| *** sdake has joined #openstack-ansible | 00:02 | |
| *** sdake has quit IRC | 00:08 | |
| *** sacharya has quit IRC | 00:11 | |
| *** sacharya has joined #openstack-ansible | 00:11 | |
| *** sacharya has quit IRC | 00:15 | |
| *** sdake has joined #openstack-ansible | 00:24 | |
| *** sdake has quit IRC | 00:25 | |
| *** Jack_Iv has joined #openstack-ansible | 00:28 | |
| *** timrc has quit IRC | 00:32 | |
| *** Jack_Iv has quit IRC | 00:32 | |
| *** hw_wutianwei has joined #openstack-ansible | 00:33 | |
| *** sdake has joined #openstack-ansible | 00:38 | |
| *** ianychoi has quit IRC | 00:40 | |
| *** jamielennox is now known as jamielennox|away | 00:41 | |
| *** ianychoi has joined #openstack-ansible | 00:44 | |
| *** severion has quit IRC | 00:51 | |
| *** jamielennox|away is now known as jamielennox | 00:58 | |
| *** v1k0d3n has joined #openstack-ansible | 00:58 | |
| *** sdake_ has joined #openstack-ansible | 01:00 | |
| *** sdake has quit IRC | 01:01 | |
| *** May-meimei has quit IRC | 01:09 | |
| *** SerenaFeng has joined #openstack-ansible | 01:09 | |
| *** SerenaFeng has quit IRC | 01:12 | |
| *** ikalnytskyi has quit IRC | 01:12 | |
| *** kukacz_ has quit IRC | 01:12 | |
| *** hachi has quit IRC | 01:12 | |
| *** jrosser has quit IRC | 01:12 | |
| *** spotz has quit IRC | 01:12 | |
| *** calebb has quit IRC | 01:12 | |
| *** mcarden has quit IRC | 01:12 | |
| *** zz_pwnall1337 has quit IRC | 01:12 | |
| *** mgagne has quit IRC | 01:12 | |
| *** palendae has quit IRC | 01:12 | |
| *** bgmccollum has quit IRC | 01:12 | |
| *** b3rn-n00dl3s has quit IRC | 01:12 | |
| *** mattt has quit IRC | 01:12 | |
| *** neillc has quit IRC | 01:12 | |
| *** mattoliverau has quit IRC | 01:12 | |
| *** jhesketh has quit IRC | 01:12 | |
| *** csmart has quit IRC | 01:12 | |
| *** darrenc has quit IRC | 01:12 | |
| *** kukacz has joined #openstack-ansible | 01:12 | |
| *** csmart has joined #openstack-ansible | 01:12 | |
| *** neillc_ has joined #openstack-ansible | 01:12 | |
| *** matt6434 has joined #openstack-ansible | 01:12 | |
| *** darrenc_ has joined #openstack-ansible | 01:12 | |
| *** bgmccollum has joined #openstack-ansible | 01:12 | |
| *** hachi has joined #openstack-ansible | 01:12 | |
| *** zz_pwnall1337 has joined #openstack-ansible | 01:12 | |
| *** sacharya has joined #openstack-ansible | 01:12 | |
| *** jrosser has joined #openstack-ansible | 01:13 | |
| *** calebb has joined #openstack-ansible | 01:13 | |
| *** palendae has joined #openstack-ansible | 01:13 | |
| *** mattt has joined #openstack-ansible | 01:13 | |
| *** jhesketh has joined #openstack-ansible | 01:13 | |
| *** palendae is now known as Guest64701 | 01:13 | |
| *** mcarden has joined #openstack-ansible | 01:13 | |
| *** Guest96631 has joined #openstack-ansible | 01:13 | |
| *** neillc_ is now known as neillc | 01:13 | |
| *** darrenc_ is now known as darrenc | 01:14 | |
| *** sacharya has quit IRC | 01:14 | |
| *** sacharya has joined #openstack-ansible | 01:14 | |
| *** Guest64701 is now known as palendae | 01:14 | |
| *** mattt has quit IRC | 01:17 | |
| *** mattt has joined #openstack-ansible | 01:17 | |
| *** b3rn-n00dl3s has joined #openstack-ansible | 01:17 | |
| *** ikalnytskyi has joined #openstack-ansible | 01:17 | |
| *** timrc has joined #openstack-ansible | 01:19 | |
| *** woodard has quit IRC | 01:22 | |
| *** xgerman has quit IRC | 01:23 | |
| *** izaakk has quit IRC | 01:23 | |
| *** xar- has quit IRC | 01:23 | |
| *** qwang has quit IRC | 01:23 | |
| *** fedruantine has quit IRC | 01:23 | |
| *** ioni has quit IRC | 01:23 | |
| *** afranc has quit IRC | 01:23 | |
| *** lbragstad has quit IRC | 01:23 | |
| *** jmccrory has quit IRC | 01:23 | |
| *** dstanek has quit IRC | 01:23 | |
| *** FrankZhang has quit IRC | 01:23 | |
| *** chris_hultin|AWA has quit IRC | 01:23 | |
| *** toan has quit IRC | 01:23 | |
| *** common has quit IRC | 01:23 | |
| *** zerick has quit IRC | 01:23 | |
| *** common has joined #openstack-ansible | 01:23 | |
| *** lbragstad_ has joined #openstack-ansible | 01:23 | |
| *** jmccrory_ has joined #openstack-ansible | 01:23 | |
| *** FrankZhang has joined #openstack-ansible | 01:23 | |
| *** afranc has joined #openstack-ansible | 01:23 | |
| *** zerick has joined #openstack-ansible | 01:23 | |
| *** ioni has joined #openstack-ansible | 01:23 | |
| *** dstanek has joined #openstack-ansible | 01:24 | |
| *** ioni is now known as Guest30067 | 01:24 | |
| *** fedruantine has joined #openstack-ansible | 01:24 | |
| *** jmccrory_ is now known as jmccrory | 01:25 | |
| *** toan has joined #openstack-ansible | 01:25 | |
| *** qwang has joined #openstack-ansible | 01:26 | |
| *** chris_hultin|AWA has joined #openstack-ansible | 01:28 | |
| *** chris_hultin|AWA is now known as chris_hultin | 01:28 | |
| *** izaakk has joined #openstack-ansible | 01:28 | |
| *** xgerman has joined #openstack-ansible | 01:33 | |
| *** qwang has quit IRC | 01:38 | |
| *** qwang has joined #openstack-ansible | 01:39 | |
| *** thorst has joined #openstack-ansible | 01:39 | |
| *** mrda has quit IRC | 01:41 | |
| *** dstanek has quit IRC | 01:43 | |
| *** dstanek has joined #openstack-ansible | 01:43 | |
| *** thorst has quit IRC | 01:44 | |
| *** mrda has joined #openstack-ansible | 01:46 | |
| *** matt6434 is now known as mattoliverau | 01:48 | |
| *** v1k0d3n has quit IRC | 01:55 | |
| *** thorst has joined #openstack-ansible | 01:56 | |
| *** Mudpuppy_ has joined #openstack-ansible | 02:03 | |
| *** Trident has quit IRC | 02:08 | |
| *** thorst has quit IRC | 02:09 | |
| *** v1k0d3n has joined #openstack-ansible | 02:19 | |
| *** xgerman has quit IRC | 02:20 | |
| *** Jack_Iv has joined #openstack-ansible | 02:28 | |
| *** Jack_Iv has quit IRC | 02:32 | |
| *** Mahe has quit IRC | 02:34 | |
| *** Mahe has joined #openstack-ansible | 02:37 | |
| *** sdake_ has quit IRC | 02:40 | |
| *** meteorfox has quit IRC | 02:42 | |
| *** b3rn-n00dl3s has quit IRC | 02:42 | |
| *** ikalnytskyi has quit IRC | 02:42 | |
| *** worstadmin has quit IRC | 02:42 | |
| *** kelv has quit IRC | 02:42 | |
| *** toan has quit IRC | 02:43 | |
| *** sdake has joined #openstack-ansible | 02:43 | |
| *** meteorfox has joined #openstack-ansible | 02:43 | |
| *** jroll has quit IRC | 02:44 | |
| *** worstadmin has joined #openstack-ansible | 02:44 | |
| *** calebb has quit IRC | 02:45 | |
| *** b3rn-n00dl3s has joined #openstack-ansible | 02:45 | |
| *** kelv has joined #openstack-ansible | 02:46 | |
| *** jroll has joined #openstack-ansible | 02:46 | |
| *** calebb has joined #openstack-ansible | 02:47 | |
| *** ikalnytskyi has joined #openstack-ansible | 02:47 | |
| *** toan has joined #openstack-ansible | 02:47 | |
| *** May-meimei has joined #openstack-ansible | 02:55 | |
| *** david-lyle has quit IRC | 02:55 | |
| *** chris_hultin has quit IRC | 02:58 | |
| *** andymccr has quit IRC | 03:01 | |
| *** andymccr has joined #openstack-ansible | 03:02 | |
| *** bgmccollum has quit IRC | 03:03 | |
| *** erikmwilson has quit IRC | 03:03 | |
| *** neith has quit IRC | 03:05 | |
| *** bgmccollum has joined #openstack-ansible | 03:06 | |
| *** neith has joined #openstack-ansible | 03:06 | |
| *** chris_hultin|AWA has joined #openstack-ansible | 03:06 | |
| *** chris_hultin|AWA is now known as chris_hultin | 03:07 | |
| *** erikmwilson has joined #openstack-ansible | 03:07 | |
| *** Jeffrey4l_ has quit IRC | 03:18 | |
| *** Mudpuppy_ has quit IRC | 03:25 | |
| *** thetrav has joined #openstack-ansible | 03:36 | |
| *** askb has quit IRC | 03:38 | |
| *** askb has joined #openstack-ansible | 03:42 | |
| openstackgerrit | Logan V proposed openstack/openstack-ansible-ceph_client: Use Jewel as default Ceph release https://review.openstack.org/417738 | 03:50 |
|---|---|---|
| *** jrobinson is now known as jrobinson-afk | 03:52 | |
| openstackgerrit | Logan V proposed openstack/openstack-ansible: ceph-ansible integration https://review.openstack.org/409407 | 03:59 |
| thetrav | is it possible to configure most of my hosts to use shared ephemeral storage (I've already done that bit) but have some hosts use local ephemeral? | 04:00 |
| thetrav | ideally in a separate zone | 04:00 |
| logan- | thetrav: yes. you can split your compute hosts inventory into two separate groups, and then use group_vars to configure different storage types on each group. then add the appropriate hypervisors to appropriate host aggregates in openstack and set flavor metadata schedule certain flavors on certain host aggregates | 04:04 |
| thetrav | ok, so is that in the OSA docs somewhere? | 04:04 |
| logan- | no | 04:04 |
| thetrav | I assume it's openstack_user_config.yml ? | 04:05 |
| logan- | it is not a standard deployment setup at all | 04:05 |
| thetrav | at least part of that | 04:05 |
| thetrav | right | 04:05 |
| thetrav | so if I'm using OSA | 04:05 |
| thetrav | can I have different setups for different compute hosts? or does it mandate that they are mostly identical? | 04:05 |
| logan- | i think you could possibly use container_vars in your openstack_user_config.yml to set storage configurations at per-host granularity if you want to avoid diving into restructuring env.d stuff | 04:06 |
| logan- | so you could have a set of hosts that have the shared storage oriented container_vars set, like rbd, and some that don't | 04:07 |
| logan- | there are some container_vars examples here openstack-ansible container_vars | 04:08 |
| logan- | er | 04:08 |
| logan- | https://github.com/openstack/openstack-ansible/blob/master/etc/openstack_deploy/openstack_user_config.yml.example#L590 | 04:08 |
| thetrav | right, I saw that as an option for cinder | 04:09 |
| thetrav | and that's definitely getting towards what I want | 04:09 |
| logan- | it is a generic dict that could be used on any host, including compute_hosts, so you could use it on the shared storage hosts to set the nova rbd vars | 04:09 |
| thetrav | however I believe setting up lvm as the ephemeral storage would give me slightly better throughput than libvirt talking to a cinder volume that just happens to be co-hosted | 04:10 |
| logan- | so your default setup would be local storage, and the shared hosts would have the shared storage vars set in container_vars to override the default behavior | 04:10 |
| thetrav | erm, kvm, not libvirt | 04:10 |
| thetrav | I would prefer to do it the other way... shared is the default, local storage is a special thing we do for people who care about performance and not about their data safety | 04:11 |
| logan- | understandable but it increases the difficulty of configuring this considerably because the OSA default is local storage | 04:12 |
| logan- | it is much easier to override the rbd behavior on the hosts that need it rather than to un-override it on your local storage hosts if you go the container_vars route | 04:12 |
| logan- | if you want to dive into env.d customizations to split your compute_hosts into two separate groups, I can share some examples from what I did on my env last month. I just rolled out this exact setup. my default hypervisor is rbd, and we just added a new hypervisor setup that uses local storage for vm disks | 04:14 |
| logan- | both are offered in the same AZ and depending on which flavor the client chooses (local or shared disk), the instance is scheduled on the appropriate hypervisor | 04:15 |
| *** xgerman has joined #openstack-ansible | 04:17 | |
| *** sacharya has quit IRC | 04:24 | |
| *** sdake has quit IRC | 04:25 | |
| *** jrobinson-afk is now known as jrobinson | 04:31 | |
| *** weezS has joined #openstack-ansible | 04:38 | |
| *** cloudnull has quit IRC | 04:45 | |
| *** cloudnull has joined #openstack-ansible | 04:47 | |
| *** ivve has joined #openstack-ansible | 05:00 | |
| *** stockfisch has joined #openstack-ansible | 05:10 | |
| *** sdake has joined #openstack-ansible | 05:12 | |
| *** thetrav has quit IRC | 05:15 | |
| *** Jack_V has joined #openstack-ansible | 05:19 | |
| *** Jack_V has quit IRC | 05:23 | |
| *** sacharya has joined #openstack-ansible | 05:24 | |
| *** shausy has joined #openstack-ansible | 05:28 | |
| *** sacharya has quit IRC | 05:30 | |
| *** weezS has quit IRC | 05:33 | |
| *** sdake has quit IRC | 05:34 | |
| *** thorst has joined #openstack-ansible | 05:38 | |
| *** phalmos has quit IRC | 05:39 | |
| *** SerenaFeng has joined #openstack-ansible | 05:39 | |
| *** adrian_otto has joined #openstack-ansible | 05:40 | |
| *** thorst has quit IRC | 05:43 | |
| *** adrian_otto has quit IRC | 05:53 | |
| *** Jack_Iv has joined #openstack-ansible | 05:59 | |
| *** Jack_Iv has quit IRC | 06:14 | |
| *** Jack_Iv has joined #openstack-ansible | 06:14 | |
| *** Jack_Iv has quit IRC | 06:19 | |
| *** SerenaFeng has quit IRC | 06:34 | |
| *** pradiprwt has joined #openstack-ansible | 06:42 | |
| pradiprwt | evrardjp : Hi , Good morning | 06:44 |
| pradiprwt | evrardjp : Now everything is working fine :) | 06:44 |
| *** rgogunskiy has joined #openstack-ansible | 06:51 | |
| *** SerenaFeng has joined #openstack-ansible | 06:52 | |
| *** Jack_Iv has joined #openstack-ansible | 06:54 | |
| *** Jack_Iv has quit IRC | 06:57 | |
| *** Jack_Iv has joined #openstack-ansible | 06:58 | |
| *** h5t4 has joined #openstack-ansible | 07:00 | |
| *** Jack_Iv has quit IRC | 07:02 | |
