
| winggundamth | odyssey4me: I confuse about Change-Id:. How I get that id? | 12:26 |
| winggundamth | I'm trying to commit and put the message right now | 12:26 |
| odyssey4me | an example discussed yesterday was a playbook that restarted all services, or just all nova services for example | 12:26 |
| odyssey4me | winggundamth you don't need to put a change id in, it will be done for you with you 'git review' | 12:27 |
| odyssey4me | it's a post commit hook that checks for one and adds it if there isn't one | 12:27 |
| winggundamth | okay | 12:27 |
| winggundamth | odyssey4me: in example here http://docs.openstack.org/infra/manual/developers.html#committing-a-change | 12:27 |
| javeriak | odyssey4me actually yes it would, something besides from the rax maas. btw do you think i should ask someone at rpc, before starting to implement something like that | 12:27 |
| winggundamth | there is Implements:. What I should put? | 12:28 |
| odyssey4me | winggundamth as this is a bug fix, there is no 'implements' - that is for blueprints | 12:28 |
| winggundamth | I see | 12:29 |
| odyssey4me | in this you only need the 'Closes-Bug: <bug #>' | 12:29 |
| *** shanec_ has joined #openstack-ansible | 12:29 | |
| odyssey4me | javeriak no, OSA is separate from RPC - RPC is only one of the stakeholders in the project | 12:29 |
| odyssey4me | will you be at the summit javeriak ? | 12:30 |
| javeriak | odyssey4me afraid not; im still abroad and cant travel at the moment | 12:30 |
| javeriak | so hope to meet you all someday though :) | 12:31 |
| javeriak | do* | 12:31 |
| odyssey4me | javeriak if you want to integrate something into RAX MaaS, then speak to RPC - if you want to implement some operations playbooks then I'd suggest registering a wishlist bug, then submitting some sort of representative sample for review | 12:32 |
| *** pjm6 has quit IRC | 12:32 | |
| javeriak | ugh.. ive been having galera issues all day, it just refuses to work; ive got one conatiner working as primary but the others wont start; have tried everything including destroying and recreating; all i keep getting is Error while getting data from donor node:" | 12:33 |
| odyssey4me | with that we can add it for discussion at the Thu meeting and figure out whether we think a more extensive spec is needed, or whether you should just go ahead - it all depends on the size of the body of work I guess | 12:33 |
| *** Oku_OS_ has quit IRC | 12:33 | |
| odyssey4me | javeriak oh dear, have you tried wiping the container and the data folder on the host of the other two - then recreating the containers? | 12:33 |
| odyssey4me | if you have one working node, the others can always join the cluster | 12:33 |
| odyssey4me | and the data will state transfer across | 12:34 |
| *** shanec_ has quit IRC | 12:34 | |
| odyssey4me | but the data is on the host, not in the container - so just removing the container doesn't always work | 12:34 |
| javeriak | odyssey4me thats a good idea, let me get a spec together and share it with you guys; the goal is to get something together that will benefit the community. I am looking into rax maas, but im going to keep that effort separate for now. | 12:34 |
| odyssey4me | excellent, thanks javeriak - looking forward to it! | 12:35 |
| javeriak | by wiping the data, you mean this right : rm -rf /openstack/node3_galera_container-3ea2cbd3/* ? | 12:35 |
| odyssey4me | javeriak yeah - stop & destroy the container, then remove that folder entirely (not just the stuff in it) | 12:36 |
| javeriak | i think that was tried, but then ive tried quite a few things today, let me give it another shot | 12:37 |
| *** shanec_ has joined #openstack-ansible | 12:39 | |
| winggundamth | oddyssey4me: got this problem http://paste.openstack.org/show/492729/ | 12:39 |
| winggundamth | I already have my account here https://review.openstack.org/ | 12:39 |
| *** neilus has joined #openstack-ansible | 12:43 | |
| *** shanec_ has quit IRC | 12:43 | |
| winggundamth | *odyssey4me | 12:43 |
| *** keedya has joined #openstack-ansible | 12:43 | |
| *** woodard has joined #openstack-ansible | 12:46 | |
| *** shanec_ has joined #openstack-ansible | 12:48 | |
| winggundamth | fixed by manually remote add gerrit | 12:49 |
| *** shanec_ has quit IRC | 12:52 | |
| *** woodard has quit IRC | 12:53 | |
| *** woodard has joined #openstack-ansible | 12:53 | |
| *** shanec_ has joined #openstack-ansible | 12:57 | |
| *** johnmilton has joined #openstack-ansible | 12:59 | |
| openstackgerrit | Jirayut Nimsaeng proposed openstack/openstack-ansible: Fix typo in swift.yml.example file https://review.openstack.org/300460 | 12:59 |
| winggundamth | yay! my first patch | 13:00 |
| odyssey4me | winggundamth well done! | 13:00 |
| winggundamth | make me do for an hour for 2 small changed :P | 13:01 |
| *** shanec_ has quit IRC | 13:01 | |
| * mhayden high fives winggundamth | 13:02 | |
| mhayden | winggundamth: it's fun, isn't it? :) | 13:02 |
| * winggundamth \- -/ | 13:02 | |
| winggundamth | haha yeah. | 13:03 |
| winggundamth | ok. I think I'm going to find a lot of typo in OSA soon :P | 13:04 |
| winggundamth | but gonna leave now. see you around guys and thanks odyssey4me for help on my first patch :) | 13:05 |
| odyssey4me | sure thing winggundamth - have a great evening! | 13:05 |
| mgariepy | hi everyone | 13:06 |
| *** shanec_ has joined #openstack-ansible | 13:06 | |
| *** javeriak has quit IRC | 13:06 | |
| mgariepy | any of you have written a testimonial on how OSA is awesome somewhere ? | 13:06 |
| admin0 | i can write it mgariepy | 13:06 |
| admin0 | OSA is *almost* awesome | 13:07 |
| mattt | haha | 13:07 |
| mattt | saw that one coming a mile away :) | 13:07 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-galera_server: Add retry logic to mysql_upgrade https://review.openstack.org/293822 | 13:07 |
| mgariepy | lol | 13:07 |
| admin0 | public endpoints on https and regions will make it awesome — the next best thing since sliced bread :) | 13:07 |
| admin0 | becaue it can then allow everyone to test it fully :) | 13:08 |
| cloudnull | morning | 13:09 |
| mattt | admin0: open source patches welcome? :) | 13:09 |
| admin0 | i will contribute to a lot of documentation | 13:09 |
| mgariepy | public endpoint in https is configurable. | 13:09 |
| admin0 | but before that i need to see it working at least once that i can reproduce | 13:09 |
| mhayden | but sliced bread is full of carbs | 13:09 |
| admin0 | which I have not been able to do so | 13:09 |
| odyssey4me | o/ cloudnull | 13:10 |
| mattt | admin0: what are you stuck on w/ https now ? | 13:10 |
| *** shanec_ has quit IRC | 13:11 | |
| odyssey4me | cloudnull I suspect that https://review.openstack.org/296525 still makes the repo-build fail, busy testing now | 13:11 |
| admin0 | without hacking, by using just say user_variables.yml, i should be able to set all public working endpoints to https:// — i can add it to the keystone as https , but haproxy is not setup properly to do so .. as OSA is awesome, i rely on it to setup my haproxy also to do the https:// termination | 13:12 |
| odyssey4me | yeah, the optional packages are still picked up for the repo build - we may have to introduce a new category that gets ignored by the py_pkgs lookup | 13:12 |
