| *** jmckind has quit IRC | 00:26 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 00:48 | |
| *** alop has quit IRC | 00:54 | |
| sigmavirus24 | odyssey4me: https://review.openstack.org/#/c/206759/ should fix https://review.openstack.org/#/c/193422/ which will let me update my review so the fed/adfs stuff can continue | 00:57 |
|---|---|---|
| *** ntpttr2 has joined #openstack-ansible | 00:58 | |
| *** annashen has joined #openstack-ansible | 01:26 | |
| *** annashen has quit IRC | 01:31 | |
| *** daneyon has quit IRC | 01:57 | |
| openstackgerrit | Ian Cordasco proposed stackforge/os-ansible-deployment: Fix Keystone URI/URL defaults https://review.openstack.org/205192 | 02:10 |
| *** ntpttr2 has quit IRC | 02:24 | |
| *** annashen has joined #openstack-ansible | 02:27 | |
| *** annashen has quit IRC | 02:32 | |
| *** ntpttr2 has joined #openstack-ansible | 02:41 | |
| *** ntpttr2 has quit IRC | 03:23 | |
| *** JRobinson__ is now known as JRobinson__afk | 03:31 | |
| *** fawadkhaliq has joined #openstack-ansible | 03:42 | |
| *** JRobinson__afk has quit IRC | 03:56 | |
| *** annashen has joined #openstack-ansible | 03:58 | |
| *** annashen has quit IRC | 04:03 | |
| *** JRobinson__afk has joined #openstack-ansible | 04:04 | |
| *** JRobinson__afk is now known as JRobinson__ | 04:04 | |
| *** weezS has joined #openstack-ansible | 04:11 | |
| *** weezS has quit IRC | 04:12 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 04:14 | |
| *** ntpttr2 has joined #openstack-ansible | 04:54 | |
| *** ntpttr2 has quit IRC | 04:58 | |
| *** ntpttr2 has joined #openstack-ansible | 04:59 | |
| *** annashen_ has joined #openstack-ansible | 04:59 | |
| *** annashen_ has quit IRC | 05:03 | |
| *** daneyon has joined #openstack-ansible | 05:05 | |
| *** daneyon_ has joined #openstack-ansible | 05:07 | |
| *** daneyon has quit IRC | 05:10 | |
| *** ntpttr2 has quit IRC | 05:16 | |
| openstackgerrit | Merged stackforge/os-ansible-deployment: Removed inaccurate documentation https://review.openstack.org/206137 | 05:52 |
| openstackgerrit | Merged stackforge/os-ansible-deployment: Change to set the container network MTU https://review.openstack.org/205618 | 05:53 |
| *** yapeng has joined #openstack-ansible | 05:59 | |
| *** annashen_ has joined #openstack-ansible | 06:00 | |
| *** annashen_ has quit IRC | 06:04 | |
| *** yapeng has quit IRC | 06:10 | |
| *** javeriak has joined #openstack-ansible | 06:50 | |
| *** annashen has joined #openstack-ansible | 07:01 | |
| openstackgerrit | Darren Birkett proposed stackforge/os-ansible-deployment: set correct swift dispersion tenant https://review.openstack.org/206913 | 07:02 |
| *** javeriak has quit IRC | 07:03 | |
| *** annashen has quit IRC | 07:05 | |
| *** javeriak has joined #openstack-ansible | 07:13 | |
| *** JRobinson__ has quit IRC | 07:14 | |
| *** javeriak has quit IRC | 07:55 | |
| *** javeriak has joined #openstack-ansible | 07:55 | |
| *** annashen_ has joined #openstack-ansible | 08:02 | |
| evrardjp | odyssey4me: thanks for notifying me (sorry for the delay, it was a busy day yesterday, back from holidays :/ ) | 08:02 |
| *** annashen_ has quit IRC | 08:06 | |
| svg | good morning evrardjp | 08:07 |
| *** daneyon_ has quit IRC | 08:07 | |
| evrardjp | good morning svg... on holidays soon or are you already back? | 08:14 |
| svg | back from monday, but only until next thursday, then off for another 2 weeks :) | 08:15 |
| svg | stayed at home now, next > La Dource France | 08:15 |
| evrardjp | La Dource? You mean La Source, close to orleans, or a small town somewhere lost in France? ;) | 08:18 |
| svg | eurgh | 08:22 |
| svg | La Douce France :) | 08:22 |
| svg | dans le Var près d' Alès | 08:23 |
| evrardjp | I just came back from the Var... Beautiful region. | 08:25 |
| svg | yes | 08:25 |
| mattt | svg: hey! do you use leseb's ceph-ansible roles for deploying ceph? | 08:26 |
| svg | nope | 08:26 |
| svg | my ceph/admin coworker doesnt want to, he manages it 'manually' | 08:27 |
| mattt | svg: fair enough | 08:30 |
| svg | also because he wants to build up experience with it, and have good control. | 08:30 |
| mattt | svg: so we want to use leseb's ceph-ansible to deploy ceph, i'm a bit concerned that we're going to run into a situation where those roles drop a ceph.conf then yours blows it away | 08:32 |
| svg | mattt: yes, that was also leseb's concern | 08:32 |
| svg | though, I only provide ceph.conf on openstack hosts, that shouldn't overwrite any on ceph hosts | 08:33 |
| mattt | svg: one scenario where that could be possible is if you have an osd running on a host (which will run outside of a container) and then you end up running nova-compute there also (which also runs outside of the container) | 08:33 |
| mattt | so you'd have both roles trying to manage /etc/ceph/ceph.conf | 08:33 |
| svg | hm, that sounds theoretically possible, but I guess that'a also something that will happen on AIO deploys or the like | 08:34 |
| mattt | heh true | 08:34 |
| *** javeriak has quit IRC | 08:34 | |
| svg | But I agree with you. This might warrant some smaller common role that provides ceph.conf - I don;t know by heart if lesen has that already | 08:35 |
| git-harry | I think the ceph-common role provides that | 08:35 |
| svg | evrardjp: how's your osad deploy building up? | 08:38 |
| mattt | svg: perhaps we can look at having the mon/osds and clients use different ceph.conf locations | 08:44 |
| mattt | that way each system can maintain its own ceph.conf file | 08:44 |
| *** javeriak has joined #openstack-ansible | 08:49 | |
| evrardjp | mattt: I think the OSAD should consume the ceph.conf deployed, instead of overwriting it | 09:02 |
| evrardjp | I don't see the point of having different ceph.conf | 09:02 |
| evrardjp | (whether it is multiple locations with the same content or not) | 09:02 |
| *** annashen has joined #openstack-ansible | 09:02 | |
| evrardjp | svg: our first osad was deployed before the holidays, at my return there are a few components failing, I'll check what's wrong | 09:03 |
| evrardjp | I'll also need to update the deployment, so I guess... run all the playbooks again | 09:03 |
| evrardjp | (I mean there were bug fixes in kilo branch that I need to upload on my deployment) | 09:04 |
| mattt | evrardjp: yeah, i think we need to be able to interface w/ a ceph that wasn't deployed w/ ansible | 09:07 |
| *** annashen has quit IRC | 09:07 | |
| evrardjp | with an ssh connection that should be enough to ansible fetch the ceph.conf | 09:07 |
| evrardjp | the concern of svg was if one of the mon was down, we still have to continue the deployment (IIRC) | 09:08 |
| mattt | evrardjp: yeah if you had clear separation of what runs where it'd be a lot more straight-forward, the problem is an AIO etc. where you've got a boat load of stuff on the same host | 09:13 |
| evrardjp | should AIO be evaluated for ceph deployment? It doesn't make sense | 09:13 |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: Keystone Federation Service Provider Configuration https://review.openstack.org/194395 | 09:13 |
| evrardjp | If you have a ceph cluster, the chance is high that you'll have more than one host for openstack ;) | 09:14 |
| mattt | evrardjp: the AIO would largely be for developmenet/gating :) | 09:17 |
| evrardjp | ofc | 09:17 |
| mattt | not sure i understand the question then | 09:18 |
