| *** jasonsb_ has joined #kolla | 00:00 | |
| *** jasonsb has quit IRC | 00:00 | |
| *** cloudnautique has quit IRC | 00:03 | |
| *** achanda has quit IRC | 00:05 | |
| *** tpot has joined #kolla | 00:08 | |
| *** jtriley has joined #kolla | 00:16 | |
| *** jtriley has quit IRC | 00:37 | |
| *** dims_ has joined #kolla | 00:38 | |
| *** jasonsb_ has quit IRC | 01:28 | |
| *** dims_ has quit IRC | 01:41 | |
| *** vinkman1 has joined #kolla | 01:43 | |
| *** vinkman1 has quit IRC | 01:47 | |
| *** vinkman has joined #kolla | 01:58 | |
| *** vinkman has quit IRC | 02:01 | |
| *** unicell has quit IRC | 02:02 | |
| *** vinkman has joined #kolla | 02:02 | |
| *** vinkman has left #kolla | 02:02 | |
| *** cemason has joined #kolla | 02:11 | |
| *** dolpher has joined #kolla | 02:15 | |
| *** devlaps has quit IRC | 02:15 | |
| *** cemason has quit IRC | 02:15 | |
| *** rhallisey has quit IRC | 02:17 | |
| *** achanda has joined #kolla | 02:24 | |
| *** devlaps has joined #kolla | 02:29 | |
| *** jasonsb has joined #kolla | 02:29 | |
| *** jasonsb has joined #kolla | 02:29 | |
| *** devlaps has quit IRC | 02:40 | |
| *** klint has joined #kolla | 02:41 | |
| *** dims_ has joined #kolla | 02:41 | |
| *** sdake has quit IRC | 02:43 | |
| *** dims_ has quit IRC | 02:47 | |
| *** vinkman has joined #kolla | 02:47 | |
| *** vinkman has quit IRC | 02:49 | |
| *** achanda has quit IRC | 02:54 | |
| *** vinkman has joined #kolla | 03:04 | |
| *** vinkman has quit IRC | 03:05 | |
| *** vinkman has joined #kolla | 03:12 | |
| *** tummy has joined #kolla | 03:15 | |
| *** kjelly has quit IRC | 03:31 | |
| *** CBR09 has joined #kolla | 03:34 | |
| CBR09 | morning all | 03:40 |
|---|---|---|
| CBR09 | ah SamYample, when kolla will release v1.0.0? | 03:42 |
| *** dims_ has joined #kolla | 03:43 | |
| *** dims_ has quit IRC | 03:48 | |
| *** alisonh has quit IRC | 03:51 | |
| *** unicell has joined #kolla | 03:52 | |
| *** cloudnautique has joined #kolla | 03:57 | |
| *** vinkman has left #kolla | 03:57 | |
| *** Hui has quit IRC | 03:58 | |
| *** alisonh has joined #kolla | 04:06 | |
| *** sdake has joined #kolla | 04:10 | |
| *** daneyon has joined #kolla | 04:21 | |
| *** daneyon_ has joined #kolla | 04:24 | |
| *** daneyon has quit IRC | 04:27 | |
| *** bmace has quit IRC | 04:40 | |
| *** bmace has joined #kolla | 04:41 | |
| *** dims_ has joined #kolla | 04:45 | |
| *** CBR09 has quit IRC | 04:49 | |
| *** dims_ has quit IRC | 04:50 | |
| *** inc0 has joined #kolla | 04:55 | |
| inc0 | good morning | 04:56 |
| inc0 | asalkeld, did you see schedule? | 04:56 |
| *** achanda has joined #kolla | 05:04 | |
| *** alisonh has quit IRC | 05:08 | |
| *** alisonh has joined #kolla | 05:11 | |
| nihilifer | morning | 05:15 |
| *** inc0 has quit IRC | 05:16 | |
| *** cemason has joined #kolla | 05:21 | |
| asalkeld | inc0 not yet | 05:24 |
| *** cemason has quit IRC | 05:25 | |
| *** tfukushima has joined #kolla | 05:38 | |
| *** tpot has quit IRC | 05:41 | |
| *** dims_ has joined #kolla | 05:47 | |
| *** kjelly has joined #kolla | 05:49 | |
| *** dims_ has quit IRC | 05:52 | |
| *** tummy has quit IRC | 06:00 | |
| *** sdake has quit IRC | 06:05 | |
| *** exploreshaifali has joined #kolla | 06:06 | |
| *** unicell1 has joined #kolla | 06:12 | |
| *** achanda has quit IRC | 06:12 | |
| *** shardy_a1k has joined #kolla | 06:13 | |
| *** unicell has quit IRC | 06:13 | |
| *** shardy_afk has quit IRC | 06:14 | |
| *** shardy_a1k has quit IRC | 06:18 | |
| *** inc0 has joined #kolla | 06:18 | |
| *** shardy_afk has joined #kolla | 06:19 | |
| *** gfidente has joined #kolla | 06:23 | |
| *** CBR09 has joined #kolla | 06:42 | |
| *** kragniz has quit IRC | 06:46 | |
| *** dims_ has joined #kolla | 06:48 | |
| asalkeld | inc0: where is the schedule? steve said he would get that done | 06:52 |
| *** dims_ has quit IRC | 06:53 | |
| *** shardy_afk is now known as shardy | 07:27 | |
| *** daneyon_ has quit IRC | 07:31 | |
| *** jmccarthy has quit IRC | 07:32 | |
| *** jmccarthy has joined #kolla | 07:33 | |
| *** daneyon has joined #kolla | 07:34 | |
| *** sdake has joined #kolla | 07:35 | |
| *** tfukushima has quit IRC | 07:44 | |
| *** openstackgerrit has quit IRC | 07:46 | |
| *** openstackgerrit has joined #kolla | 07:46 | |
| *** dims_ has joined #kolla | 07:49 | |
| inc0 | asalkeld, https://etherpad.openstack.org/p/kolla-mitaka-all-sessions | 07:52 |
| inc0 | there was mail on openstack-dev | 07:52 |
| *** dims_ has quit IRC | 07:55 | |
| *** mbound has joined #kolla | 08:00 | |
| *** sdake has quit IRC | 08:08 | |
| *** CBR09 has quit IRC | 08:13 | |
| *** athomas has joined #kolla | 08:22 | |
| *** sdake has joined #kolla | 08:36 | |
| *** kragniz_ has joined #kolla | 08:41 | |
| *** pbourke has quit IRC | 08:41 | |
| *** pbourke has joined #kolla | 08:41 | |
| *** kragniz_ is now known as kragniz | 08:41 | |
| *** tfukushima has joined #kolla | 08:48 | |
| *** dims_ has joined #kolla | 08:51 | |
| *** sdake has quit IRC | 08:56 | |
| *** dims_ has quit IRC | 08:57 | |
| *** shardy_ has joined #kolla | 08:59 | |
| *** shardy has quit IRC | 09:00 | |
| *** shardy_ has quit IRC | 09:04 | |
| *** shardy has joined #kolla | 09:05 | |
| *** dolpher has quit IRC | 09:08 | |
| asalkeld | ooo thanks inc0 | 09:21 |
| inc0 | you're up late;) | 09:21 |
| *** sdake has joined #kolla | 09:48 | |
| *** dims_ has joined #kolla | 09:49 | |
| pbourke | hey all | 09:49 |
| pbourke | has anyone done AIO *without* a registry lately | 09:50 |
| pbourke | seems to be not looking in my local cache and going straight to docker.io ... | 09:50 |
| nihilifer | yes, I did it recently, worked well | 09:50 |
| nihilifer | what namespace are you using? | 09:52 |
| pbourke | just the default, kollaglue/ | 09:54 |
| pbourke | it usually does this if I forget to change docker_pull_policy to missing | 09:55 |
| inc0 | I'm doing it all the time, I didn't use registry at all | 09:59 |
| *** sdake has quit IRC | 10:03 | |
| SamYaple | pbourke: yea AIO no registry is common for me | 10:07 |
| pbourke | https://paste.fedoraproject.org/281338/35723144/ | 10:09 |
| pbourke | maybe something to do with the fact im using a virtualenv and its confusing the configs | 10:09 |
| SamYaple | pbourke: are you sure you build the data container? | 10:11 |
| SamYaple | i often forget that with my regex | 10:11 |
| pbourke | yeah you can see it in the bottom of the above paste. I'll try building again | 10:12 |
| SamYaple | ah yea | 10:12 |
| SamYaple | what version of docker? | 10:12 |
| nihilifer | pbourke: what's yout /etc/kolla/global.yml? | 10:12 |
| nihilifer | your* | 10:12 |
| pbourke | http://paste.fedoraproject.org/281340/53359731 | 10:13 |
| pbourke | haven't changed much | 10:13 |
| *** cemason has joined #kolla | 10:15 | |
| *** exploreshaifali has quit IRC | 10:18 | |
| *** Kennan2 has quit IRC | 10:19 | |
| *** cemason has quit IRC | 10:20 | |
| *** Kennan has joined #kolla | 10:24 | |
| *** dolpher has joined #kolla | 10:24 | |
| *** jmccarthy has quit IRC | 10:29 | |
| *** jmccarthy has joined #kolla | 10:33 | |
| SamYaple | any testers around running centos willing to test out rabbitmq for me? | 10:47 |
| SamYaple | pbourke.... | 10:48 |
| SamYaple | ill buy you beer.... | 10:48 |
| pbourke | SamYaple: i'll try soon as I get a deploy up | 10:48 |
| SamYaple | fyi, it really needs multinode to test | 10:49 |
| SamYaple | but who knows it might break in AIO ;) | 10:49 |
| pbourke | that's a problem then :( | 10:49 |
| SamYaple | you should deploy kolla and then launch three vms and deploy kolla | 10:50 |
| pbourke | im stuck for hardware atm, vm's were too slow and making me depressed | 10:50 |
| pbourke | so trying to just do AIO on a desktop | 10:50 |
| pbourke | we're getting access to lab equipment soon | 10:50 |
| SamYaple | damn | 10:50 |
| SamYaple | cool | 10:50 |
| pbourke | this seems to work fine https://paste.fedoraproject.org/281365/14453385/ | 10:57 |
| pbourke | so whats wrong kolla :/ | 10:57 |
| SamYaple | what version of docker-py and docker pbourke ? | 10:58 |
| pbourke | docker-py (1.5.0) | 10:58 |
| pbourke | docker 1.8.2 | 10:59 |
| pbourke | ansible 1.9.2 | 10:59 |
| SamYaple | yea all thats good | 10:59 |
| SamYaple | that test you posed is going to pull down the data container pbourke ... | 10:59 |
| SamYaple | it doesnt have missing set | 11:00 |
| pbourke | missing is default | 11:00 |
| SamYaple | oh is it | 11:00 |
| pbourke | seems so yeah | 11:00 |
| SamYaple | remove the api btw | 11:00 |
| SamYaple | we dont use that anymore | 11:00 |
| *** dolpher has quit IRC | 11:22 | |
| pbourke | SamYaple: im getting the error you get when you dont have it ... | 11:23 |
| SamYaple | which error? | 11:23 |
| pbourke | this one https://github.com/ansible/ansible-modules-core/issues/1707 | 11:24 |
| pbourke | http://paste.openstack.org/show/476837/ | 11:24 |
| SamYaple | pbourke: oh then you need to be running something newer than 1.9.2 | 11:24 |
| SamYaple | look at that bug, it happens in 1.9.2 | 11:24 |
| pbourke | right, but officially I thought the requirement was >=1.9.2 | 11:25 |
| SamYaple | sure. but if its broke its broke | 11:25 |
| pbourke | it wasn't broke until that api line was removed though | 11:26 |
| SamYaple | the api line was only a stop gap for a different issue and had to be removed for causing an issue | 11:26 |