| mattt | admin0: show the stuff i showed you the other day didn't work? | 13:12 |
| mattt | admin0: *so the | 13:12 |
| admin0 | i am running it now :) | 13:12 |
| mattt | ok so you haven't actually done it then? | 13:13 |
| admin0 | on a fresh test cluster | 13:13 |
| mattt | unsure why you keep saying it doesn't work when you're still testing it out :) | 13:13 |
| cloudnull | odyssey4me: yes optional packages are still built, just not installed into the role based venv | 13:13 |
| mgariepy | so i might get the *almost* out of the testimonial soon then, | 13:13 |
| admin0 | mgariepy: who is the target reader ? i can write as a developer , as a sysadmin, as a operator, as a CTO .. or as what value it brings to business .. ? | 13:14 |
| admin0 | all those need different writeups | 13:14 |
| odyssey4me | cloudnull yeah, so I suspect that we may have to do something like have 'proprietary' packages which are ignored in the package lookup | 13:14 |
| admin0 | or one writeup to rule them all .. hmm.. can do | 13:14 |
| cloudnull | yup | 13:15 |
| odyssey4me | we should then be able to flip a switch to turn on the ability for the repo build to build them, which would have to be coupled with the configuration of the repo to also know where to find them | 13:15 |
| *** shanec_ has joined #openstack-ansible | 13:15 | |
| cloudnull | well. that may be complex | 13:15 |
| odyssey4me | the flipping of a switch? | 13:16 |
| mgariepy | I want to sell it to my CTO and site leads in my organisation | 13:16 |
| * cloudnull looking at the lookup plugin | 13:16 | |
| admin0 | mgariepy: i can help you with that :) | 13:16 |
| admin0 | mgariepy: (and others) please watch this video first: its friday, take a break: https://www.youtube.com/watch?v=BWPnjp12W5o | 13:16 |
| mattt | admin0: SFW ? :) | 13:17 |
| *** weshay has joined #openstack-ansible | 13:17 | |
| admin0 | watch the video please :) trust me.. the answer to mgariepy will then make sense | 13:17 |
| mattt | admin0: asking if it's safe for work :) | 13:18 |
| * cloudnull watching | 13:18 | |
| admin0 | it is completely safe for work | 13:18 |
| *** galstrom_zzz is now known as galstrom | 13:18 | |
| admin0 | actually, you can even show to your co-workers | 13:18 |
| *** shanec_ has quit IRC | 13:19 | |
| cloudnull | mgariepy: he loves the dutch :) | 13:21 |
| admin0 | so mgariepy, i am a non-dutch working for a dutch biggest largest public openstack provider with a big existing environemt and OSA has passed the “test” as a candiate for our next platform .. so all questions that might be asked have been asked and answered | 13:21 |
| odyssey4me | can I have my three minutes back? :p | 13:21 |
| admin0 | i can sell OSA to anyone :D | 13:21 |
| admin0 | so i can write you a use case for you to sell to your CTO | 13:23 |
| *** pjm6 has joined #openstack-ansible | 13:23 | |
| admin0 | how big is the organization, how many guys | 13:23 |
| mgariepy | I work in a university in Canada, to offer computational service to researcher. | 13:24 |
| openstackgerrit | Hugh Saunders proposed openstack/openstack-ansible-os_tempest: Allocaiton pool for public subnet https://review.openstack.org/300481 | 13:24 |
| *** shanec_ has joined #openstack-ansible | 13:24 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_nova: Update test dependencies https://review.openstack.org/300393 | 13:24 |
| *** Oku_OS has joined #openstack-ansible | 13:25 | |
| mattt | anyone know what creates 10.1.13.1 on the AIO for tempest connectivity to instances? | 13:25 |
| openstackgerrit | Hugh Saunders proposed openstack/openstack-ansible-os_tempest: Allocaiton pool for public subnet https://review.openstack.org/300481 | 13:26 |
| mattt | a grep is failing me and i want to avoid having to spin up an instance to check :) | 13:26 |
| mgariepy | we are maybe ~200 tech ppl working on multiple sites to serve all the researcher across canada. | 13:26 |
| admin0 | ok | 13:26 |
| *** Oku_OS_ has joined #openstack-ansible | 13:26 | |
| mgariepy | it's a big mess mostly haha | 13:26 |
| odyssey4me | mattt https://github.com/openstack/openstack-ansible-os_tempest/blob/master/defaults/main.yml#L28 | 13:27 |
| odyssey4me | mattt https://github.com/openstack/openstack-ansible-os_tempest/blob/master/tasks/tempest_resources.yml#L232-L241 | 13:27 |
| *** shanec_ has quit IRC | 13:29 | |
| mattt | odyssey4me: thanks ... i saw that but i'm wondering if we drop a 10.1.13.1 on the instance or in the container for that traffic to flow from the utility container to the neutron network | 13:29 |
| *** Oku_OS has quit IRC | 13:29 | |
| mattt | building an AIO now to see :P | 13:29 |
| odyssey4me | mgariepy any testimonial you find about Rackspace Private Cloud powered by Openstack is essentially a testimonial about OSA, as RPC-O uses OSA and adds some extra bits on top of it | 13:32 |
| *** iceyao has quit IRC | 13:32 | |
| *** openstackgerrit has quit IRC | 13:33 | |
| odyssey4me | mgariepy I'm not sure if that helps, but as OSA is a community project it doesn't have those sorts of things which companies normally put together. | 13:33 |
| *** shanec_ has joined #openstack-ansible | 13:34 | |
| *** openstackgerrit has joined #openstack-ansible | 13:34 | |
| admin0 | mgariepy: i can do a writeup promoting on WHY you must use OSA and nothing else | 13:34 |
| *** fulcrum has joined #openstack-ansible | 13:34 | |
| admin0 | but i need the weekend for it :) | 13:34 |
| *** fulcrum has quit IRC | 13:34 | |
| *** unlaudable has quit IRC | 13:36 | |
| openstackgerrit | Hugh Saunders proposed openstack/openstack-ansible: Specify allocation pool for public subnet https://review.openstack.org/300489 | 13:37 |
| hughsaunders | git-harry: https://review.openstack.org/#/q/topic:bug/1564491 | 13:37 |
| odyssey4me | hughsaunders unfortunately we don't have the multi-repo chaining implemented yet, so you'll have to test manually :/ | 13:38 |
| *** shanec_ has quit IRC | 13:38 | |
| hughsaunders | odyssey4me: yeah I <3 IRRs :/ | 13:38 |
| *** Bjoern has joined #openstack-ansible | 13:38 | |
| admin0 | mattt: keystone verified working :) | 13:39 |
| admin0 | rest — still not | 13:40 |
| *** electrofelix has joined #openstack-ansible | 13:41 | |
| admin0 | mattt: like keystone_public_endpoint: "{{ keystone_service_publicuri }}" .. can all other be added like that in the config file ? | 13:42 |
| *** shanec_ has joined #openstack-ansible | 13:43 | |
| admin0 | cinder, glance etc | 13:43 |
| mgariepy | also, how insane does it sounds if I want to keep the DB, images and volumes of an icehouse installation migrated to liberty ? | 13:44 |
| odyssey4me | FYI to all cores - jmccrory automagically cloudnull d34dh0r53 stevelle mattt hughsaunders andymccr - the SHA's have been set for the branch point for stable/mitaka and the tag 13.0.0 on all repositories. See https://review.openstack.org/300355 - this means that we're essentially now developing for Newton from now on. I ask that everyone please keep focused on improving testing and updating configs for Mitaka for now and hold back any | 13:44 |
| odyssey4me | major Newton work until after the summit. PoC reviews for summit discussions are fine to put up for review. | 13:44 |