| *** Jack_Iv has joined #openstack-ansible | 07:05 | |
| *** Jack_Iv has quit IRC | 07:05 | |
| pradiprwt | how to do some pre-configuration in horizon_local_settings.py file ? In newton branch there is no template for this file.. | 07:09 |
| pradiprwt | I want like "DISALLOW_IFRAME_EMBED = False" changes in configuration files | 07:09 |
| *** Guest30067 has quit IRC | 07:16 | |
| *** Guest30067 has joined #openstack-ansible | 07:16 | |
| *** Guest30067 is now known as ioni | 07:16 | |
| *** sacharya has joined #openstack-ansible | 07:25 | |
| *** sacharya has quit IRC | 07:30 | |
| *** jrobinson has quit IRC | 07:33 | |
| *** DeZieg has joined #openstack-ansible | 07:35 | |
| DeZieg | greetings everyone...:) | 07:37 |
| *** thorst has joined #openstack-ansible | 07:39 | |
| DeZieg | i'm successfully install openstack-ansible with All-in-One method. I created a router with public network attached. But i still have a problem, the public IP of the router can not be pinged or access from the public | 07:40 |
| DeZieg | anyone could give a clue ? | 07:41 |
| *** pcaruana has joined #openstack-ansible | 07:41 | |
| DeZieg | i'm successfully install openstack-ansible with All-in-One method. I created a router with public network attached. But i still have a problem, the public IP of the router can not be pinged or access from the public | 07:43 |
| *** thorst has quit IRC | 07:43 | |
| *** erikmwilson has quit IRC | 07:48 | |
| evrardjp | that's a good news pradiprwt | 07:48 |
| evrardjp | pradiprwt: for horizon, we cannot use config_template, but I think we template it | 07:49 |
| evrardjp | if some part of the config is missing, feel free to submit a bug or a patchset to fix it :) | 07:49 |
| *** erikmwilson has joined #openstack-ansible | 07:50 | |
| evrardjp | DeZieg: it seems your network wiring is incorrect :) | 07:50 |
| evrardjp | for this to debug, we'd basically need to know how you created your public network, on which host interface this network is wired, and if that wiring works | 07:51 |
| DeZieg | evrardjp: I'm using 10.10.2.100/24 attached on enp2s0 interface on my Node. I can ping gatewap IP address from my node | 07:55 |
| *** ethanoatetint has joined #openstack-ansible | 07:56 | |
| DeZieg | i create a router with public ip 10.10.2.204/24 attached. but i can't ping it from 10.10.2.100 | 07:57 |
| *** sacharya has joined #openstack-ansible | 07:59 | |
| *** thorst has joined #openstack-ansible | 08:00 | |
| DeZieg | Internet---Gateway(10.10.2.254/24)---Openstack Node(10.10.2.100/24)---vRouter(10.10.2.204/24) | 08:01 |
| evrardjp | that sounds wrong | 08:02 |
| evrardjp | on what bridge is linked this network on your openstack node | 08:02 |
| evrardjp | and how do you give it to neutron if you're already using it in your node? | 08:03 |
| evrardjp | it's not technically impossible, I just want to know why you've taken a difficult route | 08:03 |
| *** thorst has quit IRC | 08:04 | |
| *** sacharya has quit IRC | 08:05 | |
| evrardjp | I'd have given enp2s0 into br-flat | 08:06 |
| evrardjp | and then hop it works | 08:06 |
| evrardjp | (when you create a flat external network) | 08:07 |
| evrardjp | if that's the only source of connectivity for your openstack node | 08:08 |
| evrardjp | then you need to trick the thing a little | 08:08 |
| evrardjp | check our AIO networking | 08:08 |
| evrardjp | I don't know if it helps | 08:09 |
| *** gfa has quit IRC | 08:10 | |
| *** gfa has joined #openstack-ansible | 08:11 | |
| *** gouthamr has quit IRC | 08:11 | |
| *** gfa has joined #openstack-ansible | 08:11 | |
| *** gfa has quit IRC | 08:12 | |
| *** gfa has joined #openstack-ansible | 08:12 | |
| *** openstackgerrit has quit IRC | 08:18 | |
| DeZieg | evrardjp: all i did was follow https://developer.rackspace.com/blog/life-without-devstack-openstack-development-with-osa/ | 08:20 |
| *** openstackgerrit has joined #openstack-ansible | 08:21 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-galera_server: Allow fallback to Newton default filename https://review.openstack.org/417776 | 08:21 |
| evrardjp | DeZieg: this is the official docs for openstack-ansible AIO: http://docs.openstack.org/developer/openstack-ansible/developer-docs/quickstart-aio.html | 08:22 |
| DeZieg | it created lxc-net-bridge.cfg and osa_interfaces.cfg under /etc/network/interfaces.d | 08:22 |
| DeZieg | but it seems enp2s0 doesn't attached to any bridge.. does it sounds wrong ? | 08:24 |
| *** bsv has quit IRC | 08:24 | |
| evrardjp | yup | 08:25 |
| evrardjp | the problem here | 08:25 |
| evrardjp | you use the interface enp2s0 for your host | 08:25 |
| evrardjp | and ubuntu is not really friendly to give the nic to neutron again | 08:25 |
| evrardjp | so question | 08:25 |
| evrardjp | does enp2s0 give your connectivity to your host, or it's just another nic card? | 08:26 |
| evrardjp | if it's the only source of connectivity you need to trick ubuntu | 08:26 |
| evrardjp | wait | 08:27 |
| DeZieg | it's the only nic | 08:27 |
| evrardjp | are you only in need for flat networks, or vlans too? | 08:27 |
| evrardjp | if you only need flat, I think it could work | 08:28 |
| evrardjp | so | 08:28 |
| DeZieg | only flat for this time | 08:28 |
| evrardjp | in your network interfaces, do you have a br-flat? | 08:28 |
| evrardjp | try to have a br-flat inet static, put your IP there | 08:28 |
| evrardjp | and bridge enp2s0 | 08:29 |
| DeZieg | i see..one sec.. | 08:29 |
| evrardjp | no promises for your connectivity in neutron but it's already closer | 08:29 |
| evrardjp | else you'd need to do tricks like these: https://github.com/openstack/openstack-ansible/blob/master/etc/network/interfaces.d/aio_interfaces.cfg#L44-L68 | 08:29 |
| evrardjp | but basically | 08:29 |
| evrardjp | if you're building an AIO | 08:29 |
| evrardjp | I'm pretty sure you'd be pleased to know we have a tool for that | 08:30 |
| evrardjp | so this part of the config just above, is basically an extract on how to build an AIO | 08:30 |
| evrardjp | and all of this is explained in the official AIO docs I pasted above | 08:31 |
| evrardjp | so I'd say: go there, see if it fits your need | 08:31 |
| evrardjp | else you can try to fiddle with what I said | 08:31 |
| evrardjp | the easiest path is using this AIO tool 'though | 08:31 |
| DeZieg | which one is the AIO tool ? | 08:32 |
| DeZieg | which tool do you meant evrardjp ? | 08:34 |
| evrardjp | the scripts/bootstrap-aio.sh | 08:34 |
| evrardjp | which basically calls a role | 08:34 |
| evrardjp | bootstrap-host | 08:34 |
| DeZieg | o ya sure... i did run the scripts too | 08:35 |
| evrardjp | that configures your host as an AIO | 08:35 |
| evrardjp | it should have configured your nic, are you sure you follow the same procedure as above? | 08:35 |
| evrardjp | if that's the case, your nic should be alright | 08:35 |
| evrardjp | above -- the openstack-ansible AIO docs | 08:36 |
| *** agrebennikov has joined #openstack-ansible | 08:37 | |
| furb | evrardjp: Hi again, today back at work, all seem to run as expected. Thanks a lot. | 08:38 |
| DeZieg | i'll try the clues, and back for the results | 08:38 |
| DeZieg | thanks evrardjp :) | 08:38 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-openstack_openrc: Bring consistency to tags https://review.openstack.org/417390 | 08:39 |
| evrardjp | DeZieg: basically doing an AIO that way and running os-tempest-install will create everything you need -- it's really super simple | 08:40 |
| evrardjp | furb: yw | 08:40 |
| evrardjp | furb if your AIO node runs for a while, it's possible your mariadb will die, because OOM | 08:41 |
| furb | is there already a way to handle this? | 08:42 |
| *** agrebennikov has quit IRC | 08:42 | |
| evrardjp | on a small sized AIO, the host will kill the containers that take too much memory when OOM | 08:42 |
| evrardjp | furb: that's linux behavior you can't change that -- being on a larger node or having swap helps | 08:42 |
| evrardjp | depends on your ram | 08:42 |
| furb | AIO refers to my infra node? | 08:43 |
| evrardjp | I'm doing builds with 8GB of ram to run everything openstack -- not very much room | 08:43 |
| evrardjp | yes | 08:43 |
| evrardjp | you run everything on your infra node, right? | 08:43 |
| evrardjp | what was its size again? | 08:43 |
| *** alezzandro has joined #openstack-ansible | 08:43 | |
| evrardjp | RAM size | 08:44 |
| evrardjp | I just say that because sometimes I'm back from the week end and my AIO died | 08:44 |
| evrardjp | because OOM | 08:44 |
| evrardjp | well I could just re-start the cluster, but I'm lazy | 08:44 |
| evrardjp | or I could put more hardware to it | 08:44 |
| evrardjp | anyway it was fyi | 08:44 |
| evrardjp | openstack on a single machine is maybe a little heavy for a laptop | 08:45 |
| evrardjp | :p | 08:45 |
| DeZieg | evrardjp: you mean by following http://docs.openstack.org/developer/openstack-ansible/developer-docs/quickstart-aio.html ? | 08:45 |
| evrardjp | DeZieg: yes | 08:45 |
| furb | i have an infra node with 6gb of ram and 6gb of swap | 08:47 |
| furb | two compute nodes with about 56cpus, 120gb ram and some storage node | 08:47 |
| furb | besides this i have a devstack deployment on another node | 08:47 |
| evrardjp | the infra node AIO recommended ram size is 16GB | 08:47 |
| furb | I just was in our storage room and grabbed old hardware no one was interested in :P | 08:48 |
| evrardjp | oh yes I understand, that's how I started too :) | 08:48 |
| evrardjp | maybe you should take one compute node to be your infra | 08:48 |
| evrardjp | it has definitely lots of what you need, and you can play with redundancy there because more ram | 08:49 |
| evrardjp | :) | 08:49 |
| pradiprwt | evrardjp : default "DISALLOW_IFRAME_EMBED" is commented in horizon_local_settings.py, I want it to enable | 08:49 |
| furb | my plan is the following, testing requirements that does not need distribution and scalability on the devstack | 08:49 |
| evrardjp | pradiprwt: let me have a look one sec | 08:49 |
| pradiprwt | ok | 08:49 |
| furb | and all effects that may need it test them on my "mini-cluster" | 08:49 |
| evrardjp | furb: with OSA you maybe don't need devstack at all, cause you can select all the upstream commits you want | 08:50 |
| evrardjp | but that's another conversation :) | 08:50 |
| furb | if we think we can handle it with openstack i grab the it-guys and get a big number of user but not to old hardware | 08:50 |
| evrardjp | makes sense furb | 08:50 |
| furb | nice to read that from you evrardjp | 08:51 |
| evrardjp | haha | 08:51 |
| furb | atm i am not really sure if we can address all with cloud-products | 08:51 |
| evrardjp | pradiprwt: https://github.com/openstack/openstack-ansible-os_horizon/blob/master/templates/horizon_local_settings.py.j2#L728 could be written like https://github.com/openstack/openstack-ansible-os_horizon/blob/master/templates/horizon_local_settings.py.j2#L734 | 08:52 |
| evrardjp | pradiprwt: want me to fix that? | 08:52 |
| furb | strict tenancy and attaching some real-hardware to ones "private network" for testing may be challenging but lets start with basic stuff ;) | 08:53 |
| pradiprwt | evrardjp : no, I want a way to enable it if I want . | 08:53 |
| evrardjp | furb well not that much challenging I think, but starting with basic stuff gets you familiar with how everything works, so it all makes sense | 08:54 |
| evrardjp | pradiprwt: on which branch do you want that? Because for the moment we don't have the feature, but it's like a few lines of code to bring it | 08:55 |
| evrardjp | let me fix that | 08:55 |
| pradiprwt | evrardjp : neuton | 08:55 |
| pradiprwt | evrardjp : newton | 08:55 |
| ioni | does openstack ansible has support of changing region names? | 08:56 |
| ioni | default is RegionOne and i want a different name, but i do have already services running and also vms | 08:57 |
| ioni | i know i have to use service_region in user_variabiles | 08:58 |
| evrardjp | ioni I think it should work | 08:59 |
| evrardjp | if not it's a bug | 08:59 |
| ioni | i'm afraid that something will happen with the already running vms in RegionOne | 08:59 |
| evrardjp | oh | 09:03 |
| evrardjp | you mean live? | 09:03 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-os_horizon: Make possible to use embed iframes https://review.openstack.org/417795 | 09:03 |
| evrardjp | that's definitely harder | 09:03 |
| evrardjp | pradiprwt: ^ | 09:03 |
| evrardjp | please review this, and see if it suits your needs | 09:03 |
| evrardjp | then we can think of backporting it | 09:04 |
| evrardjp | the backport is not ensured, but at least you'll have the feature in Ocata | 09:04 |
| pradiprwt | evrardjp : can you show me file which is doing configuration changes | 09:04 |
| evrardjp | what do you mean? | 09:05 |
| pradiprwt | I will try to do some modification for my requrement | 09:05 |
| evrardjp | pradiprwt: https://review.openstack.org/#/c/417795/ | 09:05 |
| ivve | hey, im getting failure on all containers on this task: TASK [lxc_container_create : Run container veth wiring script] *****************. anything special i should be lookin for? i think i need another set of eyes on my net-configuration as i think that might be wrong.. but can't find it | 09:05 |