| mattt | the AIO supports an end-to-end install including swift | 09:19 |
| mattt | so we'd have to do something similar for ceph, and nonsensical as it seems | 09:19 |
| mattt | s/and/as/ | 09:19 |
| evrardjp | I know we can skip the AIO part... That's just sad ;) | 09:19 |
| evrardjp | can't* | 09:20 |
| svg | evrardjp: one of my latest patches (https://github.com/srvg/os-ansible-deployment/commit/4677ab0b8d7abd4997bddbfd5e5921e7832e3b86) tries to pick a mon host that is up | 09:24 |
| evrardjp | is nc installed as part of the playbook? | 09:26 |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: Add python-cinderclient to os_glance role https://review.openstack.org/206199 | 09:26 |
| evrardjp | but it seems fine | 09:26 |
| -openstackstatus- NOTICE: Currently our CI system is broken, jobs are not getting processed at all. | 09:27 | |
| *** ChanServ changes topic to "Currently our CI system is broken, jobs are not getting processed at all." | 09:27 | |
| svg | evrardjp: afaik nc is part of a base install, but that could be an issue - nc is only run on the local (ansible running) host, so that might be a dependency for setting up osad's deploy host | 09:31 |
| evrardjp | making sure it's installed doesn't cost anything ;) | 09:32 |
| odyssey4me | mattt why not do something like we do with the ssl keys - give the option to distribute the ceph.conf from the deployment host if a specific user variable is set, otherwise gather it from the ceph mon if it's there, or error out saying that it doesn't exist and one of the two options must be provided | 09:33 |
| svg | I guess there are 3 possible sources for ceph.conf: generating one from vars, pulling one from an existing ceph cluster, or generating one with the same role used by ansible ceph roles | 09:36 |
| svg | hm, 1 is not needed if we have 3 | 09:36 |
| svg | basically what odyssey4me said :) | 09:36 |
| odyssey4me | svg :) yeah, I don't think that something that configures ceph from a client standpoint should generate ceph.conf - it should collect or be given the conf by the deployer who will design it based on the existing ceph cluster deployed | 09:37 |
| evrardjp | I like the variable approach... | 09:42 |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: development-stack Doc Update https://review.openstack.org/206016 | 09:49 |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: Add sample Keystone Federation SP configuration for ADFS https://review.openstack.org/203736 | 09:50 |
| svg | I'm still wondering what local work-flow you guys use in your local osad git checkout for working with Gerrit | 09:51 |
| odyssey4me | svg personally I work from a mac and simply use the git review CLI for all edits; to test I mostly use AIO's as they're quick and largely painless to setup - I usually clone master, then cherry-pick the appropriate review on top of it, add whatever configs are needed to test and build. If I already have an AIO, I cherry-pick the review onto it, add the config and test. | 09:54 |
| odyssey4me | If I find that edits are needed, then I figure out the fixes on the test system and when it's all working right I use git diff to inspect the changes while essentially either copying them over to my local terminal for addition to the existing review as a new patch set. I often end up rewriting bits when I do this to make them more appropriate for broader consumption. | 09:57 |
| odyssey4me | svg does that provide an answer to your question? | 09:57 |
| svg | that sounds complicated, not sure I understand it actually | 09:58 |
| svg | So you don't have a permanent checkout where you work on? | 09:59 |
| svg | Say you have a review you work on, especially something that takes time (like my Ceph stuff), you don't keep some kind of branch(es) for that? | 10:00 |
| svg | Initially I used a master-ceph branch, where i kept my different commits, which I then reased into a single commit on a separate master-ceph-gerrit ranch, which I then uploaded | 10:01 |
| svg | But that didn;t work out when someone added a change to that patch, there's no simple wat to cherry-pick that single change from what gerrit provides | 10:02 |
| svg | so now I basically do the same thing, only have separate branches per uploaded patch on gerrit | 10:02 |
| odyssey4me | yeah, gerrit workflow is different to github workflow | 10:03 |
| odyssey4me | what you're doing (feature branches) makes sense for github | 10:04 |
| *** annashen_ has joined #openstack-ansible | 10:04 | |
| odyssey4me | in gerrit essentially every review stands alone and all work is a single commit | 10:04 |
| svg | in short, when I want to make an additional change, I checkout the latest patch from gerrit, like master-ceph-review27, add commits on that, then squash it one commit I then tag as master-ceph-review28, and which will be the same when i checkout that again from gerrit | 10:04 |
| odyssey4me | so if two people are tag teaming (we often do this when sharing work between the US and UK) on a review, all you do is delete your old branch and 'git review -d <review #>' to pull down the latest version | 10:04 |
| svg | I guess that makes no sense if you rely solely on gerrit to keep track of the code | 10:05 |
| svg | But as i needed to backport that into kilo, to have it tested on our setup, I need to keep that in a clear separate branch | 10:05 |
| *** andymccr has quit IRC | 10:05 | |
| *** jmccrory has quit IRC | 10:05 | |
| svg | And then perhaps I add some commits during testing and deploying it on kilo | 10:06 |
| svg | And I need to keep track of those, to add them back on master | 10:06 |
| odyssey4me | yeah - it's pretty rare for us not to... sometimes for a spike/research we'll fork the repo in github and work there... but gerrit provides a nice track record of discussion through the lifetime of the work done, so it's actually nice to just commit --amend and git review at the end of the day so that others can see where you're going with it and can comment early | 10:06 |
| svg | I made some mistakes there in the early days, which is why I keep those separate commits to have an overview of what I did | 10:07 |
| *** jmccrory has joined #openstack-ansible | 10:07 | |
| odyssey4me | for testing on kilo, I simply clone the kilo branch and cherry-pick the patch on top of it... for a longer running AIO I would rebase to remove the old version of the patch and cherry-pick the updated patch on top again | 10:07 |
| svg | add to that that I might need to patch osad for something specif to our setup, whcih would not need to go back to the ceph patch | 10:08 |
| svg | add to that that I have a private patch to maintain neutron lbaas for our setup | 10:08 |
| *** annashen_ has quit IRC | 10:08 | |
| svg | It gets complicated very fast :) | 10:08 |
| odyssey4me | yeah, which is why we'd prefer to work with you to facilitate that osad accommodates those needs in a way that means you don't have to maintain a fork | 10:09 |
| odyssey4me | it does take a bit of time to work through those and figure out a way to do it in a way that works for everyone though, as you've seen with the ceph patch :) | 10:09 |
| odyssey4me | but once it's in, it's there and we all maintain it with you :) | 10:10 |
| svg | yes :) | 10:10 |
| svg | part of all that is of course also this is a complex project wit many moving targets, and difficult to test | 10:10 |