| admin0 | mgariepy: depends on the customizations done .. i also have icehouse and tried that route .. but later on decided that using regions was the best way to go forward .. so install new OSA .. then add existing icehouse as region2 | 13:45 |
| admin0 | and offer the possibility of migration if you use swift or ceph | 13:45 |
| admin0 | or yuo need to do a test rolling upgrade | 13:45 |
| odyssey4me | admin0 you can do that today by simply adding the second region into the service catalog with a few keystone commands | 13:46 |
| admin0 | yep | 13:46 |
| odyssey4me | migration between regions is not possible with OpenStack though | 13:46 |
| mgariepy | i have a very limited set of hw. | 13:46 |
| *** markvoelker has joined #openstack-ansible | 13:47 | |
| admin0 | if you use swift globally (fededrated ) and set glance as backend to swift .. and have cinder-backups going to swift, it sounds like a possibility for inter-region migrations | 13:47 |
| admin0 | at least, that is what i am building in office to test | 13:47 |
| *** jthorne has joined #openstack-ansible | 13:47 | |
| odyssey4me | mgariepy you would have to test the migration of the db from Icehouse through each revision to Liberty - you can do this with an offline system to validate whether it completes each upgrade properly | 13:47 |
| *** shanec_ has quit IRC | 13:47 | |
| mgariepy | my plan is to shutdown all vms, and copy de DB, then install OSA over this. with probably some db updates for cinder and such. | 13:48 |
| odyssey4me | mgariepy it's likely you'll have to do some DB cleanup, like clean out keystone tokens and such | 13:48 |
| odyssey4me | in theory you should be able to jump from Icehouse to Kilo, then from Kilo to Liberty | 13:49 |
| mgariepy | why kilo then liberty ? | 13:49 |
| mgariepy | why not just liberty ? | 13:49 |
| odyssey4me | mgariepy OpenStack's testing of upgrades only implements tests for upgrading two cycles (skipping one at most) | 13:49 |
| admin0 | mgariepy: you cannot be 12 year old without being 11 year old first when you are 10 years old :D | 13:50 |
| mgariepy | lol | 13:50 |
| admin0 | well, that corresponds to the openstack release :D | 13:50 |
| *** galstrom is now known as galstrom_zzz | 13:50 | |
| *** markvoelker has quit IRC | 13:51 | |
| odyssey4me | mgariepy I would suggest building a kilo AIO, then dropping all the tables and importing your DB's, then executing setup-openstack and seeing what happens | 13:51 |
| odyssey4me | dig into the health and all that, see if there are things that could make it work better, then repeat until you have that smoothly going | 13:52 |
| *** shanec_ has joined #openstack-ansible | 13:52 | |
| odyssey4me | then use the AIO to execute the upgrade to Liberty, and see how that goes | 13:52 |
| mgariepy | ok | 13:52 |
| mgariepy | i'll do that then. | 13:52 |
| odyssey4me | once you have that worked out, switch to multi-node and repeat with what you leanred | 13:52 |
| odyssey4me | figure out the risks, etc | 13:53 |
| odyssey4me | if it all looks good you may be able to take the upgraded DB and import it straight into a running Liberty system | 13:54 |
| odyssey4me | I'm not sure though - never tried, but it seems plausible | 13:54 |
| * admin0 suggests not to do that :D | 13:58 | |
| admin0 | especially if there are clients in there | 13:58 |
| odyssey4me | of course taking backups would be a good diea | 13:58 |
| odyssey4me | and that presumes that you have the ability to take everything offline while you do the work | 13:58 |
| *** shanec_ has quit IRC | 13:58 | |
| odyssey4me | you should send an email to the operators list asking if anyone's done anything like that and see if someone can give you tips | 13:58 |
| *** openstack has quit IRC | 13:58 | |
| *** openstack has joined #openstack-ansible | 14:01 | |
| *** shanec_ has joined #openstack-ansible | 14:01 | |
| mgariepy | each of my compute have 128g of ram, 2x E5-2650 v2 and some hdd. | 14:01 |
| mgariepy | 1 240g ssd and 2 to hdd. | 14:02 |
| mgariepy | the org have another site which have aroung 40 256gb compute nodes | 14:03 |
| admin0 | then clear one or two of your compute nodes from customers and use KVM to create VMS .. one for each of your controller node .. and say 2 for your compute nodes .. so 1 server = 5 VMS which has 3 controller and 2 compute nodes, and antoher server , 8 VMS that do CEPH .. in that way you can replicate what you have now .. then using rsync or anything, replicate your physical data to virtual and do the upgrade test there — in that way yo | 14:04 |
| admin0 | have a real test case of what needs to be done, and what can be done . | 14:04 |
| *** busterswt has joined #openstack-ansible | 14:04 | |
| admin0 | and since its virtual, you can redo that process many times to get it right | 14:04 |
| mgariepy | and we are looking to have some more in a future, don't know how many exacly tho, | 14:05 |
| *** shanec_ has quit IRC | 14:05 | |
| admin0 | some more nodes, some more people, some more hardware ? :D | 14:06 |
| *** sdake has joined #openstack-ansible | 14:06 | |
| mgariepy | yeah using vms would be handy | 14:06 |
| mgariepy | but the bad thing is the current install use Fuel. | 14:07 |
| *** sigmavirus24_awa is now known as sigmavirus24 | 14:08 | |
| mgariepy | it's not easy to setup in a vm haha | 14:08 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-plugins: Add logic to NOT build "proprietary" packages https://review.openstack.org/300505 | 14:08 |
| cloudnull | odyssey4me: ^ | 14:08 |
| admin0 | mgariepy: i have done fuel + vms :) | 14:08 |
| admin0 | free fuel or paid fuel :D ? | 14:09 |
| mgariepy | free we are cheap here haha | 14:09 |
| *** iceyao has joined #openstack-ansible | 14:09 | |
| *** KLevenstein has joined #openstack-ansible | 14:09 | |
| admin0 | how many acutal customers are there ? | 14:09 |
| mgariepy | we have 0 paying customers. | 14:10 |
| mgariepy | and 200+ but maybe more like ~50 real users. | 14:10 |
| *** shanec_ has joined #openstack-ansible | 14:10 | |
| odyssey4me | cloudnull ok, let me try and understand your stream of consciousness, because I don't quite get the code | 14:10 |
| *** saneax is now known as saneax_AFK | 14:11 | |
| admin0 | well, then the best way to move forward is to consolidate and migrate customers to compute nodes and start making them free .. so that they can be re-used on the new one using ansible .. then we provide a migration path to the 50 instances | 14:11 |
| admin0 | if they are not paying, there is no SLA .. so why should you take the pain and burden and the uncertainity of the upgrade | 14:11 |
| odyssey4me | the patch is not looking for a 'proprietary' key word, but instead will check if it can find a git source? | 14:11 |
| admin0 | just do a new install with ansible and provide a migration path | 14:11 |
| cloudnull | odyssey4me: it looks for the proprietary key word | 14:11 |
| cloudnull | if found and there is no git source skip | 14:12 |
| odyssey4me | lol, I should scroll down sometimes :p | 14:12 |
| cloudnull | well, look for proprietary key word , inspect list , if package item has no git source, skip | 14:12 |
| odyssey4me | ok, so there is possibly an issue with that - for this use case it will not be a git source, it will be a private pypi index as I understand it | 14:12 |