| *** vnogin has joined #openstack-ansible | 09:05 | |
| pradiprwt | I mean ansible file which is doing configuration of horizon | 09:05 |
| evrardjp | pradiprwt: grep horizon_local_settings.py.j2 ;) | 09:06 |
| ioni | evrardjp, yeah, live | 09:06 |
| evrardjp | sorry for this: https://www.youtube.com/watch?v=eUFY8Zw0Bag | 09:06 |
| evrardjp | that was for ioni :) | 09:07 |
| ioni | :D | 09:07 |
| evrardjp | you like to live dangerously right? :p | 09:08 |
| pradiprwt | :) | 09:08 |
| ioni | i'll test it frist | 09:09 |
| ioni | first | 09:09 |
| ioni | what about to update all configs? openstack-ansible setup-openstack --tag config should work? | 09:14 |
| ioni | or do i need to pass every config, nova-config, neutron-config etc? | 09:14 |
| evrardjp | ioni, right now I'd say you should do all the tags | 09:16 |
| evrardjp | I am working on cleaning up all these tags right now, but we don't test this use case yet | 09:16 |
| *** kysse has quit IRC | 09:23 | |
| *** kysse has joined #openstack-ansible | 09:24 | |
| pradiprwt | evrardjp : How to change horizon "https" to plain http ? | 09:30 |
| evrardjp | don't we have docs for that? | 09:31 |
| pradiprwt | pre-install changes | 09:35 |
| evrardjp | pradiprwt: only horizon or you want all your public apis to be HTTP only? | 09:35 |
| *** SerenaFeng has quit IRC | 09:36 | |
| pradiprwt | only horizon | 09:36 |
| pradiprwt | evrardjp: In liberty branch we had templates to change values before installing openstack, is there any way like that../ | 09:37 |
| pradiprwt | in newton .? | 09:38 |
| evrardjp | pradiprwt: I don't know, it's maybe horizon_external_ssl: True | 09:38 |
| evrardjp | you mean user variables? | 09:38 |
| evrardjp | that still exists | 09:39 |
| evrardjp | not sure what you mean there pradiprwt | 09:39 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-openstack_openrc: Bring consistency to tags https://review.openstack.org/417390 | 09:41 |
| pradiprwt | evrardjp : actually I am updating our existing OSA which is deploying openstack liberty. | 09:42 |
| evrardjp | sill not sure what you meant above the behavior change -- yes we changed the default behavior in https | 09:43 |
| evrardjp | pradiprwt: if you're upgrading we have tools for that | 09:43 |
| *** vnogin has quit IRC | 09:44 | |
| pradiprwt | before installing we are doing some changes those changes I want in update newton ... | 09:44 |
| odyssey4me | o/ | 09:44 |
| pradiprwt | evrardjp : here is sample http://paste.openstack.org/show/594262/ | 09:45 |
| pradiprwt | this is patch I am executing before running playbook | 09:46 |
| evrardjp | pradiprwt: well you can still patch the same file | 09:46 |
| pradiprwt | like this we have some more changes like https to http | 09:46 |
| evrardjp | https://review.openstack.org/#/c/417795/ | 09:47 |
| evrardjp | pradiprwt: maybe we need to change OSA to make sure we can deploy into HTTP | 09:47 |
| pradiprwt | but on which files we are applying patches those files are not available in newton branch | 09:48 |
| pradiprwt | https://github.com/openstack/openstack-ansible/blob/liberty-eol/playbooks/roles/os_horizon/templates/horizon_local_settings.py.j2#L712 | 09:51 |
| evrardjp | pradiprwt: roles are now separated from the main repo | 09:51 |
| evrardjp | they are now in their own repo | 09:51 |
| pradiprwt | you mean like this file https://github.com/openstack/openstack-ansible-os_horizon/blob/master/templates/horizon_local_settings.py.j2#L728 | 09:53 |
| evrardjp | exactly | 09:56 |
| evrardjp | pradiprwt: I've submitted a patch there: https://review.openstack.org/#/c/417795/ | 09:56 |
| evrardjp | for master | 09:56 |
| evrardjp | this way you don't have to patch all the time | 09:56 |
| evrardjp | just review it | 09:56 |
| evrardjp | tell me if it's fine for you | 09:56 |
| pradiprwt | evrardjp : Thanks, I will review it but there should be a way if user want to modify configuration he can modify .. :) | 09:59 |
| evrardjp | you mean shipping your own file? | 10:00 |
| evrardjp | That's something we could do too | 10:00 |
| pradiprwt | yes | 10:00 |
| evrardjp | copy content lookup template, but it seems backwards | 10:00 |
| *** thorst has joined #openstack-ansible | 10:00 | |
| pradiprwt | yes, we can do something like your patch | 10:01 |
| pradiprwt | I have reviewed, I need to provide "horizon_disallow_iframe_embed" value in which file ..? | 10:02 |
| pradiprwt | Is it in "/etc/openstack_deploy/openstack_user_config.yml" file ? | 10:02 |
| evrardjp | in a user variable file | 10:03 |
| evrardjp | like user_variables.yml | 10:03 |
| evrardjp | in that same folder | 10:03 |
| evrardjp | apply the patch on your role, define the var, and tell me if it works | 10:04 |
| pradiprwt | ok | 10:04 |
| pradiprwt | How I can apply this patch in my existing repo | 10:06 |
| *** thorst has quit IRC | 10:07 | |
| pradiprwt | I think this is different repo "openstack-ansible-os_horizon" | 10:07 |
| evrardjp | it's in ansible roles | 10:09 |
| evrardjp | /etc/ansible/roles | 10:09 |
| evrardjp | horizon | 10:09 |
| evrardjp | or something like that | 10:09 |
| evrardjp | odyssey4me: andymccr could you have a look at this? | 10:10 |
| evrardjp | https://review.openstack.org/#/c/417776/ | 10:10 |
| evrardjp | hughsaunders: ^ | 10:10 |
| *** karimb has joined #openstack-ansible | 10:12 | |
| pradiprwt | evrardjp : working fine :) | 10:24 |
| pradiprwt | please add this to newton also ... | 10:25 |
| *** vnogin has joined #openstack-ansible | 10:34 | |
| evrardjp | I can suggest it, but you need to interact in the review to say you really need it, new features generally don't go into stable branches -- but this could be considered as a bug too | 10:34 |
| openstackgerrit | Merged openstack/openstack-ansible-galera_server: Allow fallback to Newton default filename https://review.openstack.org/417776 | 10:35 |
| *** asettle has joined #openstack-ansible | 10:41 | |
| pradiprwt | evrardjp : this roles are when copying, No need to do changes in stable version.. | 10:43 |
| pradiprwt | just explain me before or after which step I need to change this role ..? | 10:43 |
| *** vnogin has quit IRC | 10:45 | |
| *** stuartgr has joined #openstack-ansible | 10:47 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_nova: Update and clean up run_tests.sh https://review.openstack.org/417071 | 10:50 |
| *** DeZieg has quit IRC | 10:59 | |
| *** DeZieg has joined #openstack-ansible | 11:01 | |
| *** sacharya has joined #openstack-ansible | 11:02 | |
| *** sacharya has quit IRC | 11:06 | |
| *** alezzandro has quit IRC | 11:07 | |
| *** vnogin has joined #openstack-ansible | 11:17 | |
| *** alezzandro has joined #openstack-ansible | 11:20 | |
| *** vnogin has quit IRC | 11:21 | |
| *** mmm_c_n has joined #openstack-ansible | 11:26 | |
| evrardjp | odyssey4me: could you have a look at this: https://review.openstack.org/#/c/413767/ ? | 11:33 |
| *** karimb has quit IRC | 11:35 | |
| *** karimb has joined #openstack-ansible | 11:40 | |
| *** mmm_c_n has left #openstack-ansible | 11:40 | |
| *** sdake has joined #openstack-ansible | 11:41 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-galera_server: Remove percona repo if not used https://review.openstack.org/417832 | 11:43 |
| odyssey4me | evrardjp yep, I don't think that patch is necessary at all - unfortunately I can't review it because my ubuntu login is failing and the password reset function appears to not be working either | 11:47 |
| evrardjp | probably the leap second thing it broke the internet! | 11:47 |
| odyssey4me | I just need to verify where/how that file is used before I make a final call... I'm just trying to catch up on emails at the moment. :) | 11:48 |
| odyssey4me | if the constraints file is only ever used for a -c parameter then filtering is is entirely unnecessary | 11:50 |
| odyssey4me | if it's used as a -r parameter then it's another matter | 11:50 |
| odyssey4me | mattt ^ | 11:50 |
| *** DeZieg has quit IRC | 11:52 | |
| *** DeZieg has joined #openstack-ansible | 11:53 | |
| *** karimb has quit IRC | 11:55 | |
| evrardjp | it seems only -c | 11:57 |
| evrardjp | or pip wheel --constraints | 11:57 |
| odyssey4me | 'ray I'm finally logged in, and my emails are done. | 11:57 |
| evrardjp | -s | 11:57 |
| evrardjp | so I think you're right | 11:57 |
| evrardjp | well it doesn't hurt but it's not useful too | 11:59 |
| odyssey4me | well, it does hurt - it's more unnecessary machinery | 12:00 |
| ioni | i forgot how to list all available tags, can you guys remind me? :D | 12:00 |
| odyssey4me | if we don't need to add scripted logic, we shouldn't | 12:00 |
| *** karimb has joined #openstack-ansible | 12:00 | |
| evrardjp | --list-tags | 12:01 |
| evrardjp | IIRC | 12:01 |
| odyssey4me | evrardjp I'm not seeing any information about how to replicate https://bugs.launchpad.net/openstack-ansible/+bug/1649329 so that I can look into it | 12:01 |
| openstack | Launchpad bug 1649329 in openstack-ansible "14.0.3 repo build error on Ubuntu 14.04" [High,Fix released] - Assigned to Jesse Pretorius (jesse-pretorius) | 12:01 |
| evrardjp | odyssey4me: I agree, not worth merging | 12:01 |
| evrardjp | odyssey4me: the issue is the cloning, not really the constraints | 12:03 |
| evrardjp | *tilt I remember* | 12:04 |
| evrardjp | more details could be useful anyway | 12:04 |
| *** smatzek has joined #openstack-ansible | 12:06 | |
| odyssey4me | evrardjp see the last comment from alextricity25 - that's probably the reason why https://review.openstack.org/#/c/413767/ was submitted | 12:06 |
| *** vnogin has joined #openstack-ansible | 12:10 | |
| *** drifterza has joined #openstack-ansible | 12:11 | |
| drifterza | Hello all | 12:11 |
| evrardjp | hello drifterza, welcome back! | 12:12 |
| drifterza | Thanks evrardjp | 12:12 |
| drifterza | happy new years and all ;) | 12:12 |
| evrardjp | to you too | 12:13 |
| drifterza | Thanks. | 12:13 |
| *** vnogin has quit IRC | 12:13 | |
| *** hw_wutianwei has quit IRC | 12:15 | |
| *** thorst has joined #openstack-ansible | 12:24 | |
| *** thorst has quit IRC | 12:24 | |
| *** thorst has joined #openstack-ansible | 12:25 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_heat: Allow admin role name to be changed https://review.openstack.org/417853 | 12:26 |
| openstackgerrit | Merged openstack/openstack-ansible-os_trove: Update and clean up run_tests.sh https://review.openstack.org/417076 | 12:26 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-pip_install: Provide toggle for get-pip.py get_url task force option https://review.openstack.org/417855 | 12:28 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-tests: Prevent repeated get-pip.py fetches in role tests https://review.openstack.org/417856 | 12:28 |
| evrardjp | The repository 'http://mirror.regionone.infracloud-chocolate.openstack.org/ubuntu xenial Release' does not have a Release file. | 12:31 |
| evrardjp | that's new | 12:31 |
| *** retreved has joined #openstack-ansible | 12:34 | |
| evrardjp | isn't infracloud supposed to be dead? | 12:35 |
| *** pjm6 has quit IRC | 12:35 | |
| *** pjm6 has joined #openstack-ansible | 12:40 | |
| *** shausy has quit IRC | 12:54 | |
| *** shausy has joined #openstack-ansible | 12:55 | |
| *** jmckind has joined #openstack-ansible | 12:56 | |
| evrardjp | odyssey4me: https://review.openstack.org/#/c/408015/ | 13:01 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-os_nova: Allow override of the repo filename https://review.openstack.org/416251 | 13:02 |
| *** sacharya has joined #openstack-ansible | 13:03 | |
| *** Jack_Iv has joined #openstack-ansible | 13:03 | |
| *** Jack_Iv has quit IRC | 13:03 | |
| *** Jack_Iv has joined #openstack-ansible | 13:04 | |
| *** sacharya has quit IRC | 13:07 | |
| *** kylek3h has quit IRC | 13:07 | |
| *** Jack_Iv_ has joined #openstack-ansible | 13:08 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-os_nova: Allow override of the repo filename https://review.openstack.org/416251 | 13:08 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-repo_build: Allow override of the repo filename https://review.openstack.org/416266 | 13:09 |
| *** ethanoatetint has quit IRC | 13:10 | |
| *** marc_ab has joined #openstack-ansible | 13:10 | |
| mattt | odyssey4me: i didn't quite follow your comments there | 13:10 |
| *** Jack_Iv has quit IRC | 13:11 | |
| evrardjp | mattt: if it's only a constraint, not a requirement, it shouldn't matter | 13:11 |
| evrardjp | (simplest version) | 13:11 |
| evrardjp | so I guess the question is how to reproduce this -- length of the groups should be fine, and git clone shouldn't be a problem | 13:12 |
| mattt | evrardjp: it says in the description | 13:13 |
| mattt | evrardjp: when we remove the conf.d/swift.yml file, repo-build fails | 13:13 |
| mattt | is that an expected outcome? | 13:13 |
| *** vnogin has joined #openstack-ansible | 13:13 | |
| evrardjp | the question would be what happens if env.d is emptied | 13:14 |
| evrardjp | but still | 13:14 |
| evrardjp | I see your point too | 13:14 |
| mattt | alextricity25: are you still running into this issue deploying newton? | 13:14 |
| *** ethanoatetint has joined #openstack-ansible | 13:15 | |
| evrardjp | maybe we should just make the lookup better from the start | 13:15 |