| evrardjp | which is the best part: working as a team ;) | 10:10 |
| svg | tbh, the complexity is a bit frustrating at times; I don;t get enough time to properly work on all that | 10:11 |
| odyssey4me | yeah, as the project grows - so does the testing matrix | 10:13 |
| odyssey4me | we could do gating better - at the moment we only do one converging test case which is based on one deployment configuration | 10:13 |
| odyssey4me | but that's why we'd like to see more eyes on as many of the reviews as possible - anyone using it should look through reviews coming through to spot potential issues | 10:14 |
| odyssey4me | we have resources committed to development, testing, etc and we have QE guys and deployers filing bugs when they find something that breaks | 10:15 |
| odyssey4me | so guys like you and evrardjp are key to keep the quality up | 10:15 |
| odyssey4me | if you guys can try to take just an hour or so per day to look through reviews that would help greatly | 10:16 |
| svg | I tried that, but often reviews are hard to grasp to be able to make usefull comments - not enough openstack experience | 10:16 |
| svg | odyssey4me: tbh, an hour per day is a lot if one doesn't work full time on osad | 10:17 |
| odyssey4me | very often the things that are confusing to you confuse others, so even a review which asks the question has value | 10:17 |
| svg | I know I have been lacking on this though, but so far I still don't know if and when we will carry on with openstack | 10:18 |
| odyssey4me | and yeah, I understand - but you have a lot of experience with ansible so you could easily suggest more optimal ways of doing things from an ansible standpoint too :) | 10:19 |
| odyssey4me | svg I saw the work you did for the with_subelements loop - I'm glad it's included in v2. We've had to work around the limitations using a module. :/ | 10:30 |
| mattt | svg: i don't ansible very well, so excuse me :P but having issues with https://review.openstack.org/#/c/181957/28/playbooks/roles/ceph_client/tasks/ceph_auth.yml | 10:56 |
| mattt | svg: the task 'Verify Ceph monitors are up' is skipped, so i get a subsequent failure | 10:57 |
| mattt | if i run that task in a standalone playbook it runs fine :( | 10:57 |
| *** andymccr has joined #openstack-ansible | 11:00 | |
| *** annashen has joined #openstack-ansible | 11:04 | |
| *** annashen has quit IRC | 11:09 | |
| *** javeriak has quit IRC | 11:14 | |
| openstackgerrit | Darren Birkett proposed stackforge/os-ansible-deployment: update teardown.sh to remove mongodb https://review.openstack.org/206996 | 11:27 |
| evrardjp | odyssey4me: about ansible standpoint, you mentionned in the past that standard inventory of ansible wasn't enough for you, would you clarify this? I'm curious and I prefer going closer to the standard system of inventory and variables | 11:29 |
| svg | mattt: if that hets skiopped, that must be bc ceph_conf.global.mon_host is somehow empty? | 11:35 |
| odyssey4me | mancdaz to my knowledge the teardown script is only meant to destroy the containers and other playbook deployed stuff, not the bootstrap bits too - if it's the bootstrap bits too then I think it's somewhat lacking | 11:35 |
| mattt | svg: i've printed it out, it's populated w/ 3 hosts | 11:35 |
| mancdaz | odyssey4me mmm, I mentioned something to that effect in teh bug | 11:35 |
| svg | Also those quotes are not needed, but I do notice there's a missing space between }} and 22 | 11:36 |
| mancdaz | but having that stuff there doesn't hurt | 11:36 |
| svg | mattt: 3 hosts comma separated, right? | 11:36 |
| odyssey4me | mancdaz also note that the teardown is for AIO and multi-node deployments... the mongodb bits would only be deployed by us on in an AIO | 11:37 |
| mattt | svg: https://gist.githubusercontent.com/mattt416/ea71b2ef2fe48ae06950/raw/bd64886ad5cb5b2a656bffb21057c458ba879a2a/gistfile1.txt | 11:37 |
| odyssey4me | evrardjp I think we'd ideally like to move towards not having to use a dynamic inventory, but right now we use it for a lot of things | 11:37 |
| mancdaz | odyssey4me mmm, I mentioned something to that effect in the bug | 11:38 |
| mancdaz | I still think it's handy to have it there | 11:38 |
| odyssey4me | evrardjp the dynamic inventory allows us to have a more flexible and modular deployment structure | 11:39 |
| svg | mattt I don't understand why all three gets skipped | 11:39 |
| odyssey4me | tbh though I think we need to re-evaluate how we're using it all and how we could do better - perhaps we should do that when ansible 2 releases | 11:40 |
| mattt | svg: same, this is the task in a separate playbook: https://gist.githubusercontent.com/mattt416/b3e60d0bc3281f6e67f8/raw/2a6ecf5a789f22a7a636f05f1dcdd6fc67b797b9/gistfile1.txt | 11:40 |
| svg | odyssey4me: | 11:41 |
| odyssey4me | svg? | 11:41 |
| svg | odyssey4me: FYI I expect updates on internal inventory that should add features past v2 | 11:42 |
| odyssey4me | svg yeah, v2 is largely a refactor to provide a better platform for the next level of growth in the core I believe... so what we can do beyond v2 is start to do PR's for the extra features we want to see so that we don't have to carry them in this project | 11:43 |
| svg | odyssey4me: yes, but also, lib/ansible/inventory was not (yet) refactored in v2 | 11:44 |
| *** fawadkhaliq has quit IRC | 11:48 | |
| svg | mattt: in the separate playbook, you run against cinder_api_container, whilst the auth tasks should only run when inventory_hostname in groups['cinder_volume'] | 11:52 |
| svg | So I think it might just happens to run on those hosts for other reasons | 11:53 |
| svg | that separate playbook test might be a red herring | 11:54 |
| *** fawadkhaliq has joined #openstack-ansible | 11:55 | |
| mgariepy | good morning everyone | 11:56 |
| mattt | svg: not pinned it down, but if i change run_once: false it seems to move along | 11:56 |
| svg | hmmz, that must be it, run_once basically jist runs it on an arbitrary first host, and depending on the play_hosts that could be a wrong one | 11:57 |
| svg | meaning one that does not match those that need it, being cinder_volume | 11:58 |
| mattt | svg: both playbooks target cinder_all tho | 11:59 |
| odyssey4me | mgariepy morning - I hope you don't mind, but I fixed up your commit message to make it comply with the openstack commit msg guidelines | 11:59 |
| mgariepy | odyssey4me, no, I don't mind. | 12:01 |
| svg | I'll update my branch with this, if you confirm it solves it | 12:02 |
| mattt | svg: running into another issue now, let me just get through this to ensure they're not related | 12:02 |
| *** annashen_ has joined #openstack-ansible | 12:05 | |
| *** javeriak has joined #openstack-ansible | 12:06 | |
| *** javeriak has quit IRC | 12:08 | |
| *** javeriak has joined #openstack-ansible | 12:08 | |
| *** annashen_ has quit IRC | 12:10 | |
| *** tlian has joined #openstack-ansible | 12:10 | |
| svg | mattt: git-harry new updates so far http://sprunge.us/WPWe | 12:15 |
| svg | So there's a reason I initially didn;t push it out yet, still got some loose ends, hence the #TODO's | 12:19 |
| mattt | no worries, just making a note so i don't forget :P | 12:21 |
| *** KLevenstein has joined #openstack-ansible | 12:47 | |
| -openstackstatus- NOTICE: zuul's disks were at capacity. Space has been freed up and jobs are being re-queued. | 12:48 | |