| SamYaple | the different issue was fixed in 1.9.2 | 11:26 |
| SamYaple | and you can use 1.9.2 if you use an older version of docker and docker-py | 11:27 |
| pbourke | ok | 11:27 |
| SamYaple | there is a floating table of the crap that works together | 11:27 |
| SamYaple | best advice is run the newest stuff | 11:27 |
| SamYaple | the only _cap_ we have is docker at 1.8.2 | 11:27 |
| SamYaple | and thats only until ansible 1.9.5 | 11:28 |
| *** rhallisey has joined #kolla | 11:33 | |
| *** rhallisey has quit IRC | 11:33 | |
| *** rhallisey has joined #kolla | 11:33 | |
| *** cemason has joined #kolla | 11:38 | |
| *** inc0 has quit IRC | 11:46 | |
| SamYaple | so with the erasure pool i see about the same perforamnce when using the cache tier | 12:16 |
| SamYaple | granted the cluster itself probably can't perform as well | 12:17 |
| SamYaple | but an instance is not affected, which is cool | 12:17 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Automate setting up a cache tier https://review.openstack.org/236756 | 12:25 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Allow specifying erasure as the pool type https://review.openstack.org/236754 | 12:25 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Allow ceph pool names to be configurable https://review.openstack.org/236755 | 12:25 |
| *** klint has quit IRC | 12:58 | |
| *** jruano has joined #kolla | 13:01 | |
| jmccarthy | SamYaple: Sorry - haven't had a chance to double check that nova_consoleauth patch, I know you did - and it's merged now anyway ;) | 13:06 |
| SamYaple | no problem | 13:08 |
| SamYaple | jmccarthy: try out the rabbitmq patches, they are good | 13:08 |
| jmccarthy | SamYaple: I don't have a working kolla env at the moment sadly =/ | 13:11 |
| SamYaple | jmccarthy: sadness | 13:16 |
| SamYaple | what happened to all your environments gusy? | 13:16 |
| jmccarthy | SamYaple: I'll have another go with the Vagrant dev-env soon hopefully, the last attempt bailed on a kernel crash =/ | 13:17 |
| pbourke | I seem to fight different issues daily | 13:18 |
| SamYaple | you guys need to get your act together! | 13:18 |
| jmccarthy | SamYaple: 'you guys' Who is that now ? ;) | 13:18 |
| SamYaple | you know all you people | 13:19 |
| jmccarthy | SamYaple: Ya ! We should get with the program ! | 13:19 |
| SamYaple | ;) | 13:19 |
| jmccarthy | =) | 13:20 |
| SamYaple | ive got my dev environment I am going to be opening up to a few people with slow builds as soon as its ready | 13:20 |
| SamYaple | basically kolla underneath, openstack on top | 13:20 |
| SamYaple | so all kolla dev is done on kolla-in-kolla | 13:20 |
| pbourke | for dev docker seems no more repeatable than rpms | 13:20 |
| pbourke | im hoping the experience for end users will be smoother | 13:20 |
| SamYaple | you run dev docker? | 13:21 |
| pbourke | no I just mean docker as a build/deployment method | 13:21 |
| SamYaple | ah | 13:21 |
| *** dims_ has quit IRC | 13:30 | |
| *** dims_ has joined #kolla | 13:30 | |
| jmccarthy | Wondering: Are there folks using the vagrant-devent with virtualbox much ? | 13:30 |
| jmccarthy | s/devent/devenv | 13:31 |
| SamYaple | i think thats the majority of vagrant users... | 13:31 |
| SamYaple | oh well i dont know about kolla | 13:31 |
| jmccarthy | Yea I remember the previous incarnation of it was good, or might be my imagination since libvirt/nfs etc style came in | 13:32 |
| pbourke | jmccarthy: it hasn't changed hugely the last while | 13:45 |
| pbourke | jmccarthy: change libvirt to virtualbox at the top, that should be all | 13:46 |
| jmccarthy | I'll give it another spin hopefully soon | 13:47 |
| SamYaple | fyi | 13:48 |
| SamYaple | ive got a few patches in to start getting the requiremnens bot to propose things | 13:48 |
| SamYaple | anyone see an issue with that | 13:48 |
| *** inc0 has joined #kolla | 13:49 | |
| pbourke | how does it work | 13:53 |
| pbourke | not turning up too much so far on google | 13:54 |
| *** jruano has quit IRC | 14:05 | |
| SamYaple | pbourke: it just proposes changes | 14:09 |
| SamYaple | pbourke: so a new pathset will pop up with a requirements change and we just approve it | 14:10 |
| openstackgerrit | Paul Bourke proposed openstack/kolla: Add documentation around how to build behind a proxy https://review.openstack.org/237598 | 14:10 |
| pbourke | SamYaple: does it watch pypi for new versions is it? | 14:10 |
| *** sdake has joined #kolla | 14:10 | |
| SamYaple | no its syncs our requirements.txt with global-requirements.txt | 14:11 |
| SamYaple | its how all the projects operator | 14:11 |
| SamYaple | operate* | 14:11 |
| pbourke | ah | 14:12 |
| pbourke | sounds good | 14:12 |
| SamYaple | so the automated cachetier things is working really well | 14:18 |
| SamYaple | made it liberty compatible | 14:18 |
| SamYaple | it doesnt change xisting behaviour | 14:18 |
| *** inc0 has quit IRC | 14:23 | |
| *** vbel has joined #kolla | 14:25 | |
| vbel | good morning/evening everyone :) | 14:26 |
| SamYaple | morning vbel | 14:27 |
| vbel | got a question - hard one... I tried to revive cinder and have working iscsi lvm that survives reboots :) Now I tried to use nfsdriver and got design problem. I could get volumes created in nfs but could not attach them to libvirt. The problem is that nova-compute creates disks for libvirt - mounts nfs shares. Then it commands libvirt to start with them but libvirt runs in own mount namespace. Nothing helps to workaround it. The only solution i have | 14:32 |
| vbel | is quite destructive to current design - to run libvirt and compute in one container | 14:32 |
| SamYaple | vbel: unfortuntely we arent going to be able to fix that | 14:33 |
| SamYaple | vbel: docker 1.10 _should_ have the code to allow that, but even then i dont know we will be able to implement that | 14:34 |
| vbel | SamYaple: yes, it depends on many other things like ansible | 14:34 |
| SamYaple | no its actually all a docker problem | 14:34 |
| SamYaple | so docker 1.10 will allow non-private mount namespaces, but that still requires you to mount the nfs stuff from the host, not the container | 14:35 |
| SamYaple | i dont think youll be able to do that | 14:35 |
| vbel | so the only option if any is to run compute/libvirt in one ... :( | 14:36 |
| vbel | no nfs | 14:36 |
| SamYaple | correct | 14:36 |
| SamYaple | if I am being truthful here I believe the only real option for Kolla is going to be Ceph | 14:37 |
| vbel | SamYaple: thanks for comments! | 14:38 |
| pbourke | SamYaple: what version exactly of ansible are you using | 14:41 |
| *** bradjones has joined #kolla | 14:42 | |
| *** bradjones has quit IRC | 14:42 | |
| *** bradjones has joined #kolla | 14:42 | |
| pbourke | or anyone else who has a working setup :) | 14:50 |
| SamYaple | # ansible --version | 14:51 |
| SamYaple | ansible 1.9.4 | 14:51 |
| SamYaple | # pip freeze | grep docker-py | 14:51 |
| SamYaple | docker-py==1.5.0 | 14:51 |
| SamYaple | # docker version | 14:51 |
| SamYaple | ..... | 14:51 |
| SamYaple | Server: Version: 1.8.2 | 14:51 |
| vbel | 1.9.2 here | 14:53 |
| SamYaple | pbourke: does that help? | 14:56 |
| pbourke | SamYaple: yeah thanks. though they are the same versions Im using | 15:01 |
| pbourke | with ansible 1.9.4 Im getting: | 15:01 |
| pbourke | fatal: [localhost] => {'msg': 'ImportError: cannot import name is_executable', 'failed': True} | 15:01 |
| pbourke | different error with 2.0.x | 15:02 |
| SamYaple | if yo uwere messing around with 2.0 your system is probably foobard | 15:02 |
| pbourke | sigh | 15:09 |
| pbourke | honestly I think its a bug in ansible 1.9.4 | 15:09 |
| pbourke | trying from a venv | 15:09 |
| SamYaple | i use it.... | 15:09 |
| pbourke | same thing | 15:09 |
| pbourke | with AIO? | 15:09 |
| SamYaple | my guess is you are missing a dep | 15:09 |
| *** mpetason has joined #kolla | 15:10 | |
| SamYaple | pbourke: are you installing from source? | 15:10 |
| pbourke | yes | 15:10 |
| SamYaple | oh you arent cloning the modules probably | 15:10 |
| SamYaple | are you inititing the subrepos? | 15:11 |
| pbourke | yup | 15:11 |
| SamYaple | you should just pip install it.... | 15:11 |
| SamYaple | why are you installing from source? | 15:11 |
| pbourke | I suppose this is the issue with baremetal, you can't be sure the env isn't contaminated | 15:11 |
| pbourke | didn't know they'd released 1.9.4 to pip | 15:11 |
| pbourke | will try that next | 15:11 |
| sdake | morning | 15:11 |
| sdake | samyaple sir | 15:12 |
| sdake | slides | 15:12 |
| sdake | prep for summit | 15:12 |
| SamYaple | yea :/ | 15:12 |
| sdake | wednesday? | 15:12 |
| sdake | 4 hour slot | 15:12 |
| sdake | can you do? | 15:12 |
| *** inc0 has joined #kolla | 15:12 | |
| SamYaple | wait 4 hours right now? | 15:12 |
| sdake | wednesday | 15:12 |
| sdake | i am busy today | 15:12 |
| SamYaple | oh todays not wednesday | 15:12 |
| SamYaple | yea thats probably fine | 15:12 |
| *** kjelly_ has joined #kolla | 15:12 | |
| sdake | can you tell me your preferred time since i know you jsut time shifted your day around | 15:13 |
| sdake | we may also have another 4 hour slot on thursday as well | 15:13 |
| SamYaple | ohhhhh pbourke i know what is going on i think | 15:13 |