| evrardjp | and not have a shitload of content we don't care | 13:16 |
| evrardjp | I mean smarter | 13:16 |
| evrardjp | this way the clone would not have these tricks, and neither would require the constraints building | 13:16 |
| evrardjp | but that's a rather radical change | 13:17 |
| evrardjp | for that, cloudnull's input would be valuable | 13:17 |
| mattt | evrardjp: i'm so out of touch with osa at the moment that i can't really offer any structural advice :) | 13:17 |
| *** vnogin has quit IRC | 13:18 | |
| *** woodard has joined #openstack-ansible | 13:19 | |
| evrardjp | In other words: I saw both your points, and they are valid. I found a middle ground, but it's not an easy one :D | 13:19 |
| *** whiteveil has joined #openstack-ansible | 13:22 | |
| mattt | evrardjp odyssey4me : i may be missing something obvious here, updated the review with a comment | 13:28 |
| mattt | evrardjp odyssey4me : alextricity25 said we can use selective_git_cloning=false as a work-around, which we can certainly do | 13:29 |
| mattt | but it does sound like there is still an issue somewhere | 13:29 |
| evrardjp | that workaround is what is it: barely a workaround | 13:30 |
| evrardjp | let me build an RPC master | 13:30 |
| evrardjp | I'll see what happens | 13:30 |
| *** gouthamr has joined #openstack-ansible | 13:31 | |
| alextricity25 | evrardjp: make sure it's without swift ;) | 13:31 |
| evrardjp | lol | 13:31 |
| evrardjp | ofc | 13:31 |
| alextricity25 | ofc? | 13:32 |
| evrardjp | of course | 13:32 |
| alextricity25 | I can't keep up with yall's hipster language | 13:32 |
| mgariepy | good morning and happy new year everyone ! | 13:32 |
| evrardjp | 14.04 or 16? | 13:32 |
| evrardjp | happy new year mgariepy :) | 13:32 |
| alextricity25 | evrardjp: either should produce the same results here | 13:32 |
| evrardjp | I know but I had to chose, so I asked what you prefered... too late I picked 14! | 13:33 |
| alextricity25 | hehehe | 13:33 |
| evrardjp | I think bindep should be included in ubuntu distros | 13:33 |
| evrardjp | I'm lazy! | 13:34 |
| *** kylek3h has joined #openstack-ansible | 13:34 | |
| *** fguillot has joined #openstack-ansible | 13:34 | |
| *** asettle has quit IRC | 13:35 | |
| mattt | mgariepy: did you fall down a hole? :) | 13:35 |
| *** asettle has joined #openstack-ansible | 13:36 | |
| mhayden | morning folks | 13:36 |
| andymccr | mornin all :) and welcome back to those who have their first day back today! | 13:37 |
| *** Jack_Iv_ has quit IRC | 13:39 | |
| mattt | evrardjp: thank ya! | 13:39 |
| evrardjp | yw | 13:39 |
| *** jamielennox is now known as jamielennox|away | 13:39 | |
| evrardjp | andymccr: you forgot "lucky ones" | 13:39 |
| andymccr | nah evrardjp - the lucky ones are the ones who have not taken a break at all and have just been able to work on through :D | 13:40 |
| evrardjp | yes. Sure I agree. | 13:40 |
| *** Matias has joined #openstack-ansible | 13:41 | |
| andymccr | hahahaha | 13:41 |
| *** Matias has quit IRC | 13:41 | |
| mgariepy | mattt, lol nop, i wasn't in a hurry to come back to the office :) haha | 13:42 |
| odyssey4me | alextricity25 mattt so, prior to the holidays, I did a standard OSA AIO build without a conf.d/swift.yml file and it did the job just fine - see the compute scenario result here: https://review.openstack.org/#/c/370638/ | 13:42 |
| odyssey4me | so I imagine that the issue may have something to do with something else in the config | 13:43 |
| *** Matias has joined #openstack-ansible | 13:43 | |
| evrardjp | odyssey4me: I'm building one right now the way mattt deploys, I can debug better that way | 13:43 |
| evrardjp | probably some export variable override or something | 13:44 |
| odyssey4me | evrardjp ok cool | 13:44 |
| evrardjp | mattt: I exported DEPLOY_SWIFT=no and ./scripts/deploy.sh | 13:44 |
| evrardjp | just in case I forgot something | 13:44 |
| evrardjp | when the deploy-host role will run I will have enough data to run the lookup and see | 13:45 |
| evrardjp | I'll still run it through everything to see if the repo build fails or not | 13:45 |
| *** thorst has quit IRC | 13:45 | |
| *** thorst has joined #openstack-ansible | 13:46 | |
| mattt | evrardjp: yeah that's all i did iirc | 13:47 |
| mattt | odyssey4me: weird, then we're hitting something else | 13:47 |
| odyssey4me | yeah, I'm curious to understand how it's happening | 13:48 |
| evrardjp | what's this thing in infra that fails all my patches? infracloud-chocolate? | 13:49 |
| evrardjp | Is this a new year's present that's really terrible? | 13:49 |
| odyssey4me | it's one of the regions in infracloud | 13:49 |
| evrardjp | so infracloud still exists | 13:49 |
| odyssey4me | well yes, it was rebuilt and made active at the end of last year | 13:50 |
| *** jheroux has joined #openstack-ansible | 13:50 | |
| odyssey4me | if there's an issue I'd recommend raising it in the infra channel | 13:50 |
| *** thorst has quit IRC | 13:50 | |
| evrardjp | I will | 13:54 |
| evrardjp | mattt: I gave the name of the node rpc-no-swift, guess who didn't think he's gonna grep swift a lot? <-- \o/ | 13:54 |
| openstackgerrit | Logan V proposed openstack/openstack-ansible-os_keystone: Fix venv build fallback mechanism https://review.openstack.org/417039 | 13:57 |
| *** drifterza has quit IRC | 14:00 | |
| *** karimb has quit IRC | 14:02 | |
| *** klamath has joined #openstack-ansible | 14:02 | |
| *** klamath has quit IRC | 14:02 | |
| mattt | evrardjp: lol! | 14:02 |
| *** klamath has joined #openstack-ansible | 14:02 | |
| *** thorst has joined #openstack-ansible | 14:05 | |
| alextricity25 | evrardjp mattt odyssey4me: Looks like it's still a problem: "Collecting swift from git+file:///var/www/repo/openstackgit/swift@bf74d5a0183c9db38a6ff9bf1ab564ea80def703#egg=swift&gitname=swift&projectgroup=swift_all (from -c /var/www/repo/os-releases/14.0.4/requirements_constraints.txt (line 34))", " Cloning file:///var/www/repo/openstackgit/swift (to bf74d5a0183c9db38a6ff9bf1ab564ea80def703) to /tmp/openstack-builder/swift"], " | 14:07 |
| alextricity25 | warnings": []}" | 14:07 |
| alextricity25 | Full error here: https://gist.github.com/alextricity25/f37bbfd16c49b861dcb3050d96635c5a | 14:07 |
| mattt | thanks alextricity25 | 14:10 |
| evrardjp | odyssey4me: https://gist.github.com/evrardjp/5a13a9a24d676eb8c07383bb3cb2cfc3 | 14:12 |
| evrardjp | mattt: alextricity25 ^ | 14:12 |
| evrardjp | alextricity25: yes I just got it | 14:13 |
| odyssey4me | alextricity25 mattt evrardjp aha, so there's a rogue requirement | 14:13 |
| alextricity25 | lol | 14:13 |
| ivve | im getting fails here with a deploy, im guessing i have to be able to ssh into containers from the outside? TASK [lxc_container_create : Wait for ssh to be available] ********************* | 14:13 |
| evrardjp | odyssey4me: indeed | 14:13 |
| alextricity25 | those rebellious little requirements thing they can sneek in there | 14:13 |
| alextricity25 | s/thing/think | 14:14 |
| odyssey4me | so the question is how the requirement got there | 14:14 |
| evrardjp | odyssey4me: I'm tracking it | 14:14 |
| *** automagically_ has joined #openstack-ansible | 14:14 | |
| ivve | how does OSA set this up, should it be with iptables? i can't find anything relevant in iptables :( | 14:14 |
| evrardjp | or at least I will start tracking | 14:14 |
| ivve | im guessing i should have had "is_ssh_address:" = false in user_config under provider_networks for the br-mgmt(?) | 14:17 |
| ivve | would deployment still work? | 14:17 |
| odyssey4me | ivve by default we expect to be able to ssh into the containers on the br-mgmgt address for the container | 14:18 |
| odyssey4me | the container is already setup to allow that | 14:18 |
| *** cathrichardson has joined #openstack-ansible | 14:18 | |
| *** aleph1 is now known as agarner | 14:18 | |
| odyssey4me | so you'll ned to verify that your deployment node has access to that interface - try sshing to it yourself manually | 14:18 |
| odyssey4me | if that does work, then you'll have to troubleshoot why ansible can't | 14:18 |
| ivve | odyssey4me: hmm okay that is wierd because i set the cidr_networks: correctly and it is correctly setting up in the container | 14:19 |
| odyssey4me | if it doesn't, then you'll have to figure out why your network connectivity for the containers is broken | 14:19 |
| *** lbragstad_ is now known as lbragstad | 14:19 | |
| ivve | however there is no iptable nor a route for it to understand that, how does it access the container..? | 14:19 |
| ivve | ofc, the container can move out | 14:19 |
| odyssey4me | it will access it directly | 14:20 |
| odyssey4me | br-mgmt should be setup as a bridge which is on that CIDR | 14:20 |
| odyssey4me | no iptables involved | 14:20 |
| ivve | but that doesn't work unless you set it up from scratch (or is ansible supposed to do it for you?) i guess that is my main qquestion | 14:20 |
| odyssey4me | as per the install guide each host must have that bridge setup prior to the deployment | 14:20 |
| ivve | the bridge is there | 14:21 |
| *** cathrich_ has quit IRC | 14:21 | |
| odyssey4me | if you have it correctly setup, then ansible will create the container interface on that bridge | 14:21 |
| ivve | (otherwise container wouldn't get the veth to work?) | 14:21 |
| ivve | exactly, it did | 14:21 |
| odyssey4me | yep, so check whether you can ping the container address on that network | 14:21 |
| ivve | but can't ssh.. which i understand that it can't... because there is no route | 14:21 |
| odyssey4me | from the deployment host | 14:21 |
| ivve | i have no route, so it won't work ;( | 14:22 |
| odyssey4me | why is there no route? does your deployment host not have an interface on that network? | 14:22 |
| ivve | i understand the problem, im just not sure if it is intended or not | 14:22 |
| *** mgariepy has quit IRC | 14:22 | |
| ivve | hmm maybe i can paste some bin | 14:22 |
| ivve | easier to understand if my explaining is crap :D | 14:22 |
| ivve | the conf: https://hastebin.com/lacenopuno.bash | 14:23 |
| *** mgariepy has joined #openstack-ansible | 14:24 | |
| ivve | network on all "openstack" hosts.. but a few different ips (ofc): https://hastebin.com/mofebagome.css | 14:25 |
| odyssey4me | ivve FYI the internal and external lb vip address cannot be the same - both HTTP and HTTPS cannot be served on the same IP and same ports | 14:25 |
| evrardjp | rpc-maas | 14:25 |
| evrardjp | mattt: ^ | 14:25 |
| ivve | oh | 14:25 |
| evrardjp | maas_pip_packages | 14:25 |
| ivve | could it explain the problem? | 14:25 |
| mattt | ^^^ alextricity25 | 14:26 |
| mattt | :) | 14:26 |
| mattt | evrardjp: ok, cool ... thanks for digging into that one! | 14:26 |
| evrardjp | yw | 14:26 |
| evrardjp | let me try to rebuild without it | 14:27 |
| ivve | and finally, a container showing connectivity.. but obviously not back to it (lacking a route from deployhost): https://hastebin.com/ehamoduvew.pl | 14:27 |
| ivve | obvious error: https://hastebin.com/tonikugici.rb | 14:28 |
| ivve | putting in a route for 10.10.10.0/24 via 10.1.90.104 would probably give it connectivity? | 14:29 |
| *** rgogunskiy has quit IRC | 14:29 | |
| ivve | however that network is used over 3 hosts... | 14:29 |
| ivve | 104, 105 & 106 | 14:30 |
| *** smatzek has quit IRC | 14:30 | |
| odyssey4me | ivve ok, so typically many people use one of the infra hosts as the deploy node | 14:30 |
| ivve | nono, the deploy host is just a deployhost in my case (not sure what you mean) | 14:30 |
| odyssey4me | if you aren't doing that, then obviously you need to ensure that your deploy node has a route to the br-mgmt network | 14:30 |
| ivve | aha lol :D | 14:30 |
| odyssey4me | or more ideally has an interface on that network | 14:30 |
| ivve | so okay | 14:30 |
| ivve | next question | 14:31 |
| ivve | since i have 3 different infrahosts | 14:31 |
| ivve | wouldn't that screw things up for routes? | 14:31 |
| odyssey4me | well, no - they all have an interface on br-mgmt right? | 14:32 |
| odyssey4me | I'm not sure I understand what the problem is. | 14:32 |
| ivve | adding 10.10.10.0/24 via 10.1.90.104 and the same for 105 and 106 | 14:32 |
| odyssey4me | You can easily add a route to your deploy node, or add an interface on that CIDR and ensure that the IP is excluded. | 14:32 |
| ivve | from managementhost (10.1.90.100) i can't connect to container 10.10.10.162 (as example) on infrahost 10.1.90.104 | 14:32 |
| ivve | due to lacking route | 14:32 |
| odyssey4me | oh, I wouldn't use the server as a router - rather use an L3 switch if you can | 14:33 |
| ivve | however, i have 3 infrahosts | 14:33 |
| odyssey4me | if you can't, then rather add an interface to your deploy node on that network | 14:33 |
| odyssey4me | I guess you could also add each infra host as a gateway, but have different weights | 14:33 |
| ivve | problem is im trying to mimic a 3 infra node deployment that is close to a production environment | 14:33 |
| ivve | but i don't have physical servers | 14:33 |
| ivve | im guessing i could allow each container to get its own IP via another interface on a different net? | 14:34 |
| odyssey4me | the physical or virtual switch should be capable of having that route added | 14:34 |
| *** mathlin has quit IRC | 14:35 | |