| *** ChanServ changes topic to "Weekly Meetings: https://wiki.openstack.org/wiki/Meetings/openstack-ansible" | 12:48 | |
| *** annashen has joined #openstack-ansible | 13:06 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 13:08 | |
| *** annashen has quit IRC | 13:10 | |
| *** TheIntern has joined #openstack-ansible | 13:11 | |
| -openstackstatus- NOTICE: zuul jobs after about 07:00 UTC may need a 'recheck' to enter the queue. Look if your change is in http://status.openstack.org/zuul/ and recheck if not. | 13:25 | |
| *** yapeng has joined #openstack-ansible | 13:30 | |
| *** gparaskevas has joined #openstack-ansible | 13:31 | |
| *** javeriak has quit IRC | 13:35 | |
| gparaskevas | Hello there | 13:39 |
| gparaskevas | i am re evaluating OSAD + some custome addons i have created | 13:39 |
| gparaskevas | and i am stuck when i try to install openstack playbooks | 13:39 |
| *** jmckind has joined #openstack-ansible | 13:39 | |
| gparaskevas | it seems that keystone service tenant cannot be created for some reason | 13:39 |
| gparaskevas | i have also run the haproxy playbook | 13:40 |
| gparaskevas | i can hit both endpoints from the haproxy ip but when i try to "keystone tenant-list" | 13:42 |
| gparaskevas | i get this error Authorization Failed: Unable to establish connection to http://192.168.110.14:5000 | 13:42 |
| gparaskevas | can anyone help? | 13:42 |
| odyssey4me | gparaskevas if you do keystone --debug tenant-list then where does it get stuck, and also what tag/branch are you running? | 13:43 |
| *** Mudpuppy has joined #openstack-ansible | 13:44 | |
| gparaskevas | master | 13:44 |
| gparaskevas | let me try the firstone as well | 13:44 |
| *** markvoelker has joined #openstack-ansible | 13:44 | |
| odyssey4me | gparaskevas note that master tracks upstream master which is liberty code | 13:45 |
| gparaskevas | i know and i have been using master for about a month with now problems | 13:46 |
| gparaskevas | and now i decided to do a clean reavaluation and track time etc and i get stack here... | 13:46 |
| gparaskevas | http://paste.openstack.org/show/406226/ | 13:46 |
| gparaskevas | thats what i get with debug enabled on keystone client from keystone container | 13:47 |
| gparaskevas | why does it go for v3? | 13:50 |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: development-stack Doc Update https://review.openstack.org/206016 | 13:51 |
| *** fawadkhaliq has quit IRC | 13:53 | |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: Document required repository hosts config info https://review.openstack.org/205100 | 13:53 |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: Document required repository hosts config info https://review.openstack.org/202258 | 13:54 |
| openstackgerrit | Matthew Kassawara proposed stackforge/os-ansible-deployment: Reduce neutron configuration https://review.openstack.org/204597 | 13:54 |
| *** spotz_zzz is now known as spotz | 13:56 | |
| *** fawadkhaliq has joined #openstack-ansible | 13:56 | |
| *** evrardjp has quit IRC | 13:59 | |
| mgariepy | is there some doc to deploy extra containers ? | 13:59 |
| *** fawadkhaliq has quit IRC | 14:00 | |
| *** fawadkhaliq has joined #openstack-ansible | 14:01 | |
| gparaskevas | what extra containers do you mean? | 14:01 |
| gparaskevas | for extra servises? | 14:01 |
| mgariepy | yeah, i would like to have like ceph mons and stuff like this | 14:02 |
| gparaskevas | there is a folder in /etc/openstack_deploy/env.d | 14:02 |
| gparaskevas | there is a extra container example there | 14:03 |
| gparaskevas | an* | 14:03 |
| gparaskevas | extra_container.yml.example | 14:03 |
| mgariepy | i saw that but, it's not very clear to me hehe | 14:03 |
| mgariepy | can i add a container defenition there and run setup-hosts playbook ? | 14:03 |
| gparaskevas | sure | 14:04 |
| gparaskevas | like that | 14:04 |
| *** markvoelker has quit IRC | 14:04 | |
| gparaskevas | if you go here https://github.com/stackforge/os-ansible-deployment/tree/kilo/etc/openstack_deploy/env.d | 14:04 |
| gparaskevas | this is kilo branch and it contains the definitions for all containers so maybe this can hel you | 14:05 |
| *** evrardjp has joined #openstack-ansible | 14:06 | |
| *** annashen_ has joined #openstack-ansible | 14:07 | |
| gparaskevas | you create a ceph.yml and then you have to define also the hosts in openstacl_user_config.yml | 14:07 |
| odyssey4me | gparaskevas ah, so keystone CLI is deprecated in liberty - use the openstack CLI | 14:07 |
| gparaskevas | ok | 14:07 |
| odyssey4me | gparaskevas currently the admin endpoint in master is v3, the others are v2 | 14:08 |
| gparaskevas | ok then why does it use ansible keystone client module? | 14:09 |
| gparaskevas | is this ok? | 14:09 |
| odyssey4me | gparaskevas also, it seems like your endpoint is unreachable | 14:09 |
| gparaskevas | yes seems so | 14:09 |
| gparaskevas | let me see once again | 14:09 |
| odyssey4me | python-keystoneclient is moving towards being purely a library consumed by anything that needs to talk to keystone | 14:09 |
| odyssey4me | the other CLI's are heading in that direction slowly too | 14:10 |
| gparaskevas | so opensatck-cli uses keystone cli? | 14:10 |
| odyssey4me | gparaskevas confirm that your haproxy is running and that it can see the keystone container backends | 14:10 |
| gparaskevas | http://10.10.110.78:5000/v3 i can hit that and with no problems | 14:10 |
| *** richoid has quit IRC | 14:11 | |
| odyssey4me | gparaskevas no openstack cli uses the keystone library - the openstack cli provides a consistent CLI with the same options for formatting, help, etc | 14:11 |
| openstackgerrit | Merged stackforge/os-ansible-deployment: Add python-cinderclient to os_glance role https://review.openstack.org/206199 | 14:11 |
| *** annashen_ has quit IRC | 14:11 | |
| odyssey4me | gparaskevas it sounds to me then like your haproxy is not working right | 14:11 |
| *** jmccrory has quit IRC | 14:11 | |
| gparaskevas | i can curl from haproxy keystone endpoints with success | 14:12 |
| gparaskevas | :/ | 14:12 |
| gparaskevas | curl 192.168.110.234:5000 {"versions": {"values": [{"status": "stable", "updated": "2015-03-30T00:00:00Z", "media-types": [{"base": "application/json", "type": "application/vnd.openstack.identity-v3+json"}], "id": "v3.4", "links": [{"href": "http://192.168.110.234:5000/v3/", "rel": "self"}]}, {"status": "stable", "updated": "2014-04-17T00:00:00Z", "media-types": [{"base": "application/json", "type": "application/vnd.op | 14:12 |
| *** jmccrory has joined #openstack-ansible | 14:13 | |
| odyssey4me | so there's the problem - your client is trying to access 192.168.110.14 which is a different address | 14:14 |
| gparaskevas | yes that haproxy internal address with backend to 192.168.110.234:5000 and one mmore | 14:17 |
| gparaskevas | when i invoke openstack cli i get this | 14:18 |
| gparaskevas | WARNING: keystoneclient.auth.identity.generic.base Discovering versions from the identity service failed when creating the password plugin. Attempting to determine version from URL. ERROR: openstack Unable to establish connection to http://192.168.110.14:5000/v3.0/auth/tokens | 14:18 |
| gparaskevas | i cant reach this http://192.168.110.14:5000/v3.0 from browser though but i can reach this http://192.168.110.14:5000/v3 | 14:19 |
| gparaskevas | seems you are right | 14:20 |
| gparaskevas | i curl localhost:5000 from haproxy and i get connection refused | 14:21 |