| sdake | it usually takes about 8 hours | 15:13 |
| SamYaple | pbourke: ansible and venvs are tricky | 15:13 |
| SamYaple | they work until you try to use them :) | 15:13 |
| pbourke | I had a feeling about that ;) | 15:13 |
| SamYaple | when ansible sshs it doesnt ssh _into_ a venv | 15:13 |
| SamYaple | so all the deps you install in tehre arent accessible | 15:14 |
| pbourke | sssh though I think it's working now | 15:14 |
| pbourke | na actually its not | 15:14 |
| SamYaple | you can run it from a venv, but the deps must be on the host | 15:14 |
| inc0 | wassup? | 15:14 |
| pbourke | might have to nuke machine | 15:14 |
| SamYaple | pbourke: just do a new venv i think | 15:14 |
| SamYaple | morning inc0 | 15:14 |
| inc0 | howdy Sam | 15:14 |
| inc0 | you are morning person again? | 15:15 |
| inc0 | not nighly? | 15:15 |
| SamYaple | yea first shift | 15:15 |
| inc0 | that's why it was so boring on irc all day long | 15:15 |
| SamYaple | oh so you want drama | 15:16 |
| * SamYaple changes topic to "no lambda unless you are playing half-life" | 15:16 | |
| inc0 | we'll slowly become project which will only be active during US time | 15:17 |
| SamYaple | mmmm love build profiles | 15:18 |
| *** dwalsh has joined #kolla | 15:19 | |
| kjelly_ | Well, I am not live in US time. I hope the project will not only be active during US time. | 15:20 |
| SamYaple | kjelly_: im always arund :) | 15:20 |
| SamYaple | around even | 15:20 |
| sdake | kjelly_ i'm around lmost the time unless i'm sick | 15:21 |
| sdake | I have asthma, and kids, so when I get a cold I am out for the count | 15:21 |
| sdake | and i get colds often because of the kids | 15:21 |
| sdake | asthma + colds = way worse then healthy person + cold | 15:21 |
| inc0 | right now there is little plague in Poland | 15:22 |
| inc0 | of cold and sore throat | 15:22 |
| sdake | samyaple can you tell me 2 4 hour slot times you like 1 wed 1 thur | 15:22 |
| inc0 | first weeks of autumn... | 15:22 |
| sdake | 10am-2pm? | 15:22 |
| SamYaple | sure | 15:23 |
| sdake | thanks i'll send an invite do you have a new email address? | 15:23 |
| SamYaple | i dont think i ever talked to you on the RAX one... so no | 15:24 |
| inc0 | planning drinks guys?;) | 15:25 |
| pbourke | do you guys use whatsapp? | 15:25 |
| SamYaple | no work :( | 15:25 |
| inc0 | I used to, but not much any more | 15:26 |
| sdake | inc0 we have drinks on the ptl on friday | 15:26 |
| sdake | as long as the bill is under 1k ;) | 15:26 |
| SamYaple | who is the ptl? | 15:26 |
| sdake | aftre that you guys are on your own :) | 15:26 |
| pbourke | :D | 15:26 |
| * sdake wacks sam | 15:26 | |
| SamYaple | :) | 15:26 |
| inc0 | Sam you can be for one night if you want to buy us drinks | 15:26 |
| sdake | yes! | 15:26 |
| sdake | i liek that plan inc0 | 15:27 |
| SamYaple | make it 2 nights | 15:27 |
| sdake | sam yu can get drinks on thursday :) | 15:27 |
| SamYaple | im changing jobs and moving.... so ill buy a round. thats probably all i got | 15:27 |
| inc0 | I'm eager to try out japanese whiskeys | 15:27 |
| pbourke | is there a way to stop ansible cleaning up it's remote code | 15:27 |
| sdake | i'd like to get together for dinner if folks are interested thursday night | 15:27 |
| pbourke | so when it fails you can see what it was trying to execute | 15:27 |
| SamYaple | pbourke: ANSIBLE_KEEP_REMOTE_FILES=1 ansible-playbook blah | 15:28 |
| sdake | anyone interested int hat? | 15:28 |
| pbourke | SamYaple: thank yaw | 15:28 |
| pbourke | sdake: yes | 15:28 |
| SamYaple | yea | 15:28 |
| inc0 | int hat? | 15:28 |
| sdake | it may be pay your own way tho | 15:28 |
| inc0 | ah.. | 15:28 |
| SamYaple | so good news | 15:28 |
| SamYaple | overlayfs seems to work for building | 15:28 |
| sdake | but i'll try to find someon place not off the hook bonkers expensive | 15:28 |
| SamYaple | with docker 1.8.2 and kernel 3.19 | 15:28 |
| inc0 | what was the issue SamYaple ? | 15:28 |
| sdake | it was fixed samyaple? | 15:28 |
| SamYaple | inc0: lots of issues | 15:29 |
| inc0 | uhh | 15:29 |
| SamYaple | sdake: not completely, there are still kernel patches that need to land in kernel | 15:29 |
| SamYaple | but apparentl i can build | 15:29 |
| sdake | oh so its coming along? | 15:29 |
| inc0 | sdake, streetfood and some cool japanese garden is always an option | 15:29 |
| sdake | i mean it will be working at some point? | 15:29 |
| SamYaple | sdake: its working for me right now | 15:29 |
| SamYaple | there are just more known bugs | 15:29 |
| SamYaple | im just not hitting them aparrently | 15:29 |
| sdake | nice | 15:29 |
| SamYaple | or ubuntu backported them maybe | 15:30 |
| SamYaple | thats possible | 15:30 |
| sdake | sam can you build and push all the variants plz | 15:30 |
| SamYaple | unlikely :) | 15:30 |
| SamYaple | i dont have bandwidth for that now | 15:30 |
| sdake | oh right job change | 15:30 |
| sdake | can you buid lnad push ubuntu? | 15:30 |
| SamYaple | ive got gigabit at new place,but im not setup | 15:30 |
| inc0 | SamYaple, when do you move? | 15:30 |
| sdake | pbourke can you bild and push ol? | 15:31 |
| SamYaple | inc0: mid nov or later | 15:31 |
| SamYaple | before the end of the year | 15:31 |
| pbourke | sdake: yes | 15:31 |
| inc0 | I see, I hope we'll get to have a drink in SA | 15:31 |
| sdake | question, how do we build for the docker registry? | 15:31 |
| pbourke | sdake: do you want it pushed to dockerhub? | 15:31 |
| SamYaple | we probably will | 15:31 |
| SamYaple | sdake: what do you mean? | 15:31 |
| sdake | yes push to docker hub all variants i want - looking to distribute | 15:31 |
| sdake | well there needs credentails to push to dockerhub | 15:31 |
| SamYaple | so? | 15:32 |
| SamYaple | oh you mean host does build.py do it | 15:32 |
| sdake | what is the build.py cli to push to dockerhub? | 15:32 |
| *** jmccarthy has quit IRC | 15:32 | |
| sdake | i'll build cento right now for source and binary | 15:32 |
| SamYaple | i dont use build.py for pushing to the hub since the hub can do it in parallel | 15:32 |
| *** cloudnautique has quit IRC | 15:33 | |
| sdake | do you have a script to build for the docker hub then? | 15:33 |
| *** dims__ has joined #kolla | 15:33 | |
| SamYaple | docker images | awk '/kollaglue/ {print $1}' | xargs -P100 -n1 docker push | 15:33 |
| SamYaple | something like that | 15:33 |
| SamYaple | make sure to put an rc tag on it until its been tested | 15:34 |
| SamYaple | then retag | 15:34 |
| SamYaple | btw, we need to talk about tagging | 15:34 |
| sdake | ok sort of lost ;) | 15:34 |
| sdake | we hvae friday 2.5 hours for parking lot | 15:34 |
| sdake | then drinking afterwards | 15:34 |
| sdake | we can cover that in parking lot ok? | 15:34 |
| sdake | contributor meetup on friday | 15:34 |
| SamYaple | so were going to wait to push? | 15:34 |
| SamYaple | im confused | 15:34 |
| sdake | i wantto push now if possible | 15:35 |
| SamYaple | so we should talk about tagging | 15:35 |
| sdake | i thought yuo wanted to have a bigger discussion about how to do it going forward | 15:35 |
| SamYaple | well i mean are you just going to push with the "liberty" tag? | 15:35 |
| SamYaple | what happens when we push another? | 15:35 |
| SamYaple | liberty == latest for liberty | 15:35 |
| SamYaple | but there should still be tags | 15:35 |
| SamYaple | liberty-1 | 15:35 |
| SamYaple | liberty-2 | 15:35 |
| SamYaple | liberty-3 | 15:35 |
| *** dims_ has quit IRC | 15:36 | |
| sdake | liberty.0 | 15:36 |
| sdake | liberty-1 is for milestones | 15:36 |
| sdake | would be my recommendation | 15:36 |
| SamYaple | i dont think you can use . | 15:36 |
| sdake | or liberty-releae.0 | 15:36 |
| sdake | liberty-release-0? | 15:36 |
| SamYaple | be more programatic? liberty-release-000 | 15:37 |
| SamYaple | i doubt well ever push 999 | 15:37 |
| sdake | wfm | 15:37 |
| rhallisey | won't it just be liberty | 15:37 |
| rhallisey | stable-liberty | 15:37 |
| SamYaple | rhallisey: for the latest one, yes | 15:37 |
| SamYaple | but we need to keep a tag for every image we ever push | 15:37 |
| sdake | liberty-stable is good too | 15:37 |
| SamYaple | thats our value add | 15:37 |
| rhallisey | what do you mean by latest though | 15:37 |
| rhallisey | I thought were wern't doing that anymore | 15:38 |
| SamYaple | as in most recent | 15:38 |
| SamYaple | we arent | 15:38 |
| sdake | rhallisey i dont want stable-liberty but liberty-stable is good | 15:38 |
| SamYaple | i would also like the 1.0.0 tags, mayching our versioning | 15:39 |
| SamYaple | ideally we can use any image with any playbook through a branch, but thats unlikely to be the case | 15:39 |
| sdake | ya i guess ill just create a new gpg key | 15:39 |
| sdake | since my old one is fubared | 15:39 |
| sdake | becasue of the upgrade to the intel nvme | 15:39 |
| *** blahRus has joined #kolla | 15:41 | |
| *** shardy_ has joined #kolla | 15:41 | |
| pbourke | sdake: just ping me once you sort the keys and I can push ol | 15:41 |
| SamYaple | pbourke: he is talking about for github | 15:41 |
| SamYaple | pbourke: whats your dockerhub uername | 15:41 |
| sdake | if you need a dockerhub account pbourke let me knwo | 15:42 |
| sdake | or added | 15:42 |
| pbourke | dockerhub username is 'brk3' | 15:42 |
| SamYaple | added pbourke | 15:42 |