| ivve | yea i was under the impression that it was solved locally with haproxy/iptables/routes created by OSA :) | 14:35 |
| odyssey4me | nope, we don't modify your deployment node | 14:35 |
| ivve | but if each container gets it's own routable ip then.. | 14:35 |
| odyssey4me | well, typically the routing into or out of the environment is handles on your infrastructure routers/swithces | 14:37 |
| odyssey4me | *switches | 14:37 |
| ivve | so if i had set the "container: 10.10.10.0/24" to 10.1.90.0/24 it would work... im guessing? | 14:37 |
| odyssey4me | we don't modify those, as those are out of scope | 14:37 |
| evrardjp | I didn't follow your conversation, but please stop using the term l3 switch. For me it means nothing at all. When I hear l3 switch, I feel sad and puppies/kittens die. Please don't. | 14:38 |
| odyssey4me | you could, if you wanted to, add those to all the containers and hosts I guess... but that seems quite a lot of extra stuff to manage when you could just add the route to your switch | 14:38 |
| odyssey4me | evrardjp forgive my archaic networking knowledge | 14:38 |
| evrardjp | vendors love the term l3 switch though | 14:38 |
| odyssey4me | I learned networking in the 90's :p | 14:38 |
| ivve | its alright i understand what you mean tho | 14:38 |
| evrardjp | OSI layers precede that! | 14:38 |
| odyssey4me | what is the correct term for a switch that has routing capabilities then? | 14:39 |
| odyssey4me | ie it handles both L2 and L3 | 14:39 |
| ivve | hmm | 14:39 |
| ivve | so i need to get each individual container to get its own IP, skipping routes | 14:40 |
| ivve | that would be easiest? | 14:40 |
| odyssey4me | ivve it already has its own IP's | 14:40 |
| odyssey4me | your deployment host just needs a route to br-mgmt | 14:40 |
| odyssey4me | well, the CIDR on br-mgmt | 14:40 |
| ivve | yeah i mean right now it was a pair to the br-mgmt but its in a different cidr | 14:41 |
| odyssey4me | each container will have at least two interfaces | 14:41 |
| odyssey4me | the first (eth0) is the interface used for internet access, which is NATted through the host | 14:42 |
| odyssey4me | the second (a veth on br-mgmt) is used for management | 14:42 |
| odyssey4me | others may or may not be present depending on the container's role in the infra | 14:42 |
| ivve | problem is im trying to get this to run in a virtual environment and i have only access to this one vlan (vmware crap) | 14:43 |
| ivve | and i have no permissions except for creating new machines :P | 14:43 |
| odyssey4me | oh, well then why not just add a secondary IP address to your deployment host's NIC? | 14:43 |
| odyssey4me | a secondary IP on that CIDR | 14:43 |
| ivve | im not following now | 14:44 |
| odyssey4me | so right now you have a single NIC on a vlan and multiple CIDR's traversing it | 14:44 |
| odyssey4me | right? | 14:44 |
| *** karimb has joined #openstack-ansible | 14:44 | |
| odyssey4me | ie all hosts, including your deployment host, are on the same VLAN? | 14:45 |
| ivve | i was hoping that tunnel and storage cidr were tunneled but im guessing not now that i understand how this work | 14:45 |
| ivve | yes | 14:45 |
| ivve | (first time doing lxc+bridges) | 14:45 |
| odyssey4me | the tunnel and storage networks can be entirely isolated - you should not need to access them except from the hosts | 14:46 |
| ivve | that was my impression :) | 14:46 |
| ivve | so i had them in a different cidr, however my config might not reflect that.. im unsure | 14:46 |
| odyssey4me | ok, if they're all on the same vlan then you should just be able to add a secondary IP address to your deployment host which is an IP in the br-mgmt CIDR | 14:46 |
| odyssey4me | it will then just magically work | 14:47 |
| ivve | okay, to iron this out 's#container: 10.10.10.0/24#container: 10.1.90.0/24#g' /etc/openstack_deploy/openstack_user_config.yml ? :D | 14:48 |
| ivve | that is what is assigned to container eth1 | 14:48 |
| ivve | and that is on br-mgmt | 14:48 |
| ivve | i also added another IP for external_lb_vip | 14:49 |
| odyssey4me | heh, oh you could do that if 10.1.90.x is your host management network | 14:49 |
| ivve | it is! | 14:49 |
| odyssey4me | but, bear in mind that if you do that change you need to remove all the containers first | 14:50 |
| odyssey4me | then remove them from the inventory | 14:50 |
| odyssey4me | then you can make the change | 14:50 |
| odyssey4me | otherwise you'll end up with duplicate containers | 14:50 |
| *** shausy has quit IRC | 14:50 | |
| ivve | since deploy_node is 10.1.90.100 | 14:50 |
| ivve | and each of the infra nodes are: -"-.103 , 104 and 105 respectively | 14:50 |
| odyssey4me | yeah, that makes sense | 14:50 |
| odyssey4me | don't forget to set another IP for the external lb vip adderss | 14:50 |
| ivve | then containers would be getting addresses in the 10.1.90.0/24 cidr | 14:51 |
| ivve | yea u did that | 14:51 |
| odyssey4me | *address | 14:51 |
| ivve | all this would be so much easier with trunks :D | 14:52 |
| ivve | ye ill run the destroy container playbook | 14:52 |
| ivve | that should do the trick | 14:52 |
| odyssey4me | that'll remove the containers | 14:52 |
| odyssey4me | you can then use the inventory-manage script to remove the existing containers from the inventory | 14:53 |
| odyssey4me | alternatively you can just delete the existing inventory if you don't need it | 14:53 |
| ivve | the json files? | 14:53 |
| odyssey4me | but remove the containers from the hosts first! | 14:53 |
| ivve | in /etc/openstack_deploy/* ? | 14:53 |
| odyssey4me | yeah, the json files - and the fact cache | 14:54 |
| evrardjp | asettle: https://bugs.launchpad.net/openstack-ansible/+bug/1655019 | 14:54 |
| openstack | Launchpad bug 1655019 in openstack-ansible "cant access this link: http://docs.openstack.org/developer/openstack-ansible/install-guide/index.html" [Undecided,New] | 14:54 |
| evrardjp | this kind of bugs | 14:54 |
| evrardjp | it was another one I've seen this morning | 14:54 |
| evrardjp | but the gist is there | 14:54 |
| ivve | odyssey4me: great help man, i've gotten some great insight on how this works now :) | 14:55 |
| asettle | evrardjp: they should be marked as duplicates of another. There's no way this link can be accessed or searched, it's people that have the link bookmarked. andymccr was going to look into a redirect. | 14:55 |
| odyssey4me | ivve no problem, happy to help | 14:55 |
| ivve | hopefully i will be able to contribute to this project some day, i like it a lot :) | 14:55 |
| openstackgerrit | Merged openstack/openstack-ansible-openstack_openrc: Bring consistency to tags https://review.openstack.org/417390 | 14:55 |
| evrardjp | well that's not the same thing | 14:55 |
| *** BjoernT has joined #openstack-ansible | 14:55 | |
| evrardjp | I'll discuss that tomorrow with you if you prefer | 14:56 |
| asettle | evrardjp: it is a duplicate of this bug: https://bugs.launchpad.net/openstack-ansible/+bug/1653228 | 14:56 |
| openstack | Launchpad bug 1653228 in openstack-ansible "Install Guide not found (404)" [High,Incomplete] - Assigned to Alexandra Settle (alexandra-settle) | 14:56 |
| asettle | They are reporting the same thing. | 14:56 |
| openstackgerrit | Andy McCrae proposed openstack/openstack-ansible-os_keystone: Allow Apache ports to be specified per VHost https://review.openstack.org/417907 | 14:56 |
| evrardjp | asettle: well there are two things to do: fix the google ppl landing on 404 with a redirect (if possible) AND fix our docs that internally has wrong links | 14:58 |
| palendae | I can't get a 404ing link from a google search of 'openstack-ansible install guide' right now | 14:59 |
| asettle | evrardjp: so, I can tell you there are no internal links within our internal documentation that link to the install.html link - that was fixed here: https://bugs.launchpad.net/openstack-ansible/+bug/1652948 I need to get a backport for newton/mitaka - but we're only looking at master | 14:59 |
| openstack | Launchpad bug 1652948 in openstack-ansible "Broken links in documentation" [High,In progress] - Assigned to Andy McCrae (andrew-mccrae) | 14:59 |
| palendae | Neither bug mentions google, though | 15:00 |
| asettle | And secondly, as palendae you do not get a 404 from that google search. You are linked to either the dev docs (which has an up-to-date link) or the deploy guide | 15:00 |
| asettle | palendae: they do not. Which is why I've marked one bug as incomplete and requested further information regarding WHY they are searching using that link | 15:00 |
| asettle | As that link they have both reported does *not* exist at all. | 15:00 |
| *** galstrom is now known as galstrom_zzz | 15:00 | |
| palendae | 👍 | 15:00 |
| asettle | Google search or otherwise. | 15:00 |
| asettle | That is why I am saying that it is most likely someone who has previously book marked the link | 15:00 |
| evrardjp | asettle: this morning I found one | 15:01 |
| *** smatzek has joined #openstack-ansible | 15:01 | |
| evrardjp | let me find it again and fix it | 15:01 |
| evrardjp | I'm NOT becoming INSANE (I think) | 15:01 |
| evrardjp | :p | 15:01 |
| asettle | evrardjp: that would be helpful, but I should have fixed them all from this bug : https://bugs.launchpad.net/openstack-ansible/+bug/1652948 | 15:01 |
| openstack | Launchpad bug 1652948 in openstack-ansible "Broken links in documentation" [High,In progress] - Assigned to Andy McCrae (andrew-mccrae) | 15:01 |
| asettle | Now I feel like *i'm* insane | 15:01 |
| *** Mudpuppy_ has joined #openstack-ansible | 15:01 | |
| asettle | Although as palendae said, I can definitely verify that you cannot get to that link from a google search | 15:02 |
| palendae | Maybe some other search engine has it...but actually having someone mention how they got there would be much more helpful than gussing :) | 15:02 |
| palendae | guessing* | 15:02 |
| *** h5t4 has quit IRC | 15:02 | |
| asettle | palendae: precisely. I just haven't got any info from the reportee :( | 15:02 |
| evrardjp | asettle: got it on master | 15:02 |
| evrardjp | ok let me fix this | 15:02 |
| evrardjp | this is gonna be easier | 15:02 |
| cloudnull | mornings | 15:02 |
| palendae | Morning cloudnull | 15:03 |
| palendae | Glad to see you're done harassing your dogs | 15:03 |
| cloudnull | hahaha | 15:03 |
| cloudnull | never done. | 15:03 |
| cloudnull | but I am back to work :) | 15:03 |
| asettle | evrardjp: sure | 15:03 |
| asettle | What's the link? | 15:03 |
| *** sacharya has joined #openstack-ansible | 15:03 | |
| asettle | I'm surprised teh scrapy tool missed it | 15:03 |
| evrardjp | morning cloudnull !!!! | 15:04 |
| asettle | Anyway, I'm going back to being sick | 15:04 |
| asettle | Good luck evrardjp | 15:04 |
| evrardjp | sorry to hear that asettle | 15:04 |
| evrardjp | get well soon | 15:04 |
| evrardjp | cloudnull: happy new year | 15:05 |
| evrardjp | and nice work at home ;) | 15:05 |
| cloudnull | ohnos asettle is out spreading sickness again?! | 15:06 |
| * cloudnull runs | 15:06 | |
| cloudnull | j/k -- feel better asettle | 15:07 |
| cloudnull | o/ evrardjp | 15:07 |
| cloudnull | happy new year to you too and thanks. the home work has been fun :) | 15:07 |
| *** sacharya has quit IRC | 15:08 | |
| evrardjp | result is amazing | 15:10 |
| evrardjp | I'm waiting for you to come in Belgium for inspirations :D | 15:10 |
| evrardjp | I've continued cleaning up tags for your commit (artifacts).. lots of work ahead | 15:11 |
| openstackgerrit | Andy McCrae proposed openstack/openstack-ansible: Fix broken link in OpenStack-Ansible Newton https://review.openstack.org/417918 | 15:12 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Separate host type and container type with a dash https://review.openstack.org/407655 | 15:15 |
| *** vnogin has joined #openstack-ansible | 15:15 | |
| *** mmm_c_n_ has joined #openstack-ansible | 15:17 | |
| *** dxiri has joined #openstack-ansible | 15:17 | |
| dxiri | good morning OSA | 15:18 |
| *** mmm_c_n_ is now known as mmm_c_n | 15:18 | |
| dxiri | I have a problem while running setup hosts, wondering if you guys can help, here is the error: http://paste.openstack.org/show/594296/ | 15:18 |
| *** Jack_Iv has joined #openstack-ansible | 15:18 | |
| dxiri | I need to make this work with only a single bridge so I tried to replicate that in the config | 15:18 |
| *** vnogin has quit IRC | 15:19 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible: [Docs] Fix some 404 in navigation https://review.openstack.org/417924 | 15:22 |
| *** markvoelker has joined #openstack-ansible | 15:24 | |
| *** karimb has quit IRC | 15:25 | |
| *** karimb has joined #openstack-ansible | 15:28 | |
| andymccr | evrardjp: i think we have quite a few links that are erroneous | 15:29 |
| *** asettle has quit IRC | 15:29 | |
| palendae | Yeah, the relative links not being updated makes sense | 15:29 |
| andymccr | basically a grep for "install-guide" in the repo shows quite a few :P | 15:29 |
| evrardjp | I just did it where I saw it, but yes maybe a more thorough job needs to be done | 15:30 |
| evrardjp | put it in my review | 15:30 |
| andymccr | its harder to spot the reverse - where it references developer docs | 15:30 |
| evrardjp | :p | 15:30 |
