| openstackgerrit | Steven Dake proposed stackforge/kolla: Remove compose from documenation https://review.openstack.org/213401 | 00:03 |
|---|---|---|
| *** sdake_ has joined #kolla | 00:04 | |
| *** sdake has quit IRC | 00:07 | |
| *** ccrouch has quit IRC | 00:08 | |
| *** sdake has joined #kolla | 00:20 | |
| *** sdake_ has quit IRC | 00:23 | |
| *** achanda has quit IRC | 00:26 | |
| openstackgerrit | Steven Dake proposed stackforge/kolla: Add Ansible support for Heat https://review.openstack.org/205528 | 00:59 |
| *** weiyu_ has joined #kolla | 01:00 | |
| openstackgerrit | Steven Dake proposed stackforge/kolla: WIP: Add Ansible support for Heat https://review.openstack.org/205528 | 01:12 |
| *** erkules_ has joined #kolla | 01:26 | |
| *** erkules has quit IRC | 01:29 | |
| *** unicell1 has quit IRC | 01:52 | |
| *** unicell has joined #kolla | 01:52 | |
| *** banix has quit IRC | 02:09 | |
| *** unicell1 has joined #kolla | 02:12 | |
| *** unicell has quit IRC | 02:13 | |
| sdake | samyaple are you alive | 02:20 |
| *** weiyu_ has quit IRC | 02:22 | |
| *** weiyu_ has joined #kolla | 02:22 | |
| *** klint has joined #kolla | 02:50 | |
| SamYaple | sdake: yea | 03:18 |
| sdake | hey sam | 03:19 |
| sdake | mind taking over that bootstrap removal patches for nova and neutron | 03:19 |
| sdake | I've got a demo coming up | 03:19 |
| sdake | looks bit sloppy with that in there | 03:19 |
| SamYaple | yep | 03:19 |
| SamYaple | i was waiting to hear back form coolsvap|away but ill do it | 03:19 |
| sdake | cool thanks i doubt he would mind | 03:20 |
| sdake | i've almost got heat ported over to the new structure | 03:21 |
| SamYaple | oh | 03:21 |
| SamYaple | i have code for that... | 03:21 |
| SamYaple | it doesnt matter | 03:21 |
| sdake | is it complete? | 03:21 |
| SamYaple | its being held up by the shade modules anyway | 03:21 |
| SamYaple | yea except for the shade modules in the project | 03:22 |
| sdake | i've got the gate working for templating for binary and source | 03:22 |
| sdake | hammered on it all day | 03:24 |
| sdake | painful | 03:24 |
| SamYaple | yea i see alot of patches from you | 03:24 |
| sdake | lots of easy reviews in the commit queue | 03:24 |
| sdake | did zaqar and murano | 03:24 |
| sdake | we have 5 contianers left | 03:24 |
| sdake | and a slew of reviews outstanding | 03:25 |
| SamYaple | mandre: all these distros aren't source-based distros. There is no way to install these other applications from source and then all the deps etc etc. You would be doing a LFS build. I don't know what there is to discuss on that point to be honest | 03:26 |
| SamYaple | the openstack stuff can be installed via python/pip which is the only 'from-source' anyone cares about when refering to openstack | 03:26 |
| sdake | samyaple i have a few technical quesitons about how the ansible code currently operates, have a few minutes? | 03:37 |
| SamYaple | sure | 03:39 |
| SamYaple | but first | 03:39 |
| SamYaple | http://gliderlabs.viewdocs.io/docker-alpine/usage/ | 03:39 |
| SamYaple | look at the virtual packages section! | 03:39 |
| SamYaple | me and a coworker are building kolla images around alpine | 03:39 |
| sdake | ya alpine is the hot new hotness apparently | 03:40 |
| sdake | real2m58.870s | 03:41 |
| sdake | time spent deploying heat | 03:41 |
| sdake | so read that article | 03:41 |
| SamYaple | ok | 03:41 |
| sdake | ok questions | 03:42 |
| sdake | if a new container is avaialble in the registry and i run a deploy operation does the new image get started? | 03:42 |
| SamYaple | yes | 03:42 |
| SamYaple | the 'reloaded' flag for the ansible module/docker-py | 03:42 |
