| *** inc0 has joined #kolla | 00:16 | |
| inc0 | Wassup? | 00:16 |
|---|---|---|
| *** shadower has quit IRC | 00:23 | |
| *** shadower has joined #kolla | 00:23 | |
| SamYaple | inc0: hey | 00:31 |
| inc0 | Texas is...hot. | 00:31 |
| SamYaple | you in TX now? | 00:31 |
| inc0 | Yup | 00:31 |
| SamYaple | oh cool | 00:31 |
| SamYaple | but youre in houston | 00:32 |
| SamYaple | thats the issue | 00:32 |
| inc0 | Waiting for plane to ca | 00:32 |
| inc0 | Yeah, and just on airport | 00:32 |
| *** dolpher has joined #kolla | 00:33 | |
| SamYaple | review some patches for me | 00:33 |
| SamYaple | i have 7, including the haproxy ansible one you are author on | 00:33 |
| inc0 | Tomorrow, it's my 21st hour awake | 00:34 |
| inc0 | And few more to go | 00:34 |
| SamYaple | eh thats nothing | 00:35 |
| inc0 | Well my last trip to Romania was worse | 00:36 |
| inc0 | I need to develop ability to sleep wherever whenever | 00:37 |
| SamYaple | its called "exhaustion" | 00:45 |
| inc0 | Btw..I wonder..why do we have midcycle in ca anyway? | 00:51 |
| SamYaple | sdake | 00:51 |
| SamYaple | i wanted it in TX | 00:51 |
| inc0 | I don't think anyone from kolla lives there | 00:52 |
| SamYaple | sdake / cisco / controlled venue | 00:52 |
| SamYaple | because sdake is at cisco i think he books the venue for free and things | 00:52 |
| SamYaple | some sponsored stuff. i dont really know | 00:52 |
| inc0 | Meh... most expensive location in us isn't it? | 00:53 |
| SamYaple | my wallet says yes | 00:54 |
| SamYaple | next midcycle im going to see if RAX will host | 00:54 |
| SamYaple | they almost certainly will | 00:54 |
| inc0 | I could get Intel host on Oregon as well | 00:54 |
| * SamYaple doesn't want to travel... | 00:55 | |
| inc0 | or we all go to Slower's place and hunt mooses between sessions;) | 00:55 |
| SamYaple | illegal! | 00:56 |
| inc0 | Hey, I have 2 continents and ocean to cross, don't complain | 00:56 |
| *** dims has joined #kolla | 00:58 | |
| inc0 | Btw, I think tx is only place where they need to put sign saying explicitly that no, you cant take shotgun to plane cabin, it has to be in declared luggage | 00:59 |
| *** erkules_ has joined #kolla | 01:05 | |
| *** erkules has quit IRC | 01:08 | |
| inc0 | Ok, going for boarding | 01:16 |
| inc0 | Cyas | 01:17 |
| *** inc0 has quit IRC | 01:17 | |
| *** erkules_ has quit IRC | 01:52 | |
| *** erkules_ has joined #kolla | 01:53 | |
| *** daneyon has joined #kolla | 01:54 | |
| dolpher | I just found that the workflow for this patch is not started, https://review.openstack.org/#/c/203946/ | 01:55 |
| dolpher | Any idea how to continue? | 01:55 |
| *** sdake has joined #kolla | 01:55 | |
| sdake | yo all | 01:56 |
| sdake | look 42 poeple in channel yay :) | 01:56 |
| *** daneyon has quit IRC | 01:59 | |
| *** rhallisey has joined #kolla | 02:29 | |
| *** rhallisey has quit IRC | 02:31 | |
| SamYaple | sdake: yo | 02:48 |
| sdake | yo | 02:48 |
| SamYaple | yo yo yo | 02:48 |
| SamYaple | lots of awesomeness in the review queue | 02:48 |
| sdake | haven't looked in last couple of days - been unplugged entirely :) | 02:49 |
| SamYaple | good! | 02:49 |
| openstackgerrit | Merged stackforge/kolla: Install tooz for designate pool manager https://review.openstack.org/203946 | 02:53 |
| SamYaple | sdake: all the multinode base stuff for ansible is done | 02:53 |
| SamYaple | and tested | 02:53 |
| sdake | cool and networking works too then? | 02:53 |
| SamYaple | havent had any issues so far | 02:54 |
| sdake | nice! | 02:54 |
| SamYaple | havent really tested alot though, like vlans havent been tested multinode | 02:54 |
| SamYaple | no reason it _shouldnt_ work but you know | 02:54 |
| sdake | Im going to do a 5 minute demo of AIO ansible at the kolal midcycle | 02:54 |
| sdake | ya - we have a few months to beat the codebase into full submission ;-) | 02:54 |
| SamYaple | yep. i dont know what the itenerary is but i have a few things to show | 02:55 |
| sdake | just to bring people up to speed | 02:55 |
| SamYaple | like private registry and how it helps speed things up | 02:55 |
| sdake | typically design sessions dont have demos | 02:55 |
| SamYaple | eh informal | 02:55 |
| sdake | but I want to show the basics of where we are headed since many people attending are not day to day kolla developers | 02:55 |
| SamYaple | do we have a full list? | 02:56 |
| sdake | we have 14 sessions | 02:56 |
| SamYaple | of who is attending i mean | 02:56 |
| sdake | yes I have a list | 02:56 |
| SamYaple | is it a public list? | 02:56 |
| sdake | no | 02:56 |
| SamYaple | oh | 02:56 |
| SamYaple | well anyway yea lots of cool stuff to talk about | 02:56 |
| sdake | I dont htink its appropriate ot publish people's email addresses unless they choose to do that | 02:56 |
| SamYaple | i want to discuss the config thing and how to make it better in ansible | 02:57 |
| SamYaple | yea im not worried | 02:57 |
| SamYaple | i want to make that config peice work smoother. | 02:57 |
| sdake | which part | 02:57 |
| SamYaple | the end result is as designed, but i dont like the current ansible flow for it | 02:57 |
| SamYaple | merging all the configs to get a single one | 02:58 |
| sdake | seems to work ok to me | 02:58 |
| SamYaple | for the configs that are not configini compatible I was thinking we just have a full 100% override config. It wont merge things, but you have complete control over it if you want | 02:58 |
| sdake | the only part I have concern with is upgrade - we have a session on upgrade | 02:58 |
| SamYaple | it works great | 02:58 |
| SamYaple | i dont like the technical flow with ansible is all | 02:59 |
| SamYaple | agreed. | 02:59 |
| sdake | since upgrade is listed specifically in our mission, I think we need to tackle that properly ;-) | 02:59 |
| *** dims has quit IRC | 02:59 | |
| sdake | since pretty much nobody elses deployment tool does upgrade ;-) | 02:59 |
| SamYaple | its going to be hard to have a non-flexible plan for upgrades though since openstack upgrades change | 03:00 |
| SamYaple | for instance, if we support juno->kilo that will be easy | 03:00 |
| sdake | well we can discuss at the design summit | 03:00 |
| sdake | more brains better then less ;-) | 03:00 |
| SamYaple | if you expect to stick with upgrades on the master branch, that would be harder/impossible/impracticle | 03:00 |
| sdake | ya major upgrades are esy, but I'd like CD | 03:00 |
| sdake | I think the big tent needs to solve this problem in general | 03:01 |
| sdake | like via a tag for example | 03:01 |
| sdake | and some standardized way to handle db migrations | 03:01 |
| sdake | based upon tags int he repo or some other mechanism | 03:01 |
| SamYaple | they are working on all of that. i dont want kolla to have hackish workarounds before it lands upstream | 03:01 |
| SamYaple | i mean they already do tags | 03:01 |
| sdake | cool good to hear folks are working on it | 03:01 |
| sdake | well my sweet tomatoes is getting warm, give me 20 mins to finish my dinner pzl :) | 03:02 |
| SamYaple | oh yea youre on wierd itme | 03:02 |
| SamYaple | you can simply stop responding you know :) | 03:02 |
| sdake | ya but that is rude | 03:02 |
| SamYaple | gooo | 03:03 |
| sdake | so excited for the midcycle? | 03:07 |
| SamYaple | i am | 03:08 |
| SamYaple | i am working on push for the build script no | 03:08 |
| SamYaple | now* | 03:08 |