| * admin0 looks for mattt | 14:12 | |
| cloudnull | skipped | 14:13 |
| cloudnull | but if its a secondary index they can add that as a var to the pip_install/lockdown bits | 14:13 |
| odyssey4me | so it will work if the repo build is set to use that index to look for links? | 14:13 |
| cloudnull | at current repo-build will do nothing with those packages, unless there is a git source . | 14:14 |
| *** galstrom_zzz is now known as galstrom | 14:14 | |
| odyssey4me | ok, so to make the packages available it will never end up on the repo - it has to be made available and configured using the lock down links | 14:14 |
| cloudnull | if they have an index else where they can set that for pip lockdown to distribute | 14:14 |
| mattt | admin0: what's up ? | 14:14 |
| cloudnull | odyssey4me: yup | 14:14 |
| cloudnull | that was my thought | 14:14 |
| *** shanec_ has quit IRC | 14:15 | |
| *** weshay has quit IRC | 14:15 | |
| admin0 | mattt: thank you .. keystone_public_endpoint: "{{ keystone_service_publicuri }} worked :) | 14:15 |
| admin0 | how to do the same for others :D ? | 14:15 |
| odyssey4me | ok, I was kinda preferring that the repo server held the packages always - but this can work | 14:15 |
| *** Mudpuppy has joined #openstack-ansible | 14:15 | |
| odyssey4me | admin0 that work is still ongoing, and matt is quite busy right now so please either figure it out yourself or wait until after the summit | 14:18 |
| mattt | admin0: not sure, tore down my liberty install this morning :( i'll have to have a look in a bit | 14:18 |
| cloudnull | it'd be nice for that to go, however we have no way to know that an index will be available at the time a lookup is run | 14:18 |
| admin0 | i can happlily add your keys to my install :D | 14:18 |
| odyssey4me | I appreciate that you really want to get it done, but we also need to complete a release. | 14:18 |
| odyssey4me | cloudnull maybe something like a try and if not found then warn? | 14:18 |
| odyssey4me | meh, that sounds messy actually | 14:18 |
| cloudnull | it would have to be something done in the wheel build | 14:18 |
| admin0 | mgariepy: all my OSA installs are documented here: https://www.openstackfaq.com/ including what I have seen as the architecture that people prefer: https://www.openstackfaq.com/openstack-private-cloud-architecture/ .. so if you want it implemented, i can help | 14:18 |
| odyssey4me | ok, let's give this a go and see how that works out - I think this is a reasonable solution | 14:18 |
| *** iceyao has quit IRC | 14:18 | |
| *** pjm6 has quit IRC | 14:18 | |
| cloudnull | IMHO if it requires a proprietary closed source index / bundle of software, it up to the deployer to do that we can give them the tools, but I'd rather it not be our problem to manage | 14:19 |
| *** shanec_ has joined #openstack-ansible | 14:19 | |
| cloudnull | they can add https://github.com/openstack/openstack-ansible-pip_lock_down/blob/master/defaults/main.yml#L31 | 14:20 |
| cloudnull | and when pip goes to install things it will go to the super secrete special index | 14:20 |
| odyssey4me | yup, that's already how they've been doing it anyway :) | 14:20 |
| cloudnull | boom ! move along citizen , nothing to see here. | 14:20 |
| odyssey4me | we just need to help instrument it so that our build works without the extra index too | 14:21 |
| cloudnull | :p | 14:21 |
| odyssey4me | hughsaunders jmccrory would you mind taking a peek at https://review.openstack.org/300505 if you have a moment? | 14:21 |
| * cloudnull goes back to my pit of ironic dispare | 14:22 | |
| * admin0 throws in a rople to get cloudnull out of the pit | 14:22 | |
| admin0 | *rope | 14:22 |
| cloudnull | thanks admin0 | 14:22 |
| cloudnull | :) | 14:22 |
| * admin0 gives pizza and beer to cloudnull | 14:22 | |
| Bofu2U | someone say despair? Good morning. | 14:23 |
| cloudnull | ohai Bofu2U | 14:23 |
| Bofu2U | hai2u | 14:23 |
| admin0 | hello2u | 14:23 |
| admin0 | :D | 14:23 |
| cloudnull | hows the great server migration going ? | 14:23 |
| Bofu2U | well I got galera back to working order | 14:23 |
| admin0 | tofu 2u :D | 14:23 |
| Bofu2U | then rabbit | 14:23 |
| Bofu2U | then all the rest | 14:23 |
| Bofu2U | then my DC told me I need to show them 90% utilization on the IP's I don't yet have. | 14:24 |
| Bofu2U | soooooo working on getting some IP's now. :| | 14:24 |
| cloudnull | oh good | 14:24 |
| *** shanec_ has quit IRC | 14:24 | |
| admin0 | if you need a DC in amsterdam for cloud /me is the person to go to | 14:24 |
| admin0 | boo /me does not work like that | 14:24 |
| * admin0 is sad | 14:24 | |
| Bofu2U | I'll PM ya | 14:24 |
| Bofu2U | but we just got a cage here | 14:25 |
| Bofu2U | :-/ | 14:25 |
| cloudnull | if rabbit is being a pain, destroy the container, recreate them, redeploy rabbit. | 14:25 |
| cloudnull | openstack-ansible lxc-container-destory.yml lxc-container-create.yml rabbitmq-install.yml --limit rabbitmq_all | 14:25 |
| Bofu2U | yeah I just lxc-attach'd and rebooted it | 14:26 |
| Bofu2U | and it came back up :) | 14:26 |
| cloudnull | good | 14:27 |
| odyssey4me | heh, I didn't know you could chain playbooks on the CLI like that | 14:27 |
| cloudnull | yup | 14:27 |
| odyssey4me | they all inherit the limit? | 14:27 |
| cloudnull | yes they do | 14:27 |
| odyssey4me | that's pretty awesome | 14:27 |
| cloudnull | I do that all the time when i take the nuclear route with things | 14:28 |
| *** shanec_ has joined #openstack-ansible | 14:29 | |
| Bofu2U | sounds familiar :P | 14:29 |
| cloudnull | if you take the nuclear route w/ rabbit in liberty + just make sure you recreate all the vhosts and such | 14:31 |
| cloudnull | openstack-ansible setup-openstack.yml --tags=aodh-rabbitmq,ceilometer-rabbitmq,cinder-rabbitmq,glance-rabbitmq,heat-rabbitmq,keystone-rabbitmq,neutron-rabbitmq,nova-rabbitmq | 14:31 |
| *** Brew has joined #openstack-ansible | 14:32 | |
| *** phiche has quit IRC | 14:32 | |
| *** shanec_ has quit IRC | 14:33 | |
| cloudnull | deployments >=liberty have vhosts users and passwords for all OS services which was done for stronger security across the cluster, but requires an extra step to recover from an orbital nuke | 14:33 |
| *** phalmos has joined #openstack-ansible | 14:37 | |
| *** shanec_ has joined #openstack-ansible | 14:38 | |
| odyssey4me | cloudnull FYI, in https://etherpad.openstack.org/p/newton-cross-project-sessions there is a cross-project talk proposed at the summit to discuss splitting requirements and requirements caps to the projects and dropping the requirement for co-installability | 14:39 |
| odyssey4me | which means that we'll need to be adjusting the repo build process somewhat in Newton - assuming that this is agreed to | 14:39 |
| *** spotz_zzz is now known as spotz | 14:40 | |
| cloudnull | no real change for us. | 14:41 |
| cloudnull | if they drop global requirements we'll stop indexing it | 14:41 |
| *** markvoelker has joined #openstack-ansible | 14:41 | |
| *** ametts has joined #openstack-ansible | 14:42 | |
| odyssey4me | yeah, I suppose we just grab the bits from the service repositories and go from there | 14:42 |
| cloudnull | we already pull in the packages into a venv and distribute that | 14:42 |