| SamYaple | you can push | 15:42 |
| *** shardy has quit IRC | 15:42 | |
| pbourke | it may have scrolled past me, do we want to standarise on build.py options for release builds | 15:42 |
| SamYaple | any other cores need access? | 15:42 |
| rhallisey | pretty sure I already have it | 15:43 |
| SamYaple | inc0 coolsvap ? | 15:43 |
| SamYaple | rhallisey: yea you do | 15:43 |
| sdake | ok so what the hell is the release naming plan then :) | 15:43 |
| sdake | liberty-release-000? | 15:43 |
| SamYaple | +1 | 15:43 |
| rhallisey | ^ that's from git eh? | 15:43 |
| sdake | and the tag will be liberty-1.0.0 in git? | 15:44 |
| SamYaple | sdake: no | 15:44 |
| SamYaple | the git tag is jsut 1.0.0 | 15:44 |
| sdake | ok | 15:44 |
| SamYaple | so we really need to talk about a process for this | 15:44 |
| SamYaple | but we can fish bowl it | 15:44 |
| rhallisey | sorry if I'm behind here, but you want image tag to be liberty-release-000? | 15:44 |
| SamYaple | and retroactively fix | 15:44 |
| SamYaple | rhallisey: yes | 15:44 |
| SamYaple | it will have other tags as well | 15:45 |
| sdake | we dont have fishbowl sessions for this | 15:45 |
| sdake | we can cover on friday | 15:45 |
| sdake | but i'd like images up before friday | 15:45 |
| SamYaple | images can push with liberty-release-000 | 15:45 |
| pbourke | etherpad? | 15:45 |
| rhallisey | SamYaple, how often are we planning on building? | 15:45 |
| SamYaple | we can add other tags later | 15:45 |
| rhallisey | and pushing a new tag | 15:45 |
| SamYaple | rhallisey: i would say once a week. but itll be influnced by things like testing, tagging, security, etc | 15:45 |
| pbourke | https://etherpad.openstack.org/p/kolla-release-builds | 15:45 |
| sdake | lets etherpad guys | 15:46 |
| *** shardy_ has quit IRC | 15:46 | |
| kjelly_ | How about iberty-release-XXXXX ? XXXXX is git commit id. only five char | 15:46 |
| SamYaple | kjelly_: theres not assurance we dont build two images from the same git commit | 15:46 |
| *** shardy has joined #kolla | 15:47 | |
| inc0 | I should make dockerhub account at some point...I'll come back to you later | 15:47 |
| sdake | inc0 creating dockerhub account takes 3 minutes | 15:47 |
| sdake | do it now pelase so we can get it setup | 15:48 |
| rhallisey | SamYaple, I don't follow what you have | 15:48 |
| SamYaple | do we build images from head of stable? | 15:49 |
| inc0 | SamYaple, inc0 it is;) | 15:49 |
| SamYaple | or tag? | 15:49 |
| inc0 | I'd say head | 15:49 |
| inc0 | or have multiple tags | 15:50 |
| inc0 | latest, stable-liberty and such | 15:50 |
| SamYaple | inc0: if we build against head, how can we ensure pip installed kolla works | 15:50 |
| inc0 | tags then? | 15:50 |
| inc0 | and default tag will be stable? | 15:51 |
| SamYaple | we can build against tags | 15:51 |
| inc0 | I mean docker images tags | 15:51 |
| inc0 | not github tags | 15:51 |
| SamYaple | im confused then | 15:51 |
| SamYaple | the images we push cant be from head i dont think | 15:51 |
| rhallisey | SamYaple, got a question on 'liberty' tag. When will we ever tag liberty-release-XXX if liberty is always latest | 15:51 |
| SamYaple | rhallisey: for not wanting to run the latest for whatever reason | 15:51 |
| inc0 | guys can we discuss it tomorrow on meeting? | 15:51 |
| SamYaple | still wanting to run the playbooks without new containers being used | 15:51 |
| SamYaple | latest is borked | 15:52 |
| SamYaple | this is the immutable thing we have | 15:52 |
| rhallisey | SamYaple, but how will that get built? Everytime we build latest we also have to build say liberty-release-001. Where 002 would be the latest | 15:53 |
| SamYaple | rhallisey: well we never build "latest" we would build liberty-release-001 | 15:53 |
| SamYaple | once that is pushed and tested then we update the latest tag | 15:53 |
| rhallisey | ok I think I understand now | 15:54 |
| SamYaple | if we are never building from the head of the branch, maybe a plain 'liberty' tag doesnt make any sense | 15:58 |
| SamYaple | maybe just the version number | 15:58 |
| kjelly_ | SamYaple: When user use image with latest tag, does the user know image with latest tag is liberty-release-001 now ? | 15:58 |
| SamYaple | kjelly_: no, they would only have teh sha. they could use that to find a specific tag if they wanted | 15:59 |
| kjelly_ | SamYaple: Ok. So when user reoport bug, the user need to provide sha to let us find what images he use. Is that right? | 16:00 |
| *** CBR09 has joined #kolla | 16:03 | |
| *** cloudnautique has joined #kolla | 16:05 | |
| *** cloudnautique has quit IRC | 16:05 | |
| *** cloudnautique has joined #kolla | 16:05 | |
| *** inc0 has quit IRC | 16:05 | |
| SamYaple | kjelly_: correct | 16:10 |
| SamYaple | kjelly_: we probably need some template for them to fill out | 16:10 |
| rhallisey | that is a good point.. | 16:10 |
| sdake | ok guys i am waiting on the liberty release of rdo before tagging | 16:11 |
| sdake | apevec is moving it to the top of his work queue | 16:11 |
| sdake | he said today | 16:11 |
| rhallisey | cool | 16:11 |
| sdake | approximately 1 hour | 16:12 |
| sdake | so samyaple you said something about tagging as a rc vs tagging as a relase once we tesst? | 16:12 |
| sdake | but there are no directions in the ehterpad for that | 16:12 |
| SamYaple | its not needed if we only push up liberty-release-XXX | 16:12 |
| SamYaple | that way we can test and repush if needed | 16:12 |
| SamYaple | until we update the other tags (1.0.0) we are fine | 16:13 |
| *** unicell has joined #kolla | 16:13 | |
| *** gfidente has quit IRC | 16:13 | |
| *** gfidente has joined #kolla | 16:13 | |
| *** unicell1 has quit IRC | 16:14 | |
| sdake | for this rdo rpm repos change, we are violating the typical backport mechanism | 16:14 |
| sdake | its going straight into the liberty branch | 16:14 |
| SamYaple | why | 16:14 |
| SamYaple | is it a patch only for liberty? | 16:15 |
| sdake | because we would hae to remove https://github.com/openstack/kolla/blob/master/docker/base/Dockerfile.j2#L53-L56 | 16:15 |
| sdake | and then revert that removal | 16:15 |
| sdake | yes only for liberty | 16:15 |
| SamYaple | then its not a backport | 16:15 |
| sdake | it makes the rpms liberty specific | 16:15 |
| SamYaple | thanks fine | 16:15 |
| sdake | right | 16:15 |
| SamYaple | if its a backport we wont backport master | 16:15 |
| sdake | yes its juts not a typical workflow | 16:15 |
| SamYaple | if its not then yea theres no option | 16:15 |
| sdake | typically nothing should ever jsut jump into a branch without a backport | 16:16 |
| SamYaple | thats not true at all | 16:16 |
| sdake | but this is our probaby only exception | 16:16 |
| SamYaple | it hasnt even been the only exception so far | 16:16 |
| sdake | things shoudl go into master first if feasible | 16:16 |
| SamYaple | its far from it | 16:16 |
| sdake | to avoid regressions | 16:16 |
| SamYaple | right but master changes, well have bugs to fix in liberty that dont apply to master | 16:16 |
| sdake | ya | 16:16 |
| SamYaple | if it applies to multiple branches it goes to master first | 16:17 |
| sdake | hopefpully they will be bugs in master too, or we would hae already backportedthem :) | 16:17 |
| sdake | right we ar ein violent agreement :) | 16:17 |
| SamYaple | YEA! | 16:17 |
| CBR09 | hi all | 16:19 |
| CBR09 | I have a question | 16:20 |
| CBR09 | openstack liberty was released | 16:20 |
| sdake | this is the last bug for liberty https://bugs.launchpad.net/kolla/+bug/1508127 | 16:20 |
| openstack | Launchpad bug 1508127 in kolla "Need a Liberty-specific repo for RDO" [Critical,Confirmed] - Assigned to Steven Dake (sdake) | 16:20 |
| CBR09 | If I want build images from liberty-stable | 16:20 |
| CBR09 | what I must do? | 16:20 |
| sdake | cbr i think i jumped the gun on that email | 16:20 |
| sdake | my apologies | 16:20 |
| sdake | give me about 1 hour please | 16:21 |
| sdake | then checkout the liberty branch of kolla | 16:21 |
| sdake | and use that | 16:21 |
| sdake | we are also going to be buildling liberty for all the container base oses | 16:21 |
| sdake | so you can pull directly from the registry | 16:21 |
| sdake | SamYaple do you think we need to update liberty tags to pull liberty-release-000 for example? | 16:22 |
| sdake | in the ansible files | 16:22 |
| SamYaple | no they would pull '1.0.0' | 16:22 |
| SamYaple | and yes we should update that | 16:22 |
| SamYaple | ill pop a patch | 16:22 |
| sdake | can you get a patch in queue | 16:22 |
| sdake | thanks | 16:22 |
| CBR09 | yea, thank sdake | 16:22 |
| sdake | SamYaple please file a bug too or if you want ill file | 16:23 |
| sdake | shall I file? | 16:23 |
| SamYaple | TrivialFix? | 16:23 |
| SamYaple | this isnt a bug | 16:23 |
| sdake | ok tf is fine | 16:23 |
| sdake | the repo thing isn'ta tv so i didn't tf tag it | 16:23 |
| sdake | but this shoudl be | 16:24 |
| sdake | although please make the commit message good enough to understand what the change is about | 16:24 |
| sdake | there isa higher bar on trivialfix in the commit message | 16:24 |
| SamYaple | so we set the version anywhere? | 16:24 |
| SamYaple | do we set* | 16:25 |
| sdake | we set it for the image tag "latest" | 16:25 |
| sdake | in the globals.yml I think? | 16:25 |