| SamYaple | that means if anything is different (image hash/volumes/env) reload the image | 03:43 |
| SamYaple | which means new container | 03:43 |
| sdake | is there a way to query the current hash for every container in the deployment if we wanted that feature? | 03:43 |
| SamYaple | no ansible module for it but I could write one no problem | 03:43 |
| SamYaple | ansible isnt good at returning info though | 03:43 |
| sdake | i think that will be important | 03:43 |
| sdake | maybe not immediately | 03:44 |
| sdake | but down the road | 03:44 |
| SamYaple | i dont think you really want ansible for that though | 03:44 |
| SamYaple | ansible does things, it doesnt return things | 03:44 |
| SamYaple | lol i have a branch called 'make_sdake_happy' | 03:45 |
| sdake | is there a way to mirror an external registry with an internal one | 03:45 |
| SamYaple | sdake: what do you mean? | 03:46 |
| sdake | i mean take docker registry and load it into my internal registry | 03:46 |
| sdake | like docker pull and push in one :) | 03:46 |
| SamYaple | docker pull/push? | 03:46 |
| SamYaple | ah i dont think so | 03:46 |
| SamYaple | i have seen some talk about it but its not a priority | 03:46 |
| SamYaple | its talk related to fixing v2 registry | 03:47 |
| sdake | would docker pull followed by push do the job? | 03:47 |
| SamYaple | no | 03:47 |
| SamYaple | docker pull/docker tag/docker push | 03:47 |
| sdake | never used docker tag | 03:48 |
| SamYaple | its how you tag images | 03:48 |
| sdake | i gathered that by the name | 03:48 |
| SamYaple | just making sure | 03:48 |
| sdake | but if you pull, you pull latest | 03:48 |
| SamYaple | no | 03:48 |
| sdake | why does it need to be tagged lateset agin? | 03:48 |
| SamYaple | if you git pull with no tag it will try to pull the 'latest' tag | 03:48 |
| SamYaple | which we dont have at all | 03:48 |
| SamYaple | if you docker pull with a tag, it will pull that tag | 03:49 |
| SamYaple | you can even docker pull a hash | 03:49 |
| sdake | can you add a compute node to the inventory, run deploy, and that compute node will be deployed and operational? | 03:50 |
| SamYaple | since you have asked that question at least twice in the past 3 days, can you write down the answer? | 03:50 |
| SamYaple | 'yes' | 03:50 |
| sdake | just verifying :0 | 03:51 |
| sdake | ;) i mean | 03:51 |
| SamYaple | =O | 03:51 |
| sdake | if i kill the main ha proxy, how long does it take to recover to a redundant haproxy? | 03:51 |
| SamYaple | thats controlled by the keepalived script/values | 03:51 |
| SamYaple | hold | 03:52 |
| sdake | thnks forchecking | 03:52 |
| SamYaple | https://github.com/stackforge/kolla/blob/master/ansible/roles/haproxy/templates/keepalived.conf.j2#L3-5 | 03:52 |
| SamYaple | so it takes 2 intervals of 2 seconds of being down to switch | 03:52 |
| SamYaple | so 4 seconds | 03:52 |
| sdake | nice | 03:53 |
| sdake | ahve you tested that? | 03:53 |
| SamYaple | it takes 20 seconds of successful checks to be 'up' | 03:53 |
| SamYaple | yup | 03:53 |
| sdake | how wuld you feel about putting the ports in a separate config file | 03:54 |
| SamYaple | cant hot reload them that way | 03:54 |
| sdake | why not? | 03:54 |
| SamYaple | the hot reload only works with a single file according to inc0 | 03:54 |
| sdake | i see | 03:55 |
| sdake | thats annying :( | 03:55 |
| SamYaple | I haven't verified that, but it is haproxy, so i wouldn't doubt it | 03:55 |
| sdake | is the keepaived interval tuff configurable? | 03:55 |
| SamYaple | nope | 03:55 |
| SamYaple | the only things configurable are the ini-style configs | 03:55 |
| SamYaple | the best I can offer is a complete override file | 03:56 |