| *** shanec_ has quit IRC | 14:42 | |
| cloudnull | and that venv is built using the requirements from the project | 14:42 |
| cloudnull | i mean its likely we'll have to change some, but its not a lot of change | 14:43 |
| odyssey4me | it will mean that we should probably execute the wheel build per service though - so that multiple wheels can exist - and there'll be no need to resolve down to a single set of versions for the whole bunch | 14:43 |
| cloudnull | yup, we kinda do that already with the introduction of the op-venv-scripts | 14:44 |
| cloudnull | if they drop the co-installability requirements our build would likely be even faster. | 14:44 |
| odyssey4me | or we just build the wheels from the package lists, then create the venvs - if the wheels don't exist for the venvs then they will just pull from pypi or wherever | 14:44 |
| *** sigmavirus24 is now known as sigmavirus24_awa | 14:44 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 14:45 | |
| cloudnull | we'd just build the known git projects and take the resultant and build a venv | 14:45 |
| odyssey4me | if I build the wheel for 'aodh', does it also build all the dependent wheels? | 14:45 |
| cloudnull | yup | 14:45 |
| odyssey4me | oh, then why do we bother building the global requirements wheels? | 14:46 |
| odyssey4me | why not just do each service, then each package we have specifically listed? | 14:46 |
| *** markvoelker has quit IRC | 14:47 | |
| cloudnull | some of the services depend on packages in global requirements and they dont know it | 14:47 |
| cloudnull | its just there so the service works | 14:47 |
| odyssey4me | so how do those get into the venv? | 14:47 |
| openstackgerrit | Amy Marrich (spotz) proposed openstack/openstack-ansible: Add Ceilometer instructions to new compute node instructions https://review.openstack.org/300103 | 14:48 |
| openstackgerrit | Amy Marrich (spotz) proposed openstack/openstack-ansible: Add trusty_backports note to requirements https://review.openstack.org/300120 | 14:48 |
| odyssey4me | or is there a problem with resolving the deps of the deps? | 14:48 |
| cloudnull | for the most part we have them listed in our roles | 14:48 |
| cloudnull | the extra packages | 14:48 |
| cloudnull | so our roles do the needful | 14:48 |
| cloudnull | well they assist in that | 14:48 |
| odyssey4me | ah, but then our specifically listed packages will have wheels built too - so problem solved | 14:49 |
| cloudnull | but if global requirements is gone, the services will have to do all the things themselves | 14:49 |
| cloudnull | which means we can remove things and itll be faster | 14:49 |
| odyssey4me | interesting, maybe I'll try removing it and see what happens | 14:49 |
| cloudnull | yea please do | 14:49 |
| cloudnull | it maybe better already | 14:49 |
| * cloudnull is old and yells at kids | 14:50 | |
| cloudnull | and has not tried it recently | 14:50 |
| *** iceyao has joined #openstack-ansible | 14:51 | |
| *** lykinsbd has joined #openstack-ansible | 14:51 | |
| odyssey4me | well, if we think about it - we build from source in the roles as-is, and all we do is use upper-constraints from g-r in that build | 14:51 |
| odyssey4me | so it seems to me that this should work pretty well | 14:52 |
| odyssey4me | I'll take a crack at it | 14:52 |
| cloudnull | ++ | 14:52 |
| cloudnull | with the recent changes we've gotten the repo_build time down to about 10 min, if we can make that faster thats a win in my book | 14:52 |
| cloudnull | we branching today ? | 14:54 |
| *** neilus has quit IRC | 14:54 | |
| odyssey4me | cloudnull we'll branch whenever the release team gets a moment to do it - it'll be at these SHA's: https://review.openstack.org/300355 | 14:55 |
| cloudnull | cool | 14:56 |
| openstackgerrit | Rohan Parulekar proposed openstack/openstack-ansible-os_neutron: Nuage neutron plugin ansible changes https://review.openstack.org/296525 | 14:56 |
| odyssey4me | cloudnull FYI I've tested it and it works (the proprietary thing) | 14:59 |
| cloudnull | nice | 15:00 |
| *** thorst has quit IRC | 15:01 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_neutron: Nuage neutron plugin ansible changes https://review.openstack.org/296525 | 15:02 |
| *** thorst has joined #openstack-ansible | 15:02 | |
| *** phiche has joined #openstack-ansible | 15:03 | |
| *** electrofelix has quit IRC | 15:03 | |
| *** admin0 has quit IRC | 15:04 | |
| *** weezS has joined #openstack-ansible | 15:04 | |
| *** markvoelker has joined #openstack-ansible | 15:05 | |
| cloudnull | required watching for everyone working in the cloud https://www.youtube.com/watch?v=GyV_UG60dD4&feature=youtu.be | 15:08 |
| cloudnull | :) | 15:08 |
| cloudnull | you're all welcome | 15:08 |
| *** shanec_ has joined #openstack-ansible | 15:13 | |
| *** pjm6 has joined #openstack-ansible | 15:19 | |
| *** thorst has quit IRC | 15:22 | |
| *** pjm6 has quit IRC | 15:25 | |
| *** galstrom is now known as galstrom_zzz | 15:25 | |
| mgariepy | lol | 15:25 |
| *** Andrew_jedi has joined #openstack-ansible | 15:30 | |
| *** thorst has joined #openstack-ansible | 15:30 | |
| Andrew_jedi | Hello folks, Do we have anything in OSA which will be helpful in system maintenance? | 15:31 |
| Andrew_jedi | Gracefully stopping all openstack services etc etc | 15:31 |
| odyssey4me | Andrew_jedi there are a few things implemented (cron jobs and such) as part of the roles where they've been historically needed, but do you have something specific in mind? | 15:32 |
| Andrew_jedi | odyssey4me: Nothing as such, i have to to bring down the systems to move power circuits and i was thinking if there is anything in OSA which i can use to gracefully shutdown the services. | 15:34 |
| odyssey4me | Andrew_jedi we were discussing some operational things like convenience plays for restarting services and such and javeriak has kindly offered to put a spec together to discuss and do some work in that regard in this next cycle - see http://eavesdrop.openstack.org/irclogs/%23openstack-ansible/%23openstack-ansible.2016-04-01.log.html#t2016-04-01T12:34:54 | 15:34 |
| odyssey4me | Andrew_jedi at this stag you can use ansible to execute things across the nodes, but there's nothing specific that orchestrates a graceful shut down | 15:36 |
| Andrew_jedi | odyssey4me: Great, this will be really helpful. | 15:36 |
| Andrew_jedi | odyssey4me: Ahh okay, Thanks for letting me know :) | 15:37 |
| odyssey4me | if you could make some notes in an etherpad or something about what you expect to do, then I'm sure that we could all club in and comment or assist - then perhaps that can form the basis of something we add as an operational playbook during the cycle | 15:37 |
| dolphm | can someone point me to something, anything, that describes what an ansible "virtual module" actually is? | 15:37 |
| dolphm | cc- cloudnull because it appears you wrote such a thing | 15:37 |
| *** galstrom_zzz is now known as galstrom | 15:38 | |
| odyssey4me | hah dolphm I can give you an example - assuming that it is what I think it is | 15:38 |
| Andrew_jedi | odyssey4me: that sounds like a good idea, let me see what i can come up with | 15:39 |
