| *** zhiwei has joined #kolla | 01:37 | |
| *** erkules_ has joined #kolla | 01:54 | |
| *** erkules has quit IRC | 01:56 | |
| *** apmelton has quit IRC | 02:32 | |
| *** tobe has joined #kolla | 03:12 | |
| *** sdake has joined #kolla | 03:32 | |
| *** sdake has quit IRC | 03:36 | |
| *** sdake has joined #kolla | 03:39 | |
| openstackgerrit | Sam Yaple proposed stackforge/kolla: Add initial config function and keystone support https://review.openstack.org/192544 | 03:40 |
|---|---|---|
| sdake | samyaple can you ack that aboe bug | 03:42 |
| openstackgerrit | Sam Yaple proposed stackforge/kolla: Add initial config function and keystone support https://review.openstack.org/192544 | 03:43 |
| sdake | the one harm fixed | 03:43 |
| sdake | 196471 | 03:44 |
| openstackgerrit | Sam Yaple proposed stackforge/kolla: Ansible Mariadb w/ Galera support https://review.openstack.org/196428 | 03:45 |
| SamYaple | sdake: We need to merge in the keystone ansible stuff, even though we dont have the modules yet | 03:47 |
| sdake | for liberty 1? | 03:47 |
| SamYaple | its getting very hard to keep rebasing to continue the work | 03:47 |
| openstackgerrit | Merged stackforge/kolla: Properly set MAGNUM_API_SERVICE_PORT https://review.openstack.org/196471 | 03:47 |
| SamYaple | no sdake, it doesnt need to be in l1 | 03:47 |
| sdake | ok well lets get l1 sorted out | 03:47 |
| sdake | re merging - you mean just merge sans the modules? | 03:48 |
| SamYaple | yea | 03:48 |
| sdake | are the moduels not asl licensed yet? | 03:48 |
| SamYaple | its got changes that the other patches depend on (for the config-interal/config-exteral support0 | 03:48 |
| SamYaple | nope | 03:48 |
| sdake | any problem there or -EWEEKEND | 03:49 |
| SamYaple | probably just weekend, but i was hoping for them on thursday | 03:49 |
| SamYaple | i dont think there is a holdup | 03:49 |
| SamYaple | ive gotten IRC permssiosn from all the authors on the relicense, just need a commit to point to | 03:50 |
| SamYaple | either way, that keystone patch is useless without the DB, but i need pieces of that patch to merge the db | 03:50 |
| SamYaple | i only started with keystone to allow otehrs to do openstack ones immediatelly after merge | 03:51 |
| *** tobe has quit IRC | 03:55 | |
| *** tobe has joined #kolla | 03:56 | |
| *** tobe has quit IRC | 03:56 | |
| *** sdake_ has joined #kolla | 03:58 | |
| *** tobe has joined #kolla | 03:59 | |
| *** sdake has quit IRC | 04:02 | |
| *** tobe has quit IRC | 04:13 | |
| sdake_ | rebuidling all imges now | 04:30 |
| *** nihilifer has joined #kolla | 04:45 | |
| nihilifer | good morning | 04:45 |
| *** alisonh has quit IRC | 04:46 | |
| *** alisonh has joined #kolla | 05:13 | |
| SamYaple | morning nihilifer | 05:18 |
| *** zhiwei has quit IRC | 05:31 | |
| sdake_ | hey nih* | 05:54 |
| sdake_ | samyaple I just finished building master | 05:54 |
| sdake_ | getting ready to fire it up | 05:54 |
| sdake_ | boy we got alot of containers | 05:55 |
| sdake_ | amazing amount of work we hve done in only one cycle | 05:55 |
| sdake_ | we did more work this one cycle then almost all of kilo ;) | 05:55 |
| *** tobe has joined #kolla | 05:56 | |
| *** tobe has quit IRC | 05:58 | |
| *** tobe_ has joined #kolla | 05:58 | |
| SamYaple | sdake_: yup. lots of images | 06:01 |
| SamYaple | I am building an alternate build-all script in python | 06:02 |
| SamYaple | we can add it in without breaking anyones flow | 06:02 |
| SamYaple | it does parallel threading on the builds | 06:02 |
| sdake_ | those tools should probably be in python anyway | 06:03 |
| sdake_ | but they arein shell because that is how they started at ~10-15 lines | 06:03 |
| sdake_ | now they are several hundred | 06:03 |
| SamYaple | yup | 06:03 |
| SamYaple | i had a working parallel in shell that was pretty slick | 06:03 |
| SamYaple | but overall it was just to ugly to implement | 06:04 |
| SamYaple | so i abandonded | 06:04 |
| *** dasm|afk is now known as dasm | 06:23 | |
| SamYaple | 67 centos containers | 06:50 |
| *** athomas has joined #kolla | 06:53 | |
| *** akwasnie has joined #kolla | 07:11 | |
| *** alisonh has quit IRC | 07:45 | |
| *** mstachow has joined #kolla | 08:13 | |
| mstachow | o/ all | 08:13 |
| SamYaple | \o | 08:14 |
| *** erkules_ is now known as erkules | 08:21 | |
| *** erkules has joined #kolla | 08:21 | |
| pbourke | do you guys generally work all weekend then? | 08:30 |
| harmw | \o/ | 08:31 |
| harmw | depends pbourke :) | 08:32 |
| harmw | SamYaple: did you create an etherpad with summit topics? | 08:38 |
| harmw | *proposals | 08:38 |
| harmw | https://etherpad.openstack.org/p/kolla-summit-talk-proposals | 08:43 |
| harmw | there you go | 08:43 |
| harmw | I couldn't find the other topics in my scrollback so as always, feel free to add :) | 08:44 |
| pbourke | Im starting to look at Swift with jmccarthy so have added a tentative talk that could come out of that work | 08:56 |
| *** gfidente has joined #kolla | 08:59 | |
| *** gfidente has joined #kolla | 08:59 | |
| SamYaple | pbourke: i actually work weekends, so yea | 09:04 |
| pbourke | SamYaple: yeah I think I remember you saying you have slightly different hours | 09:13 |
