| *** tzn has joined #kolla | 00:03 | |
| *** sdake has joined #kolla | 00:07 | |
| *** tzn has quit IRC | 00:08 | |
| *** sdake_ has joined #kolla | 00:15 | |
| *** sdake has quit IRC | 00:15 | |
| asalkeld | ooo, nice : https://github.com/openstack/kolla-mesos | 00:18 |
|---|---|---|
| sdake_ | SamYaple good thing we have tht gate ! | 00:22 |
| sdake_ | asalkeld you also hae a tracker | 00:23 |
| sdake_ | feel free to go wild | 00:23 |
| asalkeld | will soon | 00:23 |
| sdake_ | the tracker is completely distributed | 00:23 |
| sdake_ | anyone can do anything they like | 00:23 |
| sdake_ | i prefer CRs muck with blueprints but it isn't strictly rquired | 00:23 |
| asalkeld | sdake_: what do you mean by a "tracker"? | 00:23 |
| sdake_ | we are still sorting out a process for dealing with launchpad blueprints | 00:23 |
| sdake_ | tracker = launchpad | 00:23 |
| asalkeld | aaa, lp project - nice! | 00:24 |
| sdake_ | who is your second person for the kolla-mesos group asalkeld | 00:24 |
| asalkeld | sdake_: i think (Kirill Proskurin) will get the most dedicated time | 00:25 |
| sdake_ | asalkeld trying to get the olla-mesos-grou psetup but its sunday | 00:25 |
| sdake_ | so you and krill then? | 00:25 |
| asalkeld | yeah | 00:25 |
| sdake_ | seem slike michael is right on track to hit core reviewer in short order | 00:26 |
| asalkeld | nihilifer: is doing a great job too (in case you want #3) | 00:26 |
| sdake_ | but your wish is my command | 00:26 |
| sdake_ | just two | 00:26 |
| sdake_ | they can always just end up joining the cr team | 00:26 |
| sdake_ | who is nihilfer's name? | 00:26 |
| sdake_ | (ddon't recall his real name) | 00:26 |
| sdake_ | who/what | 00:27 |
| asalkeld | k, i think best to go by who has the most time to dedicate | 00:27 |
| asalkeld | Michal Rostecki | 00:27 |
| sdake_ | ok makes sense | 00:27 |
| sdake_ | it mya b that machiel just gts i the cr team directly | 00:27 |
| sdake_ | asalkel i know I said in teh past I dont have a hard cap on the number of core reviewers, however, I want to ensure we keep our team diversity tag | 00:28 |
| sdake_ | which means there is a hard cap on number of crs from one company | 00:28 |
| *** mimizone has joined #kolla | 00:28 | |
| asalkeld | sounds ok'ish | 00:29 |
| sdake_ | i thin kthe limit to th eore reiewers is a bad ide aon part of the tc | 00:29 |
| sdake_ | ut there is little i can do about it | 00:29 |
| asalkeld | though i don't really think core is very important (other than actually getting code in), you can be very productive without been core | 00:30 |
| sdake_ | right it just hep move code along | 00:30 |
| sdake_ | we do a good job of reviews typically | 00:30 |
| sdake_ | the more complex they r the slower they re | 00:30 |
| sdake_ | so keep em simple :) | 00:30 |
| sdake_ | last week - i wrote off ;) | 00:31 |
| sdake_ | but the ocming weeks should be soid | 00:31 |
| asalkeld | ok, i'll start posting some idea patches soon | 00:31 |
| sdake_ | no rush | 00:32 |
| sdake_ | repo is created - spec is neraly wrapaped | 00:32 |
| sdake_ | these r eitems i wanted to rush on | 00:32 |
| sdake_ | after tht whatever happens naturally wfm | 00:32 |
| asalkeld | I am quite happy with interest from others (harmw and Zhongyue) | 00:33 |
| asalkeld | i'd really like this to be widely used | 00:33 |
| asalkeld | one thing i am trying to think about is not diverging the config options between deployments | 00:34 |
| asalkeld | tho' i am not sure how that can be best achieved | 00:35 |
| *** mimizone has quit IRC | 00:41 | |
| *** mimizone has joined #kolla | 00:43 | |
| *** jasonsb has quit IRC | 00:47 | |
| sdake_ | which config options | 00:48 |
| sdake_ | yu mean globals.yml? | 00:48 |
| openstackgerrit | Steven Dake proposed openstack/kolla: Base image changes for drop-root https://review.openstack.org/242876 | 00:50 |
| openstackgerrit | Steven Dake proposed openstack/kolla: drop root for glance https://review.openstack.org/242877 | 00:50 |
| asalkeld | sdake_: yea | 00:56 |
| asalkeld | and the way a user can customise the config | 00:56 |
| asalkeld | /etc/kolla/<service>/<project> ... | 00:57 |
| asalkeld | /etc/kolla/config ... | 00:57 |
| *** sdake_ has quit IRC | 01:13 | |
| *** mimizone has quit IRC | 01:17 | |
| *** mimizone has joined #kolla | 01:21 | |
| *** mimizone has quit IRC | 01:33 | |
| *** mimizone has joined #kolla | 01:38 | |
| *** tfukushima has joined #kolla | 01:41 | |
| *** mimizone has quit IRC | 01:48 | |
| *** tzn has joined #kolla | 02:04 | |
| SamYaple | poor asalkeld | 02:05 |
| SamYaple | well get that spec merged at some point :/ | 02:05 |
| asalkeld | ha | 02:05 |
| SamYaple | i mean we already have the repo for it sooooooo | 02:05 |
| * asalkeld just hacking away happily | 02:05 | |
| *** tzn has quit IRC | 02:09 | |
| britthouser | asalkeld - I know very little about Mesos, but wondering how much "ownership" it takes of each container? Like, can it only manage containers its deployed? Or can it take over the management of containers previously deployed, with say, ansible? | 02:15 |
| asalkeld | britthouser: i'd assume it only know about ones it starts | 02:15 |
| asalkeld | (tho' i may be wrong) | 02:15 |
| britthouser | Ok. I figured that would be the case. | 02:16 |
| *** stvnoyes has quit IRC | 02:16 | |
| *** tzn has joined #kolla | 02:16 | |
| *** stvnoyes has joined #kolla | 02:16 | |
| *** mbound has quit IRC | 02:35 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Fix bootstrap for services https://review.openstack.org/242777 | 02:40 |
| *** alisonh has quit IRC | 02:51 | |
| *** alisonh has joined #kolla | 02:56 | |
| *** sdake has joined #kolla | 03:06 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Fix bootstrap for services https://review.openstack.org/242777 | 03:10 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Remove unused cinder-data container https://review.openstack.org/242907 | 03:10 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Move data container to start.yml https://review.openstack.org/242908 | 03:10 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Move data container to start.yml https://review.openstack.org/242908 | 03:17 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Remove unused cinder-data container https://review.openstack.org/242907 | 03:17 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Fix bootstrap for services https://review.openstack.org/242777 | 03:17 |
| openstackgerrit | Angus Salkeld proposed openstack/kolla-mesos: Add a simplistic bootstrap script to install mesos (AIO) https://review.openstack.org/242911 | 03:25 |
| openstackgerrit | Angus Salkeld proposed openstack/kolla-mesos: add config generation script and some examples https://review.openstack.org/242912 | 03:25 |
| SamYaple | asalkeld: so the idea is to use ansible to do the config generation for kolla-mesos? | 03:29 |
| SamYaple | jinja2* | 03:30 |
| SamYaple | not ansible | 03:30 |
| asalkeld | yeah just jinja2 | 03:34 |
| asalkeld | ansible hostvars is going to be a problem | 03:35 |
| asalkeld | (dynamic stuff) | 03:35 |
| *** mbound has joined #kolla | 03:35 | |
| SamYaple | asalkeld: does mesos have some way to detect that stuff? | 03:36 |
| asalkeld | i need to convert that to some kind of zookeeper variables that the containers can wait on and update the config as it is discovered | 03:36 |
| asalkeld | SamYaple: mesos afaik totally stays out of the configuration game | 03:36 |
| asalkeld | except for loadbalancers and service discovery | 03:36 |
| asalkeld | (which we can use) | 03:36 |
| SamYaple | right but service discovery comes with ip detection right? | 03:37 |
| SamYaple | im not sure | 03:37 |
| SamYaple | i should play with it | 03:37 |
| asalkeld | yeah, i need to look at it too | 03:37 |
| asalkeld | i have been trying to understand how ansible works | 03:37 |
| asalkeld | (newbie to ansible) | 03:37 |
| SamYaple | eh don't worry about ansible | 03:38 |
| SamYaple | i dont think youll have to know anything about it | 03:38 |
| asalkeld | yeah, just to see how the templating works | 03:38 |
| SamYaple | we just need a way to detect that ip information in mesos | 03:38 |
| SamYaple | oh ok | 03:38 |
| SamYaple | then run `ansible localhost -m setup` | 03:38 |
| SamYaple | it will spit out all the variables it knows about | 03:38 |
| asalkeld | now he tells me... | 03:39 |
| SamYaple | i told you before :) | 03:39 |
| asalkeld | if you did it slipped by | 03:39 |
| asalkeld | :-( | 03:39 |
| SamYaple | no problem of course | 03:39 |
| SamYaple | i may not have told you this actually | 03:40 |
| SamYaple | sorry :) | 03:40 |
| *** mbound has quit IRC | 03:40 | |
| asalkeld | np, it looks like mostly ip address/ports | 03:40 |
| asalkeld | tho' one biggie is figuring how to run the bootstrap stage | 03:41 |
| asalkeld | but, I'll get there | 03:41 |
| SamYaple | I am now the proud owner of a patch that merged into nova | 03:42 |
| SamYaple | its so rare. its like a unicorn | 03:42 |
| asalkeld | well done | 03:43 |
| * asalkeld taking a break, brb | 03:43 | |
| *** klint has joined #kolla | 03:49 | |
| *** dolpher has joined #kolla | 03:51 | |
| *** tfukushima has quit IRC | 03:55 | |
| SamYaple | asalkeld: this is actually a conversation I would like to have with you | 03:58 |
| SamYaple | is there _any_ outside orchastration of any kind? | 03:58 |
| SamYaple | or is it literally "start all containers, all contaienrs must be smart with teh logics" | 03:58 |
| *** _tzn has joined #kolla | 04:02 | |
| *** tzn has quit IRC | 04:02 | |
| sdake | SamYaple note, python kolla_set_configs is not a list in sudo but an app + args | 04:02 |
| sdake | so python kolla_set_configs is as secure as it gtes... | 04:02 |
| SamYaple | if that is the case the python piece is redundat since kolla_set_configs is executable | 04:03 |
| sdake | ya i dont know why it was there in the first place | 04:03 |
| SamYaple | sdake: i got a patch merged into nova | 04:03 |
| sdake | i thought for some good reason | 04:03 |
| sdake | time to bring donuts into the office ;) | 04:04 |
| SamYaple | sdake: its leftover cruft | 04:04 |
| sdake | yes just a fyi so you learn something ;) | 04:04 |
| sdake | i will remove ethe python (if tht works with sudo) | 04:04 |
| SamYaple | sdake: yea i couldn't remember if it was space or command seperated lists | 04:05 |
| SamYaple | I thought wrong | 04:05 |
| *** tfukushima has joined #kolla | 04:06 | |
| SamYaple | its comma seperated | 04:06 |
| SamYaple | https://github.com/SamYaple/yaodu/blob/master/ansible/roles/docker_build/templates/ubuntu/neutron/sudoers.j2 | 04:06 |
| *** dims has quit IRC | 04:06 | |
| sdake | atm all our containers require tty which i htink is incorrect | 04:08 |
| *** _tzn has quit IRC | 04:09 | |
| SamYaple | were keeping the tty around for other reasons anyway | 04:09 |
| SamYaple | it makes working with the contaienrs so much more pleasent | 04:09 |
| SamYaple | but we _could_ turn off tty for the process itself | 04:09 |
| SamYaple | or at the beginning of the shell | 04:09 |
| SamYaple | the issue is if the contaienr isn't created with a tty it can't be added after the fact | 04:10 |
| SamYaple | then when you have to docker exec into the contaienr is all broken | 04:10 |
| SamYaple | you can't use stuff like less | 04:10 |
| SamYaple | or mysql | 04:10 |
| sdake | so you recommendd that !requiretty or what | 04:10 |
| sdake | ya i get the tty thing | 04:10 |
| SamYaple | oh it doesn't matter | 04:11 |
| SamYaple | you can use either | 04:11 |
| SamYaple | you want to add it or leave it off, it really doesnt matter | 04:11 |
| sdake | i'll leave my sudoers file as is i think | 04:11 |
| sdake | minus trying to remove python | 04:12 |
| SamYaple | you still need to change that second ALL to (root) | 04:12 |
| sdake | i'll see if that works or not | 04:12 |
| sdake | what does that do? | 04:12 |
| SamYaple | prevents the user from sudo'ing to a user other than what is specified | 04:12 |
| SamYaple | in this case only to root | 04:12 |