| odyssey4me | an action plugin, for example: https://github.com/openstack/openstack-ansible-plugins/blob/master/action/config_template.py / https://github.com/openstack/openstack-ansible-plugins/blob/master/library/config_template | 15:39 |
| dolphm | odyssey4me: i have a security concern with https://github.com/openstack/openstack-ansible-plugins/blob/master/library/config_template but would like to see an implementation before filing anything lol | 15:39 |
| dolphm | odyssey4me: ah, okay -- i missed the action/ one | 15:39 |
| *** daneyon has joined #openstack-ansible | 15:39 | |
| odyssey4me | those two are linked, but don't asked me how they work - that's some cloudnull magic right there | 15:40 |
| cloudnull | dolphm: ohai | 15:40 |
| odyssey4me | maybe this helps? http://ndemengel.github.io/2015/01/20/ansible-modules-and-action-plugins/ | 15:40 |
| cloudnull | virtual module is a library that is really an action plugin. | 15:40 |
| *** pjm6 has joined #openstack-ansible | 15:41 | |
| dolphm | odyssey4me: thanks, i can probably find what i need here in terms of this specific module https://github.com/openstack/openstack-ansible-plugins/blob/master/action/_v2_config_template.py | 15:41 |
| *** mikelk has quit IRC | 15:41 | |
| cloudnull | libraries are sent to the target and executed, where an action plugin is executed on the host and the resultant is sent to the target. | 15:41 |
| *** fawadkhaliq has joined #openstack-ansible | 15:42 | |
| *** iceyao has quit IRC | 15:43 | |
| odyssey4me | ah, I didn't know that | 15:45 |
| odyssey4me | good to know | 15:45 |
| *** woodard has quit IRC | 15:50 | |
| *** rohanp_ has joined #openstack-ansible | 15:52 | |
| *** weshay has joined #openstack-ansible | 15:52 | |
| *** pjm6_ has joined #openstack-ansible | 15:55 | |
| *** pjm6 has quit IRC | 15:55 | |
| *** Andrew_jedi has quit IRC | 16:12 | |
| *** weezS has quit IRC | 16:18 | |
| *** weezS has joined #openstack-ansible | 16:30 | |
| *** antonym has quit IRC | 16:31 | |
| *** openstackgerrit has quit IRC | 16:31 | |
| *** weezS has quit IRC | 16:34 | |
| *** Oku_OS_ has quit IRC | 16:39 | |
| *** antonym has joined #openstack-ansible | 16:44 | |
| *** pcaruana has quit IRC | 16:44 | |
| *** openstackgerrit has joined #openstack-ansible | 16:45 | |
| *** johnmilton has quit IRC | 16:47 | |
| *** fawadkhaliq has quit IRC | 16:50 | |
| *** fawadkhaliq has joined #openstack-ansible | 16:50 | |
| pjm6_ | Hi I'm having problem with getting dhcp in a VM from a physical network | 16:51 |
| pjm6_ | I created the network with vlan with VLAN segmentation ID | 16:52 |
| pjm6_ | I debug with tcpdump and see that the request from VM | 16:52 |
| pjm6_ | goes into the router | 16:52 |
| pjm6_ | but he didn't receive reply | 16:52 |
| pjm6_ | also the cirros and centos image (that I created) seems to have problems in getting the metadata | 16:52 |
| pjm6_ | checking http://169.254.169.254/2009-04-04/instance-id | failed to read iid from metadata. tried 20 | 16:53 |
| pjm6_ | anyone have a clue or a hint how to debug this? | 16:53 |
| *** johnmilton has joined #openstack-ansible | 17:00 | |
| openstackgerrit | Rohan Parulekar proposed openstack/openstack-ansible-os_nova: Nuage nova configuration ansible changes https://review.openstack.org/296538 | 17:01 |
| *** johnmilton has quit IRC | 17:02 | |
| spotz | odyssey4me Have you put together when the design sessions will be? I need to hit some of the heat stuff but want to be in our session for the docs and such for sure | 17:05 |
| odyssey4me | spotz I'm not sure if the final sessions are up yet - there was a ML message from ttx quite recently with the proposed final schedule and he said he'd be putting it up soon so that we can actually edit the titles and such | 17:06 |
| odyssey4me | so all we know at this time are when our slots are | 17:07 |
| spotz | odyssey4me ok cool, I know I'll only make the first part of the fishbowl, the git and gerrit session I'm running is right up against it | 17:08 |
| spotz | And yes I am already plotting everywhere I need to be and when:) | 17:09 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible: Add installation support for os_ironic https://review.openstack.org/293779 | 17:09 |
| *** Andrew_jedi has joined #openstack-ansible | 17:12 | |
| *** admin0 has joined #openstack-ansible | 17:13 | |
| *** mgoddard has quit IRC | 17:14 | |
| *** michaelgugino has joined #openstack-ansible | 17:16 | |
| michaelgugino | cloudnull: odyssey4me do either of you know how to enable the gpg keys for the ubuntu cloud archive repo? | 17:17 |
| michaelgugino | ansible is complaining, but when I try to install packages from that repo via cli manually, there is no problem | 17:17 |
| *** Bjoern has quit IRC | 17:20 | |
| openstackgerrit | Rohan Parulekar proposed openstack/openstack-ansible-os_nova: Nuage nova configuration ansible changes https://review.openstack.org/296538 | 17:23 |
| *** Andrew_jedi has quit IRC | 17:23 | |
| *** jiteka has quit IRC | 17:24 | |
| pjm6_ | In the compute node I receive the reply of DHCP request made from the VM | 17:25 |
| pjm6_ | it's possible that for some reason that reply are not being redirected to the corresponding VM? | 17:26 |
| *** johnmilton has joined #openstack-ansible | 17:28 | |
| *** weezS has joined #openstack-ansible | 17:32 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_build: Remove global-requirements build from the build process https://review.openstack.org/300589 | 17:32 |
| odyssey4me | cloudnull ^ FYI, it appears to actually work - at least for the build part | 17:32 |
| cloudnull | awesome! | 17:33 |
| michaelgugino | nvm, I got it all sorted. | 17:33 |
| cloudnull | sorry michaelgugino, pjm6_ im all over the place today | 17:34 |
| michaelgugino | no worries | 17:34 |
| *** weezS has quit IRC | 17:35 | |
| michaelgugino | this patch I'm about to push took a lot of refactoring and a little hair pulling ;) | 17:35 |
| pjm6_ | don't worry cloudnull, you've been helping a lot :) | 17:35 |
| *** jiteka has joined #openstack-ansible | 17:36 | |
| odyssey4me | alright, I'm out for the night and weekend... and I plan not to do any real work until Monday - it's time to catch up on sleep and give my back a break | 17:39 |
| odyssey4me | have a good w/end all! | 17:39 |
| mgariepy | cya odyssey4me | 17:40 |
| openstackgerrit | Michael Gugino proposed openstack/openstack-ansible-os_neutron: [WIP] Implementing neutron_openvswitch_agent https://review.openstack.org/298765 | 17:40 |
| michaelgugino | changes: +313, -58 | 17:41 |
| michaelgugino | it became pretty big pretty quickly. A lot of it is just some refactoring. | 17:41 |
| pjm6_ | cya odyssey4me | 17:42 |
| pjm6_ | anyone knows if the metadata from neutron if its not fetched by the image, that could be the cause of netwroking problems | 17:43 |
| *** Andrew_jedi has joined #openstack-ansible | 17:47 | |
| *** tiagogomes has quit IRC | 17:48 | |
| *** fawadkhaliq has quit IRC | 17:52 | |
| *** fawadkhaliq has joined #openstack-ansible | 17:52 | |
| *** markvoelker has quit IRC | 17:52 | |
| *** markvoelker has joined #openstack-ansible | 17:56 | |
| admin0 | bye odyssey4me | 17:57 |
| admin0 | pjm6_: pre-boot or post-boot metadata ? | 17:59 |
| admin0 | pre-boot is direct to api, post-boot is 169.254.169.254 proxied | 18:00 |