| sdake | its a shame martin and jeff wont be able to make it | 03:10 |
| sdake | martin is moving and jeff tieing knot | 03:10 |
| sdake | s | 03:10 |
| sdake | sss | 03:10 |
| SamYaple | yea | 03:11 |
| sdake | so when we set conditionals for the l2 agent, will we just do something like l2_agent=ovs or l2_agent=linuxbridge? | 03:25 |
| sdake | there are a whole bunch of different l2 agents we could theoretically want to implement | 03:25 |
| sdake | btw yesterday yo umentioned something about mido | 03:26 |
| sdake | is that an ovs competitor or somethiing? | 03:26 |
| SamYaple | midonet yes | 03:26 |
| SamYaple | i have no clue to be honest | 03:27 |
| SamYaple | the neutron containerized kuyr(?) project guy way on here tlaking about it | 03:27 |
| SamYaple | i asked what was involved and we can esialy implement it instead of ovs/linubridge | 03:27 |
| sdake | ya i'm sure once we figur eout how to do the 2nd l2 agent, we can add new ones easily | 03:28 |
| SamYaple | its already figured out. i jave havent tested/pushed it | 03:28 |
| SamYaple | just* | 03:28 |
| sdake | nice, you ahve linuxbridge implemented then? | 03:28 |
| SamYaple | i have the correct container starting but i havent tested teh config options to ensure they work | 03:29 |
| SamYaple | not a big linuxbridge guy | 03:29 |
| SamYaple | the logic is all there, just container testing is needed | 03:29 |
| sdake | nice - can you get a wip review up? | 03:29 |
| SamYaple | probably needs a big rebase at this point, but like i said its not much | 03:29 |
| SamYaple | yea i need to | 03:30 |
| sdake | we can probably get daneyon to look at it | 03:30 |
| SamYaple | cool | 03:30 |
| SamYaple | i should learn it too. | 03:30 |
| SamYaple | i know the flows network wise, the openstack config options are a ??? to me | 03:30 |
| sdake | but if you have the work to start with it woul probably help | 03:30 |
| sdake | like how to run one vs the other | 03:30 |
| *** coolsvap|away is now known as coolsvap | 03:30 | |
| SamYaple | oh yea the work is done, its just the config params | 03:30 |
| sdake | I can helpl get the onfig setup | 03:31 |
| SamYaple | neutron_agent=openvswitch | 03:31 |
| SamYaple | neutron_agent=linuxbridge | 03:31 |
| SamYaple | neutron_agent=midonet | 03:31 |
| SamYaple | so i just do a when conditional in the templates and on the tasks for ansible | 03:31 |
| sdake | ya so the part I dont know how to implement is getting those to run the blobs of config | 03:31 |
| SamYaple | oh thats just basic templating | 03:32 |
| sdake | oh you dont make a separate template dir? | 03:32 |
| SamYaple | i do | 03:32 |
| sdake | well if you get it into a wip I can take it over | 03:32 |
| SamYaple | ill just rebase and WIP it | 03:32 |
| sdake | sounds good | 03:32 |
| SamYaple | it may just need testing to be honest | 03:32 |
| sdake | I will tackle it next week | 03:32 |
| sdake | if you get me something to get started with :) | 03:33 |
| SamYaple | oh btw | 03:33 |
| sdake | I dont know the correct syntax for ansible to do the job | 03:33 |
| sdake | but do know how to configure linuxbridge | 03:33 |
| SamYaple | asked around #ansible and the vairable 'use_service: True' is prefered | 03:33 |
| sdake | what does that mean | 03:33 |
| SamYaple | so sometihng like 'use_heat: True' in globals.yml will work | 03:33 |
| SamYaple | use_cinder: False | 03:34 |
| SamYaple | use_trove: True | 03:34 |
| sdake | and that magically starts the service in the children? | 03:34 |
| sdake | from the inventory file? | 03:34 |
| SamYaple | nah dont worry about the implementation details. there is no magic | 03:34 |
| SamYaple | its just a normal when conditional | 03:34 |
| sdake | ok well sounds like fun for you :) | 03:35 |
| SamYaple | just that is what is prefered | 03:35 |
| sdake | cool | 03:35 |
| SamYaple | so we can turn on/off per serivce | 03:35 |
| sdake | that sounds good | 03:35 |
| SamYaple | that was my main concern | 03:35 |
| SamYaple | cool cool | 03:35 |
| sdake | my #1 priority after our big tent review is linuxbridge | 03:35 |
| SamYaple | whats the status on that | 03:35 |
| sdake | we are going to group edit it tuesday | 03:36 |
| SamYaple | cool | 03:36 |
| SamYaple | i went through the blueprint and bug list entirely yesterday | 03:36 |
| sdake | and i'll submit it | 03:36 |
| SamYaple | cleaned it all up | 03:36 |
| sdake | nice | 03:36 |
| SamYaple | i think we are down to 45 bugs, with 20 in progress | 03:36 |
| sdake | finally someone helping me for launchpad mantenance ;-) | 03:37 |
| sdake | 45 open bugs for l2 you mean? | 03:37 |
| sdake | we may ahve to bounce some of those to l3 | 03:37 |
| SamYaple | no in total | 03:37 |
| SamYaple | i adjusted all that i thought needed adjusting | 03:37 |
| sdake | cool | 03:37 |
| SamYaple | all current l2 should be doable | 03:37 |
| sdake | deadline is 30th | 03:37 |
| SamYaple | but we need to squash queue | 03:38 |
| sdake | so we should bounce any that are not critical to l3 | 03:38 |
| SamYaple | yup | 03:38 |
| SamYaple | like i said, we need to squash queue | 03:38 |
| sdake | we have a long period at the end of l3 that is spent for stabilization | 03:38 |
| SamYaple | as long as we implement templating, i think thats great | 03:39 |
| SamYaple | we need templating though | 03:39 |
| sdake | https://wiki.openstack.org/wiki/Liberty_Release_Schedule | 03:39 |
| sdake | aug 31 is feature freeze for l3 | 03:39 |
| sdake | then we have 6 weeks for stabilization + bug fixing | 03:39 |
| SamYaple | i expect templating to be 100% done before end of august | 03:40 |
| sdake | october 12th | 03:40 |
| sdake | october 12 is release date | 03:40 |
| SamYaple | we basically did all of ansible in july alone | 03:40 |
| sdake | dont forget from source ;-) | 03:40 |
| sdake | we knocked that one out too ;-) | 03:41 |
| SamYaple | _and_ from-source | 03:41 |
| SamYaple | i know we are amazing | 03:41 |
| SamYaple | i didnt even help with form-source | 03:41 |
| sdake | ya our team rocks | 03:41 |
| SamYaple | and i expected to lead that | 03:41 |
| sdake | I did a couple from source | 03:41 |
| SamYaple | but the community took over | 03:41 |
| sdake | that was a really easy subproject once the first one was done | 03:41 |
| sdake | pretty much c&p | 03:41 |
| SamYaple | yea i know, but still | 03:42 |
| SamYaple | that takes work | 03:42 |
| sdake | I think paul id the first one | 03:42 |
| sdake | which was the hard one ;-) | 03:42 |
| SamYaple | pbourke lead the whole drive | 03:42 |
| SamYaple | without him it would not have been done | 03:42 |
| sdake | I think we finished the job on all the services as well | 03:42 |
| sdake | we may have a few left over, not certain | 03:42 |
| *** daneyon has joined #kolla | 03:43 | |
| SamYaple | when we move to templating we should be able to cut our LOC in half _and_ support more distros at the same time | 03:43 |
| sdake | hey daneyon whats up bro | 03:43 |
| SamYaple | daneyon you feeling homesick? | 03:43 |
| SamYaple | Is Otto being mean? | 03:44 |
| sdake | we only hav eabout 5k lines of code | 03:45 |
| sdake | although sloccount doesn't count yaml unfortunately | 03:45 |
| sdake | for 5k lines of code, our software sure does a hell of alot | 03:45 |
| sdake | I find it fascinating what we are able to do with such a compact code base | 03:45 |
| SamYaple | indeed. and i think we can reduce LOC and complexity at the same time without losing features | 03:45 |