| sdake | cool | 04:12 |
| sdake | i'm going to stick with kolla | 04:12 |
| sdake | i don't like configs | 04:12 |
| SamYaple | sure i dont care, openstack just didnt make sense for a user in the ceph or mariadb contaienrs | 04:13 |
| sdake | right | 04:13 |
| sdake | implementation looks pretty good tho no? | 04:13 |
| sdake | interestingly ti is the sipliest one so far ;-) | 04:14 |
| SamYaple | I say yes, though I think we should have a helper file to append commands too | 04:15 |
| SamYaple | maybe | 04:15 |
| SamYaple | no that seems fine | 04:15 |
| SamYaple | i was thinking for the extend starts | 04:15 |
| SamYaple | but we can add special sudoers for the specific user | 04:16 |
| sdake | you mean for nova and neutron that erqurei root sudo? | 04:16 |
| SamYaple | no those are handled by thier rootwrap things | 04:16 |
| SamYaple | we can just move those into place | 04:16 |
| SamYaple | i was thinking of like ceph | 04:16 |
| SamYaple | with parted and sdisk comannds | 04:16 |
| sdake | maybe you can sisgn up for ceeph ;) | 04:17 |
| * sdake runs | 04:17 | |
| SamYaple | yea no problem | 04:17 |
| SamYaple | hye sdake, not realyl related | 04:17 |
| sdake | wound me | 04:17 |
| SamYaple | but lets have a talk about harpxyo really quick | 04:17 |
| sdake | i just woke up | 04:18 |
| SamYaple | no problem its simple | 04:18 |
| sdake | so don't expect rocket scientistry atm | 04:18 |
| SamYaple | right now with COPY_ONCE we cannot hot reload haproxy config | 04:18 |
| SamYaple | is this the behaviour you want? | 04:18 |
| sdake | that is a complex question | 04:18 |
| sdake | the reason we have copy_once is to preserve immutability of the container config from start through exit | 04:19 |
| SamYaple | its a yes no answer :P | 04:19 |
| sdake | so if we caan presserve that then its an easy yes | 04:19 |
| sdake | if not, it may be more a quetion of pragmatistim | 04:19 |
| sdake | i mean changing it above | 04:20 |
| SamYaple | well here is the issue, if you add another node (mariadb, nova-api,neutron-server, etc) this requires a _very_ distruptive container top start | 04:20 |
| * sdake goes backt o beed ;) | 04:20 | |
| SamYaple | stop/start* | 04:20 |
| sdake | if the restart is rolling its not disruptive right? | 04:22 |
| SamYaple | it can't be rolling | 04:22 |
| SamYaple | you have to restart haproxy | 04:22 |
| SamYaple | restarting haproxy is very distruptive | 04:22 |
| sdake | but the other tw onodes provide service? | 04:22 |
| sdake | can't you drain haproxy out | 04:22 |
| sdake | then restat it? | 04:22 |
| SamYaple | sure. but then your cluster is dead for x ammount of time | 04:23 |
| sdake | what is the hit to immutability? | 04:23 |
| SamYaple | the haproxy container may not have the same config as when it was first launched | 04:23 |
| SamYaple | thats teh hit | 04:23 |
| SamYaple | but to be realistic here, haproxy doesnt ever restart... | 04:23 |
| SamYaple | if it did everything would be dead anyway | 04:23 |
| sdake | one reason we want immutability is so the state of the system stays syncrhonized | 04:24 |
| sdake | this sounds like it could lead to desynced state | 04:24 |
| SamYaple | that doesnt make sense | 04:24 |
| sdake | just woke up | 04:25 |
| sdake | warned you :) | 04:25 |
| SamYaple | anyway. think about it because no one will ever use COPY_ONCE if they are aware of these facts | 04:25 |
| SamYaple | and I will maek them aware | 04:25 |
| SamYaple | thats an operators worst nightmare | 04:25 |
| sdake | is it not possible to take one haproxy out of service (out of the load blancers) then drain its queue, then resstart it? | 04:25 |
| SamYaple | only one haproxy is ever processing things anyway | 04:26 |
| sdake | oh wasnt' totally ware of that | 04:26 |
| SamYaple | you were | 04:26 |
| sdake | or maybe i was | 04:26 |
| sdake | just woke up like i said | 04:26 |
| SamYaple | weve had the conversation before | 04:26 |
| SamYaple | sure | 04:26 |
| sdake | struggle when i first wake up :) | 04:26 |
| SamYaple | i woke up like an hour ago | 04:27 |
| sdake | i woke up about 10 mins ago | 04:27 |
| *** tzn has joined #kolla | 04:27 | |
| SamYaple | youve been talking to me for 30 minutes | 04:27 |
| sdake | so lets say we take the hit to immutability | 04:27 |
| sdake | what is the impact? | 04:27 |
| sdake | again reason we want immutability is to make ure system state stays synchrnized | 04:28 |
| sdake | does this result in desynced behavior? | 04:28 |
| SamYaple | expand on that | 04:28 |
| SamYaple | i dont know what you are saying with sync-state | 04:28 |
| sdake | for example one nova-compute could end with qemu and one with libvirt | 04:28 |
| sdake | rhater kvm | 04:28 |
| sdake | when all should be one or the other | 04:28 |
| sdake | definately the running state shoudl match teh config for compute | 04:28 |
| SamYaple | well that isnt assured with COPY_ONCE | 04:28 |
| sdake | roger i understand | 04:29 |
| SamYaple | the entire environment isn't immutable, the container is | 04:29 |
| sdake | ya understood | 04:29 |
| sdake | but kvm if configed wil be running | 04:29 |
| SamYaple | the haproxy config deals with teh entire environemnt which is not immutable | 04:29 |
| sdake | doesn't sound like an exception to me | 04:29 |
| SamYaple | even with COPY_ONCE you can have two hosts in the environemtn, one iwth KVM, one with QEMU | 04:30 |
| sdake | imo run wild | 04:30 |
| sdake | yes i get that :) | 04:30 |
| sdake | someone invents compltely immutable infrstructure = docker 3.0 :) | 04:31 |
| SamYaple | the truth of the matter is, COPY_ALWAYS is the only way we are going to be able to do hot reloads in the future | 04:32 |
| SamYaple | and thats the future of openstack | 04:32 |
| asalkeld | +1 | 04:33 |
| SamYaple | so this idealistic COPY_ONCE thing will be more and more a pain in the ass as the default | 04:33 |
| SamYaple | unless we kinda change what it means | 04:33 |
| sdake | so my recommendation would be sort out the pragmatic approach to dealing with haproxy | 04:33 |
| sdake | and do that | 04:33 |
| sdake | i think we can change what it means | 04:33 |
| *** bmace has quit IRC | 04:33 | |
| SamYaple | sure and that means ignore COPY_ONCE for haproxy | 04:33 |
| sdake | but i will have to think more about that | 04:33 |
| SamYaple | i understand | 04:33 |
| SamYaple | its why i bring it up now | 04:34 |
| *** bmace has joined #kolla | 04:34 | |
| asalkeld | SamYaple: as far as the bootstrap stuff goes... can we run all the bootstrap tasks before any services? | 04:34 |
| SamYaple | asalkeld: in the case of openstack stuff, i think yes | 04:34 |
| openstackgerrit | Diogo Monteiro proposed openstack/kolla: Add support for ubuntu when deploying kolla with Vagrant https://review.openstack.org/242923 | 04:34 |
| SamYaple | in the case of mariadb glaera cluster not a chance | 04:34 |
| sdake | asalkeld yup | 04:34 |
| asalkeld | ok, that might make it easier | 04:34 |
| SamYaple | with rabbitmq clusterer rabbitmq doesnt really have a "bootstrap" phase | 04:35 |
| sdake | ya mariadb and galera snowflakes of the unicorn variety | 04:35 |
| asalkeld | basially db up, then all bootstrap, then all services | 04:35 |
| SamYaple | asalkeld: kind of | 04:35 |
| SamYaple | you need keystone up to register the services | 04:35 |
| SamYaple | (endponts) | 04:35 |
| SamYaple | thats part of our bootstrap stuff i think | 04:36 |
| asalkeld | ok | 04:36 |
| sdake | yup it is | 04:36 |
| *** unicell1 has joined #kolla | 04:36 | |
| sdake | so a possible new definition of copy_once is if a hot reload happens container cnofigchanges | 04:36 |
| sdake | but that is the only case | 04:36 |
| sdake | otherwise it does *not* change | 04:37 |
| SamYaple | that would be anytime the playbooks run | 04:37 |
| sdake | even if admin hacks around on the files | 04:37 |
| *** unicell has quit IRC | 04:37 | |
| SamYaple | if it lays a new config, it hot reloads | 04:37 |
| SamYaple | there would be no time when it doesnt run | 04:37 |
| SamYaple | its essentially gettting rid of the config strategy | 04:37 |
| asalkeld | hopefully we can just add a zookeeper "driver" to oslo.config so no files needed in /etc/ | 04:37 |
| SamYaple | asalkeld: that would be sweet | 04:37 |
| sdake | i am not hot on adding a java dep for base olla | 04:38 |
| sdake | etcd i could live with tho ;) | 04:38 |
| asalkeld | sdake: kazoo is a python client to zookeeper | 04:38 |
| SamYaple | sdake: the two options as I see them. Get right of COPY_ONCE/COPY_ALWAYS. make it all copy_always. Keep COPY_ONCE with the way it works not but make COPY_ALWAYS default | 04:38 |
| asalkeld | (but sure the server is java) | 04:39 |
| sdake | well at mwe dont have to worry abut hot reload | 04:39 |
| sdake | asalkeld i am talking abut the server | 04:39 |
| asalkeld | but openstack is now mostly requiring zookeeper anyway | 04:39 |
| SamYaple | sdake: some openstack stuff does hotreload already | 04:39 |
| asalkeld | dlm | 04:39 |
| SamYaple | asalkeld: +1 | 04:39 |
| SamYaple | that will help alot actually | 04:39 |
| sdake | i dobut peope will actuallly use dlm ;) | 04:39 |
| SamYaple | sdake: itll be a requirement for ha cinder | 04:40 |
| asalkeld | sdake: hahaha | 04:40 |
| asalkeld | floodgates open ... | 04:40 |
| SamYaple | https://review.openstack.org/#/c/232599/ | 04:40 |
| sdake | nobody used it in ha cluster stack | 04:40 |
| SamYaple | sdake: see last comments | 04:40 |
| SamYaple | new HA cinder will use it | 04:40 |
| sdake | dlm doesn't and never will scale | 04:41 |
| sdake | just based on experience not book reading :) | 04:41 |
| SamYaple | in this case it doesn't need to scale | 04:41 |
| SamYaple | but either way the point is it will be a requirement | 04:42 |
| asalkeld | new option: what dlm do you want | 04:42 |
| asalkeld | well will have to install something | 04:42 |
| SamYaple | Tooz is the oslo DLM project | 04:43 |
| SamYaple | it is supposed to have different backends | 04:43 |
| SamYaple | including memcached | 04:43 |
| SamYaple | and mysql | 04:44 |
| SamYaple | and obivously ZK | 04:44 |
| sdake | off to eat bbl | 04:45 |
| asalkeld | o/ | 04:47 |
| diogogmt | is there a problem if i checkout master on kolla and then build the openstack services from source using the juno or liberty release? | 04:52 |
| diogogmt | would kolla master support only the master branches of the other openstack services? or would it be backwards compatible? | 04:53 |
| sdake | so sam re hot reload | 04:54 |
| sdake | didyou grasp the concept of only change the config in a hot reload situation | 04:55 |
| sdake | (not when the admin mucks arund with files on teh system) | 04:55 |
| sdake | so hot reload = responsible for copy operation rather then container start | 04:55 |
| sdake | that preserves spirit of immutability | 04:55 |
| sdake | rather then/in addition to | 04:55 |
| SamYaple | yea thats fair | 04:57 |
| sdake | asalkeld the problem we faced with jamming all the stuff at once in a start | 04:57 |
| SamYaple | i didnt pick up on that subtly | 04:57 |
| sdake | was that keystone returns its ready before its ready | 04:57 |
| sdake | iirc we had some hacky sleep in there or something | 04:58 |
| asalkeld | sdake: ok thanks | 04:58 |
| asalkeld | at worst we can have a zookeeper variable to wait on | 04:58 |
| *** britthouser has quit IRC | 04:58 | |
| openstackgerrit | Angus Salkeld proposed openstack/kolla-mesos: Add a simplistic bootstrap script to install mesos (AIO) https://review.openstack.org/242911 | 04:59 |
| openstackgerrit | Angus Salkeld proposed openstack/kolla-mesos: add config generation script and some examples https://review.openstack.org/242912 | 04:59 |
| openstackgerrit | Angus Salkeld proposed openstack/kolla-mesos: Remove validate-all-maintainers.sh https://review.openstack.org/242927 | 04:59 |
| sdake | hey britt | 04:59 |
| sdake | asalkeld can you add your other guy here https://review.openstack.org/#/admin/groups/1168,members | 05:05 |
| asalkeld | sdake: sure | 05:06 |