| SamYaple | not configurable, but entire file is overridable | 03:56 |
| sdake | dont like that much | 03:56 |
| sdake | someone likely to fubar their system | 03:57 |
| sdake | down the road there is no other way to do the job of confuginrg just those two things? | 03:57 |
| SamYaple | we can't stop people from breaking thier system | 03:58 |
| SamYaple | if they have need to change these things, they can break thier system | 03:58 |
| SamYaple | no matter what we do they can break thier system | 03:58 |
| SamYaple | i wont make decisions around trying ot prevent people from breaking the system | 03:59 |
| SamYaple | any merge type setup on the other files will be way more likely to break the stuff than a full override file | 03:59 |
| sdake | a perfect system cannot be broken ;) | 03:59 |
| SamYaple | there is no such thing as a perfect system. i wont entertain you on this | 04:00 |
| sdake | what does an ovveride file look like for keepalived? | 04:01 |
| SamYaple | a complete override. like the template but with your options | 04:02 |
| sdake | the override template is parsed by ansible to fill in values? | 04:03 |
| SamYaple | yea its all run through ansible templates | 04:03 |
| sdake | one option is to write out the override file with just the two things we care about | 04:04 |
| *** banix has joined #kolla | 04:04 | |
| sdake | where is this template - i dont see i nsource base | 04:04 |
| SamYaple | 03:52 < SamYaple> https://github.com/stackforge/kolla/blob/master/ansible/roles/haproxy/templates/keepalived.conf.j2#L3-5 | 04:05 |
| sdake | youcoudln't do interval {{ openstack_keepalive_interval }} there? | 04:06 |
| SamYaple | sure. but i dont want that. youll never be able to template all the options. If we go down the path of templating them, we will be adding new things forever. | 04:07 |
| sdake | or olla_keepalive_interval? | 04:07 |
| SamYaple | This is the OSAD paradox | 04:07 |
| SamYaple | what about an option that isnt even o nhere? | 04:07 |
| SamYaple | what about unicast vs multicast? | 04:07 |
| SamYaple | having a full override solves everything | 04:07 |
| SamYaple | templating causes overhead | 04:07 |
| sdake | if my 10 years of experience with ha tells me anything people will want to tune those values | 04:08 |
| sdake | without having to be a rocket surgeon | 04:08 |
| SamYaple | right they will | 04:09 |
| SamYaple | and they will want to tune options we dont have | 04:09 |
| sdake | what other shit did osad end up templating in? | 04:09 |
| SamYaple | thats my point | 04:09 |
| sdake | i want to keep config options to an absolute minimum | 04:09 |
| SamYaple | you cant have it both ways | 04:10 |
| sdake | thatis why i asked about osad paradox | 04:10 |
| SamYaple | you know all about it | 04:10 |
| sdake | all the shit they hve tunable is because customers asekd for it | 04:10 |
| SamYaple | a dev-cycle to template an option | 04:10 |
| *** britthouser has joined #kolla | 04:11 | |
| SamYaple | full override == no dev-cycle | 04:11 |
| SamYaple | full operators control | 04:11 |
| SamYaple | exactly what it needs to be | 04:11 |
| SamYaple | templating shit will get us nowhere fast and no one will be happy about it | 04:11 |
| sdake | interesting | 04:11 |
| sdake | one last technical question | 04:11 |
| sdake | what is the echnical difference between external address and internal address | 04:11 |
| sdake | atm i set them to the same thing | 04:12 |
| SamYaple | as answered before, the external is the address that sets the 'public_endpoint' for openstack | 04:12 |
| *** britthou_ has joined #kolla | 04:12 | |
| SamYaple | in a public cloud this would be public addresses | 04:12 |
| sdake | sorry i mut have missed yur answer | 04:12 |