| SamYaple | ITS THE PERFECT CRIME | 03:45 |
| sdake | definately the smalleest code base I've ever written | 03:46 |
| sdake | that is a complete project | 03:46 |
| SamYaple | its great | 03:46 |
| *** bmace has quit IRC | 03:46 | |
| sdake | so sam cn you get that linuxbridge wip up this week and then you mind i take it over? | 03:46 |
| SamYaple | maybe. i am into the push code right now | 03:47 |
| SamYaple | i have alot of code in review and i dont want to be into more rebases than i have to | 03:47 |
| SamYaple | one of those patches is 4 deep | 03:48 |
| *** daneyon has quit IRC | 03:48 | |
| sdake | oh one of the linuxbridge has 4 patches? | 03:48 |
| SamYaple | no | 03:48 |
| SamYaple | the in review patches | 03:48 |
| SamYaple | they touch some files so i have to have deps for new patches | 03:48 |
| sdake | well this week is going to be slow on reviews because of midcycle I think | 03:49 |
| sdake | maybe we canget people to jam out on the reviews on monday ;-) | 03:49 |
| SamYaple | its why i am working on the push code which has no deps | 03:49 |
| sdake | i'll go hit the review queue in a few minutes | 03:49 |
| SamYaple | check out https://review.openstack.org/#/c/205612/2/ansible/roles/message-broker/templates/rabbitmq.config.j2 | 03:49 |
| SamYaple | i now speak erlang | 03:50 |
| SamYaple | and i hate it | 03:50 |
| sdake | looks like greek to me ;-) | 03:50 |
| SamYaple | it was a pita to write too | 03:50 |
| SamYaple | all to ensure binding to a single interface | 03:50 |
| *** bmace has joined #kolla | 03:58 | |
| *** dims has joined #kolla | 04:00 | |
| sdake | bmace around? | 04:03 |
| *** achanda has quit IRC | 04:03 | |
| sdake | coolsvap around? | 04:04 |
| *** Slower has quit IRC | 04:04 | |
| coolsvap | sdake, yes | 04:04 |
| sdake | coolsvap could I bribe you with my neverending devotion to take on a blueprint | 04:04 |
| sdake | it is really easy - mostly documentation change and a git rename | 04:04 |
| coolsvap | sdake, sure will do | 04:04 |
| sdake | https://blueprints.launchpad.net/kolla/+spec/kolla-compose-script | 04:05 |
| coolsvap | sure I will do that | 04:05 |
| sdake | make sure to alter the documentation | 04:06 |
| sdake | needs to be done tomorrow if possible | 04:06 |
| *** dims has quit IRC | 04:06 | |
| coolsvap | yes I will look into it on priority | 04:06 |
| sdake | if you cant get it done let me know and I'll tackle it - but I've got a full plate because of the midcycle | 04:07 |
| sdake | and our midcycle lines up with our deadlines | 04:07 |
| sdake | and not sure where diga is atm | 04:07 |
| coolsvap | diga, around? | 04:08 |
| diga | yes sdake | 04:10 |
| diga | Hi sdake | 04:10 |
| diga | coolsvap: hi | 04:10 |
| coolsvap | diga, do you think i can help you with https://blueprints.launchpad.net/kolla/+spec/kolla-compose-script | 04:11 |
| diga | sure coolsvap | 04:12 |
| sdake | 18 blueprints and 78 bugs for l2 | 04:13 |
| sdake | busy ;) | 04:13 |
| coolsvap | i will send a patch for review in sometime | 04:13 |
| diga | coolsvap: gimme some time | 04:13 |
| sdake | it needs to be done tomorrow at latest | 04:14 |
| diga | sure sdake | 04:14 |
| sdake | we are out of time | 04:14 |
| sdake | the docs are the onerous part | 04:14 |
| sdake | the git mv should be easy | 04:14 |
| SamYaple | sdake: is that the number of completed things? | 04:14 |
| sdake | we did 18 blueprints | 04:14 |
| diga | yes sdake | 04:14 |
| SamYaple | wowzers | 04:14 |
| sdake | but a couple blueprints had like 10-15 patches | 04:15 |
| diga | I will pick it on high priority, will complete it today EOD | 04:15 |
| sdake | like ansible and from source specifically | 04:15 |
| SamYaple | sdake: i submitted my review for ansible-heat :) I never submitted it last time, it was stuck in draft | 04:15 |
| coolsvap | diga, alright, i am around if needed :) | 04:15 |
| sdake | samyaple what do you mean | 04:15 |
| sdake | samyaple you mean you fixed my heat ansible patch? | 04:15 |
| SamYaple | i had written comments for your ansible heat work. i had not submitted them | 04:16 |
| sdake | oh | 04:16 |
| SamYaple | i havent taken ownership, no | 04:16 |
| diga | sure coolsvap | 04:16 |
| sdake | if you want to I wouldn't mind - I don't know how to do the domain creation | 04:16 |
| SamYaple | if you can answer the questions about the templates for the config files, ill do it | 04:16 |
| SamYaple | and i did reply on how to do the domain creation | 04:17 |
| sdake | the template works as is | 04:17 |
| SamYaple | i dont think it does, look at my comments. i think its a bit foobar | 04:17 |
| sdake | but it is missing code for domain creation so heat stack create fails | 04:17 |
| sdake | it isn't perfect | 04:17 |
| sdake | it is missing some things like the ur for wait conditions | 04:17 |
| SamYaple | specifically look here https://review.openstack.org/#/c/205528/1/ansible/roles/heat/templates/heat-engine.conf.j2 | 04:17 |
| sdake | looking | 04:17 |
| SamYaple | your stack_domain things are in thewrong section | 04:17 |
| SamYaple | i think | 04:17 |
| SamYaple | according to the docks you also need | 04:18 |
| SamYaple | heat_metadata_server_url = http://controller:8000 | 04:18 |
| SamYaple | heat_waitcondition_server_url = http://controller:8000/v1/waitcondition | 04:18 |
| SamYaple | and you are specifying rabbit in two places | 04:19 |
| SamYaple | didnt comments on those things, just saw them | 04:19 |
| SamYaple | anyway. if you can handle the configs since i dont know Heat well, i can handle the ansibleness | 04:19 |
| *** daneyon has joined #kolla | 04:20 | |
| sdake | samyaple yes i'll handle the configs | 04:22 |
| sdake | i posted a response | 04:22 |
| *** daneyon_ has joined #kolla | 04:22 | |
| *** daneyon has quit IRC | 04:22 | |
| sdake | samyaple if you would take over the review to make it ansible-correct I'll suggest what else needs to be fixed | 04:22 |
| sdake | i'd like heat to make l2 if possible | 04:23 |
| sdake | since its my main way of demoing the code | 04:23 |
| *** daneyon_ has quit IRC | 04:23 | |
| sdake | it doesn't have to be perfect - we can fix things up in l3 if needed | 04:23 |
| *** achanda has joined #kolla | 04:24 | |
| *** achanda has quit IRC | 04:26 | |
| sdake | samyaple yes those two url things need to be defined as you stated above | 04:27 |
| sdake | samyaple install form source isn't good progress its needs code review correct? | 04:29 |
| SamYaple | i think so, but i was going to check with pbourke | 04:30 |
| SamYaple | i dont know whats left to start and what has to be reviewed/merged | 04:30 |
| sdake | well I went ahead and changed it | 04:31 |
| sdake | if we are missing a few things we can address in l3 | 04:31 |
| SamYaple | alrighty | 04:32 |
| SamYaple | hey would you still be able to assist in testing that heat patch? | 04:32 |
| SamYaple | i do not into heat | 04:32 |
| sdake | absolutely | 04:32 |
| sdake | i will be happy to verify it | 04:33 |
| sdake | I have detailed domain knowledge there | 04:33 |
| SamYaple | cool. ill patch it up to what i think is working, you test it | 04:33 |
| SamYaple | we have a trove container right? | 04:33 |
| sdake | I dont think so | 04:34 |
| SamYaple | oh man. we gotta work on that | 04:34 |
| SamYaple | i like trove | 04:34 |
| sdake | i'd like folks to finish up the current containers in ansible | 04:34 |