| sdake | or somewhere else? | 16:25 |
| SamYaple | right but what about pbr and stuff | 16:25 |
| *** exploreshaifali has joined #kolla | 16:25 | |
| sdake | version 1.0.0 is in setup.cfg | 16:25 |
| SamYaple | for pip and packaging | 16:25 |
| sdake | thats fixed | 16:25 |
| SamYaple | ah ok | 16:25 |
| sdake | double check tho plz | 16:25 |
| SamYaple | we should be setting those two at the same time | 16:25 |
| CBR09 | sdake, but I want to know how to build images?, only use build.py? | 16:25 |
| sdake | cbr09 right | 16:26 |
| sdake | cbr09 but the images are not quite ready to be built yet | 16:26 |
| kjelly_ | SamYaple: launchpad seems not to support template which let user use to report bug. Maybe we need to a document how to report bugs. | 16:26 |
| sdake | samyaple so if your putting 1.0.0 in place of "latest" | 16:26 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Set version tag for Liberty https://review.openstack.org/237675 | 16:26 |
| CBR09 | Why not?, I saw openstack liberty was release few day ago | 16:27 |
| sdake | do we not want liberty in tthere? | 16:27 |
| sdake | cbr09 we hae 1 outstanding bug | 16:27 |
| sdake | we are dependent upon rdo community | 16:27 |
| SamYaple | idk sdake | 16:27 |
| sdake | if you are buliding ubuntu your fine | 16:27 |
| SamYaple | 1.0.0-liberty? | 16:27 |
| sdake | samyaple persoanlly i'd go for libeerety-1...0.0 | 16:27 |
| sdake | wow spelling! | 16:27 |
| sdake | liberty-1.0.0 | 16:27 |
| sdake | or 1.0.0-liberty | 16:27 |
| SamYaple | so we would have 1.0.0-liberty, 1.0.0-liberty-release-XXX, liberty-release-XXX ? | 16:28 |
| SamYaple | or should we just have 1.0.0-liberty, 1.0.0-liberty-release-XXX, | 16:28 |
| sdake | second option | 16:29 |
| SamYaple | agreed | 16:29 |
| sdake | [09:28:56] <SamYaple>or should we just have 1.0.0-liberty, 1.0.0-liberty-release-XXX, | 16:29 |
| SamYaple | we no | 16:29 |
| SamYaple | we need all three | 16:29 |
| SamYaple | dont we? | 16:29 |
| SamYaple | or is it not a ticking count for the entire branch anymore? | 16:29 |
| SamYaple | just the one version? | 16:29 |
| sdake | oh elt me think about that | 16:30 |
| CBR09 | sdake: so, I just run ./build.py -b ubuntu -t source, it will build ubuntu with liberty-stable, is it right? | 16:30 |
| sdake | i think that gets more complex to manage | 16:30 |
| sdake | lets have a tick per 1.0.0 1.0.1 etc | 16:30 |
| sdake | so resets to 0 each new release | 16:30 |
| *** cloudnau_ has joined #kolla | 16:30 | |
| SamYaple | sdake: maybe. we can right a little tool to query dockerhub for the next increment | 16:30 |
| SamYaple | 1.0.0-liberty-release-000 | 16:31 |
| SamYaple | 1.0.0-liberty-release-001 | 16:31 |
| sdake | the problem is if liberty-1.0.0 comes out | 16:31 |
| SamYaple | 1.0.1-liberty-release-000 | 16:31 |
| sdake | and liberty 1.0.1 comes out | 16:31 |
| sdake | then if we have a global #, there is no wayt o keep the stuff in sync | 16:31 |
| SamYaple | well the idea is we would no longer pump out versions for the 1.0.0 branch | 16:31 |
| SamYaple | but still does the above number workf or you? | 16:31 |
| sdake | samyaple could you restate which #s,s there are e so many :) | 16:32 |
| *** cloudnautique has quit IRC | 16:32 | |
| SamYaple | sdake: bottom of the etherpad | 16:32 |
| sdake | let me check pad | 16:33 |
| SamYaple | sdake: i dont follow your question | 16:35 |
| SamYaple | if we push out a 1.0.2 we are not pushing any 1.0.1 anymore | 16:35 |
| sdake | so each release retags with 1.0.z? | 16:35 |
| sdake | got that | 16:35 |
| SamYaple | yea | 16:35 |
| sdake | if we push 1.0.1-001 the n1.0.1-002, 1.0.1 gets tagged to 1.0.1-002 ? | 16:35 |
| SamYaple | yes | 16:36 |
| sdake | cool hwo do we do that - are there commands in the pad? | 16:36 |
| sdake | i dont know how to tag an existing dockerhub entry | 16:36 |
| *** asalkeld_ has joined #kolla | 16:36 | |
| sdake | hey asalkeld | 16:36 |
| SamYaple | oh well work on that | 16:36 |
| SamYaple | in not worried about that atm | 16:37 |
| sdake | ok so after summit we can figure that shit out | 16:37 |
| sdake | a script to do the job would be dandy | 16:37 |
| SamYaple | yea we can do that | 16:37 |
| *** asalkeld has quit IRC | 16:37 | |
| sdake | btw, did folks see the schedule i sent to the ml | 16:38 |
| sdake | and anyone displeased with what had to be trimmed based upon available space? | 16:38 |
| sdake | if so, we hae 2.5 hours friday we have available for either parking lot or drinking (water responsibly of course:) | 16:38 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Set version tag for Liberty https://review.openstack.org/237675 | 16:38 |
| sdake | pbourke looks good for your ack if your willing | 16:40 |
| sdake | so how do we do this initial build once we are ready since we dont have tagging sorted out? just use --tag 1.0.0-liberty? | 16:41 |
| *** achanda has joined #kolla | 16:41 | |
| SamYaple | --tag 1.0.0-liberty-release-000 | 16:41 |
| SamYaple | pnce that is validated as working we add another tag | 16:41 |
| SamYaple | if it doesnt work we repush with: | 16:41 |
| SamYaple | --tag 1.0.0-liberty-release-001 | 16:41 |
| sdake | i see | 16:42 |
| SamYaple | i figure thats cleaner than -rc | 16:42 |
| sdake | how do we add the validated tag | 16:42 |
| *** jasonsb has quit IRC | 16:42 | |
| SamYaple | dockerhub ui is one way | 16:42 |
| SamYaple | but we really need a script | 16:42 |
| SamYaple | manual is pain in the rear | 16:42 |
| sdake | can thee make a script | 16:43 |
| SamYaple | maybe | 16:43 |
| SamYaple | probably not until after summit | 16:43 |
| SamYaple | or during | 16:43 |
| sdake | ok well we will need to manually do it then | 16:43 |
| SamYaple | yea | 16:43 |
| sdake | yo usay the dockerhub ui can do it? | 16:43 |
| SamYaple | yup | 16:43 |
| SamYaple | you can add/remove tags | 16:43 |
| sdake | so we have to find all 90 containers and tag them? | 16:44 |
| * sdake groans | 16:44 | |
| SamYaple | yea :( | 16:44 |
| SamYaple | you could also retag locally and repush | 16:44 |
| SamYaple | it should go quicker, but not instant | 16:44 |
| sdake | is it possible to easily script that? | 16:44 |
| sdake | somethign hacky | 16:44 |
| SamYaple | yea i do it with a oneliner | 16:44 |
| sdake | just to speed things up for everyone | 16:44 |
| sdake | cool! | 16:45 |
| sdake | put in etherpad plz | 16:45 |
| SamYaple | there bottom of the therpad sdake | 16:47 |
| sdake | got it | 16:48 |
| SamYaple | replace oldtag and nettag | 16:48 |
| *** tummy has joined #kolla | 16:48 | |
| sdake | so quick question | 16:49 |
| sdake | if we have 1.0.0-liberty-release-000 | 16:49 |
| sdake | and we release 1.0. | 16:49 |
| sdake | 1.0.1 | 16:49 |
| sdake | would it be 1.0.1-liberty-release-001? | 16:50 |
| SamYaple | no | 16:50 |
| SamYaple | look at the list | 16:50 |
| *** unicell has quit IRC | 16:50 | |
| sdake | got it line 58-63 | 16:51 |
| *** kjelly_ has quit IRC | 16:51 | |
| sdake | apevec is testing rpms now | 16:51 |
| SamYaple | ansible is being dumb again :/ | 16:51 |
| *** sdake has quit IRC | 16:56 | |
| *** achanda has quit IRC | 17:00 | |
| SamYaple | i really dont like ansible | 17:03 |
| *** diga has joined #kolla | 17:03 | |
| SamYaple | it violates so much DRY to use it... | 17:03 |
| *** achanda has joined #kolla | 17:13 | |
| *** sdake has joined #kolla | 17:14 | |
| *** diga has quit IRC | 17:16 | |
| *** unicell has joined #kolla | 17:20 | |
| *** sdake_ has joined #kolla | 17:21 | |
| *** sdake has quit IRC | 17:23 | |
| sdake_ | http://ur1.ca/o49yp -> http://paste.fedoraproject.org/281564/44536186 | 17:24 |
| sdake_ | testing now | 17:25 |
| SamYaple | i was wrong | 17:26 |
| SamYaple | everything build but kolla-ansible | 17:26 |
| SamYaple | with overlayfs | 17:26 |
| *** dims__ has quit IRC | 17:27 | |
| *** dims_ has joined #kolla | 17:28 | |
| *** dwalsh has quit IRC | 17:28 | |
| sdake_ | wierd kolla-ansible doesn't build | 17:36 |
| SamYaple | its because of the git install | 17:36 |
| SamYaple | pip is doing stupidsocket stuff | 17:36 |
| sdake_ | SamYaple did your 1.0.0-liberty patch get acked? | 17:37 |
| SamYaple | i dont think so | 17:37 |
| sdake_ | rhallisey can you ack that patch please | 17:37 |
| SamYaple | https://review.openstack.org/#/c/237675/ | 17:38 |
| sdake_ | SamYaple do you ahve to change any of the code in build.py to push to the docker registry | 17:38 |
| sdake_ | I seem to recall in the past I had to change a flag about insecure/vs/secure | 17:39 |
| *** kragniz has quit IRC | 17:39 | |
| SamYaple | not a local registry | 17:39 |
| SamYaple | probaly the docker hub | 17:39 |
| *** achanda has quit IRC | 17:39 | |
| sdake_ | ya i mean the docker hub | 17:39 |
| SamYaple | maybe | 17:39 |
| *** kragniz has joined #kolla | 17:39 | |
| SamYaple | haveny tried | 17:39 |
| SamYaple | probably need to read the auth file | 17:39 |
| sdake_ | the auth file is read by docker-py | 17:39 |
| sdake_ | how do you push to the docker registry then? | 17:39 |
| SamYaple | is that a question | 17:39 |
| sdake_ | no its a statement | 17:40 |
| sdake_ | i am certain I have read the code | 17:40 |
| sdake_ | but i am not sure about the secure flag | 17:40 |
| sdake_ | or is that a question realting to how to pus hto docker hub? | 17:40 |