| pbourke | just always surprised at the amount of logs to read on a monday morning :) | 09:13 |
| SamYaple | ITS ALL BROKE AAAAHHH | 09:13 |
| *** alisonh has joined #kolla | 09:18 | |
| *** jmccarthy has joined #kolla | 09:26 | |
| harmw | no SamYaple that was 2 days ago :p | 09:34 |
| harmw | SamYaple: did you read my comment on your thin containers revert? | 09:36 |
| SamYaple | havent looked | 09:37 |
| SamYaple | its merged comments mean nothing | 09:37 |
| SamYaple | sucker | 09:37 |
| harmw | ehe | 09:37 |
| SamYaple | looking now | 09:37 |
| SamYaple | ah yea. i mean i could, but i dont like the idea of broken duplicate containers in the code base | 09:38 |
| SamYaple | ill be following the progress on this and we wont use a feature until it hits stable next time | 09:38 |
| harmw | agree, but what if we add a experimental/ corner | 09:38 |
| harmw | that way it's not just on your shoulders anymore, anyone can test | 09:39 |
| SamYaple | i really dont like the idea of duplicate containers | 09:40 |
| harmw | me neither, but things in /experimental don't need to be cut down by such rules imho | 09:41 |
| SamYaple | we dont have a branch, and neither does anyone else in openstack | 09:41 |
| harmw | but anyway, I'm not pushing, merely putting up for discussion because of said benefit :) | 09:41 |
| SamYaple | it would be perpetually out of date | 09:41 |
| harmw | I don't mean branching, just a new folder somewhere in the tree | 09:42 |
| SamYaple | oh. then thats even worse | 09:42 |
| SamYaple | then "experinmental" would exist in the stable code base | 09:42 |
| harmw | yep, but without any script or process touching on it | 09:43 |
| SamYaple | pbourke: hows your python? | 09:43 |
| *** athomas has quit IRC | 09:46 | |
| pbourke | SamYaple: decent | 09:51 |
| pbourke | SamYaple: or maybe a better answer is, that depends ;) | 09:52 |
| *** athomas has joined #kolla | 09:52 | |
| *** openstackgerrit has quit IRC | 09:53 | |
| *** openstackgerrit has joined #kolla | 09:53 | |
| SamYaple | ive got a new build script all in python with threads for parallel about to send for review | 09:54 |
| pbourke | go for it | 09:55 |
| harmw | he SamYaple RAX want you to work semi-fulltime on kolla? | 09:55 |
| SamYaple | im not. I only do Kolla on my downtime | 09:57 |
| SamYaple | i would love to work on it [semi]fulltime | 09:57 |
| SamYaple | that would be awesome | 09:57 |
| SamYaple | imagine what I would do wit hthat | 09:58 |
| SamYaple | mmmmmmm | 09:58 |
| harmw | you have quite some downtime on your hands already though :) | 09:58 |
| SamYaple | says you. just because im active here doesnt mean i have alot of downtime | 09:58 |
| sdake_ | red hat has two container developer + openstack job reqs open :) | 09:58 |
| SamYaple | im on the phones and i have a stupid ticket queue | 09:58 |
| harmw | haha | 09:59 |
| SamYaple | yea but... redhat | 09:59 |
| SamYaple | :P | 09:59 |
| sdake_ | just pointing it out for everyones benefit :) | 09:59 |
| harmw | sdake_: ready to stab again I see | 09:59 |
| sdake_ | my computer just beeped | 10:00 |
| sdake_ | and woe me up | 10:00 |
| sdake_ | need t o just shut it off at night | 10:00 |
| SamYaple | but no one said your name | 10:00 |
| harmw | luckily mine doesn't beep on underscores:> | 10:00 |
| harmw | and yes, that to | 10:00 |
| SamYaple | harmw: sounds like a bad regex to me | 10:00 |
| harmw | lol | 10:00 |
| *** sdake_ is now known as sdake | 10:01 | |
| harmw | sdake: got a link to those openings? | 10:01 |
| harmw | just curious | 10:02 |
| sdake | harmw let me see if i canfnd | 10:02 |
| SamYaple | what do they pay, like 200k? | 10:03 |
| sdake | http://jobs.redhat.com/jobs/descriptions/senior-software-engineer-openstack-containers-job-1-5289921 | 10:03 |
| sdake | samyaple highly unlikely ;) | 10:03 |
| harmw | senior, puh | 10:03 |
| sdake | ya thats like 8-12 years of expeience typically | 10:04 |
| SamYaple | well when the price gets up there let me know | 10:04 |
| sdake | harmw other req is http://jobs.redhat.com/jobs/descriptions/senior-software-engineer-openstack-containers-job-2-5289921 | 10:05 |
| harmw | online since april, damn | 10:06 |
| sdake | it may be onloy 1 req | 10:07 |
| sdake | since changing from 1 to 2 to 3 sows same info | 10:07 |
| harmw | both have the same content :) but anyway, how much do they typically pay for such a job? | 10:08 |
| sdake | how would I know? | 10:08 |
| harmw | oh perhaps you knew | 10:08 |
| sdake | I fid a job I like, pay is secondary | 10:08 |
| harmw | I'm just curious, thats all :) | 10:09 |
| sdake | I do but I am probably bound by some conficdnialy agreement or omething | 10:09 |
| harmw | yea well, even then I can't blindly compare figures like 200k with what I'd be getting for it here I guess | 10:10 |
| sdake | if folks would like an intro to the hiring manager, ping m oflin | 10:10 |
| sdake | another perwon o hit up would be jpeeler/rhqllisey | 10:11 |
| sdake | what I want is sam's 15% pay bonus for working 3rd shift | 10:12 |
| sdake | since that is what I/ MDOING atm :) | 10:13 |
| sdake | wow keyboard typing whie seeop bad | 10:13 |
| sdake | bought a power purifier and two cables tody | 10:14 |
| sdake | painful | 10:14 |