| sdake | but if someone wnats to take on trove thatwfm ;a)a | 04:35 |
| sdake | wfm that is :) | 04:35 |
| sdake | we can tackle that in l3 | 04:35 |
| sdake | we have a weeks worth of reviews to complete in only a couple days | 04:35 |
| SamYaple | im going to try to stay out of the ansible code for the other containers | 04:35 |
| SamYaple | unless asked to finish them | 04:35 |
| sdake | sounds good | 04:35 |
| sdake | ya if someone asks for help i hope you can :) | 04:36 |
| SamYaple | i can throw them together quickly but want otehrs to learn the code base | 04:36 |
| sdake | like this heat one for example | 04:36 |
| SamYaple | yup | 04:36 |
| sdake | I did about 90% of the job | 04:36 |
| sdake | and know how to do new ones now | 04:36 |
| sdake | it took me about 3 hours to write the code | 04:36 |
| SamYaple | really clean right? | 04:36 |
| sdake | ya super fast to implement too | 04:36 |
| SamYaple | i thought so at least | 04:36 |
| sdake | it rocks | 04:36 |
| SamYaple | yea i was happy with the turn out | 04:36 |
| SamYaple | easy to read and modify | 04:37 |
| sdake | our deployment tool totally kicks ass | 04:37 |
| SamYaple | i would hope so we tlaked/argued about it with a dozen people for a month | 04:37 |
| sdake | one thing I think we need to do is hardcode our rpc-bakcned | 04:37 |
| sdake | each service has different defaults | 04:37 |
| SamYaple | expand? | 04:37 |
| sdake | heat defaults to pqid for exmple | 04:38 |
| sdake | qpid | 04:38 |
| SamYaple | eq | 04:38 |
| SamYaple | ew* | 04:38 |
| sdake | so I set rpc_backend to rabbit | 04:38 |
| SamYaple | yep | 04:38 |
| sdake | so I think we should make that one of those things we always set in every default config | 04:38 |
| SamYaple | i would like to support qpid as an option for people, but overall i feel rabbitmq works best | 04:38 |
| sdake | since we can't rely on defaults | 04:38 |
| sdake | agree | 04:39 |
| sdake | shoudl be configurable globlly at some point | 04:39 |
| sdake | not really a priority for l2 tho ;-) | 04:39 |
| SamYaple | i think we will tweak lots of things | 04:39 |
| *** bmace has quit IRC | 04:40 | |
| SamYaple | im going to rebase heat ontop of all of the haproxy/keepalived bits | 04:40 |
| sdake | soundsgood | 04:42 |
| SamYaple | what are your thoughts on the 'use_heat: True' config option? | 04:42 |
| sdake | sounds good | 04:42 |
| SamYaple | what name? | 04:42 |
| SamYaple | use_heat? deploy_heat? | 04:42 |
| SamYaple | just 'heat' | 04:42 |
| sdake | enable_heat? | 04:42 |
| SamYaple | i like it | 04:42 |
| SamYaple | i think defaulting to True is ok for Heat | 04:43 |
| sdake | agree | 04:43 |
| SamYaple | but i would like to keep services that depend on the disk defaulted to False (like cinder/ceph/swift) | 04:43 |
| *** Slower has joined #kolla | 04:43 | |
| sdake | might argue about ceph if we plan to use that for torage backend | 04:43 |
| SamYaple | if that becomes our default then sure | 04:44 |
| sdake | for glance/nova | 04:44 |
| SamYaple | but keep in mind it has a performance hit in a big way on alot of 1GB networks | 04:44 |
| sdake | agree re cinder/swift | 04:44 |
| SamYaple | itll be faster than the container storage though | 04:44 |
| SamYaple | but you do saturate | 04:44 |
| sdake | faster then container storage = better ;-) | 04:44 |
| SamYaple | ceph is also the only way to achieve live migration btw | 04:45 |
| sdake | we can tackle ceph in l3 i think | 04:45 |
| SamYaple | i would agree. ive got an old branch with some bit ported | 04:45 |
| SamYaple | its a bit of a big thing though since it requires config. We can default to a folder-only config, but if you want to use real disks/cache tiering/erasure coding youre going to have to do some work | 04:46 |
| sdake | we are basically out of time for l2 new work | 04:47 |
| SamYaple | yup | 04:47 |
| sdake | becaue of midcycle we just hae to get what we have done merged - review loe time :) | 04:47 |
| *** zhiwei has joined #kolla | 04:49 | |
| *** bmace has joined #kolla | 04:51 | |
| coolsvap | sdake, SamYaple one query, regarding https://review.openstack.org/#/c/204939/3/docker/common/openvswitch/ovs-vswitchd/config-internal.sh | 04:55 |
| sdake | shoot | 04:56 |
| SamYaple | shoot sdake not me | 04:56 |
| sdake | what is your Q | 04:56 |
| SamYaple | i want to live | 04:56 |
| coolsvap | the OVS_LOG_FILE variable, i think according to rhallisey we can remove it from check, does it make sense to add if for exec? | 04:56 |
| coolsvap | if we dont have value for OVS_LOG_FILE? | 04:56 |
| SamYaple | we cant remove logfile if we specify -vfile in exec | 04:57 |
| SamYaple | but we could drop that i guess | 04:57 |
| SamYaple | I dont know when we will ahve a big logging push for kolla | 04:57 |
| sdake | what does -vfile do | 04:57 |
| SamYaple | tells it what file to log too | 04:57 |
| sdake | we need logging in l3 - that will be a high priority item to tackle | 04:58 |
| sdake | i'll be leading that | 04:58 |
| SamYaple | without it it might go to stdout idk | 04:58 |
| coolsvap | i think it makes sense to have a log file | 04:58 |
| SamYaple | sdake: hopefully well be discussing that at the midcycle. ive seen some recent projects do that poorly | 04:58 |
| sdake | i think we are out of time for the midcycle but we can add to parking lot | 04:58 |
| SamYaple | then well need a spec of something | 04:59 |
| SamYaple | something more than one person agrees to. logging is tricky | 04:59 |
| coolsvap | i think it make sense to discuss it on review itself, but just wanted to get some views | 04:59 |
| SamYaple | coolsvap: i have no opinion at the moment | 04:59 |
| sdake | coolsvap lets not get bogged down in logging for l2 | 05:00 |
| SamYaple | well handle logging at a later time and at that point there will be only one answer | 05:00 |
| sdake | just do something that makes sense, and we can sort out logging properly in l3 with a groan spec or something less onerous | 05:00 |
| coolsvap | sdake, SamYaple, alright | 05:00 |
| SamYaple | sdake: i am personally ok with the idea of a bindmount to the host to dump all the log files so the host can do logging however they want | 05:01 |
| sdake | my only point is we dont need to get bogged down in standardizing how we do logging 2 days before our deadine ;-) | 05:01 |
| SamYaple | i dont like forcing syslog or any logging stack at all | 05:01 |
| coolsvap | i will make a defect for the log file discussion and remove the log file for now, make sense? | 05:01 |
| SamYaple | yea i dont think we are | 05:01 |
| sdake | ithink we want logging in our implementation coolsvap so I'd not remove loggin entirely | 05:02 |
| sdake | but what we do doesn't hve to be perfect | 05:02 |
| sdake | the logging especially with ansible is highly inconsistent | 05:06 |
| sdake | some things log to files, some things lo to std out, some long to std err | 05:06 |
| sdake | etc | 05:06 |
| *** achanda has joined #kolla | 05:07 | |
| SamYaple | agreed its not been a priority | 05:09 |
| SamYaple | but a conversaion and discussion needs to be had | 05:09 |
| sdake | tomorrow I will make the schedule for our midcycle | 05:10 |
| SamYaple | I would like to see/have the option for all the logs to end up on the host in /var/log/keystone /var/log/glance etc | 05:10 |
| sdake | I'll try to fit in a design session for auditing | 05:10 |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed stackforge/kolla: Make variables in OVS container configurable https://review.openstack.org/204939 | 05:15 |