| odyssey4me | gparaskevas v3 endpoint is not v3.0 it's just v3 | 14:21 |
| gparaskevas | yes v3.0 shows the error | 14:21 |
| *** markvoelker_ has joined #openstack-ansible | 14:21 | |
| *** jwagner_away is now known as jwagner | 14:22 | |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: Reduce neutron configuration https://review.openstack.org/204597 | 14:23 |
| *** yaya has joined #openstack-ansible | 14:26 | |
| gparaskevas | Now its a different error it cannot install pip install keystone | 14:33 |
| odyssey4me | gparaskevas are you running a specific playbook? | 14:34 |
| gparaskevas | yes os-keystone-install.yml | 14:35 |
| gparaskevas | i have already run successfully setup-hosts.yml , infrastructure and haproxy | 14:36 |
| gparaskevas | but even with setup-openstack.yml i get that error... | 14:36 |
| gparaskevas | i will manually uninstall it a check again | 14:37 |
| *** markvoelker_ has quit IRC | 14:38 | |
| *** markvoelker has joined #openstack-ansible | 14:39 | |
| odyssey4me | gparaskevas you'll need to do the repo-build again if you've updated os-ansible-install to ensure that your python wheel repo rebuilds | 14:41 |
| *** yapeng has quit IRC | 14:41 | |
| gparaskevas | ok | 14:42 |
| gparaskevas | i uninstalled keystone manualy using pip and then the playbook did the installation succesfully but again at step ensure keystone service tenant it fails... | 14:42 |
| odyssey4me | gparaskevas is it trying to access the right address? you can run openstack-ansible with -vv to get more verbose output | 14:45 |
| *** markvoelker has quit IRC | 14:47 | |
| *** markvoelker has joined #openstack-ansible | 14:47 | |
| gparaskevas | how can i check what address its trying to access? | 14:47 |
| gparaskevas | how can i check what endpoint is trying to access? | 14:48 |
| gparaskevas | when i source openrc and i try openstack service list i get this "ERROR: openstack Could not determine a suitable URL for the plugin" | 14:48 |
| odyssey4me | gparaskevas the task running is 'Ensure service tenant' right? if so, this is the endpoint it's trying to reach: https://github.com/stackforge/os-ansible-deployment/blob/master/playbooks/roles/os_keystone/tasks/keystone_service_setup.yml#L50 | 15:00 |
| odyssey4me | ah tell me, do you have a user_group_*.yml in /etc/openstack_deploy/ ? | 15:01 |
| odyssey4me | jwitko this may be relevant to you too | 15:01 |
| *** gparaskevas has quit IRC | 15:01 | |
| odyssey4me | if you do have such a file, delete it - those values have been moved into the inventory to allow easier overrides | 15:02 |
| *** jwagner is now known as jwagner_away | 15:03 | |
| *** jmckind has quit IRC | 15:03 | |
| *** hiddentoken has joined #openstack-ansible | 15:03 | |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: Templateize and change Galera SST method https://review.openstack.org/207082 | 15:05 |
| *** annashen has joined #openstack-ansible | 15:07 | |
| odyssey4me | sigmavirus24 palendae d34dh0r53 mattt Please review https://review.openstack.org/207082 as we'll need to include it for the 11.1.0 release. The update to MariaDB 10 will likely only make it for 11.2.0. | 15:08 |
| *** weezS has joined #openstack-ansible | 15:09 | |
| *** hiddentoken has quit IRC | 15:09 | |
| *** annashen has quit IRC | 15:13 | |
| odyssey4me | thx | 15:13 |
| *** jmckind has joined #openstack-ansible | 15:13 | |
| *** jmckind has quit IRC | 15:13 | |
| jwitko | odyssey4me, my issue on that similar problem was that I was using a branch of osad that wasn’t ready yet | 15:13 |
| jwitko | causing library mismatches, etc | 15:14 |
| *** jmckind has joined #openstack-ansible | 15:14 | |
| odyssey4me | jwitko sure, if you want something stable then use a tagged branch | 15:14 |
| odyssey4me | there are sometimes patches due to upstream issues that need to be pulled in, but those are generally well known | 15:14 |
| jwitko | I’m sure they were, I was just uneducated. | 15:15 |
| jwitko | but I am actually up and running now | 15:15 |
| jwitko | for the most part... i’m sure errors will come up down the road | 15:15 |
| odyssey4me | we unfortunately also didn't tag point releases on the usual two week schedule after 11.0.4 as we've been trying to iron out issues for 11.1.0 which we've uncovered when merging major feature-reated changes in... so the master and kilo branches are a little offish at the moment, but getting better by the day | 15:16 |
| odyssey4me | jwitko good :) perhaps you can help others get educated now :) | 15:17 |
| *** jwagner_away is now known as jwagner | 15:17 | |
| jwitko | odyssey4me, absolutely | 15:17 |
| jwitko | odyssey4me, i notice when I’m consoled in via spice that I get a token timeout very quickly. do you know how to adjust the token expiration? | 15:17 |
| odyssey4me | jwitko offhand I have no idea | 15:18 |
| odyssey4me | jwitko I don't see any options for that here: https://github.com/stackforge/os-ansible-deployment/blob/master/playbooks/roles/os_nova/templates/nova.conf.j2#L132-L141 | 15:19 |
| odyssey4me | unless you're talking about the keystone token? | 15:19 |
| *** yaya has quit IRC | 15:20 | |
| Sam-I-Am | it'd probably be the a keystone thing | 15:20 |
| odyssey4me | in which case, from https://github.com/stackforge/os-ansible-deployment/blob/master/playbooks/roles/os_keystone/templates/keystone.conf.j2#L82 the value to set in user_variables will be keystone_token_expiration | 15:20 |
| Sam-I-Am | default token expiration? | 15:21 |
| odyssey4me | the default is pretty long though: https://github.com/stackforge/os-ansible-deployment/blob/master/playbooks/roles/os_keystone/defaults/main.yml#L40 | 15:21 |
| odyssey4me | that time is in seconds as I recall | 15:21 |
| *** richoid has joined #openstack-ansible | 15:25 | |
| openstackgerrit | Nolan Brubaker proposed stackforge/os-ansible-deployment: Document some of the common scripts https://review.openstack.org/207105 | 15:47 |
| palendae | ^ backport | 15:48 |
| *** alop has joined #openstack-ansible | 15:52 | |
| *** fawadkhaliq has quit IRC | 15:54 | |
| *** annashen_ has joined #openstack-ansible | 16:08 | |
| *** sdake has joined #openstack-ansible | 16:08 | |
| *** daneyon has joined #openstack-ansible | 16:11 | |
| *** meteorfox has quit IRC | 16:13 | |
| *** serverascode has quit IRC | 16:13 | |
| *** annashen_ has quit IRC | 16:13 | |
| odyssey4me | palendae I think that something is broken about https://review.openstack.org/202258 and https://review.openstack.org/205100 - perhaps it's worth abandoning them and re-submitting the patch as a fresh one to master and see if a gate job kicks off? | 16:27 |
| palendae | Maybe | 16:28 |
| odyssey4me | hmm, that said none of the doc patches are gating so I think we have an issue with our last gate check patch breaking this | 16:28 |
| openstackgerrit | Merged stackforge/os-ansible-deployment: Templateize and change Galera SST method https://review.openstack.org/207082 | 16:28 |
| odyssey4me | palendae can you review this backport? https://review.openstack.org/206913 | 16:29 |
| palendae | odyssey4me: Yep. done | 16:31 |
| *** TheIntern has quit IRC | 16:31 | |
| odyssey4me | palendae thx | 16:32 |
| *** capricorn_one has joined #openstack-ansible | 16:37 | |
| *** sdake has quit IRC | 16:37 | |
| *** capricorn_one has quit IRC | 16:41 | |
| *** serverascode has joined #openstack-ansible | 16:43 | |