| sdake | well atleast hifi hobby is less distructiful then C&H | 10:15 |
| SamYaple | cyanide and happiness? | 10:15 |
| sdake | yup! | 10:16 |
| SamYaple | good stuff | 10:16 |
| harmw | cookies & honey? | 10:16 |
| SamYaple | thought maybe you meant calvin and hobbes | 10:16 |
| harmw | whats holding up our L1 release sdake ? | 10:16 |
| SamYaple | sdake | 10:17 |
| harmw | aw | 10:17 |
| sdake | well I want to test that the neutron agents work | 10:17 |
| sdake | and I am not superman | 10:17 |
| harmw | my instance booted, did dhcp and did metadata the other night | 10:17 |
| harmw | so thats working | 10:17 |
| harmw | (didn't test float though) | 10:17 |
| SamYaple | sdake: lies! | 10:18 |
| sdake | samyaple your rihgt, I amsuerman ;) | 10:18 |
| harmw | amuserman! | 10:19 |
| harmw | always fun when he's around | 10:19 |
| SamYaple | amuse me amuserman! | 10:19 |
| sdake | that wass a typo | 10:19 |
| SamYaple | wass it now? | 10:19 |
| sdake | but worked out ok ;) | 10:19 |
| SamYaple | RAWR | 10:20 |
| *** shardy_ has joined #kolla | 10:25 | |
| *** shardy has quit IRC | 10:27 | |
| sdake | I bought one of these http://www.nordost.com/norse2/frey2/frey2-tonearm.php | 10:27 |
| openstackgerrit | Sam Yaple proposed stackforge/kolla: Add new build script https://review.openstack.org/196617 | 10:29 |
| SamYaple | pbourke: ^ | 10:29 |
| *** shardy_ has quit IRC | 10:31 | |
| *** shardy has joined #kolla | 10:31 | |
| *** vbel has joined #kolla | 10:37 | |
| *** dims has joined #kolla | 10:37 | |
| pbourke | SamYaple: i'll try have a few comments by the afternoon | 10:37 |
| sdake | i like in strack they still use the nato coeorcommunicating spelling | 10:38 |
| sdake | str tre | 10:38 |
| SamYaple | dude i honestly have no idea what you are saying sdake | 10:39 |
| nihilifer | SamYaple: I left some comments now ;) | 10:39 |
| SamYaple | no comments just merge! | 10:40 |
| nihilifer | :D | 10:40 |
| sdake | star trek uses nato code | 10:40 |
| sdake | "Alpha brav whisky tankgo foxtrot etc" | 10:41 |
| SamYaple | echo | 10:41 |
| SamYaple | not etc | 10:41 |
| SamYaple | n00b | 10:41 |
| sdake | hah | 10:41 |
| sdake | alpha foxtrot uniform | 10:42 |
| SamYaple | nihilifer: well foxtrot uniform you too | 10:42 |
| SamYaple | :D | 10:42 |
| pbourke | nihilifer: aw your comments match almost all my drafts ;) | 10:47 |
| nihilifer | lol | 10:48 |
| SamYaple | anyone have any boilerplate code for a pretty CLI gui? | 10:48 |
| SamYaple | mainly to wrap the status since we have several tiers and many threads | 10:48 |
| pbourke | SamYaple: personally Im not a fan of those fancy type concurrent clis as they're hard to wrap in scripts | 11:01 |
| SamYaple | pbourke: yea well 45 minutes to gate isnt acceptable | 11:02 |
| SamYaple | and since this _is_ the script wrapping things... | 11:02 |
| pbourke | SamYaple: no I mean just for results output | 11:02 |
| pbourke | SamYaple: I am a fan of the new script | 11:02 |
| SamYaple | oh i havent handled that yet | 11:02 |
| pbourke | SamYaple | anyone have any boilerplate code for a pretty CLI gui? | 11:02 |
| pbourke | I thought that's what you were asking for there | 11:03 |
| SamYaple | im going to be pushing the stdout and stderr to indvidual logs | 11:03 |
| pbourke | kk that sounds fine | 11:03 |
| SamYaple | pbourke: yea im bouncing around. i really just want a pretty status screen whith a list of images that turn green/red/yellow depending on if the pass/failed/currently building | 11:04 |
| *** shardy_ has joined #kolla | 11:18 | |
| *** shardy has quit IRC | 11:19 | |
| *** shardy_ has quit IRC | 11:24 | |
| *** shardy has joined #kolla | 11:25 | |
| *** rhallisey has joined #kolla | 11:34 | |
| *** rhallisey has quit IRC | 11:35 | |
| *** rhallisey has joined #kolla | 11:36 | |
| *** akscram has joined #kolla | 11:41 | |
| *** sdake has quit IRC | 11:43 | |
| *** rhallisey has quit IRC | 11:45 | |
| *** tobe_ has quit IRC | 11:49 | |
| *** rhallisey has joined #kolla | 11:51 | |
| SamYaple | real 11m8.713s | 11:56 |
| SamYaple | user 0m1.348s | 11:56 |
| SamYaple | sys 0m1.072s | 11:56 |
| SamYaple | thats down from 60 minutes on my machine for the single threaded | 11:56 |
| SamYaple | the gate should easily be able to gate in 10 minutes now | 11:56 |
| *** nihilifer has quit IRC | 12:43 | |
| *** shardy has quit IRC | 12:46 | |
| *** shardy has joined #kolla | 12:46 | |
| *** jruano has joined #kolla | 13:12 | |
| *** jruano has quit IRC | 13:12 | |
| *** absubram has quit IRC | 13:36 | |
| *** absubram has joined #kolla | 13:36 | |
| *** prad has joined #kolla | 13:38 | |
| *** prad has quit IRC | 13:43 | |
| *** prad has joined #kolla | 13:55 | |
| *** absubram has quit IRC | 14:11 | |
| *** absubram has joined #kolla | 14:15 | |
| *** absubram has quit IRC | 14:15 | |
| *** dims has quit IRC | 14:21 | |
| *** dims has joined #kolla | 14:24 | |
| *** sdake has joined #kolla | 14:37 | |
| sdake | morning | 14:38 |
| rhallisey | morning | 14:39 |
| *** sdake_ has joined #kolla | 14:40 | |
| *** nihilifer has joined #kolla | 14:42 | |
| *** sdake has quit IRC | 14:43 | |
| *** shardy_ has joined #kolla | 14:51 | |
| *** shardy has quit IRC | 14:51 | |