| sdake | that maks sense | 04:12 |
| sdake | i dc alot to switch to email network | 04:12 |
| SamYaple | but you have responded with 'that makes sense' before | 04:13 |
| sdake | what serves it? haproxy? | 04:13 |
| SamYaple | nothing serves the external address at all | 04:13 |
| SamYaple | that would be on the deployer to setup | 04:13 |
| SamYaple | 1:1 nat maybe | 04:13 |
| SamYaple | thats what I would do | 04:13 |
| sdake | i recall you saying that would be onthe deployer before | 04:14 |
| sdake | how does a 1;1 natwork | 04:14 |
| SamYaple | It _should_ be a dns address | 04:14 |
| SamYaple | that would be the proper way to do it | 04:14 |
| SamYaple | so its not even an ip | 04:14 |
| SamYaple | but it _can_ be an ip | 04:15 |
| sdake | dns gets converted to ip by resolver lib | 04:15 |
| *** britthouser has quit IRC | 04:15 | |
| SamYaple | on the client | 04:15 |
| SamYaple | notihng to do with the deployment | 04:15 |
| sdake | here is where i get stuck - nothing forwards that traffic to the internal network | 04:16 |
| sdake | that would be done by a 1:1 nat? | 04:16 |
| *** openstack has joined #kolla | 04:19 | |
| sdake | design makes sense | 04:20 |
| sdake | i think we need to think more about implementing a 1:1 nat or documenting how one could be implemented | 04:20 |
| sdake | otherwise people a re going to go wtf when they have their privat enetwork isolated from their endpoint network | 04:20 |
| SamYaple | i have no problem with whatever is in docs | 04:20 |
| SamYaple | we cannot implement it in kolla though... | 04:21 |
| sdake | cool I'll start a faq document | 04:23 |
| sdake | these are questions i get sked over an over | 04:23 |
| sdake | which is why im asking you | 04:23 |
| *** weiyu_ has quit IRC | 04:25 | |
| sdake | i'll consult daneyon on docs for a 1:1 nat solution | 04:26 |
| sdake | he should probably have an idea | 04:26 |
| sdake | he is a networking genius | 04:26 |
| SamYaple | im not sure what you really want to document there | 04:26 |
| sdake | how to isolate your internal network from your endpoint network essentially | 04:27 |
| sdake | that is the whole purpose of external_network | 04:27 |
| SamYaple | thats realy not whats happening | 04:27 |
| SamYaple | at all | 04:27 |
| SamYaple | you seem to have external network and public endpoint confused | 04:28 |
| sdake | possible expand pls | 04:28 |
| SamYaple | i dont know what you want me to expand on. | 04:28 |
| SamYaple | i dont know what you dont know | 04:29 |
| sdake | lets take my network for example | 04:29 |
| sdake | my public endpoints are 192.168.1.149 | 04:29 |
| sdake | my prviate network is 192.168.1.z | 04:29 |
| sdake | someone could connect to 192.168.1.100 3306 and DOS my database | 04:29 |
| sdake | 100 is my machine where kolla is running | 04:30 |
| SamYaple | on your private network, yes | 04:30 |
| SamYaple | but your public endpoint is also on your private, thats your choice | 04:30 |
| SamYaple | nothing to document there | 04:30 |
| *** diogogmt has quit IRC | 04:30 | |
| SamYaple | if you want it available make it a public and route it | 04:30 |
| SamYaple | this is basic stuff not a kolla issue | 04:30 |
| sdake | so a router handles that? | 04:31 |
| SamYaple | or any number of other methods | 04:31 |
| sdake | cool never played with a router | 04:31 |
| sdake | have no idea how they work ;-) | 04:31 |
| sdake | does the router forward traffic to the VIF blocking ports that are not pubilc endpoints? | 04:33 |
| SamYaple | i have no idea what you set your router up to do | 04:33 |
| sdake | i mean in a real deployment | 04:33 |
| SamYaple | doesnt change the answer | 04:34 |