| andymccr | im creating a PR for stable/newton - mostly because i started there | 15:30 |
| *** asettle has joined #openstack-ansible | 15:30 | |
| evrardjp | andymccr: what do you mean? | 15:30 |
| evrardjp | It doesn't seem we include the navigation in the deployment guide | 15:31 |
| andymccr | well its easy to spot the refs for install-guide, but if there is a ref in the deploy-guide that is now in the developer docs, it'll fail | 15:31 |
| andymccr | and its not easy to spot | 15:31 |
| palendae | andymccr: So relative links to dev docs from deploy-guide? | 15:32 |
| andymccr | palendae: yeah - except they arent listed as "dev-docs" because i assume when the install guide was in the same place it would just work | 15:32 |
| palendae | Yeah | 15:32 |
| andymccr | but yeah thats the use case. | 15:32 |
| andymccr | or rather error case :P | 15:32 |
| palendae | Probably 100% relative | 15:32 |
| palendae | Like ../../ | 15:32 |
| andymccr | maybe it would work to just grep "../developer-docs" inside the deploy guide path | 15:33 |
| *** spotz_zzz is now known as spotz | 15:34 | |
| *** spotz_zzz has joined #openstack-ansible | 15:35 | |
| *** Jeffrey4l_ has joined #openstack-ansible | 15:35 | |
| *** cmart has joined #openstack-ansible | 15:35 | |
| evrardjp | andymccr: in the rendered would be far easier | 15:36 |
| evrardjp | because else it could be relative | 15:36 |
| *** marst has quit IRC | 15:36 | |
| evrardjp | anyway these could be different bugs | 15:36 |
| *** vnogin has joined #openstack-ansible | 15:36 | |
| *** drifterza has joined #openstack-ansible | 15:36 | |
| *** whiteveil has quit IRC | 15:37 | |
| *** phalmos has joined #openstack-ansible | 15:37 | |
| andymccr | evrardjp: yeah true, but if we're linking to a page that doesnt exist its still a fail :P | 15:39 |
| evrardjp | I'm lost -- are you talking about https://review.openstack.org/#/c/417924/ ? | 15:39 |
| openstackgerrit | Andy McCrae proposed openstack/openstack-ansible: Fix broken link in OpenStack-Ansible Newton https://review.openstack.org/417918 | 15:39 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-os_horizon: Ensure 0755 default mode on new directories https://review.openstack.org/417929 | 15:39 |
| andymccr | evrardjp: https://bugs.launchpad.net/openstack-ansible/+bug/1652948 | 15:39 |
| openstack | Launchpad bug 1652948 in openstack-ansible "Broken links in documentation" [High,In progress] - Assigned to Andy McCrae (andrew-mccrae) | 15:39 |
| andymccr | mostly | 15:39 |
| andymccr | i think this is a complete list for newton: https://review.openstack.org/#/c/417918/ | 15:40 |
| ivve | odyssey4me: reran the setup-hosts.yml but no cigar.. containers get correct cidr for eth1 but eth0 is configured with lxc defaults 10.0.3.0/24 (not sure if this is intended). containers pair veth doesn't grab an IP on the host, just in lxc-net.. (cloud3_aodh_container-96767365 RUNNING 1 onboot, openstack 10.0.3.226, 10.1.90.148) but 96767365_eth1 does not get | 15:40 |
| ivve | an IP.. :( and i can't remember how the config should look like when it should reach out.. | 15:40 |
| *** sdake_ has joined #openstack-ansible | 15:40 | |
| evrardjp | andymccr: well my fix is for master | 15:40 |
| andymccr | evrardjp: i realise that :P | 15:41 |
| andymccr | my point is there are a tonne of failed links | 15:41 |
| andymccr | i'll basically repeat the process and fix it up so no stress | 15:41 |
| *** vnogin has quit IRC | 15:41 | |
| evrardjp | agreed, let's take them all out , just step by step | 15:41 |
| evrardjp | ok, do I need to abandon mine, or? | 15:41 |
| andymccr | evrardjp: nope dont - i'll just PR on top of it | 15:42 |
| evrardjp | ok | 15:42 |
| andymccr | ugh i guess there are some rel-notes that need to get fixed too sigh :P | 15:42 |
| andymccr | but perhaps not worth fixing those | 15:42 |
| evrardjp | wait | 15:43 |
| evrardjp | how do the others do? | 15:43 |
| *** dxiri has quit IRC | 15:43 | |
| andymccr | ? | 15:43 |
| *** sdake has quit IRC | 15:43 | |
| evrardjp | brainfart :p | 15:43 |
| *** dxiri has joined #openstack-ansible | 15:44 | |
| palendae | andymccr: A somewhat minor thing I found; would make it easier to patch in master then backport, though | 15:44 |
| *** markvoelker has quit IRC | 15:44 | |
| *** marst has joined #openstack-ansible | 15:45 | |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible: Ensure RHEL 6 STIG content is used https://review.openstack.org/417932 | 15:46 |
| andymccr | palendae: ahh nice so i just replace that in the links? | 15:47 |
| andymccr | although to be fair none of this is uniform so im wondering... | 15:47 |
| palendae | I think so; I hope replacement happens in links. https://github.com/openstack/openstack-ansible/blob/master/deploy-guide/source/conf.py#L305 might be a better var to use since it's all lower case | 15:48 |
| andymccr | for deploy guide thats going to be "draft" :/ | 15:48 |
| palendae | >.< | 15:48 |
| andymccr | i think i'll -w that pr | 15:48 |
| odyssey4me | andymccr only in master? | 15:48 |
| palendae | I think the conf.py for deploy-guide was just a copy paste from docs | 15:48 |
| andymccr | and speak to the experts | 15:48 |
| andymccr | odyssey4me: newton exists for newton | 15:49 |
| *** phalmos has quit IRC | 15:49 | |
| palendae | andymccr: Annoying, but could also put some logic in the conf.py to s/master/draft/ | 15:49 |
| andymccr | makes me think that hte newton links are half pointing to master | 15:51 |
| palendae | D'oh | 15:51 |
| andymccr | docs are making my head hurt and me feel pretty stupid :P | 15:51 |
| palendae | Should doc the documenting process for docs | 15:51 |
| andymccr | hahaha yes yes | 15:52 |
| palendae | I believe the release name var stuff was added during last midcycle | 15:52 |
| palendae | Or close to | 15:52 |
| *** whiteveil has joined #openstack-ansible | 15:52 | |
| palendae | Not that time added is very relevant | 15:52 |
| odyssey4me | andymccr so it should be easy enough to change the conf to use 'draft' if the branch is master, otherwise use the branch name | 15:52 |
| *** mmm_c_n has quit IRC | 15:52 | |
| palendae | Yeah | 15:53 |
| odyssey4me | lemme fine where we do that for the watermark | 15:53 |
| odyssey4me | hang a sec | 15:53 |
| odyssey4me | lol | 15:53 |
| odyssey4me | https://github.com/openstack/openstack-ansible/blob/master/deploy-guide/source/conf.py#L331-L333 | 15:53 |
| palendae | if branch == "master": release_name = "draft" else release_name = branch | 15:53 |
| andymccr | odyssey4me: yeah i dont think that'd be too bad. although dev-docs dont exist at draft (only deploy-guide) | 15:54 |
| *** KLevenstein has joined #openstack-ansible | 15:54 | |
| palendae | andymccr: Dev docs are a different conf.py | 15:54 |
| andymccr | oh yeah true | 15:54 |
| odyssey4me | basically that grabs the branch name, if there isn't one then it's master so the watermark is set to 'Pre-release' | 15:54 |
| palendae | https://github.com/openstack/openstack-ansible/blob/master/deploy-guide/source/conf.py vs https://github.com/openstack/openstack-ansible/blob/master/doc/source/conf.py | 15:54 |
| palendae | odyssey4me: There's also https://github.com/openstack/openstack-ansible/blob/master/deploy-guide/source/conf.py#L302-L305 :p | 15:54 |
| odyssey4me | if you use this mechanism then it's easy to just use a substitution reference in the docs which will adapt automatically. | 15:54 |
| andymccr | ok so fix in master and then backport - but going to have to ensure some consistency in links | 15:54 |
| *** deadnull has joined #openstack-ansible | 15:55 | |
| odyssey4me | https://bugs.launchpad.net/openstack-ansible/+bug/1620367 | 15:55 |
| openstack | Launchpad bug 1620367 in openstack-ansible "Doc: Links for docs outside the current doc tree do not transfer to stable branches properly" [Wishlist,Triaged] | 15:55 |
| odyssey4me | also, if we did that then we'd be implementing the solution for ^ | 15:56 |
| andymccr | that seems necessary - if im following a newton deploy-guide and any links push me to master, that seems error prone. although this link management seems quite a difficult task | 15:57 |
| palendae | Seems any logic would need to be duplicated between the two conf files | 15:57 |
| palendae | So they can link back properly | 15:57 |
| odyssey4me | yeah, unfortunately | 15:58 |
| odyssey4me | but if we do it right it'll take all the maintenance out of it | 15:58 |
| *** h5t4 has joined #openstack-ansible | 15:59 | |
| *** rpittau has joined #openstack-ansible | 15:59 | |
| andymccr | worth it | 16:00 |
| andymccr | ok | 16:00 |
| *** galstrom_zzz is now known as galstrom | 16:01 | |
| palendae | Hell | 16:01 |
| palendae | Probably worth putting the early part of the URLs in a var | 16:02 |
| andymccr | yeah id agree with that | 16:02 |
| odyssey4me | we kinda do for the git url: https://github.com/openstack/openstack-ansible/blob/master/deploy-guide/source/conf.py#L83 | 16:02 |
| odyssey4me | so yeah, that could obviously be extended a little | 16:03 |
| palendae | deploy_guide_url = 'http://docs.openstack.org/project-deploy-guide/openstack-ansible/%s' % branch_link_name | 16:03 |
| andymccr | seems good to me | 16:03 |
| palendae | or with .format for more better | 16:03 |
| andymccr | we can probably use full instead of relative too | 16:03 |
| palendae | Similar for dev docs | 16:04 |
| palendae | Using the vars for absolute everywhere would certainly be more explicit | 16:04 |
| andymccr | palendae: yeah i think so, and i think it'd help be less error prone. | 16:04 |
| palendae | Yeah | 16:04 |
| palendae | Now that they're not in the same dir structure it makes sense | 16:05 |
| palendae | Especially considering it took like a week for us to realize where these reports were originating from | 16:05 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-galera_client: Remove old repos with different filenames https://review.openstack.org/417943 | 16:05 |
| andymccr | palendae: yeah thats the thing, Andreas did a good job of pointing a whole bunch out - but that wasnt complete either | 16:06 |
| *** phalmos has joined #openstack-ansible | 16:10 | |
| openstackgerrit | Kyle L. Henderson proposed openstack/openstack-ansible-os_heat: Enable oslo_middleware proxy header parsing https://review.openstack.org/417945 | 16:10 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-os_neutron: Keep the N behavior optionally https://review.openstack.org/417946 | 16:11 |
| *** ivve has quit IRC | 16:12 | |
| *** Jack_Iv has quit IRC | 16:12 | |
| *** furb has quit IRC | 16:13 | |
| *** sacharya has joined #openstack-ansible | 16:13 | |
| *** weezS has joined #openstack-ansible | 16:14 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-os_neutron: Keep the N behavior optionally https://review.openstack.org/417946 | 16:14 |
| *** Jack_Iv has joined #openstack-ansible | 16:17 | |
| odyssey4me | evrardjp quite honestly I think the new default for Ocata should be to use static file names for any apt/yum sources we lay down | 16:17 |
| evrardjp | odyssey4me: we can't backport then | 16:18 |
| evrardjp | that is not my goal | 16:18 |
| odyssey4me | it makes upgrading a lot easier, and also ensures that if the repo url is changed, the file is simply overwritten | 16:18 |
| odyssey4me | we can still backport, it just means that in the backport we need to make it optional | 16:18 |
| *** furb has joined #openstack-ansible | 16:18 | |
| evrardjp | which is exactly that | 16:18 |
| odyssey4me | that doesn't mean that we can't make the Ocata default something more sane | 16:18 |
| evrardjp | exactly | 16:18 |
| evrardjp | let's do it after | 16:18 |
| openstackgerrit | Bjoern Teipel proposed openstack/openstack-ansible-plugins: Implementation Neutron SR-IOV https://review.openstack.org/416362 | 16:19 |
| evrardjp | the default(omit) is a good thing | 16:19 |
| odyssey4me | lol, ok so you want to do a two step process so that you have a cherry-pick backport... whereas I would simply just do a different implementation in the backport | 16:19 |
| evrardjp | I don't see the point arguing if we agree that we can change it later in master | 16:20 |
| *** Jeffrey4l_ has quit IRC | 16:22 | |
| odyssey4me | hardly arguing, it's just not evident in your commit in any way that you plan a two stage process | 16:22 |
| evrardjp | before the few commits recently it was not using filenames in master for UCA, it didn't change but now the deploy can do it if he wants | 16:22 |
| *** adrian_otto has joined #openstack-ansible | 16:22 | |
| evrardjp | and he can manage it because he knows the default filename -- I don't see what's wrong by , on top of it, being completely optional and backport compatible | 16:23 |
| *** Jeffrey4l_ has joined #openstack-ansible | 16:24 | |
| *** phalmos has quit IRC | 16:24 | |
| *** rromans has quit IRC | 16:27 | |
| *** rromans has joined #openstack-ansible | 16:27 | |
| *** schwicht has joined #openstack-ansible | 16:28 | |
| *** Matias has quit IRC | 16:30 | |
| *** deadnull has quit IRC | 16:31 | |
| *** pjm6_ has joined #openstack-ansible | 16:31 | |
| *** TxGirlGeek has joined #openstack-ansible | 16:31 | |
| *** Jack_Iv has quit IRC | 16:32 | |
| *** Jack_Iv has joined #openstack-ansible | 16:32 | |