| SamYaple | sdake_: https://review.openstack.org/#/c/196617/ | 14:56 |
| SamYaple | builds in 10 minutes on my box | 14:56 |
| SamYaple | sdake_ rhallisey jpeeler harmw I need to get the Ansible keystone stuff in since the other playbooks depend on some of these changes. We are still waiting on modules upstream to be relicensed, so this can't be 100% completed at this time | 14:58 |
| SamYaple | https://review.openstack.org/#/c/192544/ | 14:58 |
| SamYaple | please review this and vote on this to merge in specifically on the docker/* components for those not planing to use ansible | 14:58 |
| *** kevsi has quit IRC | 14:59 | |
| *** shardy_ has quit IRC | 15:00 | |
| *** shardy has joined #kolla | 15:04 | |
| *** absubram has joined #kolla | 15:04 | |
| *** vinkman has joined #kolla | 15:05 | |
| *** jruano has joined #kolla | 15:06 | |
| *** kragniz has joined #kolla | 15:07 | |
| *** jasonsb has quit IRC | 15:09 | |
| *** mattt has left #kolla | 15:10 | |
| *** pbourke has quit IRC | 15:40 | |
| *** pbourke has joined #kolla | 15:40 | |
| *** kevsi has joined #kolla | 15:56 | |
| *** jasonsb has joined #kolla | 16:25 | |
| *** rbergeron has quit IRC | 16:41 | |
| *** athomas has quit IRC | 16:46 | |
| pbourke | SamYaple: I will review this first thing in the morning | 16:51 |
| pbourke | overall it looks good but would like to try it out | 16:52 |
| SamYaple | pbourke: you wont be able to, thats the problem | 16:53 |
| SamYaple | it is however holding up the database and rabbit playbooks | 16:54 |
| pbourke | ok well I'll try my best to spot any potential problems. dont hold it back for me obviously, just saying I will give it one last check if it helps | 16:54 |
| *** jasonsb has quit IRC | 16:59 | |
| *** rbergeron has joined #kolla | 17:10 | |
| *** mfalatic has joined #kolla | 17:10 | |
| *** sdake has joined #kolla | 17:16 | |
| *** sdake_ has quit IRC | 17:20 | |
| *** jasonsb has joined #kolla | 17:24 | |
| *** dontalton has joined #kolla | 17:32 | |
| harmw | sdake: where is L1? | 17:35 |
| sdake | trying to test it | 17:35 |
| sdake | rhallisey is cinder just broke becuse of upstream packaging? | 17:36 |
| rhallisey | I think so | 17:36 |
| rhallisey | it stopped working about ~week ago | 17:36 |
| rhallisey | a bunch of stuff just changed at once | 17:37 |
| sdake | ok well I'm going to pull the trigger on the release assuming networking wors properly | 17:37 |
| sdake | and we can address teh cinder containers asap ;) | 17:37 |
| rhallisey | okie dokie | 17:37 |
| rhallisey | well they work with kilo packages :P. This new issue is puzzling | 17:38 |
| sdake | i get waiting for Keystone @ http://192.168.1.100:35357/v2.0 | 17:43 |
| sdake | in a loop in the eystone container | 17:43 |
| *** mstachow has quit IRC | 17:44 | |
| sdake | Setting up Keystone DB | 17:46 |
| sdake | File "/usr/lib/python2.7/site-packages/keystone/server/eventlet.py", line 23, in <module> | 17:46 |
| sdake | Configuring keystone.conf | 17:46 |
| sdake | Setting up PKI | 17:46 |
| sdake | from oslo_service import service | 17:46 |
| sdake | Initializing Keystone DB | 17:46 |
| sdake | Starting Keystone | 17:46 |
| sdake | ImportError: No module named oslo_service | 17:46 |
| sdake | Verifying Keystone is running | 17:46 |
| * sdake swears loudly | 17:46 | |
| rhallisey | sdake, that happens in glance-api and nova-api occasionally | 17:49 |
| sdake | well i am fixing globally | 17:50 |
| rhallisey | it's the wait_for <service> function | 17:50 |
| sdake | https://bugs.launchpad.net/kolla/+bug/1469821 | 17:50 |
| openstack | Launchpad bug 1469821 in kolla "keystone requires oslo_service" [Critical,Confirmed] - Assigned to Steven Dake (sdake) | 17:50 |
| sdake | i am pretty sure its because the dependency is missing! | 17:50 |
| rhallisey | that one is | 17:51 |
| rhallisey | I'm referring to a slightly different bug | 17:51 |
| rhallisey | fail_unless_os_service_running keystone | 17:52 |
| rhallisey | sdake, ^ that function will need to be changed in the future | 17:52 |
| openstackgerrit | Michal Rostecki proposed stackforge/kolla: Add base Nova image for Ubuntu https://review.openstack.org/196760 | 17:52 |
| sdake | yes we know what is wrong with it i dont understand why someone doesn't just fix it | 18:12 |
| sdake | harmw rebuilding master atm - taes 40 minutes | 18:12 |
| SamYaple | sdake: the alternate build script i have in review does it in 10 minutes (down from 60) on my compute | 18:21 |
| openstackgerrit | Steven Dake proposed stackforge/kolla: Install oslo.service via pip https://review.openstack.org/196777 | 18:21 |
| sdake | samyaple I hear what your syaing but that build script isn't in the repo - iwant to test what is in repo | 18:22 |
| sdake | dev stack from beginning to end | 18:22 |
| sdake | rather dev quickstart | 18:22 |
| sdake | could use some reviews on the above patch plz | 18:23 |
| sdake | its a releease blocker | 18:23 |
| *** gfidente has quit IRC | 18:23 | |
| SamYaple | sdake: i agree. im saying for future, once that alternate build script gets merged we can test for a bit and switch the gate to it | 18:24 |
| SamYaple | 10 minute gate times would be nice | 18:24 |
| sdake | wfm | 18:25 |
| sdake | why i sit faster | 18:25 |