| asalkeld | sdake: https://www.mail-archive.com/openstack-dev@lists.openstack.org/msg67560.html | 05:06 |
| asalkeld | dlm stuff | 05:06 |
| sdake | ya i saw it already | 05:06 |
| asalkeld | cos' i know you love that stuff;) | 05:06 |
| sdake | i really dont like java ;) | 05:06 |
| sdake | i have dave nightmares of dlm | 05:06 |
| *** britthouser has joined #kolla | 05:07 | |
| sdake | that and posix locking in a distributed fashion | 05:07 |
| * sdake groans | 05:07 | |
| sdake | boy dave tetsted my patience | 05:07 |
| *** tzn has quit IRC | 05:11 | |
| openstackgerrit | Angus Salkeld proposed openstack/kolla: Fix validate-all-maintainer.sh to look for Dockerfile.j2 https://review.openstack.org/242933 | 05:20 |
| asalkeld | sdake: ^ | 05:20 |
| sdake | thanks angus | 05:20 |
| asalkeld | seems some were not right | 05:20 |
| sdake | ya thats why we had the validation | 05:21 |
| sdake | apparently our gating was busted | 05:21 |
| SamYaple | im good to +2 that | 05:21 |
| SamYaple | but ive been thinking for a while that should be a variable set in build.conf | 05:21 |
| SamYaple | if other people build these images, its not kolla maintaining them | 05:22 |
| sdake | ya fil ea blueprint imo | 05:22 |
| sdake | someone new can pick it up pretty easily | 05:22 |
| SamYaple | esh blueprint? kinda heavy for that | 05:22 |
| SamYaple | bug labeled feature request i would think | 05:22 |
| sdake | for feature request bugs, supposed to use "wishlist" bugs | 05:23 |
| sdake | you can do that as well if you like | 05:23 |
| SamYaple | ok ill do wishlist | 05:23 |
| asalkeld | daddy taxi about to head off ... | 05:25 |
| sdake | SamYaple note we need to keep the python in the sudoers file or no bueno | 05:25 |
| SamYaple | sdake: no you have to take it out | 05:26 |
| SamYaple | and then not call python | 05:26 |
| sdake | oh maymbe not | 05:26 |
| sdake | nah that isn't hte problem - maybe -a isnt working correctly - still investigating | 05:27 |
| SamYaple | you can update start.sh with `kolla_set_configs` | 05:27 |
| sdake | already did that | 05:27 |
| sdake | my sudoers file is wrong - hass %openstack insteado f %kolla | 05:27 |
| sdake | need to hit teh rack1 | 05:27 |
| SamYaple | ok but `sudo -E python /usr/local/bin/kolla_set_configs` should be updated to `sudo -E kolla_set_configs` | 05:28 |
| sdake | i've already done that | 05:28 |
| SamYaple | since its in the path, we no longer need call the full path | 05:28 |
| SamYaple | ok | 05:28 |
| sdake | oh path | 05:29 |
| sdake | ya haven't done that | 05:29 |
| sdake | i don't have a secure path setting so I dont think that will work | 05:29 |
| SamYaple | hey sdake i was talking to cloudnull and we may be able to compress all the config gen stuff to a single command instead of the two step process we have right now | 05:29 |
| SamYaple | if you recall the reason i didnt go for the action module original was it wasnt compatible with ansible 2.0 | 05:30 |
| SamYaple | but he has a PR up to push something we can use | 05:30 |
| sdake | i'd rather not lock us into ansible 2.0 | 05:30 |
| sdake | until it is absolutely necessary | 05:30 |
| SamYaple | as ive already been over, it is | 05:30 |
| SamYaple | its necessary the momemnt it lands | 05:31 |
| sdake | i can run current code on onn 2.0 | 05:31 |
| sdake | the moment which lands | 05:31 |
| SamYaple | ansible 2.0 | 05:31 |
| SamYaple | there is no more bug fixes for 1.9 | 05:31 |
| SamYaple | and we need new features | 05:31 |
| cloudnull | ++ it'd be a good thing to work on that together if you guys are interested in it: https://github.com/ansible/ansible/pull/12555 | 05:31 |
| sdake | where is /usr/local/bin in the path samyaple? | 05:32 |
| sdake | (where set) | 05:32 |
| cloudnull | in the mean time you guys can work off of our 1.9 lib if you decide to go that route https://github.com/openstack/openstack-ansible/blob/master/playbooks/plugins/actions/config_template.py | 05:32 |
| SamYaple | cloudnull: i won't be backporting this work, so its strictly a looking forward action | 05:33 |
| sdake | liberty kolla is definately pinned to ansible 1.9 | 05:33 |
| cloudnull | ah. fair point. | 05:33 |
| SamYaple | ive been waiting for ansible 2.0 for quite a while for quite a few things :/ | 05:33 |
| SamYaple | like blocks | 05:33 |
| SamYaple | i wants blocks so bad | 05:33 |
| cloudnull | ++ blocks | 05:33 |
| cloudnull | itll solve a bunch of crusty bits we've been carrying | 05:34 |
| SamYaple | yup | 05:34 |
| SamYaple | cloudnull: can you pass multiple source files into the config_template stuff? | 05:34 |
| cloudnull | no, its a single source file. | 05:35 |
| *** tzn has joined #kolla | 05:35 | |
| SamYaple | ok. how would you feel about extending that? | 05:36 |
| SamYaple | list-of-source-files style | 05:36 |
| cloudnull | it essentially follows the same scheme as the template module with the ability to override items. | 05:36 |
| cloudnull | we could | 05:36 |
| SamYaple | well its not without precedent | 05:37 |
| SamYaple | there is another module that accepts either a single file or list | 05:37 |
| cloudnull | that could work checking the object type | 05:37 |
| SamYaple | yea thats how they do it | 05:37 |
| SamYaple | i cant remember whic one | 05:37 |
| cloudnull | seems sensible | 05:37 |
| *** achanda has quit IRC | 05:38 | |
| cloudnull | would you want the module to merge the listed sources and then apply the overrides? or render individually? | 05:38 |
| SamYaple | thats how we do it currently | 05:38 |
| SamYaple | the first one | 05:38 |
| SamYaple | althought i dont know that the end result would be affected... | 05:39 |
| *** tzn has quit IRC | 05:39 | |
| SamYaple | well not what we are talking about it. how does the it work with for loops and what not? | 05:39 |
| SamYaple | it can't merge them | 05:39 |
| SamYaple | it would have to render it first, would it not? | 05:40 |
| cloudnull | it renders the jinja, applys the overrides, renders jinja again incase there was jinja in the overrides and then writes the target | 05:40 |
| SamYaple | yea that should work | 05:41 |
| SamYaple | render-apply loop | 05:41 |
| SamYaple | then final render | 05:41 |
| *** tzn has joined #kolla | 05:41 | |
| cloudnull | seems like we could apply a simple input check to the config tempalte plugin > run the meat from https://github.com/openstack/kolla/blob/master/ansible/library/merge_configs.py to render the unified config > render jinja > apply overrides > render jinja > write target | 05:46 |
| cloudnull | seems like a very sensible change to the module , if your keen on working on it together | 05:47 |
| SamYaple | of course | 05:47 |
| SamYaple | opensource is opensource | 05:47 |
| *** alisonh has quit IRC | 05:47 | |
| cloudnull | that'd be great | 05:48 |
| SamYaple | so combine the bits of merge_config.py that do the multi-file stuff and run them over the functions in config_template if source is detected as type list | 05:48 |
| cloudnull | that sounds like it'd work well | 05:48 |
| SamYaple | let me hammer out a gist real quick | 05:49 |
| SamYaple | you know cloudnull, I don't think there is any reason why the config type couldn't be detected automagically | 05:56 |
| SamYaple | ini wont validate as json or yaml | 05:56 |
| SamYaple | json wont validate as yaml | 05:56 |
| cloudnull | maybe detected unless stated ? | 05:57 |
| SamYaple | or accept any valid ini, json, or yaml file and specify the output type | 05:58 |
| SamYaple | well no | 05:58 |
| SamYaple | that wouldn't work | 05:58 |
| SamYaple | ini isn't going to allow nest dict | 05:58 |
| SamYaple | so on https://github.com/cloudnull/ansible/blob/1c5993cbb06ba6839688728e2daae7b7d9f9dc04/lib/ansible/plugins/action/config_template.py#L314 | 06:02 |
| SamYaple | does this code work _without_ config_overrides? | 06:02 |
| SamYaple | just as a normal template module? | 06:03 |
| SamYaple | yea i suppose it would | 06:04 |
| cloudnull | sorry | 06:04 |
| * cloudnull watching homeland | 06:04 | |
| cloudnull | yes the module should function the same as the regular template module if no overrides are passed | 06:05 |
| SamYaple | i havent watched it since end of season 3 | 06:05 |
| SamYaple | for obvious reasons... | 06:05 |
| cloudnull | good show, season 4 was ok, season 5 seems to be getting off to a good start | 06:06 |
| SamYaple | i might watch it | 06:09 |
| *** alisonh has joined #kolla | 06:10 | |
| cloudnull | im out. | 06:13 |
| SamYaple | ill ping you with a gist later. see ya cloudnull | 06:13 |
| cloudnull | SamYaple: if you get a chance to work on that can you send it up as a pr to my feature branch so I can get it merged which should update the PR to ansible proper | 06:14 |
| SamYaple | yea im testing now | 06:14 |
| cloudnull | sweet. | 06:14 |
| SamYaple | tried to keep changes minimal | 06:14 |
| SamYaple | shouldn't affect existing code | 06:14 |
| SamYaple | (existing tasks i mean) | 06:14 |
| cloudnull | ill look for that in the AM. ttyl | 06:14 |
| SamYaple | bai | 06:15 |
| *** jasonsb has joined #kolla | 06:33 | |
| *** achanda has joined #kolla | 06:38 | |
| *** mimizone has joined #kolla | 06:45 | |
| *** achanda has quit IRC | 06:46 | |
| *** mimizone has quit IRC | 06:51 | |
| openstackgerrit | Steven Dake proposed openstack/kolla: Base image changes for drop-root https://review.openstack.org/242876 | 06:55 |
| openstackgerrit | Steven Dake proposed openstack/kolla: drop root for glance https://review.openstack.org/242877 | 06:55 |
| *** alisonh has quit IRC | 07:08 | |
| *** alisonh has joined #kolla | 07:14 | |
| sdake | SamYaple if your about can you ahve a look at those two patches | 07:19 |
| *** kjelly has quit IRC | 07:27 | |
| sdake | mandre about? | 07:32 |
| sdake | asalkeld lets try to get a vote on the spec now | 07:32 |
| sdake | your spec that is | 07:33 |
| *** sdake has quit IRC | 07:37 | |
| *** sdake has joined #kolla | 07:38 | |
| SamYaple | sdake: since when did we start -1 for needed backport tag? | 07:44 |
| SamYaple | i dont htink we should be doing that | 07:44 |
| sdake | how else will people know to backport it? | 07:45 |
| sdake | -1 = futher work needed | 07:45 |
| SamYaple | well i was using it for when tehre wasnt a bug attached | 07:45 |
| SamYaple | if it has a bug then it should be set in launchpasd | 07:45 |
| SamYaple | like everyone else | 07:45 |
| sdake | hrm... i guess that is ok although more difficult to track via git | 07:46 |
| sdake | two sources of truth = annoing | 07:46 |
| SamYaple | yea but there will _always_ be stuff that will need to be backported after it merged into master | 07:47 |
| SamYaple | maybe we thought ti didnt affect the other branch or whatever | 07:47 |
| SamYaple | in that case we would update teh bug with the new affected branch | 07:47 |
| sdake | ya although in this case we know ahead of time we need a backport | 07:48 |
| sdake | your exmpe is one where we didn't know ahead of time | 07:48 |
| sdake | feel free to start a disussion on the ml | 07:48 |
| sdake | i'd be curiosu to know what the cr team thinks on this topic | 07:48 |
| SamYaple | considering no one seems to backport or use launchpad... | 07:49 |
| sdake | i intend to solve the lack of backporting problem | 07:50 |
| sdake | it would be easier if its in the git log | 07:50 |
| sdake | i can bring up in this week's team meeting if you like | 07:50 |
| SamYaple | sure. but im going to continue to not add backport tag to patches with bugs attached like i have been doing sinec i recommened the backport tag | 07:51 |
| SamYaple | might as well tag all patches | 07:51 |
| asalkeld | https://etherpad.openstack.org/p/kolla-mesos-tasks | 07:51 |
| sdake | i'd rather keep stable stable | 07:51 |
| sdake | then backport the world | 07:51 |
| asalkeld | hi sdake | 07:52 |
| sdake | sup asalkeld | 07:52 |
| asalkeld | "let's vote on the spec..." | 07:52 |
| sdake | SamYaple can you review the drop-root patches - want to get moving on that | 07:52 |
| sdake | as in quit reving it :) | 07:52 |