| *** markvoelker has joined #openstack-ansible | 16:32 | |
| openstackgerrit | Bjoern Teipel proposed openstack/openstack-ansible-plugins: Implementation Neutron SR-IOV https://review.openstack.org/416362 | 16:33 |
| *** pjm6 has quit IRC | 16:33 | |
| *** Matias has joined #openstack-ansible | 16:35 | |
| *** pjm6 has joined #openstack-ansible | 16:35 | |
| *** Jeffrey4l__ has joined #openstack-ansible | 16:35 | |
| *** Jack_Iv has quit IRC | 16:36 | |
| *** pjm6_ has quit IRC | 16:36 | |
| *** Jack_Iv has joined #openstack-ansible | 16:36 | |
| *** Jack_Iv_ has joined #openstack-ansible | 16:37 | |
| *** Jack_Iv has quit IRC | 16:37 | |
| *** pcaruana has quit IRC | 16:38 | |
| *** Jack_Iv_ has quit IRC | 16:38 | |
| *** Jeffrey4l_ has quit IRC | 16:39 | |
| *** japestinho has joined #openstack-ansible | 16:41 | |
| *** japestinho has quit IRC | 16:44 | |
| *** adrianreza has joined #openstack-ansible | 16:45 | |
| *** dxiri has quit IRC | 16:47 | |
| *** cmart has quit IRC | 16:48 | |
| *** dxiri has joined #openstack-ansible | 16:48 | |
| *** karimb has quit IRC | 16:50 | |
| *** adrianreza has quit IRC | 16:51 | |
| *** Jeffrey4l__ has quit IRC | 16:51 | |
| *** japestinho has joined #openstack-ansible | 16:52 | |
| *** dxiri has quit IRC | 16:53 | |
| *** dxiri_ has joined #openstack-ansible | 16:53 | |
| *** ivve has joined #openstack-ansible | 16:53 | |
| *** drifterza has quit IRC | 16:55 | |
| *** david-lyle has joined #openstack-ansible | 16:57 | |
| *** mmm_c_n has joined #openstack-ansible | 16:58 | |
| *** mmm_c_n has quit IRC | 17:03 | |
| *** weezS has quit IRC | 17:03 | |
| *** weezS has joined #openstack-ansible | 17:05 | |
| *** cmart has joined #openstack-ansible | 17:06 | |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Add conditional branch names for docs links https://review.openstack.org/417976 | 17:10 |
| openstackgerrit | Darren Wardlow proposed openstack/openstack-ansible-os_nova: Add nova-lxd storage-pool option for ZFS storage backend Choosing zfs storage backend for nova-lxd requires specifying a storage pool name. This option is not in the template. This adds an option to specify the storage pool name if a variable has been po https://review.openstack.org/417977 | 17:11 |
| *** jamesdenton has joined #openstack-ansible | 17:13 | |
| openstackgerrit | Bjoern Teipel proposed openstack/openstack-ansible-plugins: Implementation Neutron SR-IOV https://review.openstack.org/416362 | 17:13 |
| *** asettle has quit IRC | 17:14 | |
| *** alezzandro has quit IRC | 17:14 | |
| *** h5t4 has quit IRC | 17:22 | |
| *** karimb has joined #openstack-ansible | 17:22 | |
| *** japestinho has quit IRC | 17:22 | |
| *** KLevenstein has quit IRC | 17:23 | |
| *** h5t4 has joined #openstack-ansible | 17:24 | |
| *** japestinho has joined #openstack-ansible | 17:25 | |
| *** jlockwood has joined #openstack-ansible | 17:27 | |
| *** furb has quit IRC | 17:28 | |
| *** weezS has quit IRC | 17:31 | |
| *** jheroux has quit IRC | 17:32 | |
| *** karimb has quit IRC | 17:34 | |
| *** jheroux has joined #openstack-ansible | 17:38 | |
| *** KLevenstein has joined #openstack-ansible | 17:38 | |
| *** lkoranda has quit IRC | 17:40 | |
| *** lkoranda has joined #openstack-ansible | 17:44 | |
| *** rmelero has joined #openstack-ansible | 17:50 | |
| alextricity25 | Does anyone know what happened to the auth_identity_uri variable used here? | 17:50 |
| alextricity25 | https://github.com/openstack/openstack-ansible-os_swift/blob/master/doc/source/configure-swift-glance.rst | 17:51 |
| alextricity25 | I can't seem to find it anywhere in the integrated repo nor the role repos | 17:51 |
| *** Guest96631 is now known as mgagne | 17:51 | |
| *** mgagne has quit IRC | 17:51 | |
| *** mgagne has joined #openstack-ansible | 17:51 | |
| *** whiteveil has quit IRC | 17:51 | |
| *** ivve has quit IRC | 17:51 | |
| odyssey4me | alextricity25 it looks like that is a doc error - see https://github.com/openstack/openstack-ansible/blob/eol-kilo/scripts/upgrade-utilities/scripts/new-variable-prep.sh#L99-L102 | 17:54 |
| odyssey4me | so basically from liberty onwards it should be using a different var | 17:55 |
| odyssey4me | hmm, actually kilo onwards | 17:55 |
| odyssey4me | that var is from juno | 17:55 |
| alextricity25 | Ah. That would explain why I don't have it :) Thanks odyssey4me | 17:55 |
| *** lkoranda has quit IRC | 17:57 | |
| cloudnull | so finally though trying to catch up from being away. where are we with things? anything specifically that we need to work on? | 17:58 |
| *** cmart has quit IRC | 18:01 | |
| *** cmart has joined #openstack-ansible | 18:02 | |
| *** DeZieg has quit IRC | 18:02 | |
| *** lkoranda has joined #openstack-ansible | 18:04 | |
| *** japestinho has quit IRC | 18:08 | |
| odyssey4me | Alright, I'm out for the night. Catch y'all tomorrow. | 18:08 |
| *** chris_hultin is now known as chris_hultin|AWA | 18:12 | |
| *** whiteveil has joined #openstack-ansible | 18:12 | |
| *** whiteveil has quit IRC | 18:14 | |
| *** drifterza has joined #openstack-ansible | 18:18 | |
| *** ivve has joined #openstack-ansible | 18:20 | |
| *** Jeffrey4l_ has joined #openstack-ansible | 18:23 | |
| *** Jack_Iv has joined #openstack-ansible | 18:26 | |
| *** sdake_ has quit IRC | 18:28 | |
| *** sdake has joined #openstack-ansible | 18:29 | |
| *** thorst has quit IRC | 18:30 | |
| *** Jack_Iv has quit IRC | 18:30 | |
| *** thorst has joined #openstack-ansible | 18:31 | |
| *** sdake_ has joined #openstack-ansible | 18:31 | |
| *** sdake has quit IRC | 18:33 | |
| *** thorst has quit IRC | 18:35 | |
| *** KLevenstein has quit IRC | 18:35 | |
| *** whiteveil has joined #openstack-ansible | 18:36 | |
| *** dxiri_ has quit IRC | 18:36 | |
| *** sdake has joined #openstack-ansible | 18:39 | |
| *** drifterza has quit IRC | 18:41 | |
| *** sdake_ has quit IRC | 18:42 | |
| *** whiteveil_ has joined #openstack-ansible | 18:46 | |
| *** chris_hultin|AWA is now known as chris_hultin | 18:46 | |
| *** lkoranda has quit IRC | 18:47 | |
| *** thorst has joined #openstack-ansible | 18:49 | |
| *** whiteveil has quit IRC | 18:49 | |
| *** whiteveil_ is now known as whiteveil | 18:49 | |
| *** joelparker has joined #openstack-ansible | 18:49 | |
| *** KLevenstein has joined #openstack-ansible | 18:50 | |
| *** lkoranda has joined #openstack-ansible | 18:51 | |
| *** cmart has quit IRC | 19:04 | |
| *** joelparker has left #openstack-ansible | 19:07 | |
| openstackgerrit | Logan V proposed openstack/openstack-ansible-ceph_client: Use Jewel as default Ceph release https://review.openstack.org/417738 | 19:12 |
| *** ianychoi has quit IRC | 19:14 | |
| *** rmelero has quit IRC | 19:15 | |
| *** dxiri has joined #openstack-ansible | 19:15 | |
| *** rmelero has joined #openstack-ansible | 19:16 | |
| *** phalmos has joined #openstack-ansible | 19:17 | |
| *** stuartgr has quit IRC | 19:18 | |
| *** cmart has joined #openstack-ansible | 19:18 | |
| *** TxGirlGeek has quit IRC | 19:19 | |
| *** rmelero has quit IRC | 19:20 | |
| *** phalmos has quit IRC | 19:22 | |
| *** weezS has joined #openstack-ansible | 19:23 | |
| *** david-lyle has quit IRC | 19:28 | |
| mgariepy | new release of ospurge if someone is interested https://github.com/openstack/ospurge/commit/3a612efd536b721851351cfcd55198aa98e2f2a8 | 19:29 |
| *** rmelero has joined #openstack-ansible | 19:34 | |
| *** KLevenstein has quit IRC | 19:35 | |
| *** MasterOfBugs has joined #openstack-ansible | 19:36 | |
| *** vnogin has joined #openstack-ansible | 19:37 | |
| *** Jack_Iv has joined #openstack-ansible | 19:41 | |
| *** Jack_Iv has quit IRC | 19:43 | |
| *** vnogin has quit IRC | 19:44 | |
| *** phalmos has joined #openstack-ansible | 19:47 | |
| *** whiteveil_ has joined #openstack-ansible | 19:47 | |
| *** cmart has quit IRC | 19:48 | |
| *** pcaruana has joined #openstack-ansible | 19:50 | |
| *** whiteveil has quit IRC | 19:50 | |
| *** whiteveil_ is now known as whiteveil | 19:50 | |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Restructure lib directory to install via pip https://review.openstack.org/418076 | 19:52 |
| *** TxGirlGeek has joined #openstack-ansible | 19:54 | |
| *** karimb has joined #openstack-ansible | 19:55 | |
| *** TxGirlGeek has quit IRC | 19:57 | |
| *** weezS has quit IRC | 20:02 | |
| *** weezS has joined #openstack-ansible | 20:05 | |
| *** furb has joined #openstack-ansible | 20:05 | |
| *** phalmos has quit IRC | 20:06 | |
| *** cmart has joined #openstack-ansible | 20:06 | |
| *** gouthamr has quit IRC | 20:06 | |
| *** phalmos has joined #openstack-ansible | 20:06 | |
| *** KLevenstein has joined #openstack-ansible | 20:06 | |
| *** jlockwood has quit IRC | 20:10 | |
| *** dxiri has quit IRC | 20:13 | |
| *** weezS has quit IRC | 20:14 | |
| *** phalmos has quit IRC | 20:15 | |
| *** phalmos has joined #openstack-ansible | 20:16 | |
| *** automagically_ has quit IRC | 20:17 | |
| *** dxiri has joined #openstack-ansible | 20:21 | |
| *** dxiri has quit IRC | 20:32 | |
| *** dxiri has joined #openstack-ansible | 20:32 | |
| *** david-lyle has joined #openstack-ansible | 20:37 | |
| *** TxGirlGeek has joined #openstack-ansible | 20:59 | |
| *** jamielennox|away is now known as jamielennox | 21:00 | |
| *** dxiri has quit IRC | 21:00 | |
| *** dxiri has joined #openstack-ansible | 21:00 | |
| *** timrc has quit IRC | 21:01 | |
| *** weezS has joined #openstack-ansible | 21:01 | |
| openstackgerrit | Merged openstack/openstack-ansible-ops: Sort out the hosts.json. https://review.openstack.org/417469 | 21:06 |
| *** jlockwood has joined #openstack-ansible | 21:06 | |
| *** vnogin has joined #openstack-ansible | 21:17 | |
| *** timsim has joined #openstack-ansible | 21:20 | |
| *** KLevenstein has quit IRC | 21:21 | |
| *** dfflanders has joined #openstack-ansible | 21:21 | |
| *** galstrom is now known as galstrom_zzz | 21:22 | |
| *** weezS has quit IRC | 21:22 | |
| *** smatzek has quit IRC | 21:23 | |
| *** v1k0d3n has quit IRC | 21:24 | |
| *** stockfisch has quit IRC | 21:26 | |
| *** jlockwood1 has joined #openstack-ansible | 21:26 | |
| *** jlockwood2 has joined #openstack-ansible | 21:27 | |
| *** jlockwood1 has quit IRC | 21:27 | |
| *** ivve has quit IRC | 21:28 | |
| *** jlockwood has quit IRC | 21:28 | |
| *** KLevenstein has joined #openstack-ansible | 21:31 | |
| *** whiteveil has quit IRC | 21:33 | |
| *** dxiri_ has joined #openstack-ansible | 21:34 | |
| *** Jeffrey4l__ has joined #openstack-ansible | 21:34 | |
| *** Jeffrey4l_ has quit IRC | 21:35 | |
| *** vnogin has quit IRC | 21:35 | |
| *** dxiri has quit IRC | 21:36 | |
| *** weezS has joined #openstack-ansible | 21:36 | |
| *** dxiri_ has quit IRC | 21:38 | |
| *** dxiri has joined #openstack-ansible | 21:38 | |
| *** cmart has quit IRC | 21:39 | |
| *** cmart1 has joined #openstack-ansible | 21:39 | |
| *** fguillot has quit IRC | 21:44 | |
| *** fguillot has joined #openstack-ansible | 21:44 | |
| *** retreved has quit IRC | 21:46 | |
| *** cmart1 has quit IRC | 21:49 | |
| *** timrc has joined #openstack-ansible | 21:49 | |
| *** Mudpuppy_ has quit IRC | 21:52 | |
| *** cmart has joined #openstack-ansible | 22:00 | |
| *** whiteveil has joined #openstack-ansible | 22:00 | |
| *** fguillot has quit IRC | 22:01 | |
| *** adrian_otto has quit IRC | 22:10 | |
| *** jheroux has quit IRC | 22:11 | |
| *** jmckind has quit IRC | 22:14 | |
| *** agrebennikov has joined #openstack-ansible | 22:20 | |
| agrebennikov | hey cloudnull are you around by chance? | 22:20 |
| *** mattoliverau has quit IRC | 22:21 | |
| *** matt6434 has joined #openstack-ansible | 22:21 | |
| *** matt6434 is now known as mattoliverau | 22:22 | |
| *** darrenc is now known as darrenc_afk | 22:23 | |
| *** galstrom_zzz is now known as galstrom | 22:31 | |
| cloudnull | agrebennikov: I am | 22:32 |
| cloudnull | hows it ? | 22:32 |
| agrebennikov | cloudnull, great | 22:33 |
| agrebennikov | 22:33 | |
| agrebennikov | look at this please | 22:34 |
| agrebennikov | https://github.com/openstack/openstack-ansible-os_horizon/blame/3502e699253069460602f8ba5290fed0c6bd2368/templates/horizon_local_settings.py.j2#L210 | 22:34 |
| agrebennikov | just wanted to ask - if I see commits like that - who is responsible for the line in fact then? | 22:34 |
| agrebennikov | just to clarify - talking about this issue https://bugs.launchpad.net/openstack-ansible/+bug/1655167 | 22:35 |
| openstack | Launchpad bug 1655167 in openstack-ansible "Horizon variable uses missing file horizon_cacert_pem" [Undecided,New] - Assigned to Andrey Grebennikov (agrebennikov) | 22:35 |
| agrebennikov | I mean - maybe I have to talk to the feature maintainer but not sure how I find proper person (if it is not you) | 22:35 |
| cloudnull | ah. | 22:36 |
| cloudnull | I wont claim to be the feature maintainer. but I'm happy to talk about the issue. | 22:36 |
| agrebennikov | oh, sure | 22:36 |
| agrebennikov | do I just fix it? | 22:36 |
| cloudnull | I did however do a lot of work on the horizon role . so maybe that's me :) | 22:37 |
| agrebennikov | asking you first because there is a lot of this kind of commits | 22:37 |
| agrebennikov | "making the code galaxy style" | 22:37 |
| agrebennikov | whatever | 22:37 |
| cloudnull | yea. that was me . | 22:37 |