| SamYaple | Also have a section for the templating support when we get around to that | 18:25 |
| SamYaple | I think I can have it done by L2 based on how easy it was in my test case | 18:25 |
| SamYaple | Faster builds, no symlink speghetti | 18:25 |
| SamYaple | w00t | 18:25 |
| sdake | or we could stay focused on ansible ;) | 18:26 |
| sdake | and do that in l3 | 18:26 |
| SamYaple | im blocked on ansible! but yea im sure ill have time | 18:26 |
| SamYaple | i can blow through the ansible stuff | 18:26 |
| sdake | ya | 18:26 |
| SamYaple | and help others to understand it | 18:27 |
| sdake | the entire repository is bloced atm | 18:27 |
| sdake | until we get a release out that works | 18:27 |
| SamYaple | sdake wanna unblock me? https://review.openstack.org/#/c/192544/ | 18:27 |
| sdake | nah lets not change anything except for critical bug fixes | 18:27 |
| sdake | and TRY HARD to get a release out today | 18:27 |
| *** openstackgerrit has quit IRC | 18:30 | |
| *** openstackgerrit has joined #kolla | 18:30 | |
| sdake | man dealing with delorean is a huge pita | 18:31 |
| sdake | install from source would be far superior as the dependencies would work | 18:31 |
| *** jasonsb has quit IRC | 18:33 | |
| SamYaple | sdake: indeed I have been saying just that | 18:33 |
| SamYaple | its much better all around | 18:33 |
| sdake | pacages still add value - not dropping those | 18:34 |
| sdake | but they make building liberty a huge pita | 18:34 |
| SamYaple | im not asking you to drop anything, you can handle the packages just as soon as i get my source installs :D | 18:34 |
| sdake | some of our mq/ha/db components will be difficult to build from src | 18:35 |
| SamYaple | no those wont be from source | 18:35 |
| SamYaple | only the openstack stuff | 18:35 |
| SamYaple | I would like an all from source container, but thats another time | 18:36 |
| SamYaple | jpeeler: if you check the neutron-openvswitch-agent youll notice it is missing %%KOLLA_TAG%% | 18:42 |
| SamYaple | this script cataches that (though the error message needs to be.... changed) where the bash one does not | 18:42 |
| jpeeler | yeah i didn't look too hard | 18:42 |
| SamYaple | i need to add a patch for neutron-openvswitch-agent | 18:42 |
| sdake | samyaple can you get a bug in for that problem | 18:43 |
| sdake | anda patch plz | 18:43 |
| SamYaple | but if you want to test it, just add the :%%KOLLA_TAG%% to the FROM line in the openvswitch agent | 18:43 |
| jpeeler | glad it uncovered a real problem though | 18:43 |
| jpeeler | right | 18:43 |
| SamYaple | sdake: I can, but i need to sleep, it doesnt block builds or need to exist for L1 | 18:43 |
| SamYaple | I will file a bug and do it this evening sdake | 18:43 |
| * jpeeler can do it | 18:43 | |
| sdake | jpeeler thanks :) | 18:44 |
| SamYaple | or jpeeler can do it | 18:44 |
| SamYaple | :[] | 18:44 |
| *** nihilifer has quit IRC | 18:46 | |
| openstackgerrit | Jeff Peeler proposed stackforge/kolla: Correct FROM line in openvswitch-agent https://review.openstack.org/196797 | 18:53 |
| jpeeler | seriously what's up with launchpad not being updated with the review links anymore | 18:53 |
| jpeeler | immediately marked in progress, so it's not lagging behind | 18:54 |
| sdake | harmw can you review https://review.openstack.org/#/c/196797/ plz | 18:54 |
| sdake | jpeeler I am not sure why that is happening best to ak on #openstack-infra | 18:55 |
| jpeeler | might as well, it's been doing that for a few days | 18:55 |
| sdake | is keystone woring for you jpeeler? | 19:04 |
| sdake | i am not getting a 35357 bounded port | 19:04 |
| jpeeler | i don't know, is it port 35357 or 5000? | 19:06 |
| openstackgerrit | Merged stackforge/kolla: Install oslo.service via pip https://review.openstack.org/196777 | 19:10 |
| harmw | sdake: done | 19:13 |
| sdake | https://bugs.launchpad.net/kolla/+bug/1469851 | 19:13 |
| openstack | Launchpad bug 1469851 in kolla "backtrace from keystone" [Critical,Confirmed] | 19:13 |
| sdake | can you guys tae a look at that backtrace and tell me what you think | 19:13 |
| harmw | wtf thats ugly | 19:17 |
| harmw | god I hate delorean, keystone was a-ok last friday | 19:17 |
| harmw | (from scratch) | 19:17 |
| sdake | ERROR keystone % {'service': service, 'base': ServiceBase}) | 19:18 |
| harmw | but that doesn't answer your question :) | 19:18 |
| sdake | possibly a config option | 19:18 |
| harmw | is there a service endpoint named keystone? | 19:18 |
| harmw | (just a wild shot) | 19:18 |
| harmw | it sucks bigtime LP wraps the lines at X chars btw | 19:19 |
| jpeeler | heh agreed | 19:19 |
| *** shardy has quit IRC | 19:22 | |
| *** shardy has joined #kolla | 19:23 | |
| sdake | essentially oslo_service was added as a dependency for keystone, but keystone running in eventlet mode was not tested | 19:32 |
| sdake | devstack runs keystone in wsgi mode | 19:32 |
| sdake | eventlet is being deprecated in M cycle. | 19:33 |
| sdake | I suspect keystone's versoin of service was different then the oslo version in some subtle way | 19:33 |
| sdake | service.py | 19:34 |
| sdake | before the oslo dependency was created | 19:34 |
| sdake | so pretty much we need to port keystone to wsgi I guess | 19:34 |