| sdake | every time you revise it it wipes out the +2 votes | 07:53 |
| SamYaple | sdake: sure if you review all of the patches i have in the queue | 07:53 |
| asalkeld | sure | 07:53 |
| sdake | i am going to start reviewing noow actually sam | 07:53 |
| sdake | you saw my first two just now ;) | 07:53 |
| asalkeld | sdake: see > https://etherpad.openstack.org/p/kolla-mesos-tasks | 07:53 |
| asalkeld | place for ideas and what we are up to | 07:53 |
| sdake | working top to bottom atm SamYaple | 07:53 |
| SamYaple | sdake: thats... weird | 07:54 |
| SamYaple | you should be going the other direction | 07:54 |
| SamYaple | oldest first | 07:54 |
| sdake | i am using asalkeld's review pge | 07:54 |
| sdake | its not actualy top to bottom | 07:54 |
| sdake | it helps prioritize reviews | 07:54 |
| asalkeld | it is handy;) | 07:55 |
| sdake | https://review.openstack.org/#/dashboard/?foreach=%28project%3Aopenstack%2Fkolla+OR+project%3Aopenstack%2Fkolla%2Dmesos%29+status%3Aopen+AND+NOT+label%3AWorkflow%3C%3D%2D2+label%3AVerified%3E%3D1%2Cjenkins&title=Kolla+Review+Inbox&My+Patches=owner%3Aself+status%3Aopen&Team+Starred=starredby%3Aasalkeld+OR+starredby%3A%22Sam+Yaple+%3Csam%40yaple.net%3E%22+OR+starredby%3Asdake&Waiting+for+final+approve=NOT+label%3AWorkflow%3E%3D1+NOT+label%3AWorkflow%3C%3D%2D1+ | 07:55 |
| sdake | NOT+owner%3Aself+label%3ACode%2DReview%3E%3D2+NOT+label%3ACode%2DReview%2D1+branch%3Amaster&Bug+fixes=topic%3A%22%5Ebug%2F.%2A%22&1+Week+Without+Feedback=NOT+label%3AWorkflow%3E%3D1+NOT+label%3AWorkflow%3C%3D%2D1+NOT+owner%3Aself+NOT+label%3ACode%2DReview%3C%3D2+age%3A7d+branch%3Amaster&No+Negative+Feedback=NOT+label%3AWorkflow%3E%3D1+NOT+label%3AWorkflow%3C%3D%2D1+NOT+owner%3Aself+NOT+label%3ACode%2DReview%3C%3D%2D1+NOT+label%3ACode%2DReview%3E%3D2+limit%3A | 07:55 |
| sdake | 50+branch%3Amaster | 07:55 |
| sdake | grn | 07:55 |
| asalkeld | lol | 07:55 |
| sdake | asalkeld can you get a proper link for SamYaple to bookmark? | 07:55 |
| asalkeld | i shorten it | 07:55 |
| asalkeld | https://goo.gl/Esg053 | 07:55 |
| sdake | thanks | 07:56 |
| sdake | SamYaple I rview no feedbak within 1 wee, followed by no negative feedback first | 07:56 |
| asalkeld | sdake: if you want to tweak that you put this http://paste.openstack.org/show/478290/ into "dashboards/kolla.dash" in the gerrit-dash-creator | 07:58 |
| asalkeld | https://github.com/stackforge/gerrit-dash-creator | 07:58 |
| asalkeld | you can come up with other ideas for the "team starred" | 07:59 |
| *** sdake_ has joined #kolla | 08:03 | |
| *** achanda has joined #kolla | 08:05 | |
| *** sdake has quit IRC | 08:06 | |
| *** akwasnie has joined #kolla | 08:16 | |
| *** zigo has quit IRC | 08:16 | |
| SamYaple | sdake_: im not sure you are seeing all the patches with that | 08:17 |
| sdake_ | SamYaple you may be right | 08:18 |
| sdake_ | i am not 100% convinced it is generating correct output but I ike the concept | 08:18 |
| SamYaple | ok well ive got other patches (backports and blueprints) do you see those? | 08:19 |
| *** zigo has joined #kolla | 08:19 | |
| openstackgerrit | Merged openstack/kolla: Fix validate-all-maintainer.sh to look for Dockerfile.j2 https://review.openstack.org/242933 | 08:26 |
| *** tzn has quit IRC | 08:34 | |
| *** kjelly has joined #kolla | 08:41 | |
| *** kproskurin has joined #kolla | 08:46 | |
| *** mimizone has joined #kolla | 08:48 | |
| *** vbel has joined #kolla | 08:49 | |
| openstackgerrit | Merged openstack/kolla: HAProxy does not have murano-api configuration https://review.openstack.org/242379 | 08:49 |
| openstackgerrit | Merged openstack/kolla: Add missing murano-api to site.yml https://review.openstack.org/242380 | 08:49 |
| *** mimizone has quit IRC | 08:53 | |
| sdake_ | SamYaple yup i just went through them i think | 08:53 |
| sdake_ | SamYaple asalkeld's thing doesn't show the stable branch unfortunately | 08:54 |
| sdake_ | i htink your bacports need a rebase or something - they dont seem to be going in | 08:54 |
| SamYaple | AJaeger just told me to use gertty | 08:54 |
| SamYaple | oh my | 08:54 |
| SamYaple | i have been missing out | 08:54 |
| SamYaple | sdake_: likely they have dependant patches that arent merged | 08:55 |
| sdake_ | possible | 08:55 |
| *** diga has joined #kolla | 08:55 | |
| sdake_ | what is gertty | 08:55 |
| SamYaple | cli gerrit | 08:55 |
| *** shardy has joined #kolla | 08:59 | |
| SamYaple | am I no core in kolla-mesos sdake_ ? | 09:01 |
| SamYaple | where do i even check the core list for a project | 09:02 |
| sdake_ | one moment i'll fix that | 09:02 |
| asalkeld | sdake_: I could add stable if you want | 09:02 |
| asalkeld | easy enough | 09:02 |
| sdake_ | asalkeld ya add kolla-core asalkeld | 09:02 |
| sdake_ | i meant to do that today and got sidetracked | 09:02 |
| sdake_ | SamYaple to seee it, go to review.openstack.org | 09:03 |
| sdake_ | click "people" | 09:03 |
| sdake_ | click list groups | 09:03 |
| sdake_ | filter: kolla-mesos-cored | 09:03 |
| sdake_ | SamYaple ok your all set | 09:04 |
| asalkeld | added now https://review.openstack.org/#/admin/groups/1168,members | 09:04 |
| *** tzn has joined #kolla | 09:06 | |
| asalkeld | sdake_: SamYaple https://goo.gl/Oi11bi | 09:08 |
| asalkeld | with stable/liberty ^ | 09:08 |
| openstackgerrit | Merged openstack/kolla-mesos: Remove validate-all-maintainers.sh https://review.openstack.org/242927 | 09:10 |
| *** achanda has quit IRC | 09:10 | |
| sdake_ | thanks asakleld | 09:10 |
| sdake_ | asalkeld | 09:11 |
| SamYaple | asalkeld: I have moved on to gertty | 09:11 |
| *** tzn has quit IRC | 09:11 | |
| SamYaple | i can never go back | 09:11 |
| asalkeld | np | 09:11 |
| SamYaple | its a rule | 09:11 |
| SamYaple | i didnt make it | 09:11 |
| SamYaple | it just is one | 09:11 |
| asalkeld | (totally doesn't worry me) | 09:11 |
| *** gfidente has joined #kolla | 09:13 | |
| sdake_ | what is geerty | 09:13 |
| *** mbound has joined #kolla | 09:13 | |
| SamYaple | 08:56:51 < sdake_> what is gertty | 09:15 |
| SamYaple | 08:57:19 < SamYaple> cli gerrit | 09:15 |
| *** openstack has joined #kolla | 09:20 | |
| SamYaple | kproskurin: nope i accidentally did +2-1 | 09:21 |
| SamYaple | i meant to approve it | 09:21 |
| kproskurin | ok, tx :-) | 09:21 |
| sdake_ | kproskurin that happens sometimes | 09:21 |
| sdake_ | inevery project | 09:21 |
| kproskurin | I know, just to be sure, may be a small misstake in docs or something | 09:25 |
| *** tfukushima has quit IRC | 09:31 | |
| *** sdake_ has quit IRC | 09:36 | |
| *** athomas has joined #kolla | 09:40 | |
| *** tfukushima has joined #kolla | 09:45 | |
| *** jmccarthy has joined #kolla | 09:47 | |
| *** akwasnie1 has joined #kolla | 10:03 | |
| *** akwasnie has quit IRC | 10:06 | |
| *** tzn has joined #kolla | 10:07 | |
| *** tzn has quit IRC | 10:13 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: DO NOT MERGE: gate testing https://review.openstack.org/242763 | 10:24 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: DO NOT MERGE: gate testing https://review.openstack.org/242763 | 10:27 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Incorrect parsed variable name https://review.openstack.org/242768 | 10:34 |
| *** bradjones has joined #kolla | 10:37 | |
| *** bradjones has quit IRC | 10:37 | |
| *** bradjones has joined #kolla | 10:37 | |
| SamYaple | mandre: are you happy | 10:40 |
| SamYaple | im just going to remove all changes | 10:40 |
| SamYaple | abandon all patches | 10:40 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Make the database json variable more readable https://review.openstack.org/242024 | 10:42 |
| *** jmccarthy has quit IRC | 10:44 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Make the database json variable more readable https://review.openstack.org/242024 | 10:46 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Incorrect parsed variable name https://review.openstack.org/242768 | 10:46 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Move data container to start.yml https://review.openstack.org/242908 | 10:47 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Make the database json variable more readable https://review.openstack.org/242024 | 10:47 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Incorrect parsed variable name https://review.openstack.org/242768 | 10:47 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Remove unused cinder-data container https://review.openstack.org/242907 | 10:47 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Fix bootstrap for services https://review.openstack.org/242777 | 10:47 |
| *** diga has quit IRC | 10:48 | |
| *** mimizone has joined #kolla | 10:50 | |
| openstackgerrit | Merged openstack/kolla: Incorrect parsed variable name https://review.openstack.org/242768 | 10:55 |
| *** mimizone has quit IRC | 10:55 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: We now have a logs publisher https://review.openstack.org/243031 | 10:59 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: DO NOT MERGE: gate testing https://review.openstack.org/242763 | 11:04 |
| *** tzn has joined #kolla | 11:10 | |
| *** jmccarthy has joined #kolla | 11:13 | |
| *** tzn has quit IRC | 11:15 | |
| *** dims has joined #kolla | 11:15 | |
| *** athomas has quit IRC | 11:16 | |
| *** pbourke has quit IRC | 11:16 | |
| *** pbourke has joined #kolla | 11:16 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: DO NOT MERGE: gate testing https://review.openstack.org/242763 | 11:18 |
| *** athomas has joined #kolla | 11:20 | |
| *** akwasnie1 has quit IRC | 11:22 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: DO NOT MERGE: gate testing https://review.openstack.org/242763 | 11:24 |
| *** tfukushima has quit IRC | 11:32 | |
| *** akwasnie has joined #kolla | 11:36 | |
| *** tzn has joined #kolla | 12:11 | |
| *** tzn has quit IRC | 12:16 | |
| *** cemmason has joined #kolla | 12:37 | |
| openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: Add a simplistic bootstrap script to install mesos (AIO) https://review.openstack.org/242911 | 12:51 |
| *** mimizone has joined #kolla | 12:52 | |
| openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: Add a simplistic bootstrap script to install mesos (AIO) https://review.openstack.org/242911 | 12:55 |
| *** mimizone has quit IRC | 12:57 | |
| *** dims has quit IRC | 12:59 | |
| *** rhallisey has joined #kolla | 13:00 | |
| mandre | SamYaple: there is no half measure | 13:02 |
| *** tzn has joined #kolla | 13:11 | |
| SamYaple | mandre: im the one who commits | 13:12 |
| mandre | i didn't get your comment, you abandoned all the groups fixes? | 13:16 |
| mandre | or were there already in master? | 13:16 |
| *** tzn has quit IRC | 13:16 | |
| SamYaple | mandre: kinda there was a bigger issue that requires changing the way that whole section was done | 13:17 |
| SamYaple | ive rebased that entire series onto another patchseriees | 13:17 |
| SamYaple | in other news, we now have an expiremental gate | 13:19 |
| SamYaple | multinode | 13:19 |
| mandre | yay | 13:23 |
| mandre | which one is it? | 13:24 |
| mandre | what does dsvm stand for? | 13:24 |
| SamYaple | mandre: its convention | 13:25 |
| SamYaple | it stands for devstackvm | 13:25 |
| SamYaple | project-config asked me to add it rather than update our existing regex | 13:25 |
| mandre | ha, found the multinode gate :) | 13:28 |
| mandre | maybe soon they'll ask to rename all gates to kvm -- kollavm | 13:29 |
| *** macsz has joined #kolla | 13:32 | |
| *** akwasnie has quit IRC | 13:33 | |
| *** akwasnie has joined #kolla | 13:37 | |
| *** diogogmt has quit IRC | 13:40 | |
| *** diogogmt has joined #kolla | 13:42 | |
| SamYaple | that wont be confusing | 13:45 |
| *** akwasnie_ has joined #kolla | 13:47 | |
| *** akwasnie_ has quit IRC | 13:48 | |
| *** akwasnie_ has joined #kolla | 13:51 | |
| *** akwasnie_ has quit IRC | 13:52 | |
| *** klint has quit IRC | 14:02 | |
| *** thumpba has joined #kolla | 14:08 | |
| *** diogogmt has quit IRC | 14:10 | |
| *** thumpba has quit IRC | 14:17 | |
| *** macsz has quit IRC | 14:18 | |