| openstackgerrit | Merged openstack/kolla: Set version tag for Liberty https://review.openstack.org/237675 | 17:40 |
| *** jpeeler has quit IRC | 17:41 | |
| *** cloudnau_ has quit IRC | 17:42 | |
| sdake_ | SamYaple my question was how did you push ubuntu for liberty-3 to the docker hub? | 17:42 |
| sdake_ | not using ubild.py? | 17:42 |
| *** jasonsb has joined #kolla | 17:42 | |
| SamYaple | sdake_: docker push | 17:42 |
| *** cloudnautique has joined #kolla | 17:43 | |
| SamYaple | docker iamges | awk '/kollaglue/ {print $1}' | xargs -n1 -P100 docker push | 17:43 |
| sdake_ | you pushed each iamge one y one? | 17:43 |
| sdake_ | oh I see | 17:43 |
| SamYaple | soemthign like that | 17:43 |
| sdake_ | what is -P100? | 17:43 |
| sdake_ | and n1? | 17:43 |
| *** jpeeler has joined #kolla | 17:43 | |
| *** jpeeler has joined #kolla | 17:43 | |
| SamYaple | parallel 100 | 17:44 |
| SamYaple | basically push them all at once | 17:44 |
| *** dwalsh has joined #kolla | 17:44 | |
| sdake_ | can you geet that in thee pad so people know, I tend to ues build.py when I've done it in the past but had to hack the code base a bit | 17:44 |
| * sdake_ hates generating gpg keys | 17:47 | |
| *** achanda has joined #kolla | 17:47 | |
| *** sdake_ is now known as sdake | 17:47 | |
| sdake | ubuntu source keystone has had 116 pulls | 17:53 |
| sdake | that was all rc3 | 17:53 |
| sdake | gpg key created again | 17:58 |
| sdake | damn designate isn't buliding | 18:06 |
| openstackgerrit | Steven Dake proposed openstack/kolla: Use the liberty release RDO RPMS https://review.openstack.org/237722 | 18:09 |
| sdake | rhallisey/samyaple/pbourke/jpeeler can i get acks on that please ^^ | 18:09 |
| rhallisey | roger | 18:10 |
| *** achanda has quit IRC | 18:11 | |
| openstackgerrit | Steven Dake proposed openstack/kolla: Use the liberty release RDO RPMS https://review.openstack.org/237722 | 18:11 |
| sdake | it has been tested btw | 18:11 |
| *** gfidente is now known as gfidente|afk | 18:11 | |
| sdake | may have one more major bug | 18:12 |
| sdake | looks like dont ack | 18:13 |
| sdake | may not build from source | 18:13 |
| SamYaple | sdake: https://review.openstack.org/#/c/236938/ | 18:15 |
| SamYaple | lets drop that in berfore yo utag | 18:15 |
| SamYaple | looks like the last liberty patch | 18:15 |
| sdake | let me read please | 18:15 |
| *** jpeeler has quit IRC | 18:15 | |
| sdake | but likely :) | 18:15 |
| SamYaple | no blind obediance! | 18:15 |
| *** jpeeler has joined #kolla | 18:15 | |
| sdake | anything else? | 18:17 |
| SamYaple | looks like all in the queue for libery | 18:17 |
| sdake | anything else needing backports that was missed? | 18:18 |
| sdake | you said you have a tool to test for that | 18:18 |
| SamYaple | not that had a flag | 18:18 |
| SamYaple | 'tool' | 18:18 |
| SamYaple | it was a oneliner | 18:18 |
| sdake | got it | 18:18 |
| sdake | so if something got missed becaue of lack of flag, i'm ok with that | 18:18 |
| SamYaple | pretty much | 18:19 |
| SamYaple | i think we should have a potential-backport tag as well | 18:19 |
| SamYaple | so we dont miss things | 18:19 |
| SamYaple | we can tag all patches we think might need a backport as well as ones we know need them | 18:19 |
| openstackgerrit | Merged openstack/kolla: add "registry" flag to "tools/build.py" https://review.openstack.org/236938 | 18:20 |
| sdake | how that is typiclly done is bakcport-potential in launchpad | 18:25 |
| SamYaple | ? | 18:26 |
| sdake | using a launchpad tag | 18:26 |
| sdake | that said I like the git based workflow we are using | 18:26 |
| sdake | i find it infintely more easy to track then launchpad\ | 18:26 |
| *** diogogmt has joined #kolla | 18:28 | |
| SamYaple | so we are all on the same page, we need to switch to ansible 2.0 for mitaka | 18:28 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Automate setting up a cache tier https://review.openstack.org/236756 | 18:30 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Allow specifying erasure as the pool type https://review.openstack.org/236754 | 18:30 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Allow ceph pool names to be configurable https://review.openstack.org/236755 | 18:30 |
| *** mpetason has quit IRC | 18:31 | |
| *** achanda has joined #kolla | 18:33 | |
| *** diogogmt has quit IRC | 18:34 | |
| *** mbound has quit IRC | 18:35 | |
| *** diogogmt has joined #kolla | 18:35 | |
| sdake | ok guys I've got the final patch building both binary and source rdo of centos now | 18:36 |
| sdake | I have doctor papoointment for about 1 hour then I will be back | 18:36 |
| sdake | and then we will tag and start the building process | 18:36 |
| sdake | bbiaf :) | 18:36 |
| *** achanda has quit IRC | 18:39 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Automate setting up a cache tier https://review.openstack.org/236756 | 18:39 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Allow specifying erasure as the pool type https://review.openstack.org/236754 | 18:39 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Allow ceph pool names to be configurable https://review.openstack.org/236755 | 18:39 |
| *** diogogmt has quit IRC | 18:44 | |
| *** exploreshaifali has quit IRC | 18:44 | |
| *** diogogmt has joined #kolla | 18:45 | |
| *** sdake has quit IRC | 18:47 | |
| *** athomas has quit IRC | 18:47 | |
| *** pbourke_ has joined #kolla | 18:48 | |
| *** achanda has joined #kolla | 18:48 | |
| pbourke_ | https://www.oracle.com/corporate/pressrelease/oracle-openstack-4-oracle-linux-101915.html | 18:49 |
| pbourke_ | \o/ | 18:49 |
| SamYaple | pbourke_: you know alot about Kolla, you should apply there | 18:50 |
| SamYaple | pbourke_: how much as this been change from vanilla kolla? | 18:50 |
| pbourke_ | SamYaple: very little really | 18:51 |
| SamYaple | cool | 18:51 |
| pbourke_ | SamYaple: main change would be mysqlcluster instead of mariadb | 18:51 |
| SamYaple | thatll be good in the long run | 18:51 |
| pbourke_ | kolla goes commercial, you guys are all now sell outs | 18:51 |
| SamYaple | wheres my money | 18:52 |
| rhallisey | lol | 18:52 |
| SamYaple | call me what you will i gots to get paid | 18:52 |
| pbourke_ | oh, and the cli | 18:53 |
| pbourke_ | which we're v. interested in people looking at | 18:53 |
| pbourke_ | source should be up very soon if not already | 18:53 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: In gate functional testing https://review.openstack.org/233993 | 18:53 |
| SamYaple | pbourke_: bring to kolla :( | 18:53 |
| SamYaple | need cli | 18:53 |
| pbourke_ | we def will | 18:54 |
| SamYaple | ok guys that gate patch can go in. i have to have this (and the backport) merge before we can actually get infra to give us new tests | 18:55 |
| SamYaple | validate it doesnt break existing and merge please :) | 18:56 |
| bmace | the cli source should be up on oss.oracle.com imminently, but if you want to see something before that here are the cli docs: http://docs.oracle.com/cd/E64747_01/E64749/html/osusg-kollacli-commands.html | 19:01 |
| *** shardy has quit IRC | 19:04 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: In gate functional testing https://review.openstack.org/233993 | 19:05 |
| *** shardy has joined #kolla | 19:06 | |
| *** thumpba has joined #kolla | 19:09 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: In gate functional testing https://review.openstack.org/233993 | 19:15 |
| *** achanda has quit IRC | 19:23 | |
| *** sdake has joined #kolla | 19:25 | |
| SamYaple | bmace: pbourke_ yea you should make that in kolla.... | 19:26 |
| SamYaple | need those things bad | 19:26 |
| *** daneyon has quit IRC | 19:28 | |
| bmace | SamYaple: that is definitely the plan! :) | 19:31 |
| *** mbound has joined #kolla | 19:36 | |
| *** sdake has quit IRC | 19:40 | |
| *** mbound has quit IRC | 19:41 | |
| *** CBR09 has quit IRC | 19:44 | |
| *** sdake has joined #kolla | 19:51 | |
| sdake | yo folks | 19:53 |
| sdake | whats up | 19:53 |
| sdake | my computer failed while i was at dr office | 19:54 |
| sdake | so hvae to rebuild again groan | 19:54 |
| *** thumpba has quit IRC | 19:56 | |
| *** shardy has quit IRC | 19:56 | |
| *** shardy has joined #kolla | 19:58 | |
| *** tpot has joined #kolla | 20:01 | |
| *** daneyon has joined #kolla | 20:05 | |
| *** daneyon has quit IRC | 20:11 | |
| *** achanda has joined #kolla | 20:24 | |
| *** achanda has quit IRC | 20:25 | |
| openstackgerrit | Steven Dake proposed openstack/kolla: Increase Ceph RPM gpg key timeout to 90 seconds https://review.openstack.org/237768 | 20:28 |
| sdake | please ack ^^!! | 20:28 |
| openstackgerrit | Steven Dake proposed openstack/kolla: Use the liberty release RDO RPMS https://review.openstack.org/237722 | 20:30 |
| sdake | SamYaple jpeeler rhallisey pbourke can you folks ack the above | 20:30 |
| SamYaple | you should point it to a mirror | 20:31 |
| SamYaple | ceph.com goes down all the time | 20:31 |
| SamYaple | thats why i use a keyserver in ubuntu sdake | 20:31 |
| sdake | i was just following the directions | 20:31 |
| sdake | is there a mirror? | 20:31 |
| sdake | I didn't see a mirror in their documentation | 20:31 |
| SamYaple | all kinds of mirrors | 20:31 |
| SamYaple | its not there | 20:31 |
| SamYaple | normal gpg mirrors | 20:32 |
| SamYaple | all over the place | 20:32 |