| sdake | I have no idea hwo to do that | 19:34 |
| openstackgerrit | Merged stackforge/kolla: Correct FROM line in openvswitch-agent https://review.openstack.org/196797 | 19:34 |
| sdake | logging.warning(_('Running keystone via eventlet is deprecated as of Kilo ' | 19:39 |
| sdake | 'in favor of running in a WSGI server (e.g. mod_wsgi). ' | 19:39 |
| sdake | 'Support for keystone under eventlet will be removed in ' | 19:39 |
| sdake | 'the "M"-Release.')) | 19:39 |
| sdake | so pretty much we need to do that work now | 19:39 |
| sdake | (port to mod_wsgi) whatever the hell that is | 19:39 |
| sdake | http://docs.openstack.org/developer/keystone/apache-httpd.html | 19:40 |
| vinkman | That is just the keystone api is accesses via Apache rather than than the old way… | 19:41 |
| sdake | yup that is what we need, httpd access rather then eventlet accdess | 19:52 |
| sdake | and then keystone container will work | 19:52 |
| sdake | but I dont know how to actually do that | 19:52 |
| vinkman | http://docs.openstack.org/kilo/install-guide/install/yum/content/keystone-install.html | 19:53 |
| vinkman | scroll down to 'To configure the Apache HTTP server' | 19:54 |
| *** shardy has quit IRC | 19:55 | |
| jpeeler | we don't have an apache service yet, right? | 19:56 |
| jpeeler | it sounds like an apache container needs creating and then keystone/horizon modified to use it. might be easier to just fix eventlet | 19:57 |
| sdake | no we jus tinstall apapache in the keystone container | 19:58 |
| sdake | if we fix eventlet in keystone, it could take weeks before releases are available jpeeler | 19:58 |
| sdake | plus its being deprecated | 19:58 |
| sdake | horizon already installs http+wsgi mod (the container) | 19:58 |
| jpeeler | patch it, i thought you were trying to release today | 19:58 |
| sdake | I am | 19:58 |
| sdake | I dont know what to patch | 19:59 |
| sdake | all I know is as it stands it will never work | 19:59 |
| sdake | tbh i think apache is past of least resistance here | 19:59 |
| sdake | plus we wont have to do it later | 19:59 |
| sdake | but I dont know how to do that either | 19:59 |
| jpeeler | i can't get my dev env going, so i can't help much yet | 19:59 |
| jpeeler | but yes, you're right if we do it now we won't have to later! | 19:59 |
| sdake | ya i just got mine going 2 days ago | 19:59 |
| sdake | finicky software | 20:00 |
| vinkman | as a stopgap you can just also install apache in the keystone container, make sure that port 80 is off and be on your merry way… until an apache can container can be extracted to its own container for use by multiple svcs later | 20:00 |
| sdake | vinman thanks I'll try out that apache documentation | 20:00 |
| jpeeler | this may or may not be helpful | 20:01 |
| sdake | we are not extracting apache out | 20:01 |
| jpeeler | https://github.com/openstack/keystone/blob/master/httpd/wsgi-keystone.conf | 20:01 |
| *** shardy has joined #kolla | 20:01 | |
| vinkman | Yup that is the exact wsgi config you need for Keystone... | 20:01 |
| vinkman | One note though, the python script does not contain the #!/usr/bin/python at the top and caused some grieve… | 20:04 |
| *** jasonsb has joined #kolla | 20:18 | |
| *** nihilifer has joined #kolla | 20:20 | |
| *** jruano has quit IRC | 20:26 | |
| *** dontalton has quit IRC | 20:26 | |
| harmw | sdake: got it working yet? | 20:27 |
| sdake | no but close to finished with the implementation | 20:28 |
| sdake | so hang tight | 20:28 |
| harmw | meh, me 2 :) | 20:28 |
| harmw | building, lets see what happens | 20:29 |
| sdake | what should %{GLOBAL} be? | 20:30 |
| *** jasonsb has quit IRC | 20:30 | |
| vinkman | you mean in the Vhost definition? | 20:31 |
| vinkman | if so, it should stay %{GLOBAL} wsgi_mod will do the replace on use… | 20:32 |
| sdake | I get error: | 20:33 |
| vinkman | at least that how I understood it… | 20:33 |
| sdake | WSGIDaemonProcess keystone-public processes=5 threads=1 user=keystone group=keystone display-name=%{GROUP} | 20:33 |
| sdake | saying that is incorrect syntax | 20:33 |
| vinkman | At build or runtime? | 20:35 |
| vinkman | that is very odd… the syntax is correct for the apache mod… | 20:36 |
| sdake | AH00526: Syntax error on line 5 of /etc/httpd/conf.d/wsgi-keystone.conf: | 20:37 |
| sdake | Invalid command 'WSGIDaemonProcess', perhaps misspelled or defined by a module not included in the server configuration | 20:37 |
| harmw | sdake: you didn't install mod_wsgi | 20:37 |
| sdake | RUN yum -y install openstack-keystone \ | 20:38 |
| sdake | python-keystoneclient \ | 20:38 |
| sdake | httpd \ | 20:38 |
| sdake | httpd-mod-wsgi \ | 20:38 |
| sdake | && yum clean all | 20:38 |
| harmw | httpd-mod-wsgi doesn't exist ;) | 20:38 |
| harmw | its mod_wsgi | 20:38 |
| sdake | that may explain why our horizon container doesn't work | 20:38 |
| sdake | harmw you got it working then? | 20:40 |
| harmw | stuck on a 401 or something | 20:40 |
| harmw | 403 actually | 20:41 |
| harmw | (and running out of daytime..) | 20:41 |
| sdake | http://ur1.ca/myjso -> http://paste.fedoraproject.org/237831/14356106 | 20:44 |
| harmw | interesting | 20:47 |
| harmw | I can give it a quick review if you get it online in the next 2 minutes :) | 20:48 |