| *** jasonsb has joined #kolla | 05:18 | |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed stackforge/kolla: Make variables in OVS container configurable https://review.openstack.org/204939 | 05:22 |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed stackforge/kolla: Update service helper scripts for variable changes https://review.openstack.org/203935 | 05:49 |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed stackforge/kolla: Update service helper scripts for variable changes https://review.openstack.org/203935 | 05:54 |
| sdake | jasonsb there? | 05:55 |
| jasonsb | yessir | 05:55 |
| jasonsb | sdake: yessir | 05:56 |
| *** unicell1 has joined #kolla | 05:57 | |
| *** unicell has quit IRC | 05:57 | |
| sdake | jasonsb are you making it to the midcycle this week? | 06:00 |
| *** dims has joined #kolla | 06:03 | |
| *** dims has quit IRC | 06:09 | |
| *** daneyon has joined #kolla | 06:12 | |
| *** daneyon has quit IRC | 06:17 | |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed stackforge/kolla: Implement kolla-compose script https://review.openstack.org/205915 | 06:33 |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed stackforge/kolla: Implement kolla-compose script https://review.openstack.org/205915 | 06:39 |
| openstackgerrit | Digambar proposed stackforge/kolla: Implement kolla-compose script https://review.openstack.org/205916 | 06:46 |
| openstackgerrit | Digambar proposed stackforge/kolla: Implement kolla-compose script https://review.openstack.org/205916 | 06:49 |
| *** achanda has quit IRC | 07:07 | |
| *** achanda has joined #kolla | 07:12 | |
| *** jmccarthy has quit IRC | 07:37 | |
| *** jmccarthy has joined #kolla | 07:38 | |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed stackforge/kolla: Make variables in OVS container configurable https://review.openstack.org/204939 | 07:48 |
| *** athomas has joined #kolla | 07:52 | |
| *** athomas has quit IRC | 07:53 | |
| *** gfidente has joined #kolla | 07:54 | |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed stackforge/kolla: Update service helper scripts for variable changes https://review.openstack.org/203935 | 07:56 |
| *** daneyon has joined #kolla | 08:01 | |
| *** athomas has joined #kolla | 08:05 | |
| *** daneyon has quit IRC | 08:06 | |
| *** achanda has quit IRC | 08:08 | |
| *** akwasnie has joined #kolla | 08:08 | |
| *** achanda has joined #kolla | 08:09 | |
| *** erkules_ is now known as erkules | 08:16 | |
| *** erkules has quit IRC | 08:16 | |
| *** erkules has joined #kolla | 08:16 | |
| *** coolsvap is now known as coolsvap|away | 08:21 | |
| *** coolsvap|away is now known as coolsvap | 08:29 | |
| *** achanda has quit IRC | 08:31 | |
| *** coolsvap is now known as coolsvap|away | 08:31 | |
| *** achanda has joined #kolla | 08:33 | |
| *** achanda has quit IRC | 08:34 | |
| *** coolsvap|away is now known as coolsvap | 08:43 | |
| *** dolpher has quit IRC | 08:54 | |
| *** diga has quit IRC | 08:58 | |
| *** diga has joined #kolla | 08:59 | |
| openstackgerrit | Sam Yaple proposed stackforge/kolla: Add Ansible support for Heat https://review.openstack.org/205528 | 08:59 |
| sdake | evening | 09:01 |
| SamYaple | evening | 09:01 |
| SamYaple | heat should be good | 09:01 |
| SamYaple | but idk you test | 09:01 |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed stackforge/kolla: Add build from source for gnocchi https://review.openstack.org/205952 | 09:03 |
| sdake | i will test - just woke up with old man syndrome in middle of night getting ready to go back to bed | 09:04 |
| SamYaple | yea i cant sleep right now either | 09:05 |
| SamYaple | but thats partly due to the fact i am at work | 09:05 |
| *** dims has joined #kolla | 09:06 | |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed stackforge/kolla: Update Gnocchi service Dockerfiles https://review.openstack.org/205956 | 09:06 |
| *** dims has quit IRC | 09:11 | |
| openstackgerrit | Paul Bourke proposed stackforge/kolla: Swift Containers https://review.openstack.org/202171 | 09:16 |
| *** bradjones has joined #kolla | 09:42 | |
| *** bradjones has quit IRC | 09:42 | |
| *** bradjones has joined #kolla | 09:42 | |
| *** daneyon has joined #kolla | 09:50 | |
| *** daneyon has quit IRC | 09:55 | |
| *** dims has joined #kolla | 10:04 | |
| *** dims_ has joined #kolla | 10:06 | |
| *** zhiwei has quit IRC | 10:07 | |
| openstackgerrit | Sam Yaple proposed stackforge/kolla: Removes hashing from merge_configs.py https://review.openstack.org/205990 | 10:09 |
| *** dims has quit IRC | 10:10 | |
| *** loth has joined #kolla | 10:18 | |
| *** loth1 has quit IRC | 10:21 | |
| *** vbel has quit IRC | 10:21 | |
| *** vbel has joined #kolla | 10:23 | |
| *** gfidente has quit IRC | 10:28 | |
| *** gfidente has joined #kolla | 10:32 | |
| *** gfidente has joined #kolla | 10:32 | |
| *** coolsvap is now known as coolsvap|away | 10:35 | |
| *** sdake has quit IRC | 10:42 | |
| openstackgerrit | Sam Yaple proposed stackforge/kolla: Properly use groups throught Ansible https://review.openstack.org/205999 | 10:43 |
| *** sdake has joined #kolla | 10:46 | |
| sdake | diga tahnks for knocking out that blueprint ;-) | 10:47 |
| sdake | samyaple is this a dup or already solved? https://blueprints.launchpad.net/kolla/+spec/add-ovs-vsctl-tool-to-compute-images | 10:47 |
| SamYaple | its solved | 10:48 |
| SamYaple | didnt realize someone wrote a spec for it | 10:48 |
| sdake | its all goodk I reoed the bluerpint | 10:50 |
| openstackgerrit | Merged stackforge/kolla: Update init-runonce https://review.openstack.org/205247 | 10:51 |
| sdake | removved | 10:51 |
| sdake | pbourke ping | 10:51 |
| pbourke | sdake: hi | 10:51 |
| sdake | is swift in "needs code review" ratherthen "goodprogress" now? | 10:52 |
| pbourke | sdake: I have a PS incoming to address harmw's comments, so still I guess goodprogress till that lands | 10:52 |
| pbourke | expect that soon though, just testing it atm | 10:53 |
| sdake | cool make sure to set to needs review when you finish that review | 10:53 |
| sdake | bedtime befoe flighttime | 10:54 |
| sdake | night all :) | 10:54 |
| openstackgerrit | Merged stackforge/kolla: Implement kolla-compose script https://review.openstack.org/205916 | 10:54 |
| SamYaple | night sdake | 10:54 |
| openstackgerrit | Paul Bourke proposed stackforge/kolla: Implement Swift Containers https://review.openstack.org/202171 | 11:00 |
| SamYaple | hey pbourke i know i have a ton of patches to review, but they are all actually important! https://review.openstack.org/#/c/205612/ | 11:03 |
| SamYaple | that is the base of all of them | 11:03 |
| SamYaple | all of them have a dep on that | 11:03 |
| SamYaple | how do you like that rabbitmq.config? | 11:06 |
| SamYaple | good right? | 11:06 |
| SamYaple | readable not complicated at all | 11:06 |
| pbourke | yeah I'm basically taking your word on that one :p | 11:10 |
| SamYaple | dude erlang is not fun | 11:11 |
| SamYaple | one of those paramaters gets pased to the kernel and requires a kernel way of specifying the ip address ({10,10,10,10}) | 11:12 |
| SamYaple | it doesnt even translate it | 11:12 |
| SamYaple | it only validates its correct kernel syntax | 11:12 |
| SamYaple | i never thought i would have to dig into a basic kernel function to config rabbit | 11:13 |