| *** sdake has joined #openstack-ansible | 16:53 | |
| *** tlian has quit IRC | 16:53 | |
| *** javeriak has joined #openstack-ansible | 16:55 | |
| *** tlian has joined #openstack-ansible | 17:00 | |
| *** meteorfox has joined #openstack-ansible | 17:00 | |
| *** jmckind has quit IRC | 17:04 | |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: Add python-cinderclient to os_glance role https://review.openstack.org/207130 | 17:08 |
| *** annashen has joined #openstack-ansible | 17:09 | |
| odyssey4me | palendae sigmavirus24 ^ backport | 17:11 |
| *** annashen has quit IRC | 17:13 | |
| *** javeriak has quit IRC | 17:15 | |
| *** annashen has joined #openstack-ansible | 17:19 | |
| openstackgerrit | Matthew Kassawara proposed stackforge/os-ansible-deployment: development-stack Doc Update https://review.openstack.org/206016 | 17:26 |
| Sam-I-Am | lets see how that looks | 17:26 |
| odyssey4me | Sam-I-Am great changes - just one whitespace nit | 17:34 |
| Sam-I-Am | odyssey4me: yeah, those happen - no linter | 17:35 |
| Sam-I-Am | i just fixed the part of the doc that seemed relevant to the patch | 17:35 |
| Sam-I-Am | the rest of it needs attention, but i'll wait until this merges | 17:35 |
| odyssey4me | Sam-I-Am yeah, unfortunately we seem to have broken our doc merges :/ none of our docs end up in a gate check, so they never merge | 17:36 |
| odyssey4me | we'll have to figure it out some time soon | 17:36 |
| Sam-I-Am | those only apply to /doc though, right? | 17:36 |
| Sam-I-Am | not the random rst stuff | 17:37 |
| odyssey4me | Sam-I-Am not sure - might also be *.rst | 17:37 |
| Sam-I-Am | palendae: ^ | 17:37 |
| openstackgerrit | Matthew Kassawara proposed stackforge/os-ansible-deployment: development-stack Doc Update https://review.openstack.org/206016 | 17:41 |
| *** annashen has quit IRC | 17:42 | |
| *** annashen has joined #openstack-ansible | 17:46 | |
| Sam-I-Am | odyssey4me: bettar now | 17:48 |
| Sam-I-Am | i think | 17:48 |
| odyssey4me | thanks Sam-I-Am - looks great! | 17:51 |
| Sam-I-Am | cool | 17:51 |
| Sam-I-Am | progress! | 17:51 |
| palendae | odyssey4me, Sam-I-AmThat's f'in weird... | 18:01 |
| palendae | Pretty sure it is *.rst and doc/ | 18:01 |
| Sam-I-Am | where's the gate job live? | 18:03 |
| Sam-I-Am | err, i guess rather whats broken | 18:03 |
| palendae | Really not sure what's broken, but that definition lives in infra's project defintions | 18:04 |
| *** prad has quit IRC | 18:04 | |
| odyssey4me | yup, we have some regex in there but I haven't had time to dig into it | 18:06 |
| odyssey4me | I think the last change to file the docimpact bugs has broken this | 18:06 |
| odyssey4me | or something else has changed which broke what used to work | 18:06 |
| palendae | https://github.com/openstack-infra/project-config/blob/master/zuul/layout.yaml#L1575 | 18:08 |
| palendae | I think the change for the docimpact stuff was after | 18:08 |
| palendae | Mine's been refusing to merge since before that | 18:09 |
| palendae | I wonder if it's cause the whole job is skipped? | 18:09 |
| palendae | Might be safer to move the ignore into our gating scripts :( | 18:09 |
| odyssey4me | palendae shall we set this old bug to won't fix? https://bugs.launchpad.net/openstack-ansible/+bug/1410357 | 18:11 |
| openstack | Launchpad bug 1410357 in openstack-ansible trunk "Bonds should list their slave interfaces" [Low,Confirmed] | 18:11 |
| palendae | odyssey4me: Yeah | 18:11 |
| odyssey4me | it's a host networking setup item, which we don't do | 18:11 |
| palendae | Right | 18:11 |
| palendae | Back when I thought we might | 18:11 |
| palendae | btw, you can sleep while cloudnull_afk's gone :( | 18:12 |
| palendae | :)* | 18:12 |
| odyssey4me | lol palendae hardly, now I get to do his job too :p | 18:13 |
| odyssey4me | miguelgrinberg mattt is this purely a juno issue, or is it relevant for kilo/master? https://bugs.launchpad.net/openstack-ansible/+bug/1412762 | 18:14 |
| openstack | Launchpad bug 1412762 in openstack-ansible trunk "Disable unneeded heat resource-types" [Low,Confirmed] - Assigned to Matt Thompson (mattt416) | 18:14 |
| *** TheIntern has joined #openstack-ansible | 18:17 | |
| *** javeriak has joined #openstack-ansible | 18:17 | |
| *** prad has joined #openstack-ansible | 18:20 | |
| *** annashen has quit IRC | 18:21 | |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: Re-enable building of heat_repo_plugins https://review.openstack.org/207155 | 18:22 |
| odyssey4me | palendae sigmavirus24 d34dh0r53 ^ backport | 18:23 |
| *** annashen has joined #openstack-ansible | 18:24 | |
| *** prad has quit IRC | 18:24 | |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: Set default inventory file for inventory-manage.py https://review.openstack.org/207158 | 18:25 |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: Set default inventory file for inventory-manage.py https://review.openstack.org/207158 | 18:25 |
| *** prad has joined #openstack-ansible | 18:27 | |
| openstackgerrit | Matthew Kassawara proposed stackforge/os-ansible-deployment: update teardown.sh to remove mongodb https://review.openstack.org/206996 | 18:28 |
| *** javeriak has quit IRC | 18:30 | |
| miguelgrinberg | odyssey4me: regarding that heat bug, I'm not sure it is relevant anymore. These resources are harmless, they just don't work because they represent stuff that isn't installed. | 18:30 |
| odyssey4me | miguelgrinberg so you're thinking it's a won't fix? | 18:31 |
| *** openstackgerrit has quit IRC | 18:31 | |
| miguelgrinberg | yeah, we are all for not adding custom patches, so probably the best is to let them be | 18:31 |
| *** openstackgerrit has joined #openstack-ansible | 18:32 | |
| miguelgrinberg | this is certainly something we can handle on the RPC side if there is interest | 18:32 |
| palendae | Is anyone aware of what this user_group_vars.yml file is? I don't see it in-tree... https://github.com/stackforge/os-ansible-deployment/blob/master/scripts/run-upgrade.sh#L155 | 18:34 |
| odyssey4me | so is that an upstream patch requirement, or would it be something like a task to remove them after the environment is built? | 18:34 |
| palendae | https://github.com/stackforge/os-ansible-deployment/tree/master/etc/openstack_deploy doesn't contain it | 18:34 |
| odyssey4me | palendae that should be there any more - user_group_vars.yml has been moved to playbooks/inventory/group_vars/hosts.yml | 18:35 |
| odyssey4me | palendae https://review.openstack.org/204320 | 18:36 |
| odyssey4me | it's way better there | 18:36 |
| palendae | shouldn't, you mean? | 18:36 |
| odyssey4me | if anything, we should remove user_group_vars.yml from /etc/openstack_deploy/ if it's there | 18:36 |
| odyssey4me | palendae yep, sorry | 18:36 |
| palendae | np | 18:36 |
| palendae | Ok, so the upgrade script is failing because neither the /etc/ one nor the in-tree are there | 18:36 |
| odyssey4me | palendae heh, interestingly I challenged cloudnull_afk on the lack of inclusion of something in the upgrade script in the master patch | 18:37 |
| palendae | Breaking downstream :( | 18:39 |
| odyssey4me | yep | 18:39 |
| palendae | Not sure how anyone's upgrade tests are wroking | 18:40 |
| palendae | working* | 18:40 |
| odyssey4me | they're using previously build servers? | 18:41 |
| odyssey4me | instead of a clean start | 18:41 |
| odyssey4me | that's an easy hack workaround though, or a known issue note | 18:41 |