| sdake | i dont havea router in my home setup - just cable modem and wireless "router" | 04:34 |
| sdake | it does have a nat built in and port forwarding | 04:34 |
| sdake | i guess i could assign openstack ports to my vif | 04:35 |
| sdake | and set my openstack external to my public ip | 04:35 |
| sdake | if that is the model, then i got it | 04:35 |
| sdake | standard networking | 04:36 |
| sdake | is that the model? | 04:36 |
| SamYaple | there is no model. you can do that yes | 04:37 |
| SamYaple | its a deployers decision | 04:37 |
| sdake | cool thanks for bearing with me :) | 04:38 |
| sdake | your right not an impelemtnation issue for kolla to sort out | 04:41 |
| sdake | definately a documentation issue | 04:41 |
| sdake | basically the only answer I can see is nat | 04:42 |
| sdake | boy networking is hard | 04:44 |
| sdake | but as my wife says, "your company didn't build its empire because networking is easy." | 04:44 |
| sdake | samyaple here is the templating gate: https://review.openstack.org/#/c/213538/ | 04:48 |
| sdake | 16 and 13 min runtimes | 04:48 |
| sdake | do you know why a knife must have a sheath? | 04:57 |
| sdake | victory got heat stack-list working | 04:59 |
| openstackgerrit | Sam Yaple proposed stackforge/kolla: Remove start.yml abstraction for kolla-ansible https://review.openstack.org/213600 | 05:04 |
| *** daneyon_ has quit IRC | 05:13 | |
| *** daneyon has joined #kolla | 05:14 | |
| openstackgerrit | Sam Yaple proposed stackforge/kolla: Remove Ansible abstraction for Nova https://review.openstack.org/211445 | 05:15 |
| sdake | say samyaple should the registry have a data container associated with it | 05:18 |
| sdake | glance-registry | 05:18 |
| *** weiyu has joined #kolla | 05:18 | |
| SamYaple | for what? | 05:18 |
| sdake | i dont know | 05:18 |
| sdake | i have no idea what it does | 05:18 |
| sdake | mandre mentioned it | 05:19 |
| SamYaple | only if it has persistent data | 05:19 |
| SamYaple | where is the persistent data stored, api or registry? | 05:20 |
| sdake | no idea | 05:21 |
| sdake | http://docs.openstack.org/juno/config-reference/content/section_glance-registry.conf.html | 05:22 |
| *** banix has quit IRC | 05:23 | |
| sdake | looks lik eit uses edb for storing images | 05:24 |
| sdake | samyaple can you review this please | 05:30 |
| sdake | https://review.openstack.org/#/c/213412/ | 05:30 |
| *** achanda has joined #kolla | 05:30 | |
| *** weiyu has quit IRC | 05:30 | |
| sdake | and https://review.openstack.org/#/c/213401/ | 05:30 |
| sdake | TIA | 05:30 |
| *** coolsvap|away is now known as coolsvap | 05:32 | |
| sdake | yo coolsvap | 05:36 |
| coolsvap | sdake, hi | 05:36 |
| coolsvap | sorry was away traveling | 05:36 |
| coolsvap | just landed in seattle this evening | 05:36 |
| sdake | nice | 05:36 |
| sdake | i like seattle | 05:36 |
| coolsvap | was at the containerday | 05:36 |
| coolsvap | my first time | 05:36 |
| sdake | how was it | 05:37 |
| coolsvap | it was nice, do you know dave nielsen? | 05:37 |
| sdake | doesn't ring a bell | 05:37 |
| coolsvap | good lightning talks | 05:37 |
| sdake | what was his about | 05:37 |
| coolsvap | and unconferences | 05:37 |
| sdake | anything on openstack deployment? | 05:37 |
| coolsvap | he organized it, i think he's kinda doing the container days at different cities | 05:38 |
| coolsvap | very litte | 05:38 |
| coolsvap | the docker volume-driver implementation -> flocker | 05:38 |
| coolsvap | from clusterhq | 05:38 |
| coolsvap | but their openstack integration is at very basic state | 05:39 |
| coolsvap | they are still playing around vagrant | 05:39 |
| coolsvap | but good concept for stateful applications on docker | 05:39 |