| pjm6_ | well I think that is post-boot | 18:02 |
| pjm6_ | post-boot (after VM being turned on, right? | 18:02 |
| *** Andrew_jedi has quit IRC | 18:04 | |
| pjm6_ | well I do something | 18:04 |
| pjm6_ | it was adding an private network | 18:04 |
| pjm6_ | well I try something different | 18:09 |
| pjm6_ | (creating a private network and attach 2 networks to them) | 18:09 |
| pjm6_ | and i can't ping each other | 18:09 |
| pjm6_ | but teh VM name appeared | 18:10 |
| *** woodard has joined #openstack-ansible | 18:10 | |
| *** Brew has quit IRC | 18:13 | |
| admin0 | huh :D | 18:16 |
| admin0 | calm down | 18:16 |
| admin0 | rephrase | 18:16 |
| admin0 | think me as 10 year old | 18:16 |
| * admin0 is 10 year old | 18:16 | |
| pjm6_ | lool sorry for my bad explanation :( | 18:16 |
| pjm6_ | so | 18:17 |
| pjm6_ | first I created a VM and connected to an external network (with VLAN segmentation ID 3) | 18:17 |
| admin0 | ok | 18:17 |
| pjm6_ | that VLAN is an existing network that has a DHCP server | 18:17 |
| admin0 | ok | 18:17 |
| pjm6_ | when VM sends DHCP request from compute node | 18:17 |
| pjm6_ | I can receive it | 18:17 |
| admin0 | ok | 18:17 |
| pjm6_ | from the router of that router vlan 3 | 18:18 |
| admin0 | ok | 18:18 |
| pjm6_ | and the reply is sended back to the compute node | 18:18 |
| admin0 | ok | 18:18 |
| pjm6_ | But the VM | 18:18 |
| pjm6_ | don't get it | 18:18 |
| pjm6_ | 18:54:57.752513 IP 10.2.0.254.bootps > 10.2.0.200.bootpc: BOOTP/DHCP, Reply, length 301 | 18:19 |
| admin0 | did u plug in and plug out the ports multi times ? | 18:19 |
| pjm6_ | the compute node receives this, but the VM can't receive the DHCP | 18:19 |
| pjm6_ | in the VM ? | 18:19 |
| admin0 | yes | 18:19 |
| pjm6_ | the detach and attach interface? | 18:19 |
| admin0 | yes | 18:19 |
| pjm6_ | yes, at some point I did it | 18:19 |
| admin0 | :D | 18:19 |
| pjm6_ | but now for testing the networking | 18:19 |
| pjm6_ | I created another node and created a private network, right? | 18:19 |
| rohanp_ | could this be an issue with certain ports being blocked by default on the compute? | 18:20 |
| pjm6_ | hmm like ip tables or something? | 18:20 |
| admin0 | if you have a direct dhcp (shared network ) . create a VM .. see if that works .. then aother is create a router, private network and then attach it to that to see if it gets private part | 18:20 |
| rohanp_ | yea | 18:20 |
| stevelle | Looks like the release team is starting the process for our mitaka branch \o/ | 18:20 |
| admin0 | well if you plugin and out many times , your ports securty groups etc might get mismatch | 18:20 |
| admin0 | i have seen this many times before | 18:21 |
| pjm6_ | what I do next (for lower testing) | 18:21 |
| pjm6_ | was creating an private network (vxlan with ID 24) | 18:21 |
| pjm6_ | attach two computers to that network | 18:21 |
| admin0 | VM ( not booted ) will see eth0 .. and add eth1 and eth2 silently — you can check my ip a .. but if you check in db , that eht0 mac is not there .. only eth2 (last one is ) and the security groups don’t allow that | 18:21 |
| admin0 | but if you have trouble with new VM on new router, then problem is elsewhere | 18:22 |
| admin0 | VM ( not rebooted ) i meant | 18:22 |
| pjm6_ | hmmm I see what you're saying, because of that I deleted the devices of eth | 18:22 |
| pjm6_ | to reset them | 18:22 |
| admin0 | check ip a | 18:22 |
| admin0 | inside the VM | 18:22 |
| admin0 | you might see eth0 in ifconfig .. but you might find eth1 eth2 there | 18:22 |
| pjm6_ | a do in the two configs | 18:23 |
| pjm6_ | and have eth0 | 18:23 |
| *** jiteka has quit IRC | 18:23 | |
| *** eil397 has joined #openstack-ansible | 18:23 | |
| admin0 | so in a summary .. you have a new VM .. with just eth0 and DHCP is not working ? | 18:24 |
| pjm6_ | Yes | 18:26 |
| pjm6_ | and now for debugging what I did | 18:27 |
| pjm6_ | was creating an private network (with vxlan) | 18:27 |
| pjm6_ | and the two VMs and ping each other :\ | 18:27 |
| openstackgerrit | Michael Gugino proposed openstack/openstack-ansible: Add neutron_openvswitch_agent to env.d files https://review.openstack.org/300605 | 18:28 |
| stevelle | cloudnull: for adding an irr os_project in infra, I presume these roles are not published to pypi and I can skip that step in the infra manual. correct? | 18:29 |
| admin0 | pjm6_: so VMs get created ? they get IPs ? but cannot ping each other ? | 18:31 |
| admin0 | or they get created, but ip is not there ? | 18:31 |
| *** markvoelker has quit IRC | 18:31 | |
| pjm6_ | They get created and ip is there | 18:32 |
| pjm6_ | and the metadata problem is resolved | 18:32 |
| pjm6_ | (because the DHCP agent add the IP that he need) | 18:32 |
| openstackgerrit | Michael Gugino proposed openstack/openstack-ansible-os_neutron: [WIP] Implementing neutron_openvswitch_agent https://review.openstack.org/298765 | 18:35 |
| pjm6_ | I have to go, cya later | 18:36 |
| *** pjm6_ has quit IRC | 18:37 | |
| *** keedya has quit IRC | 18:45 | |
| *** elo has quit IRC | 18:49 | |
| *** markvoelker has joined #openstack-ansible | 18:53 | |
| *** markvoelker has quit IRC | 18:53 | |
| *** markvoelker has joined #openstack-ansible | 18:54 | |
| *** markvoelker has quit IRC | 18:54 | |
| *** markvoelker has joined #openstack-ansible | 18:54 | |
| *** fawadkhaliq has quit IRC | 19:05 | |
| *** fawadkhaliq has joined #openstack-ansible | 19:06 | |
| *** fawadkhaliq has quit IRC | 19:07 | |
| *** fawadkhaliq has joined #openstack-ansible | 19:13 | |
| *** KLevenstein has quit IRC | 19:16 | |
| *** KLevenstein has joined #openstack-ansible | 19:16 | |
| alextricity25 | Hey cloudnull. Have you ever had your multi-node-aio boxes spontanously stop NAT'ing instance traffic to the internet? | 19:20 |
| alextricity25 | I woke up this morning and suddenly my instances couldn't hit the outside. | 19:21 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-openstack_hosts: conditionally include the scsi_dh kernel module https://review.openstack.org/300621 | 19:22 |
| cloudnull | alextricity25: no i've not | 19:23 |
| cloudnull | i've had the same box up for about 3 weeks now and its not died in a fire yet | 19:23 |
| cloudnull | stevelle: what that ? | 19:24 |
| *** KLevenstein has quit IRC | 19:24 | |
| stevelle | cloudnull: http://docs.openstack.org/infra/manual/creators.html#give-openstack-permission-to-publish-releases | 19:24 |
| stevelle | skip this? | 19:24 |
| cloudnull | ah yes. | 19:25 |
| cloudnull | only for things being published to pypi | 19:25 |
| stevelle | cloudnull: golden, thx | 19:25 |
| cloudnull | thank you sir | 19:25 |
| cloudnull | odyssey4me: we should consider backporting https://review.openstack.org/300621 all the way back to kilo | 19:26 |
| cloudnull | ubuntu 14.04 lts now has 4.4 available to it and that will cause a hard stop | 19:27 |
| michaelgugino | ugh, another kernel version patch. It's going to get messy | 19:28 |
| michaelgugino | I'm hopping off. I'll be traveling next week, so not sure if I'll be able to make any meetings. But reviews on my patches for OVS stuff would be appreciated. I intend on submitting some documentation similar to what has been done with plumgrid in another patch set. I will probably be able to write a few small changes to my patches next week. | 19:30 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-openstack_hosts: conditionally include the scsi_dh kernel module https://review.openstack.org/300621 | 19:31 |