| *** annashen has quit IRC | 18:42 | |
| palendae | This is a clean server | 18:43 |
| odyssey4me | and that's why you're hitting it :) | 18:43 |
| palendae | I guess I can tough it | 18:43 |
| palendae | touch* | 18:43 |
| palendae | All our upgrade tests are clean servers | 18:43 |
| palendae | AIOs are always lcean | 18:43 |
| odyssey4me | what I'm saying is that the tests by others might be using previous rc's | 18:43 |
| palendae | So you're saying our upgrade tests are invalid | 18:43 |
| palendae | Sure but that file still isn't there | 18:43 |
| *** sdake has quit IRC | 18:43 | |
| palendae | Well, we don't have any upgrade testing, so w/e | 18:44 |
| odyssey4me | palendae the file would have been there in rc's previous to that merge | 18:45 |
| palendae | I'm saying that merge should have removed that check | 18:45 |
| palendae | But whatever | 18:45 |
| palendae | I'm gonna work around it because that's all going to take longer | 18:45 |
| *** fawadkhaliq has joined #openstack-ansible | 18:45 | |
| *** sdake has joined #openstack-ansible | 18:46 | |
| sigmavirus24 | hey odyssey4me miguelgrinberg so we chatted privately this morning about the keystone v3 in the catalog versus keystone v2 in the catalog. Can we kick a discussion of that off on the mailing for transparency? | 18:48 |
| *** klindgren_ has joined #openstack-ansible | 18:48 | |
| *** TheIntern has quit IRC | 18:49 | |
| klindgren_ | Question -for you guys - since libguestfs is not available via pip/python-wheels how are you handling get that into your nova installs? Just using the system deps and installing into the lxc container? | 18:49 |
| odyssey4me | sigmavirus24 yeah, good idea - do you want to kick that off? | 18:50 |
| *** jmckind has joined #openstack-ansible | 18:50 | |
| odyssey4me | klindgren_ we install via apt on the compute nodes, yes: https://github.com/stackforge/os-ansible-deployment/blob/master/playbooks/roles/os_nova/defaults/main.yml#L294 | 18:52 |
| klindgren_ | would you guys want to install via pip/python-wheel or indifferent? | 18:52 |
| sigmavirus24 | odyssey4me: I think I understand what you were proposing but I also hadn't had enough coffee this morning and I don't want to confuse anyone | 18:55 |
| sigmavirus24 | Because I'm kind of confused myself | 18:55 |
| odyssey4me | klindgren_ we might, not sure really - it'd be a way to get more updated stuff quicker, which we like :) | 18:55 |
| *** fawadk has joined #openstack-ansible | 18:55 | |
| odyssey4me | sigmavirus24 sure, I need a few more hours in the day to put together a mailing list mail though - let me add that to my to-do list and try to find a gap | 18:56 |
| klindgren_ | odyssey4me, kk - I am opening a mailing lost post re: getting libguestfs on pypi | 18:56 |
| *** fawadkhaliq has quit IRC | 18:56 | |
| klindgren_ | https://bugzilla.redhat.com/show_bug.cgi?id=1075594 | 18:56 |
| openstack | bugzilla.redhat.com bug 1075594 in libguestfs "RFE: python-libguestfs should be available on pypi" [Unspecified,New] - Assigned to rjones | 18:56 |
| klindgren_ | list* | 18:56 |
| odyssey4me | klindgren_ I'll keep an eye out for it :) the source building crowds are growing, so I think you'll get support for it | 18:57 |
| sigmavirus24 | odyssey4me: thanks | 18:57 |
| klindgren_ | we are need it because we are doing venvs on some of our services to work around some python stuff - but it sucks as its not fun to handle things in a venv when not installing via pip | 18:59 |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: Cleanup unused functions in swift_rings.py https://review.openstack.org/207174 | 18:59 |
| sigmavirus24 | klindgren_: what does "Licensing problem with the PyPI website mean"? | 18:59 |
| klindgren_ | *shrug* that comes directly from the bug report from R JOnes | 18:59 |
| sigmavirus24 | s/ mean"/" mean/ | 18:59 |
| sigmavirus24 | klindgren_: yeah I saw that, I'm just trying to figure out what licensing problems there are | 18:59 |
| klindgren_ | yea - I looked at the code as far as I can tell its lgplv2 | 19:00 |
| klindgren_ | so I don't think it *should* be an issue | 19:00 |
| sigmavirus24 | cdecimal isn't hosted directly on PyPI (and neither are a bunch of other packages I can think of that no one actually uses) but those are philosophical and regional copyright issues | 19:00 |
| odyssey4me | klindgren_ that makes sense, and there are more and more people looking to install services into venvs to prevent package dependency hell | 19:00 |
| klindgren_ | yep yep - then you jsut get venv relocation hell :-) | 19:01 |
| odyssey4me | klindgren_ hahaha, pot-ay-to pot-ah-to | 19:01 |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: Remove {{ from "with_items" and "when" statements https://review.openstack.org/207175 | 19:02 |
| klindgren_ | the only issue we ran into with nova-computei n a venv besides libguestfs was the fact that the path to rootwrap is hardocded | 19:02 |
| klindgren_ | where in neutron its a config option | 19:03 |
| sigmavirus24 | klindgren_: I don't often relocate virtualenvs, but when I do, I've already had an entire bottle of whiskey to convince me it was a good idea =P | 19:03 |
| odyssey4me | sigmavirus24 +1 | 19:03 |
| klindgren_ | sigmavirus24, we do it only on build as the directory we build the venv may not be the same dir that someone wants to install into | 19:04 |
| klindgren_ | I guess I oculd always build in the dir that someone want to install the venv into - but *shrug* | 19:04 |
| sigmavirus24 | klindgren_: I'm mostly teasing | 19:05 |
| sigmavirus24 | But seriously, relocatable virtualenv support has always been .... shoddy at best | 19:05 |
| klindgren_ | sigmavirus24, yep yep | 19:05 |
| sigmavirus24 | Have you seen the fact that there was a PR in virtualenv to rewrite the entire thing from scratch? | 19:06 |
| klindgren_ | also complaint about pip install nova not isntalling config files/templates - but just python code. Where as neutron installs everything... | 19:06 |
| klindgren_ | sigmavirus24, I have | 19:06 |
| sigmavirus24 | klindgren_: did you test it? | 19:06 |
| sigmavirus24 | I was not brave enough | 19:06 |
| klindgren_ | I did not test the PR | 19:06 |
| klindgren_ | I have bigger problems already :-) | 19:06 |
| sigmavirus24 | lol | 19:07 |
| sigmavirus24 | klindgren_: you can just not do openstack | 19:08 |
| sigmavirus24 | there are other jobs out there | 19:08 |
| odyssey4me | hahaha, someone is a sucker for punishment: https://github.com/skamithi/vagrant-osad | 19:08 |
| sigmavirus24 | =P | 19:08 |
| * klindgren_ use to do cloudstack :-) | 19:08 | |
| sigmavirus24 | odyssey4me: lol | 19:11 |
| palendae | Props to him for figuring all that out, I don't think I've heard any questions from this person | 19:12 |
| sigmavirus24 | palendae: yeah | 19:13 |
| odyssey4me | palendae sigmavirus24 he logged a bug with a fix attached :) | 19:14 |
| palendae | Cool | 19:14 |
| palendae | odyssey4me: It really looks like you're going through *every bug* right now. Seriously, go to bed | 19:14 |
| palendae | They'll be there tomorrow | 19:15 |
| odyssey4me | palendae it's only 8pm :p | 19:16 |
| *** fawadkhaliq has joined #openstack-ansible | 19:16 | |
| palendae | Well stop working :) | 19:16 |
| odyssey4me | I have a drink in hand, so this isn't work. :p | 19:16 |
| *** fawadk has quit IRC | 19:19 | |