| *** macsz has joined #kolla | 14:18 | |
| *** absubram has quit IRC | 14:20 | |
| *** thumpba has joined #kolla | 14:22 | |
| *** macsz has quit IRC | 14:22 | |
| *** thumpba has quit IRC | 14:22 | |
| *** macsz has joined #kolla | 14:23 | |
| *** thumpba has joined #kolla | 14:25 | |
| *** sdake has joined #kolla | 14:35 | |
| *** dwalsh has joined #kolla | 14:36 | |
| *** dims has joined #kolla | 14:36 | |
| *** thumpba has quit IRC | 14:37 | |
| *** sdake_ has joined #kolla | 14:37 | |
| sdake_ | morning | 14:39 |
| SamYaple | morning sdake | 14:40 |
| SamYaple | EVERYTHIGNS BROKEN | 14:40 |
| SamYaple | im kidding | 14:40 |
| *** sdake has quit IRC | 14:40 | |
| SamYaple | ive had sugar | 14:40 |
| sdake_ | SamYaple please review https://review.openstack.org/#/c/242877/ | 14:41 |
| sdake_ | pretty please | 14:41 |
| sdake_ | with sugar on top | 14:41 |
| *** thumpba has joined #kolla | 14:44 | |
| SamYaple | sdake_: reviewed | 14:47 |
| SamYaple | -1 | 14:47 |
| SamYaple | well on the initial patch | 14:47 |
| sdake_ | thanks | 14:48 |
| *** sdake_ is now known as sdake | 14:48 | |
| *** sdake_ has joined #kolla | 14:50 | |
| *** sdake has quit IRC | 14:53 | |
| *** tfukushima has joined #kolla | 14:53 | |
| *** tfukushima has quit IRC | 14:53 | |
| *** tfukushima has joined #kolla | 14:53 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: DO NOT MERGE: gate testing https://review.openstack.org/242763 | 14:54 |
| *** mimizone has joined #kolla | 14:54 | |
| sdake_ | SamYaple btw yes i hae etewted the seurity changes | 14:55 |
| sdake_ | in both ha and non-ha setup | 14:55 |
| sdake_ | i am testing the 440 change now | 14:55 |
| sdake_ | and then will have a new review up | 14:55 |
| *** thumpba has quit IRC | 14:56 | |
| *** sdake_ is now known as sdake | 14:56 | |
| *** mimizone has quit IRC | 14:58 | |
| *** mimizone has joined #kolla | 15:01 | |
| *** thumpba has joined #kolla | 15:04 | |
| *** mimizone has quit IRC | 15:06 | |
| *** mimizone has joined #kolla | 15:06 | |
| *** achanda has joined #kolla | 15:10 | |
| *** akwasnie has quit IRC | 15:11 | |
| *** tzn has joined #kolla | 15:13 | |
| *** thumpba has quit IRC | 15:14 | |
| sdake | SamYaple note the gate caught a patchc that I thought was good | 15:14 |
| sdake | which was actually bad | 15:14 |
| sdake | a few days ago | 15:14 |
| sdake | so yay ;) | 15:14 |
| *** macsz has quit IRC | 15:15 | |
| openstackgerrit | Steven Dake proposed openstack/kolla: Base image changes for drop-root https://review.openstack.org/242876 | 15:15 |
| openstackgerrit | Steven Dake proposed openstack/kolla: drop root for glance https://review.openstack.org/242877 | 15:15 |
| sdake | SamYaple more sugar on top can you review those again tia !#! | 15:16 |
| *** openstackgerrit has quit IRC | 15:17 | |
| *** openstackgerrit has joined #kolla | 15:17 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: DO NOT MERGE: gate testing https://review.openstack.org/242763 | 15:17 |
| SamYaple | sdake: yay | 15:17 |
| *** tzn has quit IRC | 15:17 | |
| SamYaple | sdake: btw for the multinode gate I think we still need to do the dind stuff | 15:17 |
| SamYaple | reason being it will allow us to have a docker container for each and every service as if it were its own node | 15:18 |
| SamYaple | if we had that right now 5 bugs wouldnt exist | 15:18 |
| *** thumpba has joined #kolla | 15:19 | |
| *** absubram has joined #kolla | 15:19 | |
| sdake | SamYaple sounds fine | 15:19 |
| sdake | SamYaple can you review https://review.openstack.org/#/c/242876/ and its child | 15:19 |
| SamYaple | sdake: already did | 15:20 |
| SamYaple | child was just a rebase | 15:20 |
| sdake | thanks | 15:21 |
| * sdake made it through the sam hurdle yay | 15:21 | |
| sdake | pbourke can you look at that patch | 15:21 |
| sdake | mandre/harmw as well ^ | 15:21 |
| sdake | SamYaple are you going to add a ded chicken test to deploy | 15:23 |
| sdake | (in other words vm boot operation) | 15:23 |
| *** gfidente is now known as gfidente|afk | 15:25 | |
| SamYaple | sdake: waiting on seeing what pbourke is doing | 15:28 |
| pbourke | just reading it now | 15:28 |
| SamYaple | we may want to reuse the already existing tempest tests that devstack uses | 15:28 |
| SamYaple | thoughts pbourke? | 15:28 |
| pbourke | still trying to grok it | 15:28 |
| SamYaple | oh btw guys, I got us a logs publisher | 15:28 |
| sdake | pbourke he means tempest not my securit ypatch | 15:28 |
| SamYaple | now we can have a logs/ directory next to console.html | 15:29 |
| SamYaple | so im going to be seperating the tasks out to make our gate more debuggable | 15:29 |
| SamYaple | (dump all the service logs in there too) | 15:29 |
| pbourke | SamYaple: i haven't done much since the intial wip patch for tempest. We just got our lab env up here so I'm getting set up in that. Once I do so tempest is first on my list | 15:29 |
| *** cloudnautique has quit IRC | 15:29 | |
| *** jruano has joined #kolla | 15:30 | |
| *** mimizone has quit IRC | 15:30 | |
| SamYaple | pbourke: ok, look at just reusing the same tempest tests that everyone else uses if possible | 15:30 |
| SamYaple | i dont know if it is | 15:30 |
| sdake | pbourke eta? | 15:30 |
| sdake | if tempest is a few months out | 15:30 |
| sdake | a simple boot test may make sense now | 15:30 |
| sdake | if its a week out it does not | 15:30 |
| pbourke | hopefully can get something up by the end of the week | 15:30 |
| sdake | cool that would be fantasti | 15:31 |
| sdake | c | 15:31 |
| sdake | pbourke my ptch uses alot of sudo magic | 15:31 |
| sdake | if you have qs feel free to ask in irc | 15:32 |
| SamYaple | sdake: even right now we have no ram in the gates | 15:32 |
| pbourke | hmm yeah im half wondering should I ask you to comment it | 15:32 |
| SamYaple | i still have to trim the fats of containers | 15:32 |
| pbourke | or should I just rtfm | 15:32 |
| pbourke | or both | 15:32 |
| sdake | i am happy to comment sudoers | 15:32 |
| sdake | the rest shuld probably stay uncommented i think | 15:32 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: DO NOT MERGE: gate testing https://review.openstack.org/242763 | 15:32 |
| sdake | otherwise we end up copy8ing those comments to all the files | 15:33 |
| *** thumpba has quit IRC | 15:34 | |
| sdake | pbourke doc incoming sec | 15:36 |
| pbourke | sdake: have a query on one of the existing comments also, have added a comment | 15:36 |
| pbourke | thanks | 15:37 |
| SamYaple | blindly approve! | 15:37 |
| sdake | wuld prefer no rubber stamping on this one thanks | 15:38 |
| *** tzn has joined #kolla | 15:38 | |
| *** thumpba has joined #kolla | 15:41 | |
| *** thumpba has quit IRC | 15:42 | |
| openstackgerrit | Steven Dake proposed openstack/kolla: Base image changes for drop-root https://review.openstack.org/242876 | 15:43 |
| openstackgerrit | Steven Dake proposed openstack/kolla: drop root for glance https://review.openstack.org/242877 | 15:43 |
| *** blahRus has joined #kolla | 15:43 | |
| sdake | pbourke can you hit thoe again plz :) | 15:44 |
| openstackgerrit | Steven Dake proposed openstack/kolla: Base image changes for drop-root https://review.openstack.org/242876 | 15:46 |
| openstackgerrit | Steven Dake proposed openstack/kolla: drop root for glance https://review.openstack.org/242877 | 15:46 |
| sdake | pbourke i had a typo (corected) | 15:46 |
| pbourke | was going to say :) | 15:46 |
| pbourke | docs looking good | 15:46 |
| *** achanda has quit IRC | 15:48 | |
| openstackgerrit | Steven Dake proposed openstack/kolla: Base image changes for drop-root https://review.openstack.org/242876 | 15:50 |
| openstackgerrit | Steven Dake proposed openstack/kolla: drop root for glance https://review.openstack.org/242877 | 15:50 |
| sdake | pbourke ok that looks ready for review now, pleae let me know ;) | 15:51 |
| pbourke | patch is clever | 15:51 |
| sdake | take 3 | 15:51 |
| sdake | blame samyaple for cleverness | 15:51 |
| sdake | although blame me for the sudo config ;) | 15:51 |
| pbourke | do we want a TODO to remove the root ALL= ... line once all services are migrated? | 15:52 |
| pbourke | or are we keeping it in | 15:52 |
| sdake | you can see my other takes | 15:52 |
| sdake | pbourke i am not sure yet tbh | 15:52 |
| sdake | i think we can jsust keep it in our mental map ;) | 15:52 |
| sdake | we may need it later | 15:52 |
| sdake | i just am not sure if wecan remove it when we finish this security work | 15:52 |
| *** dolpher has quit IRC | 15:53 | |
| *** signed8bit has joined #kolla | 15:54 | |
| pbourke | sdake: have one more question | 15:56 |
| sdake | shoot | 15:56 |
| pbourke | and a typo ;) | 15:56 |
| pbourke | why leave the kolla_* files executable by any group other than kolla? | 15:56 |
| *** thumpba has joined #kolla | 15:56 | |
| pbourke | i.e. you're chmod'ing them 755 and not using chgrp kolla in the Dockerfile | 15:56 |
| sdake | i could do that although it doesn't really add any value | 15:57 |
| pbourke | i guess not | 15:57 |
| sdake | the files are publically avilalbeo n the internets to look for ways to crack them | 15:57 |
| *** achanda has joined #kolla | 15:57 | |
| SamYaple | pbourke: also any user running will be in the kolla group | 15:57 |
| sdake | its intereting how we actuallly get *more* security with this model then in a bar emetal environment | 15:57 |
| SamYaple | but 755 is the typical perms for something in /usr/bin | 15:57 |
| pbourke | SamYaple: only the service one is in that group no? | 15:58 |
| pbourke | e.g. glance | 15:58 |
| sdake | because you ahve o break out of the container and then break into root to break out of the container into the host os | 15:58 |
| SamYaple | pbourke: which is the only running processes user | 15:58 |
| sdake | the levels of breakout required are much more difficult to pull off imo | 15:58 |
| SamYaple | pbourke: so any level of breaking into it will allow kolla group user to hit the file | 15:58 |
| pbourke | sdake: one quick typo in start.sh - usually wouldn't care but its in base so might as well fix it now | 15:58 |
| pbourke | then +2 | 15:58 |
| sdake | SamYaple good point i hadn't thought of that | 15:58 |
| sdake | sure | 15:59 |
| sdake | which word am I looking for? | 15:59 |
| pbourke | writable | 15:59 |
| pbourke | ive commented the review | 15:59 |
| *** athomas has quit IRC | 16:00 | |
| openstackgerrit | Steven Dake proposed openstack/kolla: Base image changes for drop-root https://review.openstack.org/242876 | 16:01 |
| openstackgerrit | Steven Dake proposed openstack/kolla: drop root for glance https://review.openstack.org/242877 | 16:01 |
| SamYaple | pbourke: and you say its hard to get a patch by ME | 16:01 |
| pbourke | it is :p | 16:02 |
| sdake | you guys are a bit picky | 16:02 |
| sdake | although I think with kolla repo we should be picky | 16:02 |
| SamYaple | im not picky :( | 16:02 |
| sdake | since the code is failry mature | 16:02 |
| sdake | edspecially in our commonly run codebase | 16:02 |
| sdake | i'd prefer for kolla-mesos we be less picky up front | 16:02 |
| sdake | as wwell as kolla-cli | 16:02 |
| sdake | since after those are implemented i expect they will be reimplemented | 16:03 |
| SamYaple | i point out something is broken or its a race or its a security thing, but im all about the "i dont care if its mispelled, but fix it if you rebase" comments | 16:03 |
| SamYaple | im not picky :( | 16:03 |
| sdake | it would be nice if we didn't hae mispellings int eh code bse, but commit logs i am ok with ;) | 16:03 |
| * sdake needs a deathstar | 16:04 | |
| SamYaple | i dnot speel gudd, but i code da bomb | 16:04 |
| pbourke | i wouldn't hold up a patch for misspelling, just pointed it out as sdake was here ready to fix | 16:04 |
| *** diogogmt has joined #kolla | 16:04 | |
| sdake | might as well get it right folks | 16:04 |
| SamYaple | define right... | 16:04 |
| sdake | defect free | 16:04 |
| *** cloudnautique has joined #kolla | 16:04 | |
| *** cloudnautique has quit IRC | 16:05 | |