| sdake | i'd rather follow the docs as much as possible and if we relaly run into problems we can adjust in the future | 20:32 |
| sdake | this close to release atleaste | 20:33 |
| SamYaple | well this is the type of problem im talking about | 20:33 |
| sdake | its just slow because they are using git for their key distribution | 20:33 |
| SamYaple | yes. most everyone keeps thier release keys there like that. | 20:34 |
| SamYaple | you still use a keyserver | 20:34 |
| sdake | i am open to changing after release if we have some soak time | 20:34 |
| sdake | id' like to stick to the docs for the short term if thats aok with you :) | 20:34 |
| SamYaple | its fine with me. it doesnt touch ubuntu | 20:35 |
| sdake | cool can you ack then :) | 20:35 |
| sdake | tia | 20:35 |
| SamYaple | would never do that there because it is going to break | 20:35 |
| SamYaple | its been down for several days before :/ | 20:35 |
| sdake | i can file a tech debt research bug if you like | 20:35 |
| SamYaple | i only care because i dont want to make this voting if you have that as an external dep | 20:36 |
| sdake | i htink in general we need to find solutions to these proboelms for hte voting gates | 20:36 |
| sdake | oh we were thinking along the same lines | 20:36 |
| SamYaple | yea but its only centos that i dont want voting. i dont have these issues with ubuntu :) | 20:37 |
| sdake | i'm going to try to get mirrors going for all our external repos internally | 20:37 |
| SamYaple | docker is the big "we need a mirror" thing | 20:37 |
| SamYaple | or at least for ubuntu | 20:37 |
| SamYaple | they have thier stuff together for the other mirrors | 20:37 |
| sdake | the ptl changed for infra | 20:37 |
| sdake | so i need to have a f2f discussion about if they can get some mirrors ssetup for our external kolla repos | 20:38 |
| sdake | we have a 40 minute session on gating | 20:38 |
| sdake | at summit as well | 20:38 |
| SamYaple | if we can get docker mirrored thats all we really need | 20:38 |
| SamYaple | the delorean stuff is always going to be unstable | 20:38 |
| SamYaple | ill have to see alot of improvement to make that gating | 20:39 |
| SamYaple | voting* | 20:39 |
| *** bkosick has joined #kolla | 20:42 | |
| SamYaple | fyi sdake https://review.openstack.org/#/c/237745/ | 20:43 |
| sdake | samyaple good news, clarkb said infra would allow us to have mirrors on their gear | 20:47 |
| *** achanda has joined #kolla | 20:47 | |
| sdake | if we did the puppet work to set up the mirrors | 20:47 |
| SamYaple | cool | 20:48 |
| sdake | I dont know puppet but surely we can find someone to do that work | 20:48 |
| SamYaple | thatll fix the docker issue | 20:48 |
| SamYaple | we already use the infra pip wheels | 20:48 |
| SamYaple | i suppose cloud-archive could be mirrored | 20:48 |
| SamYaple | that would be it for ubuntu, docker and cloud-archive | 20:48 |
| *** achanda has quit IRC | 20:53 | |
| *** bkosick has quit IRC | 20:56 | |
| *** jasonsb has quit IRC | 21:00 | |
| *** achanda has joined #kolla | 21:07 | |
| pbourke_ | SamYaple: do you really need to add a new profile for the gate | 21:11 |
| pbourke_ | or could you use a combo of the existing ones | 21:12 |
| SamYaple | pbourke_: i could use a combo except i dont want an extra image we dont care about to break the gate | 21:12 |
| SamYaple | mongodb for example | 21:12 |
| SamYaple | the deploy gate will be the only voting gates | 21:13 |
| pbourke_ | what about 'default', could we just change that to match the gate? | 21:13 |
| bmace | all the code is up on oss.oracle.com now. for the cli that is here: https://oss.oracle.com/git/?p=openstack-kollacli.git;a=summary | 21:13 |
| SamYaple | pbourke_: i dont know why anyone cares about default to be honest, but it was explicitly asked to only have those that ansible deploys by default with no options | 21:14 |
| SamYaple | idk like to have a profile that is specific to the gate to avoid breakage | 21:14 |
| SamYaple | i dont want someone chaning default ad breaking the gate by accident | 21:14 |
| pbourke_ | fair enough | 21:15 |
| *** achanda has quit IRC | 21:15 | |
| pbourke_ | are the current failures related to the patch? if not I'll +2 | 21:17 |
| SamYaple | idk has the gate returned | 21:18 |
| sdake | yo merge my patches!! | 21:18 |
| pbourke_ | never! | 21:18 |
| sdake | https://review.openstack.org/237722 https://review.openstack.org/#/c/237768/ | 21:18 |
| SamYaple | pbourke_: wait for the gate | 21:18 |
| sdake | rhallisey pbourke ^^ | 21:18 |
| SamYaple | dont blind merge please. wait for the gate | 21:19 |
| sdake | oh the gate hasn't finished yet? | 21:19 |
| sdake | one of my patches wont build | 21:19 |
| rhallisey | yes looking good. I'll leave the +a until the gate | 21:19 |
| SamYaple | pbourke_: nah those failures are due to the ceph stuff | 21:19 |
| SamYaple | pbourke_: you can +2 the gate job if you want | 21:20 |
| SamYaple | pbourke_: https://review.openstack.org/#/c/237768 | 21:22 |
| SamYaple | can you remove your +a | 21:22 |
| SamYaple | need to wait on the gate | 21:22 |
| sdake | ya sam is righ twe should be waiting for the gates | 21:23 |
| sdake | especially right prior to release | 21:23 |
| sdake | i'll rebase the rdo patch after the ceph patch is backported | 21:24 |
| pbourke_ | done, but I thought if the gate passes it will merge, if not it wont? | 21:24 |
| sdake | pbourke our gates are non-voting | 21:24 |
| SamYaple | ^ | 21:24 |
| sdake | they need to be voting for that behavior to happen | 21:24 |
| sdake | they tell you something worked or didn't work | 21:24 |
| sdake | they dont actually gate :( | 21:24 |
| SamYaple | lol pbourke_ your +a not +2 | 21:24 |
| sdake | week before summit and infra is overloaded ;-) | 21:25 |
| sdake | on the plus side we think infra will now support the abiilty to mirror the stuff we need mirrored so we can have oting gates! | 21:28 |
| pbourke_ | such as? | 21:29 |
| SamYaple | all the repos | 21:29 |
| sdake | every repo we install in the base image | 21:30 |
| sdake | pypi is already mirrored | 21:30 |
| sdake | because then we will hae access and control over the mirrors | 21:31 |
| sdake | so if the gate fails we can fix the mirrors | 21:31 |
| sdake | that will permit us to make the gate voting | 21:31 |
| sdake | which will be a huge deal | 21:31 |
| pbourke_ | nice | 21:33 |
| SamYaple | speaking of gate voting, this needs to be approved and backported for that to happen https://review.openstack.org/#/c/233993/ | 21:33 |
| sdake | that wwill essentially allow us to get rid of build regressions | 21:33 |
| sdake | which is one big step | 21:33 |
| *** achanda has joined #kolla | 21:35 | |
| sdake | SamYaple https://review.openstack.org/#/c/233993/14/tests/test_build.py why remove the failure count? | 21:37 |
| SamYaple | sdake: I explained it | 21:38 |
| sdake | oh i missed that | 21:38 |
| sdake | let me see if i can find it in the patchset | 21:38 |
| SamYaple | https://review.openstack.org/#/c/233993/12/tests/test_build.py | 21:38 |
| SamYaple | i apparently explained it in a comment | 21:38 |
| sdake | got it | 21:39 |
| sdake | would you mind telling me what that unmanted image means? | 21:39 |
| sdake | unmatched | 21:39 |
| sdake | does it mean something built that shoudln't have? | 21:40 |
| SamYaple | i means the regex ignored the image | 21:40 |
| sdake | right ok your correct in your assessment then | 21:41 |
| sdake | is this ready for +2 then even with the gates all red? | 21:41 |
| sdake | otherwise patch looks good | 21:41 |
| sdake | just not sure why gates are bustola | 21:41 |
| SamYaple | sdake: yea they are read for your ceph patch reasons and designate failing the gate | 21:42 |
| SamYaple | red* | 21:42 |
| sdake | got it ok souns good then | 21:42 |
| sdake | https://review.openstack.org/#/c/233993/ needs another review folks | 21:43 |
| sdake | one day the gates will vote - soon can't happen fast enough :) | 21:43 |
| sdake | (one day soon that is:) | 21:43 |
| SamYaple | sdake: its all borked! | 21:46 |
| SamYaple | https://review.openstack.org/#/c/237722/ | 21:46 |
| sdake | ya | 21:47 |
| sdake | i just saw that | 21:47 |
| *** achanda has quit IRC | 21:47 | |
| sdake | that means openstack infra is busted | 21:48 |
| sdake | something with zuul i think | 21:48 |
| SamYaple | btw with our build gates ive noticed we see issue 3-4 days before they get talked about upstream | 21:48 |
| SamYaple | across all the projects | 21:49 |
| SamYaple | well see a nova issue days before it gets talked about | 21:49 |
| SamYaple | same with designate just now | 21:49 |
| sdake | ya we will become the front line for quality control in openstack over time if we make our gates reliable | 21:52 |
| SamYaple | it also means were likely to be fixing other projects perpetually | 21:54 |
| *** dwalsh has quit IRC | 21:56 | |
| *** achanda has joined #kolla | 21:57 | |
| *** docaedo has joined #kolla | 21:59 | |
| sdake | ok folks i dont know how to fix the gate - so we can either merge those two patches and backport them (and I've manually tested both of them) | 21:59 |
| sdake | or we can stall our release/build until we can find someone to unbreak the gate | 22:00 |
| SamYaple | considering they are building now we can wait | 22:00 |