| *** weezS has quit IRC | 19:22 | |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: Allow Swift middleware to be set via a variable https://review.openstack.org/181560 | 19:23 |
| *** Apsu has quit IRC | 19:24 | |
| *** weezS has joined #openstack-ansible | 19:24 | |
| *** fawadk has joined #openstack-ansible | 19:24 | |
| *** Apsu has joined #openstack-ansible | 19:24 | |
| *** fawadkhaliq has quit IRC | 19:26 | |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: Fix repo section in example config file https://review.openstack.org/207189 | 19:29 |
| *** jmckind has quit IRC | 19:31 | |
| *** weezS has quit IRC | 19:32 | |
| odyssey4me | palendae d34dh0r53 sigmavirus24 backport to juno: https://review.openstack.org/207158 | 19:41 |
| *** annashen has joined #openstack-ansible | 19:42 | |
| *** jmckind has joined #openstack-ansible | 19:43 | |
| openstackgerrit | Merged stackforge/os-ansible-deployment: Re-enable building of heat_repo_plugins https://review.openstack.org/207155 | 19:43 |
| sigmavirus24 | odyssey4me: is that etherpad with the IdP/SP instructions still current? | 19:43 |
| odyssey4me | sigmavirus24: I don't think so. The attribute structure was changed. | 19:44 |
| sigmavirus24 | cool | 19:44 |
| palendae | odyssey4me: Pretty sure juno that backport needs to be rpc_inventory.json | 19:45 |
| odyssey4me | palendae whoops, if it is can you revise it please? I've just been trawling for orphaned patches. | 19:46 |
| openstackgerrit | Nolan Brubaker proposed stackforge/os-ansible-deployment: Set default inventory file for inventory-manage.py https://review.openstack.org/207158 | 19:49 |
| openstackgerrit | Nolan Brubaker proposed stackforge/os-ansible-deployment: Set default inventory file for inventory-manage.py https://review.openstack.org/207158 | 19:50 |
| odyssey4me | sweet, ,thanks palendae | 19:51 |
| odyssey4me | d34dh0r53 ^ | 19:51 |
| d34dh0r53 | good catch palendae | 19:52 |
| *** KLevenstein has quit IRC | 19:57 | |
| *** KLevenstein_ has joined #openstack-ansible | 19:58 | |
| *** KLevenstein_ is now known as KLevenstein | 19:58 | |
| *** fawadk has quit IRC | 20:02 | |
| *** yaya has joined #openstack-ansible | 20:06 | |
| openstackgerrit | Nolan Brubaker proposed stackforge/os-ansible-deployment: Remove reference to missing user_group_vars file https://review.openstack.org/207204 | 20:07 |
| openstackgerrit | Merged stackforge/os-ansible-deployment: update teardown.sh to remove mongodb https://review.openstack.org/206996 | 20:15 |
| *** openstackgerrit has quit IRC | 20:16 | |
| *** openstackgerrit has joined #openstack-ansible | 20:17 | |
| *** TheIntern has joined #openstack-ansible | 20:19 | |
| *** KLevenstein has quit IRC | 20:19 | |
| *** KLevenstein has joined #openstack-ansible | 20:19 | |
| *** yaya has quit IRC | 20:32 | |
| *** hiddentoken has joined #openstack-ansible | 20:32 | |
| *** jmckind has quit IRC | 20:33 | |
| openstackgerrit | Miguel Grinberg proposed stackforge/os-ansible-deployment: Wrapper script to perform K2K federated login https://review.openstack.org/203859 | 20:44 |
| *** TheIntern has quit IRC | 20:49 | |
| *** yapeng has joined #openstack-ansible | 20:50 | |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: update teardown.sh to remove mongodb https://review.openstack.org/207214 | 20:53 |
| odyssey4me | ^backport d34dh0r53 palendae sigmavirus24 | 20:53 |
| *** yaya has joined #openstack-ansible | 20:54 | |
| *** yapeng has quit IRC | 20:56 | |
| *** jwagner is now known as jwagner_away | 20:58 | |
| *** prometheanfire has quit IRC | 21:05 | |
| odyssey4me | FYI: http://lists.openstack.org/pipermail/openstack-operators/2015-July/007809.html and http://lists.openstack.org/pipermail/openstack-dev/2015-July/070936.html | 21:10 |
| openstackgerrit | Nolan Brubaker proposed stackforge/os-ansible-deployment: Remove reference to missing user_group_vars file https://review.openstack.org/207204 | 21:16 |
| *** prometheanfire has joined #openstack-ansible | 21:17 | |
| *** hiddentoken has quit IRC | 21:30 | |
| *** Mudpuppy has quit IRC | 21:33 | |
| *** Mudpuppy has joined #openstack-ansible | 21:33 | |
| *** annashen has quit IRC | 21:39 | |
| *** annashen has joined #openstack-ansible | 21:43 | |
| *** yaya has quit IRC | 21:45 | |
| *** Mudpuppy has quit IRC | 21:54 | |
| *** Mudpuppy_ has joined #openstack-ansible | 21:54 | |
| *** prad has quit IRC | 21:59 | |
| *** markvoelker has quit IRC | 22:02 | |
| odyssey4me | sigmavirus24 heh, if the admin & public endpoints are v3 but the internal is v2, glance is happy | 22:04 |
| * sigmavirus24 squints | 22:05 | |
| odyssey4me | so typically public/internal are set to the same, right - unless there's an ssl offloader or whatever | 22:05 |
| odyssey4me | all the services are set to use the internal (v2) endpoint | 22:06 |
| odyssey4me | but in this case I'm setting the public endpoint specifically to v3 for horizon's use, and it's working | 22:06 |
| odyssey4me | doing a full tempest run now | 22:06 |
| *** daneyon has quit IRC | 22:11 | |
| odyssey4me | sigmavirus24 oh yes, happy days | 22:13 |
| *** d34dh0r53 has quit IRC | 22:13 | |
| *** odyssey4me has quit IRC | 22:13 | |
| *** d34dh0r53 has joined #openstack-ansible | 22:13 | |
| *** odyssey4me has joined #openstack-ansible | 22:14 | |
| *** Mudpuppy_ has quit IRC | 22:15 | |
| *** d34dh0r53 has quit IRC | 22:15 | |
| *** d34dh0r53 has joined #openstack-ansible | 22:16 | |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: Keystone Federation Service Provider Configuration https://review.openstack.org/194395 | 22:19 |
| *** d34dh0r53 has quit IRC | 22:23 | |
| *** d34dh0r53 has joined #openstack-ansible | 22:23 | |
| *** odyssey4me has quit IRC | 22:24 | |
| *** odyssey4me has joined #openstack-ansible | 22:25 | |
| *** KLevenstein has quit IRC | 22:27 | |
| *** d34dh0r53 has quit IRC | 22:27 | |
| *** d34dh0r53 has joined #openstack-ansible | 22:27 | |
| *** sigmavirus24 has quit IRC | 22:31 | |
| *** d34dh0r53 has quit IRC | 22:32 | |
| *** sigmavirus24 has joined #openstack-ansible | 22:32 | |
| *** d34dh0r53 has joined #openstack-ansible | 22:33 | |
| *** abitha has joined #openstack-ansible | 22:38 | |
| *** sdake has quit IRC | 22:40 | |
| *** sdake has joined #openstack-ansible | 22:40 | |
| *** sdake has quit IRC | 22:42 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 22:43 | |
| openstackgerrit | Merged stackforge/os-ansible-deployment: update teardown.sh to remove mongodb https://review.openstack.org/207214 | 22:52 |
| openstackgerrit | Merged stackforge/os-ansible-deployment: Remove reference to missing user_group_vars file https://review.openstack.org/207204 | 22:52 |
| *** markvoelker_ has joined #openstack-ansible | 22:57 | |
| *** spotz is now known as spotz_zzz | 22:59 | |
| *** daneyon has joined #openstack-ansible | 23:09 | |
| *** daneyon_ has joined #openstack-ansible | 23:10 | |
| *** daneyon has quit IRC | 23:13 | |
| *** JRobinson__ has joined #openstack-ansible | 23:16 | |
| *** annashen has quit IRC | 23:28 | |
| *** daneyon_ has quit IRC | 23:29 | |
| openstackgerrit | Merged stackforge/os-ansible-deployment: Keystone Federation Identity Provider Configuration https://review.openstack.org/194259 | 23:41 |
| *** darrenc is now known as darrenc_afk | 23:47 | |
| *** daneyon has joined #openstack-ansible | 23:53 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!