| sdake | spelling error = defect | 16:05 |
| SamYaple | boo | 16:05 |
| SamYaple | comments can be mispelled | 16:05 |
| SamYaple | code is funny | 16:05 |
| *** cloudnautique has joined #kolla | 16:05 | |
| SamYaple | alright guys | 16:06 |
| *** athomas has joined #kolla | 16:06 | |
| SamYaple | brainstorm for me | 16:06 |
| SamYaple | sdake: pbourke | 16:06 |
| SamYaple | ive got an ip | 16:06 |
| SamYaple | it exists on the machine | 16:06 |
| SamYaple | how do i determine 100% of the time which interface it is on? | 16:06 |
| SamYaple | machine could have 1 or 20 interfaces | 16:07 |
| *** thumpba has quit IRC | 16:07 | |
| pbourke | grep | 16:08 |
| *** macsz has joined #kolla | 16:08 | |
| SamYaple | /kick pbourke | 16:08 |
| SamYaple | this is art | 16:08 |
| SamYaple | i want eloquence | 16:08 |
| pbourke | you can thank me later | 16:08 |
| sdake | ip grep | 16:08 |
| sdake | unix pipes exist for a good reason ;-) | 16:08 |
| SamYaple | you guys are the worst | 16:09 |
| SamYaple | when im done itll be beautiful | 16:09 |
| SamYaple | youll see | 16:09 |
| sdake | i really dont know any otheer way | 16:09 |
| sdake | not that there isn't one | 16:09 |
| SamYaple | ill probably walk the interfaces with netifaces | 16:10 |
| sdake | SamYaple can yu ack https://review.openstack.org/#/c/242876/ | 16:10 |
| pbourke | SamYaple: move to windows you can use some com object to query the registry | 16:10 |
| sdake | oh if yu wan tto write code yes that will work | 16:10 |
| sdake | SamYaple i can show you how its done in c ;) | 16:10 |
| *** vbel has quit IRC | 16:10 | |
| SamYaple | sdake: i was a C guy long before python | 16:10 |
| sdake | python is probbly similar | 16:10 |
| britthouser | I have some code that does that SamYaple | 16:11 |
| britthouser | lemme dig it up | 16:11 |
| sdake | morning britthouser | 16:11 |
| britthouser | morning sdake! | 16:11 |
| sdake | britthouser do you know bharaththiruveedula <bharath_ves@hotmail.com> | 16:12 |
| *** thumpba has joined #kolla | 16:12 | |
| britthouser | don't think so? | 16:12 |
| sdake | britthouser cool thanks | 16:14 |
| *** tfukushima has quit IRC | 16:14 | |
| sdake | SamYaple waiting for gate good idea thanks | 16:19 |
| britthouser | SamYaple - ok so its not exact, but I think you can make it work. My code finds the interface whose IP matches a particular subnet. But I think you just change the last IF statement todo what you want: | 16:21 |
| britthouser | http://paste.openstack.org/show/478338/ | 16:21 |
| *** thumpba has quit IRC | 16:24 | |
| *** thumpba has joined #kolla | 16:26 | |
| SamYaple | thanks britthouser, I just threw this together http://paste.openstack.org/show/478341/ | 16:26 |
| britthouser | I would call it "find_ipv4_interface", but otherwise looks good! =) | 16:27 |
| SamYaple | britthouser: im about to collpase it into a oneliner, ill prbably cahcne the name to 1 letter | 16:28 |
| britthouser | you lost me there, but yes!? =) | 16:29 |
| sdake | F | 16:29 |
| SamYaple | F | 16:30 |
| *** thumpba_ has joined #kolla | 16:30 | |
| *** thumpba has quit IRC | 16:30 | |
| *** tzn has quit IRC | 16:34 | |
| *** Hui has joined #kolla | 16:38 | |
| Hui | Good morning | 16:38 |
| *** sdake has quit IRC | 16:45 | |
| *** diogogmt has quit IRC | 16:45 | |
| *** dwalsh has quit IRC | 16:47 | |
| SamYaple | morning Hui | 16:48 |
| *** thumpba_ has quit IRC | 16:49 | |
| openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: Add a simplistic bootstrap script to install mesos (AIO) https://review.openstack.org/242911 | 16:50 |
| openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: Add a simplistic bootstrap script to install mesos (AIO) https://review.openstack.org/242911 | 16:51 |
| *** mfalatic has joined #kolla | 16:52 | |
| Hui | SamYaple, I will test neutron patch now | 16:53 |
| *** macsz has quit IRC | 16:54 | |
| *** daneyon has joined #kolla | 16:55 | |
| *** daneyon_ has joined #kolla | 16:56 | |
| *** thumpba has joined #kolla | 16:56 | |
| *** tzn has joined #kolla | 16:57 | |
| *** tzn has quit IRC | 16:59 | |
| *** daneyon has quit IRC | 16:59 | |
| *** tzn has joined #kolla | 17:02 | |
| *** thumpba has quit IRC | 17:03 | |
| *** _tzn has joined #kolla | 17:04 | |
| *** tzn has quit IRC | 17:04 | |
| *** thumpba has joined #kolla | 17:05 | |
| *** _tzn has quit IRC | 17:05 | |
| *** tzn has joined #kolla | 17:06 | |
| *** _tzn has joined #kolla | 17:07 | |
| *** tzn has quit IRC | 17:07 | |
| *** achanda has quit IRC | 17:09 | |
| *** _tzn has quit IRC | 17:09 | |
| *** tzn has joined #kolla | 17:09 | |
| *** cloudnautique has quit IRC | 17:10 | |
| *** nihilifer has quit IRC | 17:18 | |
| *** ashishjain has joined #kolla | 17:19 | |
| *** thumpba has quit IRC | 17:20 | |
| ashishjain | Hi | 17:20 |
| ashishjain | I am trying to setup kolla development environment aio using the bare metal deployment. | 17:21 |
| ashishjain | I have got few questions on some of the configuration parameters | 17:21 |
| ashishjain | What is the usage of "kolla_internal_address: "10.10.10.254"" | 17:22 |
| ashishjain | My host has got an IP address of 10.0.0.8 and I have set up the kolla_internal_address: "10.0.0.254 | 17:23 |
| ashishjain | Is this correct? | 17:23 |
| *** daneyon has joined #kolla | 17:23 | |
| *** thumpba has joined #kolla | 17:24 | |
| ashishjain | another question is on the network interface..I use a wireless network and the name of the interface is wlan0, I hope wireless network should not be a problem and I am using the configuration parameter as network_interface: "wlan0" | 17:24 |
| *** daneyon_ has quit IRC | 17:26 | |
| *** dwalsh has joined #kolla | 17:30 | |
| *** cloudnautique has joined #kolla | 17:31 | |
| *** unicell1 has quit IRC | 17:34 | |
| *** daneyon_ has joined #kolla | 17:44 | |
| *** nihilifer has joined #kolla | 17:45 | |
| *** daneyon has quit IRC | 17:46 | |
| *** sdake has joined #kolla | 17:47 | |
| ashishjain | sdake: hi | 17:49 |
| sdake | yo | 17:49 |
| ashishjain | need some help | 17:49 |
| ashishjain | I think I have misconfigured some of the configuration parameters in globals.yml | 17:50 |
| *** jruano has quit IRC | 17:50 | |
| ashishjain | the sample config says : "kolla_internal_address: "10.10.10.254"" | 17:50 |
| ashishjain | However my network is 10.0.0.0/24 | 17:50 |
| ashishjain | so I have set it up as "10.0.0.254" | 17:50 |
| ashishjain | the comment in the globals.yml says set kolla_internal_address as one of the unused ip address in the network and hence I have done the same | 17:51 |
| ashishjain | My host ip is 10.0.0.8 | 17:51 |
| ashishjain | Now I am seeing lot of errors in rsyslog | 17:52 |
| ashishjain | where mysql is unable to connect to 10.0.0.254 | 17:52 |
| ashishjain | so my question is what is kolla_internal_address, does it has to be host_ip or a unused ip? | 17:53 |
| SamYaple | ashishjain: 10.0.0.254 is a VIP | 17:53 |
| SamYaple | keepalived assigns it to one node | 17:53 |
| SamYaple | is this an AIO? | 17:53 |
| ashishjain | SamYaple: yes this is an AIO | 17:54 |
| *** ssurana has joined #kolla | 17:54 | |
| SamYaple | is keepalived and haproxy running | 17:54 |
| ashishjain | yes both of them are running | 17:54 |
| SamYaple | is the VIP assigned to your interface? | 17:55 |
| *** thumpba has quit IRC | 17:56 | |
| SamYaple | Hui: ohh i know whats happening | 17:56 |
| SamYaple | ugh ansible | 17:56 |
| ashishjain | SamYaple: which interface will get the VIP? | 17:56 |
| SamYaple | Hui: its run_once and the when conditional thats doing things | 17:56 |
| SamYaple | Hui: ill convert to delegate_to i suppose | 17:57 |
| sdake | ashishjain the regular interface gets the vip | 17:57 |
| sdake | the vip should be an unallocated ip addres | 17:57 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: DO NOT MERGE: gate testing https://review.openstack.org/242763 | 17:57 |
| ashishjain | sdake: it indeed is unassigned | 17:57 |
| sdake | SamYaple this is passing the gate - please kick a +a on it :) https://review.openstack.org/#/c/242876/ | 17:57 |
| ashishjain | and how do I check whether the VIP is assigned to the interface? | 17:58 |
| ashishjain | I am unable to ping this ip address | 17:59 |
| sdake | ashishjain lets focus on symptoms, what problem do you have atm? | 17:59 |
| ashishjain | 10.0.0.254 | 17:59 |
| sdake | i see | 17:59 |
| sdake | yo ucan run docker logs haproxy | 17:59 |
| sdake | it will show you what it has done | 17:59 |
| *** jmccarthy has quit IRC | 18:00 | |
| sdake | passte your globals.yml plz | 18:00 |
| ashishjain | sure one second | 18:02 |
| *** jruano has joined #kolla | 18:02 | |
| *** ashishjain has quit IRC | 18:02 | |
| openstackgerrit | Merged openstack/kolla: Base image changes for drop-root https://review.openstack.org/242876 | 18:02 |
| *** ashishjain has joined #kolla | 18:03 | |
| ashishjain | sdake: here it is http://paste.openstack.org/show/478370/ | 18:04 |
| openstackgerrit | Merged openstack/kolla: drop root for glance https://review.openstack.org/242877 | 18:04 |
| *** unicell has joined #kolla | 18:07 | |
| *** kproskurin has quit IRC | 18:09 | |
| *** sdake has quit IRC | 18:09 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: DO NOT MERGE: gate testing https://review.openstack.org/242763 | 18:13 |
| Hui | SamYaple, yes, delegate_to should solve it | 18:14 |
| *** sdake has joined #kolla | 18:15 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: DO NOT MERGE: gate testing https://review.openstack.org/242763 | 18:18 |
| ashishjain | sdake: Here is the globals.yml http://paste.openstack.org/show/478370/ | 18:18 |
| sdake | SamYaple thanks bro | 18:19 |
| sdake | paste not working for me for some reason ashishjain | 18:19 |
| sdake | oh just came up let me look | 18:19 |
| *** tzn has quit IRC | 18:20 | |
| sdake | ashishjain run docker ps -a and paste it please | 18:20 |
| sdake | after an ansible deploy | 18:20 |
| ashishjain | sdake: here it is http://paste.openstack.org/show/478375/ | 18:21 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: DO NOT MERGE: gate testing https://review.openstack.org/242763 | 18:22 |
| sdake | you shouldn't be able to ping the haproxy ip | 18:22 |
| sdake | however, try this | 18:22 |
| sdake | telnet 10.0.0.254 3306 | 18:22 |
| sdake | that should work | 18:22 |
| sdake | rather i am not sure if you can ping the haproxy ip | 18:22 |
| sdake | it may or may not work | 18:22 |
| sdake | but that telnet will tell you if haproyx is working or not | 18:22 |
| sdake | what symptoms do you have, it looks like you have a good deploy to me | 18:23 |
| sdake | does source openrc-example followed by tools/init-runonce after customization not work? | 18:23 |
| *** ashishjain_ has joined #kolla | 18:24 | |
| ashishjain_ | sdake: it does not work | 18:25 |
| sdake | what doesn't work? | 18:25 |
| *** tzn has joined #kolla | 18:25 | |
| sdake | psate your openrc | 18:25 |
| ashishjain_ | sdake: this kolla_internal_address is haproxy ip? | 18:25 |
| sdake | yes kolla_internal_address become the haproxy ip address | 18:25 |
| ashishjain_ | ok | 18:26 |
| ashishjain_ | sdake: here it is http://paste.openstack.org/show/478377 | 18:27 |
| sdake | do not use 254 | 18:28 |
| ashishjain_ | okay | 18:28 |
| sdake | oh wait | 18:28 |
| sdake | ya that should be correct | 18:28 |
| sdake | v2.0 is not supported tho | 18:28 |
| sdake | do not use v2.0 | 18:28 |
| ashishjain_ | okay, than v3.0? | 18:29 |
| sdake | export OS_AUTH_URL=http://broked.selfip.net:5000 | 18:29 |
| sdake | http://10.0.0.254:5000 should do | 18:29 |
| ashishjain_ | okay will try but you know there is no ip 10.0.0.254 | 18:29 |
| sdake | kolla creates one | 18:30 |
| sdake | you specified it in your globals.yml | 18:30 |
| ashishjain_ | okay but shouldn't this ip be pingable? | 18:30 |
| sdake | i think you can connec tto it via the telnet test if youwant | 18:30 |
| sdake | but we dont set up a ping on that ip i think | 18:30 |
| sdake | i am not sure how ping would interact, as I said | 18:30 |