| cloudnull | a lot of that was me removing the rackspace-isms from teh roles | 22:37 |
| agrebennikov | that's fine, but in the future - if I see commits like that do I have to always ask you first? | 22:37 |
| cloudnull | I'd say just fix it, if it's broken. | 22:37 |
| cloudnull | no. | 22:38 |
| agrebennikov | but how do I know the name then? | 22:38 |
| cloudnull | I'm a firm believer of "see a bug, fix a bug". | 22:38 |
| cloudnull | the name of the maintainer ? | 22:38 |
| agrebennikov | yeah | 22:38 |
| agrebennikov | because maybe there was another one before | 22:38 |
| agrebennikov | and you say refactored it to be more "galaxy stuyle" | 22:39 |
| agrebennikov | see what I mean? | 22:39 |
| cloudnull | I'd have to ping other cores however I don't believe we have set maintianers on the various roles. | 22:39 |
| agrebennikov | so all random? | 22:39 |
| cloudnull | random meaning a lot of folks have wroked on many of the roles? | 22:39 |
| agrebennikov | yeah, kind of :) | 22:40 |
| cloudnull | it's fairly diverse, so yeah, kinda random. | 22:40 |
| agrebennikov | and backwards - nobody is focused on a particular role | 22:40 |
| agrebennikov | ok | 22:40 |
| agrebennikov | that's fine | 22:40 |
| agrebennikov | but regarding this one | 22:41 |
| *** jamesdenton has quit IRC | 22:41 | |
| agrebennikov | is it ok to just hardcode local file? | 22:41 |
| agrebennikov | in the template | 22:41 |
| agrebennikov | lets say, /etc/horizon/keystone-ca.pem | 22:41 |
| agrebennikov | whatever | 22:41 |
| cloudnull | I think that's changing a bit as some folks have spent more time on particular roles but yes I do agree that there have been a lot of hands in almost all of the roles. | 22:41 |
| *** adrian_otto has joined #openstack-ansible | 22:42 | |
| agrebennikov | otherwise it will be more vars | 22:42 |
| agrebennikov | which is unlikely | 22:42 |
| cloudnull | yes. i'd say so. I think having the known target file is better overall. | 22:42 |
| *** KLevenstein has quit IRC | 22:42 | |
| agrebennikov | or | 22:42 |
| agrebennikov | probably it is going to be Always keystone CA | 22:43 |
| cloudnull | and I don't really see the benifit in having a configurable target file for a CA cert | 22:43 |
| agrebennikov | well, it is | 22:43 |
| agrebennikov | because I'm configuring a CA for keystone | 22:43 |
| agrebennikov | *keystone's cert | 22:43 |
| agrebennikov | so horizon should be using it | 22:43 |
| agrebennikov | to be able to connect to keystone properly | 22:43 |
| cloudnull | right so if that file is provided it'd always be dropped as "/etc/horizon/keystone-ca.pem" | 22:43 |
| cloudnull | which i think is a good thing | 22:44 |
| agrebennikov | no, wait. What I just thought about - I can just use that file which is copied as the part of keystone playbook | 22:44 |
| agrebennikov | into | 22:44 |
| agrebennikov | sec | 22:44 |
| cloudnull | agrebennikov: have you looked into anchor? | 22:45 |
| agrebennikov | sorry? | 22:45 |
| agrebennikov | (I'm probably not familiar yet with this one) | 22:45 |
| agrebennikov | it it a exported resources? | 22:45 |
| agrebennikov | (from puppet world) | 22:45 |
| cloudnull | https://github.com/openstack/anchor | 22:45 |
| cloudnull | you were talking about setting up a CA and while it's not totally realted to that bug it is CA related and potentially helpful. | 22:47 |
| cloudnull | we don't have a role for it | 22:47 |
| agrebennikov | err... I guess it it way more complicated :) | 22:47 |
| agrebennikov | but I definitely have to read that | 22:47 |
| cloudnull | yea, maybe more so than you need. | 22:47 |
| agrebennikov | since right now I'm working on wrapping all services into ssl | 22:47 |
| cloudnull | I think in the future it'd be nice to have something like that as an option to internally secure a cloud | 22:47 |
| agrebennikov | and from the CA standpoint it may be helpful | 22:48 |
| cloudnull | nice ! | 22:48 |
| agrebennikov | nice..... but I need to turn all them behind apache | 22:48 |
| agrebennikov | which is pain in the ass | 22:48 |
| cloudnull | indeed it si | 22:48 |
| cloudnull | *is | 22:48 |
| agrebennikov | so what I was going to say | 22:48 |
| agrebennikov | sec | 22:48 |
| agrebennikov | :) | 22:48 |
| cloudnull | are you using mod-wsgi directly or something like uwsgi? | 22:49 |
| agrebennikov | https://github.com/openstack/openstack-ansible-os_keystone/blob/master/tasks/keystone_ssl_user_provided.yml#L40 | 22:49 |
| agrebennikov | mod | 22:50 |
| agrebennikov | for now | 22:50 |
| agrebennikov | so may I just rely on the keystone's var? | 22:50 |
| agrebennikov | or it is forbidden? | 22:50 |
| agrebennikov | since they are separate plays | 22:50 |
| *** kylek3h has quit IRC | 22:51 | |
| agrebennikov | because the file {{ keystone_ssl_ca_cert }} will be exactly what I need | 22:51 |
| cloudnull | I'd change the var name to be namespaced | 22:51 |
| agrebennikov | where? | 22:51 |
| cloudnull | but that task should be all that's needed. | 22:51 |
| agrebennikov | in horizon? | 22:51 |
| cloudnull | yes. | 22:51 |
| agrebennikov | but how do I make sure that this particular var is defined when I run horizon play> | 22:52 |
| agrebennikov | ? | 22:52 |
| agrebennikov | becase I'm pretty sure I cannot run the tasks from other plays :/ | 22:53 |
| agrebennikov | (maybe I'm explaining wromg though :/) | 22:53 |
| agrebennikov | *wromg | 22:54 |
| agrebennikov | *wrong | 22:54 |
| cloudnull | if keystone_ssl_ca_cert == horizon_ssl_ca_cert then set that in your user_variables.yml file. something like horizon_ssl_ca_cert: "{{ keystone_ssl_ca_cert }}" -- in the future we can put in a PR into https://github.com/openstack/openstack-ansible/blob/master/playbooks/inventory/group_vars/all.yml should we feel like those vars will normally be the same when defined. | 22:54 |
| *** jperry has joined #openstack-ansible | 22:54 | |
| agrebennikov | aahh | 22:54 |
| agrebennikov | yeah, makes sense | 22:54 |
| agrebennikov | so then the bug is obsolete? | 22:55 |
| cloudnull | is there a task within the horizon role that will ship the ca cert? | 22:56 |
| cloudnull | like the one within the keystone role ? | 22:56 |
| * cloudnull looking | 22:56 | |
| agrebennikov | nope.... aah, yeah, it will not go into container anyway :) | 22:56 |
| agrebennikov | I keep missing conteinerized point :/ | 22:56 |
| cloudnull | https://github.com/openstack/openstack-ansible-os_horizon/blob/988006d22d9e357e988ae13008d3137c84473e67/tasks/horizon_ssl_user_provided.yml#L36 | 22:57 |
| agrebennikov | no, wait, this is not the same | 22:57 |
| agrebennikov | it goes into apache config, and potentially it may be another CA | 22:57 |
| agrebennikov | which may be responsible for only horizon name | 22:57 |
| agrebennikov | it is only responsible for horizon server cert | 22:58 |
| *** TxGirlGeek has quit IRC | 22:58 | |
| *** david-lyle has quit IRC | 22:59 | |
| *** darrenc_afk is now known as darrenc | 23:00 | |
| cloudnull | it looks like https://github.com/openstack/openstack-ansible-os_keystone/blob/b2e661d32bbd3fc1a61a570a23845dee894aa3f4/tasks/keystone_ssl_user_provided.yml#L40-L50 and https://github.com/openstack/openstack-ansible-os_horizon/blob/988006d22d9e357e988ae13008d3137c84473e67/tasks/horizon_ssl_user_provided.yml#L36-L44 accomplish the same things. | 23:00 |
| cloudnull | which could be used to drop "/etc/horizon/keystone-ca.pem" if the variables were set accordingly. | 23:01 |
| cloudnull | or maybe im just confused :) | 23:02 |
| cloudnull | it is my first day back from holidy | 23:02 |
| agrebennikov | ah, cool :) | 23:02 |
| cloudnull | **holiday | 23:02 |
| agrebennikov | now you have to be able to work 3 times more efficient ;) | 23:02 |
| cloudnull | i think it's more like 3x longer. | 23:03 |
| cloudnull | :p | 23:03 |
| cloudnull | or slower. | 23:03 |
| agrebennikov | so those 2 things | 23:03 |
| agrebennikov | they look same, but that is only the way you set up apache | 23:03 |
| agrebennikov | they both go to sites-enabled/XXX.conf | 23:03 |
| agrebennikov | under SSLCACertificateFile | 23:04 |
| agrebennikov | but in each case this file maybe different | 23:04 |
| agrebennikov | since it is only used for allowing the client to validate server cert | 23:04 |
| agrebennikov | which is also can be different | 23:04 |
| agrebennikov | this is why you have 2 different vars - horizon_user_ssl_ca_cert and keystone_user_ssl_ca_cert | 23:05 |
| *** jperry has quit IRC | 23:05 | |
| agrebennikov | which you set up separately in the user_vars | 23:05 |
| cloudnull | hum. ok so then I think we'd want another task to drop that file in place should the variable be set. | 23:07 |
| cloudnull | I suspect that this task would be used to drop other ca certs ? | 23:08 |
| cloudnull | one for nova, ceilometer, glance, etc? | 23:08 |
| agrebennikov | this is what I currently have to do in order to make it working | 23:08 |
| cloudnull | or does horizon only need the one? | 23:08 |
| agrebennikov | I already made local change into each play | 23:09 |
| *** jrobinson has joined #openstack-ansible | 23:09 | |
| agrebennikov | both keystone CA file and I need to add cafile = <> into each keystoneauthtoken section | 23:09 |
| agrebennikov | and openrc needs to be changed everywhere | 23:10 |
| agrebennikov | there is also one weird thing | 23:10 |
| agrebennikov | since you guys don't use packages | 23:11 |
| agrebennikov | requests lib is using its own CA bundle | 23:11 |
| agrebennikov | while in ubuntu for example it is using /etc/ssl/certs/ca-certificate.crt | 23:11 |
| *** sdake has quit IRC | 23:11 | |
| agrebennikov | and I can extend it with command | 23:11 |
| *** asettle has joined #openstack-ansible | 23:12 | |
| agrebennikov | while with pip requests it is a hell | 23:12 |
| cloudnull | what is it using when installed by pip ? | 23:12 |
| cloudnull | sigmavirus: ^ -cc | 23:12 |
| agrebennikov | I guess we already discussed that :) | 23:12 |
| agrebennikov | is is something local to the venv | 23:12 |
| cloudnull | ah ok. | 23:13 |
| agrebennikov | http://stackoverflow.com/questions/10667960/python-requests-throwing-up-sslerror | 23:13 |
| agrebennikov | something like this | 23:13 |
| agrebennikov | root@host:~/# python -c "import requests; print requests.certs.where()" | 23:13 |
| agrebennikov | /usr/local/lib/python2.7/dist-packages/requests/cacert.pem | 23:13 |
| agrebennikov | and it is not extendable | 23:13 |
| agrebennikov | so I have to add extra var to openrc in order to extend it | 23:14 |
| agrebennikov | but anyeay, that's all fine | 23:14 |
| agrebennikov | let me go propese a change | 23:14 |
| agrebennikov | *propose | 23:15 |
| cloudnull | sigmavirus: do you know of a why to change that in a way that would be more "supportable"? maybe something to allow the requests to use the system bundle instead of the vendored one? | 23:17 |
| cloudnull | agrebennikov: I really had no idea requests was doing that . | 23:18 |
| agrebennikov | this is what generic requests is doing | 23:18 |
| cloudnull | I would've tought it was using the system bundle. | 23:18 |
| agrebennikov | and each distro has to rebuild it | 23:18 |
| cloudnull | hum... | 23:19 |
| agrebennikov | well, this is what I spent 2 days for :D | 23:19 |
| agrebennikov | figuring out why it doesn't work as expected | 23:19 |
| agrebennikov | you never have this issue while using packages | 23:20 |
| cloudnull | are you adding the REQUESTS_CA_BUNDLE var to the openrc ? | 23:20 |
| agrebennikov | but now since my customer is very concerned about security I have to implement ssl carefully | 23:20 |
| agrebennikov | yes, this is what I have to do in order to make cli clients working | 23:20 |
| agrebennikov | and ansible openstack commands as well | 23:21 |
| cloudnull | can we just symlink /usr/local/lib/python2.7/dist-packages/requests/cacert.pem -> /etc/ssl/certs/ca-certificate.crt ? | 23:21 |
| cloudnull | well not to the crt | 23:21 |
| agrebennikov | emm...... well.... maybe :) | 23:21 |
| cloudnull | but you get what I mean ... | 23:21 |
| cloudnull | well . idk if that'd be any better really. | 23:22 |
| agrebennikov | don't see any difference in fact | 23:22 |
| agrebennikov | I still have to set up cafile in each config | 23:22 |
| *** asettle has quit IRC | 23:22 | |
| *** phalmos has quit IRC | 23:23 | |
| *** asettle has joined #openstack-ansible | 23:23 | |
| *** retreved has joined #openstack-ansible | 23:26 | |
| *** sdake has joined #openstack-ansible | 23:26 | |
| *** asettle has quit IRC | 23:27 | |
| *** dxiri has quit IRC | 23:28 | |
| *** sdake has quit IRC | 23:28 | |
| *** BjoernT has quit IRC | 23:29 | |
| *** sdake has joined #openstack-ansible | 23:29 | |
| *** galstrom is now known as galstrom_zzz | 23:32 | |
| *** thorst has quit IRC | 23:32 | |
| *** thorst has joined #openstack-ansible | 23:32 | |
| *** thorst has quit IRC | 23:37 | |
| cloudnull | yea. hum.. | 23:38 |
| cloudnull | ok. well bbl. | 23:38 |
| * cloudnull driving home | 23:38 | |
| *** Mudpuppy_ has joined #openstack-ansible | 23:40 | |
| *** marst has quit IRC | 23:42 | |
| *** schwicht has quit IRC | 23:44 | |
| *** ianychoi has joined #openstack-ansible | 23:49 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!