| coolsvap | https://github.com/ClusterHQ/flocker | 05:40 |
| coolsvap | apart from that there were basic talks on docker | 05:42 |
| sdake | familiar with it | 05:42 |
| coolsvap | hmm | 05:43 |
| sdake | samyaple you ever tried out bluebox'es thing? | 05:46 |
| sdake | or looked at it | 05:46 |
| * coolsvap sleeping now or i will not get over the jet lag | 05:51 | |
| coolsvap | and will sleep in the conf tomorrow :) | 05:52 |
| sdake | enjoy | 05:54 |
| *** coolsvap is now known as coolsvap|away | 05:55 | |
| *** weiyu has joined #kolla | 06:09 | |
| sdake | working heat -> http://ur1.ca/ngjg5 -> http://paste.fedoraproject.org/255729/43979200 | 06:13 |
| sdake | great stack delete doesn't do anything | 06:16 |
| * sdake shakes fists | 06:16 | |
| openstackgerrit | Steven Dake proposed stackforge/kolla: WIP: Add Ansible support for Heat https://review.openstack.org/205528 | 06:26 |
| sdake | samyaple your patch appears busted http://ur1.ca/ngjlc -> http://paste.fedoraproject.org/255733/97929931 | 06:30 |
| *** akscram has quit IRC | 06:32 | |
| *** akscram has joined #kolla | 06:33 | |
| sdake | bedtime | 06:43 |
| sdake | later | 06:43 |
| *** sdake_ has joined #kolla | 06:44 | |
| *** sdake has quit IRC | 06:47 | |
| *** inc0 has joined #kolla | 06:56 | |
| inc0 | good morning kolla | 06:56 |
| SamYaple | morning inc0 | 06:58 |
| sdake_ | hey inc0 | 06:58 |
| sdake_ | inc0 5 containers need jinja-2ing :) | 06:58 |
| inc0 | hehe, allright, I have 3 hrs but should be enough for them;) | 06:59 |
| sdake_ | atleat a few ;) | 06:59 |
| inc0 | can I have link to etherpad please? | 06:59 |
| sdake_ | takes aout 10-15 mins per i think | 06:59 |
| sdake_ | yup sec | 06:59 |
| sdake_ | https://etherpad.openstack.org/p/kolla-dockerfile-template | 07:00 |
| inc0 | last week I was fighting with internal stuff, few commits would be good for my mental health | 07:00 |
| sdake_ | make that 6 | 07:00 |
| *** ccrouch has joined #kolla | 07:01 | |
| sdake_ | goal with these is they match existing implementation, not necessarily work | 07:01 |
| sdake_ | mongodb being the exception | 07:02 |
| sdake_ | since that is a dependency for ceilometer | 07:02 |
| inc0 | lol, I'd feel bad with commiting whatever not working;) | 07:02 |
| sdake_ | well you can't treally test them | 07:02 |
| sdake_ | since config-internal is long gone | 07:02 |
| sdake_ | but you can et that they build ;) | 07:02 |
| inc0 | hmm, but these 6 have people assigned | 07:03 |
| inc0 | bigger question tho | 07:04 |
| inc0 | did anyone actually setup kolla-kilo openstack? | 07:05 |
| inc0 | sdake_, you've mentioned that people needs that | 07:05 |
| sdake_ | you mean besides me? | 07:05 |
| -openstackstatus- NOTICE: Gerrit is currently under very high load and may be unresponsive. infra are looking into the issue. | 07:05 | |
| inc0 | well I didn't know you did that, but good to know | 07:05 |
| *** ccrouch has quit IRC | 07:05 | |
| inc0 | if I've tried source build with different tarball, I ran into problems | 07:06 |
| sdake_ | daneyon gave live demo at ods | 07:06 |
| sdake_ | the images in docker registry work | 07:06 |
| sdake_ | with the kilo banch ;-) | 07:06 |
| inc0 | I need these built locally tho | 07:06 |
| inc0 | did you use source build or change repos for binary? | 07:07 |
| SamYaple | inc0: he is talking about the kilo branch | 07:07 |
| sdake_ | kilo only had binary building | 07:07 |
| SamYaple | not master | 07:07 |
| SamYaple | we dont do kilo in master | 07:07 |
| sdake_ | bedtime | 07:08 |
| sdake_ | later | 07:08 |
| inc0 | I see, you mean https://github.com/stackforge/kolla/tree/stable/kilo ? | 07:08 |
| *** sdake_ has quit IRC | 07:08 | |