| sdake | I can guarantee you telnet 10.0.0.254 3306 would be workign | 18:31 |
| sdake | if kolla was working | 18:31 |
| sdake | working as in making a connection | 18:31 |
| ashishjain_ | sdake: no it is not working | 18:31 |
| ashishjain_ | :( | 18:31 |
| sdake | firewall? | 18:31 |
| sdake | what distro are you running? | 18:31 |
| sdake | ubuntu? | 18:32 |
| sdake | it is possible your firewall is interacting in the deployment | 18:32 |
| ashishjain_ | sdake: yes it is ubunut | 18:33 |
| ashishjain_ | infact | 18:33 |
| ashishjain_ | You know when I try my host ip and 3306 as telnet it works | 18:33 |
| sdake | yes of course that should work | 18:34 |
| ashishjain_ | which is telnet 10.0.0.8 3306 | 18:34 |
| sdake | the vip should also map to that address | 18:34 |
| ashishjain_ | okay | 18:34 |
| ashishjain_ | so looks like the VIP part didnot go well | 18:34 |
| sdake | however your using a wlan - so not sure how that will behave | 18:34 |
| sdake | i don't run kolla on laptops | 18:34 |
| sdake | it may be vips and wlan doesn't work correctly | 18:34 |
| *** ashishjain has quit IRC | 18:34 | |
| sdake | you can always run in a vm - that will definately work ;) | 18:34 |
| ashishjain_ | okay | 18:35 |
| ashishjain_ | so shall I use the vagrant dev environment? | 18:35 |
| sdake | asalkeld if I bribe yu with more votes, would you please please fix the heat tempalte for kolla? :) | 18:35 |
| sdake | a fourth option is just to boot a plain jain vm | 18:35 |
| sdake | i think vgrant is too complex | 18:35 |
| sdake | but if you want to try it out, go for it ;) | 18:35 |
| ashishjain_ | na I think I will go ahead with "plain jain vm :)" | 18:36 |
| ashishjain_ | okay so shall I just boot up a VM with 10G RAM and 100G HDD using say virtual box with just one network interface ? | 18:37 |
| sdake | two interfaces | 18:37 |
| sdake | 16 gb ram is what i'd go with | 18:37 |
| sdake | but yup, that will 100% work | 18:37 |
| ashishjain_ | :( I got 12G only in my laptop | 18:37 |
| sdake | 10gb will do but not give you much ram to do much work | 18:38 |
| sdake | before giving up tho | 18:38 |
| sdake | i'd like to understand what doens't work | 18:38 |
| sdake | can we do a bit of debug on your haproxy? | 18:38 |
| ashishjain_ | Okay sure | 18:38 |
| sdake | paste docker logs haproxy | 18:38 |
| ashishjain_ | http://paste.openstack.org/show/478378 | 18:41 |
| *** signed8bit is now known as signed8bit_ZZZzz | 18:41 | |
| *** sdake_ has joined #kolla | 18:43 | |
| ashishjain_ | sdake: here are the logs http://paste.openstack.org/show/478378 | 18:43 |
| *** signed8bit_ZZZzz is now known as signed8bit | 18:44 | |
| *** sdake has quit IRC | 18:45 | |
| *** jasonsb has quit IRC | 18:45 | |
| *** achanda has joined #kolla | 18:45 | |
| *** achanda has quit IRC | 18:45 | |
| *** diogogmt has joined #kolla | 18:48 | |
| *** jruano has quit IRC | 18:50 | |
| *** signed8bit is now known as signed8bit_ZZZzz | 18:57 | |
| *** ashishjain has joined #kolla | 18:58 | |
| *** bmace has quit IRC | 19:01 | |
| sdake_ | ashishjain_ apologies got distracted | 19:02 |
| sdake_ | ashishjain_ docker exec haproxy ls /var/log | 19:02 |
| ashishjain_ | sdake: here are the logs http://paste.openstack.org/show/478378 | 19:02 |
| sdake_ | you are looking for the haproxy log | 19:02 |
| sdake_ | right | 19:02 |
| ashishjain_ | yes these are the haproxy logs | 19:02 |
| sdake_ | then docker exec haproxy cat /var/log/haproxy.log | 19:02 |
| sdake_ | whatever the file is called | 19:02 |
| sdake_ | i am not sure if its alled haproxy.log or not | 19:02 |
| ashishjain_ | i ran the command docker logs haproxy | 19:03 |
| *** signed8bit_ZZZzz is now known as signed8bit | 19:03 | |
| sdake_ | yes run docker exec haproxy ls /var/log | 19:04 |
| sdake_ | it will list some files | 19:04 |
| sdake_ | one of them is the haproxy log | 19:04 |
| sdake_ | pick the one that looks closest, it is probably /var/log/haproxy.log | 19:04 |
| sdake_ | then ue docker exec to cat it | 19:04 |
| sdake_ | then paste that | 19:04 |
| ashishjain_ | sdake there is no such log as haproxy.log | 19:05 |
| ashishjain_ | one more thing is haproxy running inside a container? | 19:06 |
| sdake_ | yes | 19:06 |
| ashishjain_ | yes it is running inside a container | 19:06 |
| sdake_ | SamYaple where i sthe haproxy log stored? I dont see it in my deployment | 19:07 |
| *** bmace has joined #kolla | 19:14 | |
| openstackgerrit | Steven Dake proposed openstack/kolla: Add registry information to quickstart.rst https://review.openstack.org/243273 | 19:14 |
| *** sdake_ is now known as sdake | 19:15 | |
| sdake | ashishjain_ not sure where the haproxy log is, suggest syncing up with SamYaple when he is around - but he isn't around atm it appears | 19:16 |
| ashishjain_ | sdake: sure I will do that | 19:16 |
| ashishjain_ | sdake: shall I retry the deployment anyways the containers are locally available | 19:17 |
| ashishjain_ | do you think that may help? | 19:17 |
| ashishjain_ | I think I will have to destory all the docker containers and it would probably clean up the complete machine? | 19:18 |
| *** sdake has quit IRC | 19:20 | |
| *** sdake has joined #kolla | 19:21 | |
| *** signed8bit has quit IRC | 19:23 | |
| *** cloudnau_ has joined #kolla | 19:26 | |
| *** thumpba has joined #kolla | 19:26 | |
| *** jasonsb has joined #kolla | 19:28 | |
| *** cloudnautique has quit IRC | 19:28 | |
| *** jasonsb has quit IRC | 19:30 | |
| *** thumpba has quit IRC | 19:32 | |
| *** tzn has quit IRC | 19:34 | |
| *** sdake has quit IRC | 19:35 | |
| *** tzn has joined #kolla | 19:36 | |
| *** sdake has joined #kolla | 19:36 | |
| SamYaple | sdake: inc0 didnt do those logs | 19:45 |
| SamYaple | sdake: loggging needs alot of work | 19:45 |
| sdake | it used to bee logged | 19:46 |
| sdake | are the logs being droped silenty now? | 19:46 |
| *** jasonsb has joined #kolla | 19:46 | |
| sdake | if so that neeeds to be rectified | 19:46 |
| sdake | SamYaple can i assign ceph USER work to you? | 19:49 |
| sdake | and neutron is availbale as well ;) | 19:50 |
| sdake | jpeeler mind taking ironic on https://blueprints.launchpad.net/kolla/+spec/drop-root | 19:51 |
| sdake | i dont know if anyone else has actually deployed it correctly | 19:51 |
| sdake | britthouser why dont you take keepalived :) | 19:52 |
| britthouser | Sure. I was thinking about taking Zaqar too? | 19:53 |
| britthouser | or Tom Sawyering someone on my team todo it. =P | 19:54 |
| *** tzn has quit IRC | 19:54 | |
| sdake | you wouldn't be able to test it so zaqar would be really easy task for someone :) | 19:55 |
| sdake | (since it doesns't involve testing) | 19:55 |
| britthouser | I know just the guy! =) | 19:56 |
| sdake | it probably takes 5 minutes of code work | 19:56 |
| sdake | for the ones with ansible plabooks those take longer becaue you have to deploy | 19:56 |
| *** achanda has joined #kolla | 19:56 | |
| sdake | bmace can i interest you in a container to securify: https://blueprints.launchpad.net/kolla/+spec/drop-root | 19:57 |
| bmace | that isn't much of a sales pitch sdake ;) | 19:58 |
| sdake | i could work on this project-config patch for python-kollaclient or i could work on secufiying continers ;) | 19:59 |
| sdake | how is that for a sales pitch :) | 19:59 |
| bmace | hrm, sounds more like blackmail, but not completely ineffective :) | 20:00 |
| sdake | take haproxy and keepalived /me love you forever | 20:00 |
| sdake | keepalived and haproxy may have to run as root - not sure | 20:00 |
| *** tzn has joined #kolla | 20:01 | |
| sdake | not really blackmail - more reality -c an only work on one thing at a time :) | 20:01 |
| *** cfarquhar has quit IRC | 20:01 | |
| sdake | but dont ssweat it, i'll get to kollacli project-config no matter what | 20:01 |
| bmace | i would actually need a good environment to test to make sure things are working appropriately and i always turn off haproxy in my AIO environment | 20:01 |
| *** ashishjain_ has quit IRC | 20:01 | |
| bmace | so those are probably the worst choices for me | 20:01 |
| sdake | oh ya i want it tested | 20:02 |
| *** dwalsh has quit IRC | 20:02 | |
| sdake | just pick a couple that you want to do if you dont mind ;) | 20:02 |
| sdake | swift would be good too, I dont know if people know how that works | 20:02 |
| bmace | yeah, i'll take a look at what is available. | 20:02 |
| *** _tzn has joined #kolla | 20:02 | |
| sdake | i'm going to do horizon now | 20:02 |
| *** tzn has quit IRC | 20:02 | |
| *** ashishjain has quit IRC | 20:02 | |
| sdake | to see if its a special snowflake given its httpd'ness | 20:03 |
| *** _tzn has quit IRC | 20:03 | |
| kfox1111 | whats the status of 1.1? | 20:03 |
| kfox1111 | I know deadlines in open source arn't a thing. just looking for a ballpark estimate. | 20:04 |
| sdake | kfox1111 define what you want for 1.1 please | 20:04 |
| kfox1111 | well, centos 1.0 containers aren't on the hub, right? | 20:04 |
| kfox1111 | due to a mariadb issue? | 20:04 |
| *** jasonsb has quit IRC | 20:04 | |
| sdake | kfox1111 atm working on getting security rolled into the containers - but hatwont be backported | 20:04 |
| sdake | so for mariadb, we can release 1.0.1 | 20:04 |
| sdake | and build and push to registry | 20:05 |
| sdake | we can do that now | 20:05 |
| sdake | (with an appropriate test of stable | 20:05 |
| kfox1111 | cool. ok. I'd like to test that out. :) | 20:05 |
| *** tzn has joined #kolla | 20:05 | |
| kfox1111 | SamYaple also mentioned doing some work with ansible to be able to start the mariadb containers after pwoer failure. would that script work with a 1.0.1 release or would it need a 1.1? | 20:06 |
| openstackgerrit | Michal Rostecki proposed openstack/kolla: [WIP] Use trusts in heat.conf https://review.openstack.org/236198 | 20:06 |
| kfox1111 | I'm hoping to get a keystone cluster up this week intended to provide production keystone services for multiple regions. | 20:07 |
| *** cfarquhar has joined #kolla | 20:07 | |
| *** cfarquhar has quit IRC | 20:07 | |
| *** cfarquhar has joined #kolla | 20:07 | |
| kfox1111 | was hoping to use kolla for it. | 20:07 |
| *** _tzn has joined #kolla | 20:09 | |
| *** tzn has quit IRC | 20:11 | |
| *** _tzn has quit IRC | 20:14 | |
| openstackgerrit | Steven Dake proposed openstack/kolla: Add registry information to quickstart.rst https://review.openstack.org/243273 | 20:16 |
| bmace | all the good ones, that i could test fairly easily, are taken :/ | 20:18 |
| sdake | kfox1111 that work i think has already been backported so it would be in 1.0.1 | 20:18 |
| *** ashishjain has joined #kolla | 20:18 | |
| sdake | i guess we have to tag 1.1.0 as a result | 20:18 |
| kfox1111 | nice. :) | 20:18 |
| sdake | SamYaple can you cnfirm if clustering has merged nto stable? | 20:18 |
| sdake | clusterer that is | 20:19 |
| kfox1111 | If Iunderstood correctly, clusterer's for rabbit. | 20:19 |
| sdake | oh | 20:19 |
| kfox1111 | I think SamYaple mentioned needing something for mariadb too, but I can't remember. | 20:19 |
| sdake | i a mpretty sure its for mariadb | 20:19 |
| sdake | nope its rabbitmq | 20:20 |
| sdake | i'm not sure what is neded for mariadb | 20:20 |
| sdake | so what yur looking for in 1.1 is lights out support then, in summary? | 20:20 |
| kfox1111 | yeah. | 20:20 |
| sdake | unliekly to be done his week | 20:21 |
| sdake | i dont knwo what is involved for maraidb | 20:21 |
| sdake | i'll sync up with sam | 20:21 |
| sdake | he is in and out atm | 20:21 |
| sdake | is there somethign else your waiting on? | 20:21 |
| kfox1111 | ok. thanks. I'll go with raw services then and look at upgrading to kolla containers when 1.1's ready. | 20:21 |
| kfox1111 | no, just that I think. | 20:21 |
| sdake | ok well i thnk we are looing at 1.1 then 1.2 | 20:22 |
| sdake | because 1.2 is bigger = including upgrade capability | 20:22 |
| sdake | 1.1 will be upgradeable to 1.2 | 20:22 |
| sdake | 1.0 will not be upgradeable to 1.1 | 20:22 |
| *** ashishjain has quit IRC | 20:23 | |