| SamYaple | pbourke: have you actually tested swift and its working without having real underlying disks? | 11:26 |
| openstackgerrit | Paul Bourke proposed stackforge/kolla: Implement Swift Containers https://review.openstack.org/202171 | 11:26 |
| SamYaple | or are oyu doing loopback stuff? | 11:26 |
| pbourke | SamYaple: using data vols | 11:27 |
| diga | sdake: welcome! | 11:27 |
| pbourke | or vol | 11:27 |
| SamYaple | pbourke: right but that means nothing to swift, it wil lstill complain about not having an actual backing disk | 11:27 |
| pbourke | SamYaple: mount_check: false | 11:27 |
| SamYaple | was curious how you got it to proceed without a real disk | 11:27 |
| SamYaple | ah | 11:27 |
| SamYaple | ok | 11:27 |
| pbourke | I have a note in there that its not for production etc. | 11:27 |
| pbourke | but whos going to run swift aio in production anyway :) | 11:28 |
| SamYaple | :/ | 11:28 |
| pbourke | I think ive addressed all harmw's comments so setting the bp to needs review | 11:28 |
| SamYaple | pbourke: you uploaded another patch and squashed my comments fyi | 11:28 |
| pbourke | oh | 11:28 |
| SamYaple | question about build-switch-ring.py while i have you | 11:29 |
| SamYaple | is this the only way to build the ring that we have built in? | 11:29 |
| pbourke | All guides I've seen have used a simple wrapper (in bash). I chose to use Python to make specifying the config data easier | 11:30 |
| SamYaple | but that requires (if i read this right) a seperate non swift json-type file, right? | 11:31 |
| pbourke | well, env variable | 11:31 |
| SamYaple | what do you mean? | 11:32 |
| pbourke | its the same as all the other services config-internal, just the env value is json | 11:32 |
| SamYaple | so you have to build a json file as the environment variable? | 11:33 |
| pbourke | yes | 11:33 |
| SamYaple | that is different than all our other variables | 11:33 |
| SamYaple | im not sure im ok with that change | 11:33 |
| SamYaple | what prevents normal variables for this? | 11:34 |
| pbourke | ok just to be sure you understand, it's not actually a file | 11:34 |
| pbourke | its just a json formatted string | 11:34 |
| SamYaple | i do | 11:34 |
| pbourke | I can't see how else to represent this data in a way that's easy to parse | 11:34 |
| pbourke | using env variables | 11:34 |
| diga | sdake: I need some help on this BP - https://blueprints.launchpad.net/kolla/+spec/kolla-ansible-script | 11:36 |
| SamYaple | im caution about introducing a new file format that isnt part of openstack | 11:36 |
| pbourke | we can have a chat about it, as always I'm open to improvement | 11:36 |
| SamYaple | yea im not offering a better way at the moment. im just thinking | 11:36 |
| SamYaple | i totally get why you went down this path | 11:36 |
| SamYaple | is this capable of all the options? zones, number of replics, replication-zone? | 11:38 |
| *** daneyon has joined #kolla | 11:39 | |
| pbourke | i can't tell you it supports every option as I don't know what those are. I wrote it based on my knowledge of Swift at the time | 11:40 |
| pbourke | looking now at least zones is missing | 11:40 |
| SamYaple | i realllllly dont like this json requirement. its going to prevent us from using this for config-external and i want to share as much code s possible | 11:42 |
| SamYaple | I would prefer simply a formated string | 11:42 |
| SamYaple | <zone>-<ip>:<port>/<dev> <weight> | 11:43 |
| SamYaple | then you could pass that directly | 11:43 |
| SamYaple | it would support replication zone by just updating the string | 11:43 |
| pbourke | config-external can have a playbook to run swift-ring-builder itself, or maybe use shade | 11:43 |
| SamYaple | no these arent apis | 11:43 |
| SamYaple | it has to be done in the container | 11:43 |
| pbourke | ah | 11:43 |
| pbourke | docker exec ? | 11:44 |
| *** daneyon has quit IRC | 11:44 | |
| SamYaple | yea but it wont be that straightforward | 11:45 |
| SamYaple | ill tell you what, ill make oyu a deal | 11:45 |
| SamYaple | im ok with this for now, but when config external comes around can this format be open to change? | 11:45 |
| SamYaple | I dont want the endusers to have to craft json strings | 11:45 |
| pbourke | of course, but I dont want to go with something sub-standard for now either :/ I'll have a think about it and see if it can be changed easily right now | 11:47 |
| pbourke | the above was written fairly quickly as a piece of the larger task at hand | 11:47 |
| SamYaple | I dont think this is sub-standard, but you ahve to realize that you are trying to add flexibility and still limiting options | 11:48 |
| SamYaple | missing replication zones and what not | 11:48 |
| SamYaple | we just need to determine what we are going to want to maintian longterm | 11:48 |
| SamYaple | what happens if swift changes the cli? now we have versioned files | 11:49 |
| SamYaple | its just not awesome is all | 11:49 |
| SamYaple | that being said, perhaps it is best to just extend this script since it is very clean | 11:49 |
| SamYaple | let me sleep on it | 11:50 |
| pbourke | sure, we don't have to decide right this moment | 11:51 |
| *** dwalsh has joined #kolla | 11:59 | |
| *** dwalsh has quit IRC | 12:06 | |
| *** inc0 has joined #kolla | 12:35 | |
| inc0 | good morning everyone | 12:37 |
| inc0 | sdake, up already? | 12:44 |
| *** jruano has joined #kolla | 12:50 | |
| *** dolpher has joined #kolla | 12:57 | |
| SamYaple | morning inc0 | 13:05 |
| SamYaple | sdake has been up and down all night | 13:05 |
| SamYaple | you in cali? | 13:05 |
| inc0 | yup | 13:05 |
| inc0 | observing a sunrise | 13:05 |
| inc0 | and reviewing your patches as promised | 13:05 |
| inc0 | so far 2 * -1 :P | 13:05 |
| SamYaple | lots of patches | 13:05 |
| SamYaple | eh i got a core +2 i dont care | 13:06 |
| inc0 | but you need second core | 13:06 |
| inc0 | and they see my comments already | 13:06 |
| SamYaple | im kidding | 13:06 |
| inc0 | I know | 13:07 |
| inc0 | but the problem is pretty big with single ip stuff | 13:07 |
| SamYaple | i disagree | 13:07 |
| SamYaple | and this has been tested working multinode | 13:07 |
| inc0 | I've just added second ip to an interface | 13:08 |
| inc0 | and ansible shows just a first one | 13:08 |
| inc0 | which means it will break with keepalived | 13:08 |
| SamYaple | yup. thats by design | 13:08 |
| SamYaple | no | 13:08 |
| inc0 | how so? | 13:08 |
| *** achanda has joined #kolla | 13:08 | |
| SamYaple | how would it? | 13:08 |
| SamYaple | rabbitmq doesnt bind to the keepalived vip | 13:08 |
| SamYaple | haproxy does | 13:08 |
| *** diogogmt has quit IRC | 13:09 | |
| inc0 | hmm...good point. However, are we 100% sure that keepalived will add second ip? not first one? | 13:09 |
| inc0 | always? | 13:10 |
| SamYaple | added ips always appears second on the interface | 13:10 |
| SamYaple | the ip index never changes. unless you remove the first ip it will stay the first ip | 13:10 |
| inc0 | ok then, works for me | 13:11 |
| *** diogogmt has joined #kolla | 13:11 | |
| SamYaple | yea i use this string everywhere. it has a huge weakness for multiple ips for sure but to me thats more a documentation issue than anything | 13:11 |
| SamYaple | there is simply no other way to do this dynamically | 13:11 |
| SamYaple | although should you come across one, we can implement it very very quickly | 13:12 |