| SamYaple | yea | 07:08 |
| SamYaple | no one is deploying kilo from kilo branch | 07:08 |
| openstackgerrit | Sam Yaple proposed stackforge/kolla: Remove start.yml abstraction for kolla-ansible https://review.openstack.org/213600 | 07:11 |
| SamYaple | jeez i did that like 30 minutes ago | 07:11 |
| openstackgerrit | Sam Yaple proposed stackforge/kolla: Add heat to docker_templates https://review.openstack.org/213620 | 07:32 |
| openstackgerrit | Sam Yaple proposed stackforge/kolla: Add memcached to docker_templates https://review.openstack.org/213620 | 07:33 |
| SamYaple | jeez this stuff is taking for ever to appear | 07:39 |
| openstackgerrit | Sam Yaple proposed stackforge/kolla: Add mongodb to docker_templates https://review.openstack.org/213623 | 07:40 |
| inc0 | meh | 07:49 |
| inc0 | I'll wait with reviews till they solve this | 07:49 |
| SamYaple | ye aits unusable at the moment | 07:49 |
| *** shardy has joined #kolla | 08:09 | |
| *** achanda has quit IRC | 08:10 | |
| *** achanda has joined #kolla | 08:20 | |
| *** jmccarthy has quit IRC | 08:28 | |
| *** shardy_ has joined #kolla | 08:33 | |
| *** shardy has quit IRC | 08:34 | |
| *** achanda has quit IRC | 08:35 | |
| *** shardy_ has quit IRC | 08:38 | |
| *** shardy has joined #kolla | 08:39 | |
| *** jmccarthy has joined #kolla | 08:46 | |
| *** gfidente has joined #kolla | 08:48 | |
| *** ccrouch has joined #kolla | 08:50 | |
| *** ccrouch has quit IRC | 08:54 | |
| *** tfukushima has joined #kolla | 08:58 | |
| *** inc0 has quit IRC | 09:00 | |
| *** tfukushima has quit IRC | 09:02 | |
| *** inc0 has joined #kolla | 09:02 | |
| *** tfukushima has joined #kolla | 09:04 | |
| *** athomas has joined #kolla | 09:23 | |
| *** inc0 has quit IRC | 09:59 | |
| pbourke | SamYaple: hey got a minute? | 10:07 |
| *** tfukushima has quit IRC | 10:11 | |
| *** tfukushima has joined #kolla | 10:12 | |
| SamYaple | pbourke: yea whats up | 10:16 |
| pbourke | SamYaple: just wondering if I could bounce some of the keystone config Im trying for swift off you. Its driving me demented, have spent almost two days now trying to debug it | 10:16 |
| *** dims has joined #kolla | 10:17 | |
| SamYaple | ok | 10:17 |
| pbourke | SamYaple: far as I can see its the same as what I was using for conifg-internal, but the proxy can't seem to authenticate itself | 10:17 |
| pbourke | SamYaple: I'll put some pastebins together of the various parts | 10:17 |
| -openstackstatus- NOTICE: review.openstack.org (aka gerrit) is going down for an emergency restart | 10:19 | |
| *** ChanServ changes topic to "review.openstack.org (aka gerrit) is going down for an emergency restart" | 10:19 | |
| pbourke | SamYaple: register.yml - http://paste.openstack.org/show/418992/ | 10:23 |
| pbourke | SamYaple: defaults.yml - http://paste.openstack.org/show/419000/ | 10:23 |
| pbourke | SamYaple: proxy-server.j2 - http://paste.openstack.org/show/419001/ | 10:24 |
| pbourke | SamYaple: the conf template is the piece I find the most confusing, the combinations in examples online seem endless and no matter what I put in there I get messages about it been deprecated | 10:24 |
| pbourke | SamYaple: If I put "admin" in for all the user/pass items it seems to work, but that's not what we want | 10:25 |
| openstackgerrit | wanghao proposed stackforge/kolla: Add image tag specification in dev-quickstart.md https://review.openstack.org/213652 | 10:26 |
| SamYaple | try swiftoperator rather than username swift | 10:26 |
| pbourke | even though I'm registering with swift? | 10:26 |
| SamYaple | this isn't about what youre registering | 10:27 |
| SamYaple | its about other conf options you arent setting properly | 10:27 |