| SamYaple | i kicked of a recheck about 10 minutes ago | 22:00 |
| sdake | oh they are registered now? | 22:01 |
| sdake | maybe nodeepool went out toe lunch | 22:01 |
| SamYaple | the jobs are running according to zuul | 22:01 |
| *** achanda has quit IRC | 22:02 | |
| asalkeld_ | o/ | 22:08 |
| SamYaple | \o | 22:09 |
| sdake | yo asalkeld | 22:09 |
| sdake | did that schedule help? | 22:09 |
| sdake | asalkeld i added a session just for you and tripleo - third prty integration - hope you can make that one | 22:10 |
| asalkeld_ | sdake: thanks, yip | 22:10 |
| *** achanda has joined #kolla | 22:12 | |
| *** pbourke_ has quit IRC | 22:17 | |
| *** sdake has quit IRC | 22:18 | |
| *** sdake has joined #kolla | 22:18 | |
| SamYaple | great | 22:23 |
| SamYaple | a nova bug introduced in rc3 :/ | 22:23 |
| sdake | as in bust? | 22:25 |
| sdake | i tested openstack friday night working great then :) | 22:26 |
| SamYaple | no | 22:26 |
| SamYaple | regression in ceph cloning of raw images | 22:26 |
| sdake | does that affect us or our demo? | 22:26 |
| SamYaple | we can work around it | 22:27 |
| SamYaple | im getting a patch to nova now | 22:27 |
| sdake | gate failed | 22:30 |
| sdake | ubuntu gate | 22:30 |
| sdake | and centos source | 22:30 |
| sdake | nova compute | 22:31 |
| * sdake groans | 22:31 | |
| SamYaple | sdake: that means ceph isnt working | 22:31 |
| sdake | lookign at the logs the ceph key was imported | 22:33 |
| SamYaple | based on whats failing the only thing they have in common is ceph packages | 22:34 |
| kfox1111 | so, which version is the kollaglue containers going to use? rdo snapshot of liberty? | 22:34 |
| sdake | 2015-10-20 22:08:50.420 | INFO:kolla.cmd.build:nova-compute:[91mError unpacking rpm package python-requests-2.7.0-1.el7.noarch | 22:35 |
| sdake | 2015-10-20 22:08:50.421 | INFO:kolla.cmd.build:nova-compute:[0m | 22:35 |
| sdake | 2015-10-20 22:08:50.421 | INFO:kolla.cmd.build:nova-compute: | 22:35 |
| sdake | 2015-10-20 22:08:50.421 | INFO:kolla.cmd.build:nova-compute:error: unpacking of archive failed on file /usr/lib/python2.7/site-packages/requests/packages/chardet: cpio: rename | 22:35 |
| kfox1111 | oh. nm. looks like rdo does have a final release. | 22:35 |
| sdake | kfox1111 liberty branch has a patch up to stablize on a specific version | 22:36 |
| sdake | samyaple it isn't ceph | 22:36 |
| *** jongchoi has joined #kolla | 22:36 | |
| *** jongchoi has quit IRC | 22:36 | |
| sdake | kfox1111 but its not merged yet because of gating issues | 22:37 |
| kfox1111 | k. | 22:37 |
| kfox1111 | thx. | 22:37 |
| sdake | samyaple my patch fixes the issue | 22:37 |
| sdake | there is another ceph issue | 22:37 |
| sdake | SamYaple what is wrong here: http://logs.openstack.org/68/237768/1/check/gate-kolla-build-images-ubuntu-source/fa02a52/console.html#_2015-10-20_22_13_48_940 | 22:38 |
| sdake | is it designate? | 22:38 |
| sdake | why change the api in a stable branch a few days after the releaes | 22:39 |
| * sdake growls | 22:39 | |
| sdake | samyaple i'm good with a merge on the ceph timer patch | 22:39 |
| sdake | there is some other issue that needs addressing | 22:40 |
| sdake | something to do with installing python-requests.2.7.0 | 22:40 |
| SamYaple | oh yea python-requests gets pulled in by ceph | 22:40 |
| sdake | pip uninstall python-requests if from source then? | 22:41 |
| sdake | in nova-compute? | 22:41 |
| sdake | let me file a bug | 22:41 |
| * sdake UGGGGHS | 22:41 | |
| sdake | samyaple here is the link http://logs.openstack.org/68/237768/1/check/gate-kolla-build-images-centos-source/5d0286b/console.html#_2015-10-20_22_07_54_082 | 22:43 |
| openstackgerrit | Merged openstack/kolla: Increase Ceph RPM gpg key timeout to 90 seconds https://review.openstack.org/237768 | 22:43 |
| sdake | https://bugs.launchpad.net/kolla/+bug/1508236 | 22:44 |
| openstack | Launchpad bug 1508236 in kolla "CentOS from source on Ceph fails to build nova-compute" [Critical,Confirmed] - Assigned to Steven Dake (sdake) | 22:44 |
| openstackgerrit | Merged openstack/kolla: Use the liberty release RDO RPMS https://review.openstack.org/237722 | 22:44 |
| sdake | brb | 22:44 |
| openstackgerrit | Steven Dake proposed openstack/kolla: Increase Ceph RPM gpg key timeout to 90 seconds https://review.openstack.org/237800 | 22:47 |
| *** sdake_ has joined #kolla | 22:48 | |
| *** sdake has quit IRC | 22:49 | |
| SamYaple | sdake_: https://review.openstack.org/#/c/237799/ | 22:51 |
| SamYaple | that is the ceph issue | 22:51 |
| sdake_ | looking | 22:51 |
| SamYaple | that code only gets excersided in a few select cases | 22:52 |
| SamYaple | basically only when the image type is raw | 22:52 |
| *** tummy has quit IRC | 22:52 | |
| SamYaple | so its unsuprising it wasnt caught to be honest :/ | 22:52 |
| sdake_ | doesn't qcow2 convert to raw? | 22:53 |
| SamYaple | yea but its not a CoW clone | 22:53 |
| SamYaple | if the image is qcow2 it actually downloads to the compute then reuploads to ceph | 22:53 |
| SamYaple | if its raw its a CoW clone, no additional sapce | 22:54 |
| SamYaple | oh you mean on upload? | 22:56 |
| kfox1111 | I thought they had a patch in that would convert qcow to raw on ingest into glance? | 22:56 |
| SamYaple | no | 22:56 |
| SamYaple | that would have to be a glance patch | 22:56 |
| SamYaple | but either way thats unrelated to this bug | 22:57 |
| kfox1111 | yeah. I thought kilo or liberty had that. | 22:57 |
| kfox1111 | k. | 22:57 |
| SamYaple | if its raw in ceph, you cannot boot an instance | 22:57 |
| SamYaple | well if its raw and you have CoW cloning enabled then you cannot boot an instance | 22:57 |
| SamYaple | thats the default soooo | 22:57 |
| SamYaple | do we know anyone on the nova team? | 22:58 |
| SamYaple | alright im going away for a bit | 22:59 |
| sdake_ | samyaple what is this ERROR:__main__:openstack-base:openstack-base-archive: no such file or directory | 23:03 |
| sdake_ | there is no openstack-base-archive file in the repo | 23:05 |
| sdake_ | a patch that merged today | 23:06 |
| sdake_ | or yesterday tahat is | 23:06 |
| * sdake_ groans | 23:06 | |
| sdake_ | pbourke around??!! | 23:06 |
| *** dolpher has joined #kolla | 23:09 | |
| SamYaple | sdake_: what are you talkign about | 23:10 |
| sdake_ | master is broken | 23:10 |
| sdake_ | ERROR:__main__:openstack-base:Error'd with the following message | 23:10 |
| sdake_ | ERROR:__main__:openstack-base:openstack-base-archive: no such file or directory | 23:10 |
| sdake_ | [sdake@bigiron kolla]$ tools/build.py -t source nova-compute | 23:11 |
| sdake_ | oh wait | 23:11 |
| sdake_ | kolla-build.conf | 23:11 |
| SamYaple | yup | 23:11 |
| sdake_ | in etc now | 23:11 |
| sdake_ | that is probaby it | 23:11 |
| SamYaple | you have an outdated one | 23:11 |
| sdake_ | yup fixed thanks :) | 23:12 |
| sdake_ | trying to fi this ceph from source problem | 23:12 |
| sdake_ | so atleast our shit builds :( | 23:12 |
| *** daneyon has joined #kolla | 23:12 | |
| sdake_ | daneyon what day do you depart for japan? | 23:13 |
| daneyon | Sun evening... arrive Tues morning. | 23:13 |
| sdake_ | I htink I like friday | 23:14 |
| sdake_ | but not sure | 23:14 |
| sdake_ | leave friday that is | 23:14 |
| *** daneyon_ has joined #kolla | 23:34 | |
| *** daneyon has quit IRC | 23:37 | |
| sdake_ | bmace ping | 23:38 |
| *** blahRus has quit IRC | 23:39 | |
| * asalkeld_ leaving tomorrow this time | 23:41 | |
| *** dims_ has quit IRC | 23:45 | |
| *** dims has joined #kolla | 23:46 | |
| *** cemason has quit IRC | 23:47 | |
| openstackgerrit | Steven Dake proposed openstack/kolla: Make Ceph install from Source on RPM https://review.openstack.org/237816 | 23:50 |
| sdake_ | jpeeler rhallisey samyaple need an ack on that one ^ | 23:50 |
| sdake_ | https://review.openstack.org/#/c/237800/ needs an ack | 23:54 |
| sdake_ | then i think we are good for release | 23:54 |
| sdake_ | that last change also needs to be backported | 23:55 |
| openstackgerrit | Steven Dake proposed openstack/kolla: Make Ceph install from Source on RPM https://review.openstack.org/237816 | 23:55 |
| docaedo | Wondering if I'm doing something wrong - I've tried building the images locally a few times, and always seem to hang here: | 23:56 |
| docaedo | INFO:__main__:gnocchi-api:Removing intermediate container beb0f8a42ed8 | 23:56 |
| docaedo | INFO:__main__:gnocchi-api:Successfully built 80acc3067dfd | 23:56 |
| docaedo | INFO:__main__:gnocchi-api:Built | 23:56 |
| docaedo | Did this with kolla from trunk as well as stable/liberty | 23:56 |
| SamYaple | docaedo: its not hanging. its still building/pushing | 23:57 |
| SamYaple | it will crash if something breaks i assure you | 23:57 |
| SamYaple | i highly doubt anything is just hanging | 23:58 |
| docaedo | SamYaple: excellent :) I will be patient then and just let it be | 23:58 |
| docaedo | SamYaple: because you know, I thought everything with containers was just INSTANT! | 23:58 |
| docaedo | SamYaple: :) | 23:58 |
| SamYaple | that can be improved in the scrpipt for more feedback if youre interested :) | 23:58 |
| docaedo | some indication that it's not stalled would not be bad, but I'm not going to nitpick, I'm really excited about the progress you've made | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!