| kfox1111 | upgradability would be important too. but I was thinking there was enough pieces in there already to do it, although manually. | 20:24 |
| kfox1111 | I've done enough manual upgrades taht I wasn't too worried about it. | 20:25 |
| sdake | possible although we want it to be all automatic with minimal steps | 20:25 |
| kfox1111 | yeah. that would be awesome. :) | 20:25 |
| sdake | what it would be at this point is undocumented and unknown | 20:25 |
| sdake | i can't promise 1 step upgrade, not yet, but we will have minimal steps | 20:26 |
| *** cloudnautique has joined #kolla | 20:26 | |
| sdake | atleast from our 3 hour design session it sseemed possible to hit it in only a few steps from the consumer side | 20:26 |
| kfox1111 | nice. :) | 20:27 |
| *** thumpba has joined #kolla | 20:28 | |
| kfox1111 | even if its complicated, I doubt it will be more complicated then I already have had to do. :) | 20:28 |
| *** cloudnau_ has quit IRC | 20:28 | |
| sdake | yes but atm its sort of documented or atleast understood | 20:28 |
| sdake | when you wrap everything in containers the complexxity from a "figure it out from scratch" perspective goes way up | 20:28 |
| *** thumpba has quit IRC | 20:28 | |
| *** thumpba has joined #kolla | 20:29 | |
| sdake | so SamYaple i would be satisified at limiting 1.1 to lights out recovery | 20:29 |
| sdake | SamYaple if you think there are other things we need to cover in 1.1 let me know | 20:29 |
| sdake | i thnk we can cover those other things in backports to 1.2 | 20:29 |
| sdake | becaue i think at people are really witing on the lights out recovery solutions | 20:30 |
| sdake | while the upgrades are "down the road" | 20:30 |
| sdake | I dont recall what else we had slated for 1.1 | 20:30 |
| kfox1111 | yeah. most of the upgrade stuff is either, tweak config files, do db upgrade scripts, or stateless things. | 20:30 |
| kfox1111 | the config stuf isn't too bad so long as I'm careful with tracking the git changes, and then folding the changes back into the overrides as needed. | 20:31 |
| kfox1111 | the db upgrade stuff I can do manually, and the stateless stuff is just replacing a container with a newer one probably? | 20:31 |
| *** thumpba has quit IRC | 20:33 | |
| openstackgerrit | Steven Dake proposed openstack/kolla: Add registry information to quickstart.rst https://review.openstack.org/243273 | 20:35 |
| britthouser | sdake for 1.1 we had slated: ovsdb container, rabbit clusterer, no pre-emption on keepalived, pin ceph to specific release on centos, and kolla to write its current version to /etc/koll/release | 20:36 |
| britthouser | at the bottom of https://etherpad.openstack.org/p/kolla-mitaka-upgrade | 20:36 |
| *** ashishjain has joined #kolla | 20:36 | |
| britthouser | apparently I missed the mariadb lights out when making that list. =) | 20:37 |
| sdake | ya i dont recall tht either | 20:37 |
| sdake | what is preemption on keepalived? | 20:37 |
| sdake | ceph is pinned on centos | 20:37 |
| sdake | i wasn't feeling wlel during that session, quesy so may have not mentioned that during the session | 20:37 |
| sdake | sam is working on a version write I think I saw a review for one | 20:38 |
| sdake | those all seem reasonable to tackle this week for 1.1 minus the preemption for keepalived, no idea what that is | 20:38 |
| britthouser | IIRC, SamYaple said that after you had keepalive do a failover from primary to standby, when primary comes back up, it takes back over primary role causing another uncessary failover. | 20:38 |
| sdake | oh, i was late to that session | 20:38 |
| britthouser | IOS call this pre-emption | 20:39 |
| britthouser | not sure what keepalive calls it. | 20:39 |
| sdake | got it | 20:39 |
| sdake | well i dont know how to fix that:) | 20:39 |
| britthouser | in IOS its a one liner | 20:39 |
| sdake | not sure it matters for 1.1 - not a compatibility breaking change | 20:39 |
| britthouser | 'no pre-empt' | 20:39 |
| britthouser | no idea for keepalive | 20:39 |
| sdake | what is ios | 20:40 |
| britthouser | IOS - the software that run on cisco routers | 20:40 |
| sdake | i thought thatss what you meant | 20:40 |
| ashishjain | sdake: the issue seems to have resolved now and I am able to use the ip addr 10.0.0.254 | 20:40 |
| britthouser | sorry...that is where my VRRP knowledge comes from. not very keep on keepalive implementation. | 20:41 |
| ashishjain | sdake: I am just trying to check the sanity of my installation, by running few neutron,nova commands | 20:41 |
| ashishjain | when I run the command "neutron agent-list" | 20:42 |
| sdake | ashishjain what was the problem? | 20:42 |
| ashishjain | I get the result as "The service catalog is empty" which should not happen | 20:42 |
| sdake | yes you may need to redeploy now that you hae fixed your problem, your cluster probablly didn' bootstrap properly | 20:42 |
| sdake | why that happens - i dont know, what did you fix to get your system operational? | 20:43 |
| ashishjain | sdake: I modified my home router and added a static ip addr for my mac address | 20:43 |
| sdake | ok, well go ahead and run tools/cleanup-containers | 20:43 |
| ashishjain | earlier I had done the same using /etc/network/interfaces | 20:43 |
| sdake | and tools/cleanup-images | 20:43 |
| sdake | then redeploy | 20:44 |
| ashishjain | sdake: okay I will redeploy it thanks | 20:44 |
| sdake | make sure to cleanup your broken env first | 20:45 |
| ashishjain | yeah | 20:45 |
| sdake | because bootstrap wont run twice | 20:45 |
| sdake | and you need it to run the firsst time :) | 20:45 |
| sdake | and dont deploy on wlan imo :) | 20:45 |
| ashishjain | tools/cleanup-images will remove all the existing images? | 20:47 |
| ashishjain | not using wlan may be a problem will just give it a try and if not than will try to use the physical cable | 20:48 |
| sdake | i am not sure what it does ash in master | 20:48 |
| sdake | that is what my copy does | 20:48 |
| sdake | are ou using stable or master? | 20:49 |
| sdake | ya cleanup-images probably doesnt need to be run | 20:49 |
| sdake | you can avoid that ashishjain | 20:49 |
| sdake | i run that because i do ha deploys and it makes my environment more correct | 20:49 |
| sdake | but if you deploy a registry in a container, it will also remove that container, which is probably not what you want :) | 20:50 |
| ashishjain | master I guess so as I have cloned the repo "https://git.openstack.org/openstack/kolla" | 20:50 |
| ashishjain | no I am not using the docker registry option currently | 20:50 |
| sdake | yup your using master | 20:50 |
| *** achanda has quit IRC | 20:50 | |
| ashishjain | alright clean up is done and I will kickstart the deploy again | 20:51 |
| openstackgerrit | Steven Dake proposed openstack/kolla: Add registry information to quickstart.rst https://review.openstack.org/243273 | 20:53 |
| *** jasonsb has joined #kolla | 20:53 | |
| *** akwasnie has joined #kolla | 20:54 | |
| openstackgerrit | Merged openstack/kolla: Remove unused cinder-data container https://review.openstack.org/242907 | 20:56 |
| *** tzn has joined #kolla | 20:59 | |
| *** tzn has quit IRC | 21:00 | |
| *** jasonsb has quit IRC | 21:00 | |
| *** jasonsb has joined #kolla | 21:02 | |
| *** daneyon has joined #kolla | 21:05 | |
| *** daneyon has quit IRC | 21:06 | |
| *** daneyon has joined #kolla | 21:07 | |
| *** daneyon_ has quit IRC | 21:08 | |
| *** _tzn has joined #kolla | 21:12 | |
| *** thumpba has joined #kolla | 21:12 | |
| *** sdake has quit IRC | 21:12 | |
| *** sdake_ has joined #kolla | 21:13 | |
| *** daneyon_ has joined #kolla | 21:17 | |
| *** mimizone has joined #kolla | 21:18 | |
| *** daneyon has quit IRC | 21:20 | |
| *** gfidente|afk has quit IRC | 21:22 | |
| *** mimizone has quit IRC | 21:26 | |
| *** jasonsb_ has joined #kolla | 21:27 | |
| *** ashishjain has quit IRC | 21:29 | |
| *** thumpba has quit IRC | 21:29 | |
| *** mimizone has joined #kolla | 21:30 | |
| *** jasonsb has quit IRC | 21:30 | |
| *** thumpba has joined #kolla | 21:33 | |
| *** thumpba has quit IRC | 21:34 | |
| *** thumpba has joined #kolla | 21:36 | |
| *** thumpba has quit IRC | 21:38 | |
| *** thumpba has joined #kolla | 21:39 | |
| *** achanda has joined #kolla | 21:40 | |
| *** achanda has quit IRC | 21:41 | |
| *** thumpba has quit IRC | 21:42 | |
| *** mimizone has quit IRC | 21:44 | |
| sdake_ | mandre about? | 21:44 |
| sdake_ | harmw about? | 21:44 |
| harmw | sdake_: make it quick :) | 21:51 |
| harmw | shoot | 21:51 |
| *** mimizone has joined #kolla | 21:52 | |
| asalkeld | i am tempted to make a joke about not putting yourself in harm's way ( harmw ) ;) | 21:55 |
| *** jasonsb has joined #kolla | 21:55 | |
| harmw | please don't | 21:57 |
| harmw | :p | 21:57 |
| openstackgerrit | Merged openstack/kolla: Add registry information to quickstart.rst https://review.openstack.org/243273 | 21:58 |
| *** jasonsb_ has quit IRC | 21:59 | |
| rhallisey | lol | 22:01 |
| sdake_ | asalkeld interested in taking a copule security workitems on this security bluepprint? | 22:04 |
| asalkeld | sure, where is it? | 22:04 |
| sdake_ | https://blueprints.launchpad.net/kolla/+spec/drop-root | 22:06 |
| sdake_ | look at lat patch - shuld be easy enough to copy | 22:06 |
| asalkeld | k | 22:06 |
| sdake_ | just add your name to work item list | 22:06 |
| sdake_ | tia | 22:06 |
| sdake_ | want to crank this stuff out by friday | 22:06 |
| sdake_ | then we can get cracking on harder work - ugprades ! | 22:06 |
| asalkeld | sure | 22:08 |
| asalkeld | took 3 | 22:08 |
| *** rhallisey has quit IRC | 22:11 | |
| *** achanda has joined #kolla | 22:12 | |
| sdake_ | cool thanks | 22:15 |
| sdake_ | ceph hard one | 22:16 |
| sdake_ | may want samyaple to tackle that one | 22:16 |
| sdake_ | since he is our on-hand ceph go-to guy :) | 22:16 |
| asalkeld | ok, i'll swap | 22:18 |
| sdake_ | still a couple unasigned workitems in that blueprint above if anyone is looking for some work to tackle ;) | 22:23 |
| *** sdake_ is now known as sdake | 22:34 | |
| *** dims_ has joined #kolla | 22:34 | |
| *** dims has quit IRC | 22:35 | |
| *** sdake_ has joined #kolla | 22:45 | |
| *** sdake has quit IRC | 22:47 | |
| sdake_ | asalkeld di you author https://review.openstack.org/#/c/242911/ | 22:47 |
| sdake_ | i see michal did some work there | 22:48 |
| sdake_ | asalkeld we dont permit authors to +2 their changes if they were at all involved the patch createion process | 22:49 |
| sdake_ | even if they are only a co-author line | 22:49 |
| *** absubram has quit IRC | 22:55 | |
| *** mimizone has quit IRC | 22:58 | |
| *** _tzn has quit IRC | 23:00 | |
| *** tzn has joined #kolla | 23:02 | |
| *** tzn has quit IRC | 23:04 | |
| *** achanda has quit IRC | 23:05 | |
| *** mimizone has joined #kolla | 23:07 | |
| *** tzn has joined #kolla | 23:09 | |
| *** tzn has quit IRC | 23:09 | |
| *** tzn has joined #kolla | 23:10 | |
| *** blahRus has quit IRC | 23:14 | |
| asalkeld | sdake_: sry, wow heat is way more relaxed | 23:16 |
| sdake_ | no reaon to apologize | 23:16 |
| sdake_ | just slightly more strict | 23:16 |
| sdake_ | it is because we are more young | 23:16 |
| *** sdake_ is now known as sdake | 23:17 | |
| asalkeld | seems upside down | 23:17 |
| *** akwasnie has quit IRC | 23:17 | |
| sdake | yes I get your argument :) | 23:18 |
| sdake | a young project should be more relaxed ;) | 23:18 |
| sdake | vs a mature project | 23:18 |
| *** jasonsb has quit IRC | 23:25 | |
| openstackgerrit | Steven Dake proposed openstack/kolla: Drop root for heat https://review.openstack.org/243350 | 23:25 |
| *** sdake_ has joined #kolla | 23:29 | |
| *** sdake has quit IRC | 23:29 | |
| *** achanda has joined #kolla | 23:37 | |
| *** sdake_ has quit IRC | 23:37 | |
| *** achanda has quit IRC | 23:38 | |
| *** sdake has joined #kolla | 23:38 | |
| openstackgerrit | Hui Kang proposed openstack/kolla: Fix neutron bootstrap https://review.openstack.org/242777 | 23:54 |
| *** mbound has quit IRC | 23:58 | |
| *** mimizone has quit IRC | 23:58 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!