| openstackgerrit | Steven Dake proposed stackforge/kolla: Convert keystone from eventlet to httpd https://review.openstack.org/196854 | 20:48 |
| sdake | besides sayins tee sucks, recommend exact commands | 20:49 |
| sdake | also it would be nice if I knew where ot copy that from the local filesystem rather then getting it via curl | 20:49 |
| harmw | ok | 20:49 |
| harmw | gimme a few secs | 20:49 |
| sdake | jpeeler your review appreciated as well - this isn't quite ready to go but it does appear to work | 20:50 |
| jpeeler | pretty sure it goes to / | 20:52 |
| jpeeler | sdake: it looks fine aside from the tee weirdness | 20:54 |
| sdake | no whati mean is i am downloading from git.openstack.org | 20:55 |
| sdake | that same file is probably somewhere on teh local disk | 20:55 |
| sdake | could you tell me where? | 20:55 |
| jpeeler | well i think it goes to / | 20:55 |
| jpeeler | but curl doesn't write to filesystem by default, right? | 20:55 |
| harmw | sdake: reviewed | 20:55 |
| sdake | the rpm installs the file, right? | 20:55 |
| sdake | i dont comprehend why they recommend using git to get the file when rpm should hav eit | 20:56 |
| sdake | i prefer to copy it around from the local filesystem | 20:56 |
| harmw | the -1 is just for using git, the file is in /usr as is noted in the comment :) | 20:56 |
| jpeeler | well look at that :) | 20:56 |
| harmw | although I only did a quick glimpse | 20:56 |
| harmw | feel free to proof me wrong | 20:57 |
| harmw | weird, though, I can't realy tell what I did different | 20:57 |
| harmw | oh crap | 20:57 |
| harmw | sdake: that keystonewsgi config should go in /common/bla | 20:58 |
| sdake | hmm let me think on that one | 20:58 |
| sdake | keystone.wsgi is the binary ? | 20:59 |
| harmw | python | 20:59 |
| harmw | be back in 10, to catch up and then I'm off | 21:00 |
| *** nihilifer has quit IRC | 21:02 | |
| sdake | good news i am able to get to the prompt to run init-runonce with the patch ;) | 21:07 |
| *** jasonsb has joined #kolla | 21:09 | |
| harmw | woohoo! | 21:13 |
| harmw | in that case I can safely take off :) | 21:13 |
| sdake | need a +2 out of you | 21:14 |
| sdake | hang tight 5 mins plz | 21:14 |
| sdake | new patch coming | 21:14 |
| sdake | that fixe the busteness of the original | 21:14 |
| harmw | argh | 21:18 |
| harmw | oh well | 21:18 |
| harmw | running out of minutes here sdake | 21:19 |
| sdake | ok harmw take care then | 21:19 |
| harmw | just hurry up | 21:20 |
| harmw | :) | 21:20 |
| sdake | sed command to replace /var/log/apache2 with /var/log/httpd | 21:20 |
| sdake | paying for beer for answer ;) | 21:20 |
| harmw | 's,/var/log/apache2,/var/log/httpd,' | 21:20 |
| sdake | what is the full command line | 21:21 |
| harmw | sed -i -r <insert regex here> <filename> | 21:21 |
| harmw | nah, -r is probably not even needed | 21:21 |
| harmw | test it, without -r | 21:22 |
| harmw | but where do you need to replace that anyway? the wsgi httpd.conf we ship | 21:22 |
| harmw | *ourselves | 21:22 |
| sdake | i am using the one that comes in the distro packaging | 21:23 |
| harmw | ofcourse | 21:23 |
| harmw | clean | 21:23 |
| harmw | nice | 21:23 |
| openstackgerrit | Steven Dake proposed stackforge/kolla: Convert keystone from eventlet to httpd https://review.openstack.org/196854 | 21:25 |
| sdake | harmw did you sort out why you were getting a 403 after looking at my version? | 21:26 |
| harmw | no, since yours is up for review I took care of reviewing that instead | 21:27 |
| sdake | now for int-runonce ;) | 21:27 |
| harmw | what up with all those RUN's | 21:27 |
| * sdake shudders | 21:27 | |
| sdake | half the community thinks we should compress runs, other half doesn't - i am in the other half doesn't | 21:27 |
| sdake | my take is that is a docker problem to solve | 21:28 |
| sdake | ;) | 21:28 |
| harmw | pff | 21:28 |
| sdake | what we have now is more readable | 21:28 |
| sdake | vs a big one command line operation | 21:28 |
| sdake | also each command fails individually without having to track down the bad command | 21:28 |
| vinkman | for what it is worth I vote for multiple RUNs one RUN makes debugging aproblem just a huge pain… :D | 21:29 |
| harmw | no it doesn't :p | 21:30 |
| harmw | sdake: +2 in yo pocket | 21:30 |
| sdake | winning | 21:30 |
| sdake | jpeeler ;) | 21:30 |
| harmw | nightnight people | 21:30 |
| sdake | take care harmw | 21:30 |
| sdake | hopefully no more problems | 21:30 |
| harmw | oh btw, you said something about horizon | 21:30 |
| harmw | in regards to mod_wsgi | 21:31 |
| sdake | yup | 21:31 |
| sdake | it uses httpd-mod-wsgi | 21:31 |
| sdake | rather then mod_wsgi | 21:31 |
| harmw | ok, perhaps that should be fixed as well | 21:31 |
| sdake | there is no such package | 21:31 |
| sdake | agree | 21:31 |
| harmw | indeed | 21:31 |
| sdake | that may be why horizon is not woring atm | 21:31 |
| sdake | i'll look at that next | 21:31 |
| harmw | don't know, probably not since without mod_wsgi it should be complaining about wrong config bits | 21:31 |
| harmw | like you had yourself | 21:32 |
| vinkman | BTW, I am curious, you have not disabled port 80 in keystone, how is horizon going to start on the same box? we are using host networking for both, right? | 21:32 |