| inc0 | if we'll use interface name there, I don't think there is a way | 13:13 |
| *** dolpher has quit IRC | 13:13 | |
| *** dolpher has joined #kolla | 13:13 | |
| SamYaple | assume we dont use the interface name, how do we bind to any specificaddress? | 13:13 |
| *** achanda has quit IRC | 13:13 | |
| *** britthouser has joined #kolla | 13:14 | |
| inc0 | unless we know it, it's not doable. I'm ok with interface tho | 13:14 |
| inc0 | although we should point out this somewhere | 13:14 |
| SamYaple | yea dynamically is the key here. we basically have to make the requirement we only support a single ip on the interface. keepalived should never break that by adding | 13:15 |
| *** britthouser has quit IRC | 13:15 | |
| *** diogogmt has quit IRC | 13:15 | |
| inc0 | that simply removing haproxy to make full active-passive isn't supporte | 13:15 |
| inc0 | d | 13:15 |
| SamYaple | inc0: it wont be | 13:15 |
| SamYaple | that breaks alot of things | 13:15 |
| *** britthouser has joined #kolla | 13:15 | |
| inc0 | yeah | 13:15 |
| SamYaple | there is no gaurantee that all the servies are running on one node | 13:15 |
| inc0 | you don't have to convince me;) just thinking loudly | 13:16 |
| *** britthou_ has quit IRC | 13:16 | |
| inc0 | well...metaphorically speaking | 13:16 |
| inc0 | as I don't emit voice really, just typing | 13:16 |
| SamYaple | im just responding to show i have been there and thought your thoughts. I think we are both aligned on this issue to be honest | 13:17 |
| vbel | good morning/evening! Quick question - in ansible mode where does it run keystone and user endpoints population in db? | 13:17 |
| vbel | trying to match config-internal.sh parts | 13:17 |
| SamYaple | vbel: the initial user and endpoint creation for keystone is dont in the start.sh of the keystone container. after that ansible has modules to creat ethe users and endpoints | 13:18 |
| *** dolpher has quit IRC | 13:18 | |
| inc0 | SamYaple, should we use shade for this stuff? | 13:19 |
| vbel | SamYaple: I see, thanks! I have met ansible/library/kolla_keystone_service.py | 13:19 |
| SamYaple | inc0: oh jeez dont get me started | 13:20 |
| vbel | :) shhh, no more ansible | 13:20 |
| SamYaple | we only use shade since the ansible 2.0 modules that we _will_ be using use shade | 13:20 |
| SamYaple | i couldnt import them here for licensing reasons so i did the dirty and wrote my own | 13:21 |
| inc0 | damn I wish I have job which would allow me to just sit and code stuff | 13:26 |
| *** daneyon has joined #kolla | 13:28 | |
| *** daneyon has quit IRC | 13:32 | |
| openstackgerrit | Paul Bourke proposed stackforge/kolla: Implement Swift Containers https://review.openstack.org/202171 | 13:36 |
| SamYaple | inc0: thats called a programmer | 13:38 |
| inc0 | my job title says that, thay lied to me | 13:39 |
| SamYaple | my job doesnt involve anything coding related! | 13:39 |
| inc0 | but doesn't involve meetings as well right? | 13:40 |
| SamYaple | only because im 3rd shift | 13:40 |
| SamYaple | thast the only reason | 13:40 |
| SamYaple | when i was on first lots of meetings | 13:40 |
| *** dwalsh has joined #kolla | 13:42 | |
| *** pbourke has quit IRC | 13:45 | |
| *** pbourke has joined #kolla | 13:45 | |
| *** coolsvap|away is now known as coolsvap | 13:47 | |
| *** gfidente has quit IRC | 13:53 | |
| *** gfidente has joined #kolla | 13:57 | |
| pbourke | SamYaple: wrt https://bugs.launchpad.net/kolla/+bug/1476293 (ansible must run as root), what operations specifically are we running that *have* to have root | 13:57 |
| openstack | Launchpad bug 1476293 in kolla "Ansible must be run as root for ansible deploy" [Undecided,Incomplete] | 13:57 |
| *** jtriley has joined #kolla | 14:02 | |
| SamYaple | pbourke: sysctl, directory creation, docker (unless youre added to the docker group), package installs | 14:09 |
| pbourke | k I'm going to close the above bug | 14:09 |
| pbourke | actually no I'll add sudo to the docs | 14:10 |
| SamYaple | neither of those are solutions | 14:10 |
| SamYaple | i dont think you understand the issue | 14:10 |
| *** prad has joined #kolla | 14:11 | |
| openstackgerrit | Sam Yaple proposed stackforge/kolla: Properly use groups throught Ansible https://review.openstack.org/205999 | 14:13 |
| openstackgerrit | Sam Yaple proposed stackforge/kolla: Rename role name to service name https://review.openstack.org/205798 | 14:13 |
| *** absubram has joined #kolla | 14:13 | |
| pbourke | the issue from my perspective is I run 'ansible-playbook ...' as shown in the docs and it falls over until I run it with sudo | 14:14 |
| *** jruano has quit IRC | 14:25 | |
| pbourke | harmw: ping | 14:26 |
| inc0 | allright, I'm going to get some breakfast | 14:30 |
| inc0 | cya guys | 14:30 |
| *** inc0 has quit IRC | 14:31 | |
| SamYaple | pbourke: while that is correct, there is a current patch that addresses the need to run ansible-playbooks with so in the review queue | 14:33 |
| SamYaple | so updating the docs and closing bugs is the wrong course of action here | 14:33 |
| pbourke | why didnt you say so | 14:35 |
| pbourke | Not seeing it? | 14:35 |
| bmace | morning gents. | 14:35 |
| pbourke | bmace: morning | 14:35 |
| *** jtriley has quit IRC | 14:42 | |
| pbourke | SamYaple: if you get a moment, I've tried to summarise the problems around the swift ring generation @ https://etherpad.openstack.org/p/swift-ring-generation | 15:02 |
| pbourke | would really appreciate if you can better help me understand the potential issues you see with the current approach | 15:03 |
| *** diogogmt has joined #kolla | 15:06 | |
| *** achanda has joined #kolla | 15:11 | |
| *** achanda has quit IRC | 15:15 | |
| sdake | see folks at the midcycle | 15:16 |
| sdake | i'm off on travel ;) | 15:16 |
| *** daneyon has joined #kolla | 15:16 | |
| jasonsb | sdake: yessir i'll be there | 15:17 |
| bmace | have a good trip sdake and talk to you tomorrow :) | 15:17 |
| *** jasonsb has quit IRC | 15:18 | |
| *** dims_ has quit IRC | 15:21 | |
| *** daneyon has quit IRC | 15:21 | |
| *** dims has joined #kolla | 15:21 | |
| *** sdake has quit IRC | 15:23 | |
| *** jtriley has joined #kolla | 15:24 | |
| *** jtriley has quit IRC | 15:29 | |
| *** kevsi has quit IRC | 15:31 | |
| *** jtriley has joined #kolla | 15:42 | |
| *** daneyon has joined #kolla | 15:48 | |
| *** daneyon_ has joined #kolla | 15:51 | |
| *** daneyon has quit IRC | 15:53 | |
| *** unicell has joined #kolla | 15:56 | |
| *** unicell1 has quit IRC | 15:58 | |
| *** jtriley has quit IRC | 16:02 | |
| *** jtriley has joined #kolla | 16:06 | |
| *** unicell has quit IRC | 16:21 | |
| *** akwasnie has quit IRC | 16:32 | |
| *** jasonsb has joined #kolla | 16:34 | |
| *** unicell has joined #kolla | 16:48 | |
| *** inc0 has joined #kolla | 16:48 | |
| inc0 | Slower, around? | 16:49 |
| *** achanda has joined #kolla | 16:58 | |
| *** coolsvap is now known as coolsvap|away | 17:06 | |
| *** jasonsb has quit IRC | 17:07 | |
| *** prad has quit IRC | 17:20 | |
| *** jasonsb has joined #kolla | 17:21 | |
| *** dwalsh has quit IRC | 17:21 | |
| *** inc0 has quit IRC | 17:32 | |
| *** sdake has joined #kolla | 17:33 | |
| *** dwalsh has joined #kolla | 17:36 | |
| *** prad has joined #kolla | 17:37 | |
| sdake | yar | 17:43 |
| *** sdake has quit IRC | 17:45 | |
| *** sdake has joined #kolla | 17:57 | |
| *** dwalsh has quit IRC | 17:58 | |
| harmw | pbourke: yes | 18:04 |