| spotz | safe travels michaelgugino | 19:31 |
| cloudnull | michaelgugino: cheers safe travels | 19:32 |
| michaelgugino | ty | 19:32 |
| cloudnull | ill try to get to another review later today | 19:32 |
| cloudnull | thanks for all the work you've put into that btw | 19:32 |
| michaelgugino | most welcome. We need this feature in our org as much as anyone. | 19:33 |
| cloudnull | i think its close :) | 19:33 |
| michaelgugino | me too. Anyway, cya | 19:34 |
| *** michaelgugino has quit IRC | 19:34 | |
| admin0 | have a great weekend | 19:34 |
| *** johnmilton has quit IRC | 19:35 | |
| cloudnull | odyssey4me: nice job on this https://review.openstack.org/#/c/300589/ -- im giving it a spin but thats awesome that it completed the build in <5 min | 19:39 |
| cloudnull | http://logs.openstack.org/89/300589/1/check/gate-openstack-ansible-repo_build-ansible-func-ubuntu-trusty/216b23c/console.html#_2016-04-01_17_52_26_496 | 19:40 |
| cloudnull | clone + build + venv is around 7 min | 19:41 |
| *** fawadkhaliq has quit IRC | 19:43 | |
| *** fawadkhaliq has joined #openstack-ansible | 19:44 | |
| *** fawadkhaliq has quit IRC | 19:45 | |
| *** fawadkhaliq has joined #openstack-ansible | 19:46 | |
| spotz | cloudnull odyssey4me very nice:) | 19:47 |
| *** elo has joined #openstack-ansible | 19:48 | |
| *** mkrish004c has joined #openstack-ansible | 19:53 | |
| mkrish004c | Hi guys, i am installing liberty HA openstack via ansible, i am able to do successfully. but the neutron agent-list and nova service list status is toggling up and down | 19:55 |
| mkrish004c | is this a problem with ha proxy settings or services issue | 19:55 |
| shanec_ | mkrish004c: hi, i've seen that when ntp isn't syncing time correctly | 19:57 |
| *** shanec_ is now known as shanec | 19:57 | |
| *** fawadkhaliq has quit IRC | 20:04 | |
| *** fawadkhaliq has joined #openstack-ansible | 20:05 | |
| *** mkrish004c has quit IRC | 20:06 | |
| cloudnull | ++ to what shanec said. | 20:06 |
| *** galstrom is now known as galstrom_zzz | 20:07 | |
| *** sdake has quit IRC | 20:17 | |
| *** mkrish004c has joined #openstack-ansible | 20:19 | |
| mkrish004c | @shanec Thanks a lot, saved my time :) | 20:19 |
| *** b3rnard0 is now known as b3rnard0_away | 20:20 | |
| stevelle | mkrish004c: pun intended? | 20:20 |
| *** admin0 has quit IRC | 20:23 | |
| *** jayc has quit IRC | 20:32 | |
| *** sdake has joined #openstack-ansible | 20:33 | |
| *** markvoelker has quit IRC | 20:37 | |
| *** mkrish004c has quit IRC | 20:41 | |
| *** rohanp_ has quit IRC | 20:50 | |
| *** galstrom_zzz is now known as galstrom | 20:56 | |
| *** admin0 has joined #openstack-ansible | 20:57 | |
| *** elo has quit IRC | 21:00 | |
| *** Nepoc has quit IRC | 21:04 | |
| *** Nepoc has joined #openstack-ansible | 21:05 | |
| *** phiche has quit IRC | 21:09 | |
| *** retreved has quit IRC | 21:16 | |
| *** thorst has quit IRC | 21:17 | |
| *** shanec has quit IRC | 21:35 | |
| *** kong has joined #openstack-ansible | 21:36 | |
| *** thorst has joined #openstack-ansible | 21:38 | |
| *** busterswt has quit IRC | 21:39 | |
| *** jthorne has quit IRC | 21:40 | |
| *** thorst has quit IRC | 21:42 | |
| *** galstrom is now known as galstrom_zzz | 21:49 | |
| *** jayc has joined #openstack-ansible | 21:51 | |
| *** markvoelker_ has joined #openstack-ansible | 22:01 | |
| *** fawadkhaliq has quit IRC | 22:05 | |
| *** fawadkhaliq has joined #openstack-ansible | 22:05 | |
| *** admin0 has quit IRC | 22:10 | |
| *** spotz is now known as spotz_zzz | 22:12 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 22:13 | |
| *** sdake_ has joined #openstack-ansible | 22:13 | |
| *** sdake has quit IRC | 22:13 | |
| *** admin0 has joined #openstack-ansible | 22:17 | |
| *** sdake has joined #openstack-ansible | 22:19 | |
| *** woodard has quit IRC | 22:20 | |
| *** sdake_ has quit IRC | 22:21 | |
| *** ametts has quit IRC | 22:28 | |
| *** karimb has quit IRC | 22:32 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-repo_build: Add the galera_client dependency to the repo build https://review.openstack.org/300686 | 22:38 |
| cloudnull | if cores are still around we need that reviewed please ^ | 22:39 |
| cloudnull | odyssey4me: -cc | 22:39 |
| *** markvoelker_ has quit IRC | 22:39 | |
| admin0 | the change is something i can vote:D but i am not a contrib yet | 22:39 |
| admin0 | go sleep :D | 22:40 |
| cloudnull | hahaha. | 22:40 |
| cloudnull | you can vote :) | 22:40 |
| cloudnull | non-cores are welcome too | 22:40 |
| cloudnull | fastest way to being core is reviewing patches :) | 22:40 |
| admin0 | aha :) | 22:41 |
| admin0 | let me try | 22:41 |
| *** scarlisle has quit IRC | 22:44 | |
| *** phalmos has quit IRC | 22:44 | |
| *** fawadkhaliq has quit IRC | 22:44 | |
| *** fawadkhaliq has joined #openstack-ansible | 22:45 | |
| admin0 | i see cherry-pick, rebase, follow-up :) | 22:45 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-repo_build: Add the galera_client dependency to the repo build https://review.openstack.org/300686 | 22:45 |
| openstackgerrit | Sashi Dahal proposed openstack/openstack-ansible-repo_build: +1 https://review.openstack.org/300688 | 22:50 |
| admin0 | ouch .. did i break anything there ? | 22:51 |
| *** bbmbx has quit IRC | 22:56 | |
| *** admin0 has quit IRC | 22:57 | |
| *** keedya has joined #openstack-ansible | 22:58 | |
| *** mgoddard has joined #openstack-ansible | 23:01 | |
| *** fawadkhaliq has quit IRC | 23:04 | |
| *** fawadkhaliq has joined #openstack-ansible | 23:04 | |
| *** mgoddard has quit IRC | 23:11 | |
| stevelle | cloudnull: starting a test of that now | 23:15 |
| cloudnull | admin0 looks like you submitted a new change | 23:16 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-repo_build: Add the galera_client dependency to the repo build https://review.openstack.org/300686 | 23:18 |
| cloudnull | stevelle: thanks , no code change there. just updating the test to include the galera_client_drop_config_file: false option | 23:18 |
| stevelle | cloudnull: well, still applying the cherry picks to the env | 23:18 |
| stevelle | the build should start momentarilly. | 23:19 |
| cloudnull | cool | 23:19 |
| cloudnull | thanks btw | 23:19 |
| *** iceyao has joined #openstack-ansible | 23:19 | |
| stevelle | yaya, also including https://review.openstack.org/#/c/300589/ in the test | 23:19 |
| cloudnull | ha. thats what lead me to this :) | 23:19 |
| stevelle | same | 23:19 |
| cloudnull | that change didnt cause this | 23:19 |
| stevelle | I was investigating and got lost down a log hole ... looked up and saw you had a patch already | 23:20 |
| cloudnull | ah , so you were seing the same? | 23:20 |
| stevelle | but I couldn't test Jesse's because of this failure | 23:20 |
| stevelle | yes, all builds fail on the galera server role | 23:20 |
| cloudnull | yup, same things | 23:20 |
| *** saneax_AFK is now known as saneax | 23:26 | |
| *** jayc has quit IRC | 23:44 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!