| harmw | and since it doesn't do that... is it even using wsgi? | 21:32 |
| harmw | vinkman: right | 21:32 |
| harmw | sed -i -r 's,(Listen 80),#\1,' /etc/httpd/httpd.conf | 21:33 |
| harmw | asuming it is a capital L | 21:33 |
| sdake | ok so add that command? | 21:33 |
| harmw | yup, give it a go | 21:33 |
| harmw | sed -i -r 's,^(Listen 80),#\1,' /etc/httpd/httpd.conf | 21:34 |
| harmw | thats cleaner | 21:34 |
| harmw | my brain should've been asleep like an hour ago | 21:34 |
| harmw | starting to show :p | 21:34 |
| harmw | sdake: you need my vote, right? | 21:35 |
| sdake | yes building | 21:35 |
| harmw | damn | 21:35 |
| harmw | ok | 21:35 |
| vinkman | I think that should be /etc/httpd/conf/httpd.conf | 21:36 |
| vinkman | at leas that is where the one on my system lives… | 21:36 |
| harmw | oh ok, didn't give that much thought :) | 21:36 |
| sdake | yup | 21:37 |
| harmw | done building? | 21:38 |
| sdake | yes testing now | 21:38 |
| harmw | curl :80 as well | 21:38 |
| harmw | to verify | 21:38 |
| sdake | hoirzon will have port 80 i think | 21:39 |
| harmw | yes | 21:39 |
| harmw | oh your building/starting up everything? | 21:39 |
| sdake | conn refused | 21:39 |
| harmw | ok | 21:39 |
| openstackgerrit | Steven Dake proposed stackforge/kolla: Convert keystone from eventlet to httpd https://review.openstack.org/196854 | 21:39 |
| harmw | thats good, from keystone | 21:39 |
| harmw | (so it was a capital L) | 21:40 |
| sdake | what was capital L | 21:40 |
| harmw | the listen | 21:40 |
| harmw | you've replaced | 21:41 |
| sdake | i cut and paste yoru command | 21:41 |
| sdake | is it wrong? | 21:41 |
| vinkman | No it is right… | 21:41 |
| harmw | no, well, I wasn't sure if httpd.conf was using either uppercase or lowercase Listen directives :)( | 21:41 |
| harmw | sdake: there you go | 21:41 |
| harmw | I'm realy off now | 21:42 |
| harmw | realy, realy | 21:42 |
| sdake | enjoy sleepy time :) | 21:42 |
| harmw | for about 6hrs, yes | 21:42 |
| sdake | jpeeler can you ack that patch | 21:42 |
| jpeeler | done | 21:44 |
| sdake | sweet | 21:44 |
| sdake | couple hours and I'll announce release assuing my tests of the network functionality go well | 21:44 |
| sdake | (need to wait for the patch to hit the repo) | 21:44 |
| sdake | only 3 days late ;( | 21:44 |
| vinkman | rather late and working than on time and broken … :D Just my $.0001 | 21:45 |
| sdake | vinkman thats my theory as well | 21:46 |
| sdake | although it eats into our liberty 2 schedule | 21:46 |
| sdake | that was the most challenging technical work I've done all cycle ;-( | 21:46 |
| jpeeler | the delorean packages make me wonder how we'll handle it long term | 21:47 |
| vinkman | lol… Cool… I haven't done any (yet I hope) :D | 21:47 |
| sdake | jpeeler you need a gate on dependnecies | 21:47 |
| sdake | and then I think you are good to go | 21:47 |
| jpeeler | in delorean? | 21:47 |
| sdake | yes | 21:47 |
| jpeeler | who can we talk to about that, derekh? | 21:47 |
| sdake | need to verify requirements.txt matches the requires list | 21:47 |
| sdake | no idea - I don't work at rht anymore :) | 21:48 |
| *** rhallisey has quit IRC | 21:48 | |
| sdake | ask around - your manager probably good first step there | 21:48 |
| jpeeler | yeah, it'll help ease the pain quite a lot | 21:48 |
| sdake | Configuring nova public key and quotas. | 21:50 |
| sdake | ERROR (EndpointNotFound): publicURL endpoint for compute service not found | 21:50 |
| * sdake groans | 21:50 | |
| *** vinkman has quit IRC | 21:52 | |
| *** vinkman has joined #kolla | 21:53 | |
| sdake | wierd that start the public urls registered | 21:55 |
| sdake | https://bugs.launchpad.net/kolla/+bug/1469891 | 21:59 |
| openstack | Launchpad bug 1469891 in kolla "some cinder services enter restarting state" [High,Confirmed] - Assigned to Ryan Hallisey (rthall14) | 21:59 |
| * sdake growls | 21:59 | |
| sdake | it sure was grand of the oslo pps to change all of their oslo namespace shit right at librty 1 | 21:59 |
| * sdake growls some more | 21:59 | |
| *** sdake_ has joined #kolla | 22:01 | |
| *** sdake has quit IRC | 22:04 | |
| *** daneyon has joined #kolla | 22:15 | |
| *** daneyon_ has joined #kolla | 22:16 | |
| *** absubram has quit IRC | 22:19 | |
| openstackgerrit | Merged stackforge/kolla: Convert keystone from eventlet to httpd https://review.openstack.org/196854 | 22:20 |
| *** daneyon has quit IRC | 22:20 | |
| *** prad has quit IRC | 22:31 | |
| *** dims_ has joined #kolla | 22:55 | |
| *** dims has quit IRC | 22:59 | |
| *** daneyon_ has quit IRC | 23:00 | |
| *** dims_ has quit IRC | 23:00 | |
| *** sdake has joined #kolla | 23:02 | |
| sdake | just created 3 vms with heat | 23:04 |
| sdake | now to see if they worked :) | 23:04 |
| *** sdake_ has quit IRC | 23:05 | |
| *** akwasnie has quit IRC | 23:21 | |
| *** vinkman has quit IRC | 23:30 | |
| *** rbergeron has quit IRC | 23:39 | |
| *** rbergeron has joined #kolla | 23:39 | |
| *** mandre has joined #kolla | 23:43 | |
| *** jasonsb_ has joined #kolla | 23:45 | |
| *** jasonsb has quit IRC | 23:45 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!