| *** sdake has quit IRC | 18:09 | |
| *** sdake has joined #kolla | 18:10 | |
| *** gfidente has quit IRC | 18:11 | |
| *** dwalsh has joined #kolla | 18:11 | |
| *** sdake has quit IRC | 18:23 | |
| SamYaple | pbourke: ill take a look! | 18:25 |
| SamYaple | pbourke: responded to ether | 18:42 |
| openstackgerrit | Michal Rostecki proposed stackforge/kolla: Add MariaDB image for Ubuntu https://review.openstack.org/205615 | 18:55 |
| SamYaple | reviews people! lets stop the queue | 18:58 |
| SamYaple | got like 10 patches that need to land for L2 | 18:58 |
| *** mfalatic has joined #kolla | 18:58 | |
| *** inc0 has joined #kolla | 19:25 | |
| bmace | my +1 power does you no good SamYaple and I expect many cores are in transit :/ | 19:26 |
| SamYaple | bmace: reviews are always good :) | 19:26 |
| SamYaple | i need all the eyes I can get! | 19:27 |
| bmace | i looked at a bunch this morning. a lot of ansible stuff which i am super happy is making progress :) | 19:27 |
| SamYaple | bmace: all the core ansible stuff is up for review and working AIO and multinode :) | 19:29 |
| SamYaple | 2m 35s deployment from start to finish! | 19:29 |
| bmace | SamYaple: i got the ansible all in one running on Friday. it was a pretty smooth process overall. i liked the output / speed a LOT more than docker compose | 19:30 |
| SamYaple | indeed | 19:30 |
| bmace | though i would prefer penguins over cows ;) | 19:30 |
| SamYaple | plus in the end your containers are named something reasonable! | 19:30 |
| SamYaple | docker restart keystone | 19:30 |
| SamYaple | not docker restart compose_keystone_1 | 19:30 |
| bmace | yeah, that is nice too. are you going to have 2 keystones on a single node? probably not | 19:30 |
| bmace | or your config is going to be really odd / complicated. | 19:31 |
| inc0 | SamYaple, have you seen any good comparason of OSAD vs Kolla? Because I keep answering the same questions | 19:33 |
| inc0 | especially now after Intel RAX agreement | 19:34 |
| * SamYaple remains silent | 19:34 | |
| *** achanda has quit IRC | 19:47 | |
| *** inc0 has quit IRC | 19:57 | |
| *** inc0 has joined #kolla | 19:57 | |
| *** jasonsb has quit IRC | 20:04 | |
| *** inc0 has quit IRC | 20:14 | |
| *** achanda has joined #kolla | 20:15 | |
| *** diogogmt_ has joined #kolla | 20:19 | |
| *** diogogmt has quit IRC | 20:22 | |
| *** diogogmt_ is now known as diogogmt | 20:22 | |
| *** jasonsb has joined #kolla | 20:32 | |
| *** inc0 has joined #kolla | 20:45 | |
| *** sdake has joined #kolla | 20:58 | |
| sdake | is samyaple around? | 20:59 |
| sdake | anyone in sjc want to get some dinner tonight? | 20:59 |
| inc0 | mememe | 21:04 |
| inc0 | are you already here sdake? | 21:04 |
| inc0 | I think Sam will land around 6pm | 21:05 |
| sdake | do you like indian food | 21:10 |
| sdake | we are having indian food for dinner | 21:10 |
| sdake | rather I am ,and your free to join me :) | 21:10 |
| sdake | ya im in sjc atm | 21:10 |
| sdake | sam said he gets in at 7:30 | 21:11 |
| sdake | this place has rockin indian food | 21:11 |
| sdake | its a little expensive, but I'll pick up the tab | 21:11 |
| inc0 | indian food is among my favorites;) which one is it? | 21:11 |
| sdake | its called amber | 21:12 |
| sdake | far none best indian restuarant in the us | 21:12 |
| sdake | its a chain but it still rocks ;) | 21:12 |
| inc0 | yeah, my friend from valley also said to go there | 21:12 |
| inc0 | and she's an indian so I believe her 100% | 21:12 |
| sdake | there are a couple other really good indian places but this place is absolutely fantastic | 21:12 |
| sdake | which hotel are you located? | 21:13 |
| inc0 | holiday inn silicon valley - 1350 first north street | 21:13 |
| sdake | oh i think your right near me | 21:13 |
| sdake | if your into cigars, I'm smoking cigars outside - welcome to walk over ;) | 21:13 |
| inc0 | I'm more into scotch, but I won't say no to a good cigar | 21:14 |
| sdake | these are ok, but they are cigarllos | 21:14 |
| sdake | 2$ a cigar | 21:14 |
| inc0 | only tabbaco I smoke really | 21:14 |
| sdake | ya me too | 21:14 |
| sdake | never smoked a cigarette actually | 21:14 |
| inc0 | where are you then? | 21:14 |
| inc0 | me neither, not a single one in my life;) | 21:14 |
| sdake | me either | 21:15 |
| sdake | wierd huh? :) | 21:15 |
| sdake | let me check the name of my hotel sec | 21:15 |
| *** achanda has quit IRC | 21:15 | |
| sdake | HOTEL | 21:16 |
| sdake | Name: Avatar Hotel Santa Clara Location: Santa Clara, CA | 21:16 |
| sdake | Address: 4200 Great America Parkway, Santa Clara, CA 95054 USA | 21:16 |
| sdake | when you get there, go to the front desk | 21:16 |
| sdake | tell him you are meeting someone staing at the hotel and he is in the smoking area | 21:16 |
| sdake | ask him to give you a map of where teh smoking section is | 21:17 |
| sdake | my cell is 602-292-1068 but I dont have it on me atm | 21:17 |
| inc0 | it's not that close, 5 miles | 21:17 |
| sdake | but I'll probably go back to the hotel to get it in a bit | 21:17 |
| sdake | oh 5 miles | 21:17 |
| sdake | groan | 21:17 |
| sdake | oh well was a good thought | 21:17 |
| sdake | if your into uber, its prob a 5-10$ ride :) | 21:17 |
| inc0 | well, I can go there anyway | 21:17 |
| inc0 | yeah I have to install it after all. But you need internet for that right? | 21:17 |
| sdake | yes you need internet - you probably dont have that i guess | 21:18 |
| inc0 | need to get US sim later | 21:18 |
| sdake | a cab ride will be more expensive, probably 20$ USD | 21:18 |
| inc0 | I can get a train and walk short distance | 21:18 |
| sdake | but we can leave here for dinner and i'll pick up dinner + cabs | 21:19 |
| sdake | I think this hotel has a bar as well | 21:19 |
| sdake | remember ask the front desk where to smoing section is | 21:19 |
| sdake | its not easy to find | 21:19 |
| inc0 | sure, I should be there in approx an hour | 21:20 |
| sdake | i he asks the guest name, tell him "Steven Dake" | 21:20 |
| sdake | sounds good I may go plug in or switch laptops | 21:20 |
| inc0 | if there will be any additional problems, Ill call you) | 21:20 |
| sdake | so if i'm not here immediately I will be in 5 mins | 21:20 |
| sdake | soudns good | 21:20 |
| sdake | see you then ;) | 21:20 |
| inc0 | cya | 21:20 |
| sdake | i'll just be here chain smoking :) | 21:21 |
| *** inc0 has quit IRC | 21:23 | |
| *** jtriley has quit IRC | 21:24 | |
| *** Windir has quit IRC | 21:26 | |
| *** Windir has joined #kolla | 21:34 | |
| *** ccrouch has quit IRC | 21:42 | |
| sdake | slower rhallisey around? | 21:58 |
| *** ChanServ sets mode: +o sdake | 22:00 | |
| *** sdake changes topic to "Kolla-Palooza midcycle July 28th, July 29th - More information available @ https://wiki.openstack.org/wiki/Sprints/KollaLibertySprint - Channel is Logged @ http://eavesdrop.openstack.org/irclogs/%23kolla/" | 22:03 | |
| *** unicell has quit IRC | 22:08 | |
| *** unicell has joined #kolla | 22:08 | |
| *** achanda has joined #kolla | 22:15 | |
| *** prad has quit IRC | 22:28 | |
| *** dwalsh has quit IRC | 22:33 | |
| *** sdake has quit IRC | 22:48 | |
| *** absubram has quit IRC | 23:11 | |
| *** diogogmt has quit IRC | 23:44 | |
| *** jtriley has joined #kolla | 23:51 | |
| *** absubram has joined #kolla | 23:53 | |
| *** absubram has quit IRC | 23:54 | |
| *** absubram has joined #kolla | 23:55 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!