| SamYaple | if admin works, swiftoperator will | 10:27 |
| SamYaple | ill show you why | 10:27 |
| SamYaple | *** | 10:28 |
| SamYaple | no user sorry, role | 10:28 |
| SamYaple | I do question why oyu have an issue with admin though | 10:29 |
| openstackgerrit | wanghao proposed stackforge/kolla: Add image tag specification in dev-quickstart.md https://review.openstack.org/213652 | 10:29 |
| pbourke | don't we register the swift user for the services to validate tokens with keystone? | 10:30 |
| *** dims has quit IRC | 10:30 | |
| pbourke | hmm I seem to have found a working combination | 10:30 |
| *** dims has joined #kolla | 10:31 | |
| pbourke | http://paste.openstack.org/show/419003/ | 10:31 |
| pbourke | was sure I tried that before | 10:31 |
| SamYaple | make sure to drop the v2.0 on that auth_uri line | 10:32 |
| SamYaple | oh yea | 10:32 |
| SamYaple | admin isn't the tenant | 10:32 |
| openstackgerrit | Merged stackforge/kolla: Update readme to match current state https://review.openstack.org/213412 | 10:32 |
| SamYaple | its service | 10:32 |
| pbourke | yeah | 10:33 |
| pbourke | ok thanks, sometimes just going through what you have with someone seems to help :) | 10:33 |
| SamYaple | yea | 10:34 |
| SamYaple | youll want to use the new plugin too | 10:35 |
| SamYaple | not that deprecated old one | 10:35 |
| pbourke | I really do | 10:35 |
| pbourke | just need to find up to date docs | 10:35 |
| *** dims_ has joined #kolla | 10:36 | |
| SamYaple | here ya go: http://paste.openstack.org/show/419018/ | 10:36 |
| SamYaple | you wont find docs to help you | 10:36 |
| *** akwasnie has joined #kolla | 10:37 | |
| *** dims has quit IRC | 10:38 | |
| pbourke | thanks! | 10:38 |
| SamYaple | its what i dooooo | 10:39 |
| pbourke | by the time its documented it will be deprecated again :) | 10:39 |
| SamYaple | lurn2source | 10:40 |
| SamYaple | but seriously I have given up on google and openstack | 10:40 |
| SamYaple | source is way quicker | 10:40 |
| pbourke | good advice | 10:42 |
| SamYaple | fyi this is what it looks like | 10:43 |
| SamYaple | https://github.com/openstack/cinder/blob/master/cinder/common/config.py | 10:43 |
| SamYaple | most are common like that | 10:43 |
| SamYaple | also list deprecated options | 10:43 |
| *** ChanServ changes topic to "Kolla IRC meetings on Wednesday - Agenda @ https://wiki.openstack.org/wiki/Meetings/Kolla - IRC channel is *LOGGED* @ http://eavesdrop.openstack.org/irclogs/%23kolla/" | 10:46 | |
| -openstackstatus- NOTICE: Gerrit restart has resolved the issue and systems are back up and functioning | 10:46 | |
| *** dims_ has quit IRC | 10:50 | |
| *** dims has joined #kolla | 10:51 | |
| *** inc0 has joined #kolla | 10:57 | |
| *** dims has quit IRC | 10:59 | |
| *** daneyon has quit IRC | 10:59 | |
| *** daneyon has joined #kolla | 11:00 | |
| *** dims has joined #kolla | 11:00 | |
| *** dims has quit IRC | 11:06 | |
| *** dims has joined #kolla | 11:08 | |
| *** weiyu has quit IRC | 11:10 | |
| *** tfukushima has quit IRC | 11:10 | |
| *** dims_ has joined #kolla | 11:12 | |
| *** dims has quit IRC | 11:12 | |
| openstackgerrit | Alicja Kwasniewska proposed stackforge/kolla: Add Dockerfile template for swift https://review.openstack.org/213669 | 11:16 |
| *** dims_ has quit IRC | 11:17 | |
| *** dims has joined #kolla | 11:22 | |
| *** dims has quit IRC | 11:27 | |
| *** dims has joined #kolla | 11:29 | |
| *** dims_ has joined #kolla | 11:34 | |
| *** dims has quit IRC | 11:35 | |
| *** dims_ has quit IRC | 11:39 | |
| *** dims has joined #kolla | 11:40 | |
| *** shardy_ has joined #kolla | 11:43 | |
| *** shardy has quit IRC | 11:45 | |
| *** shardy_ has quit IRC | 11:49 | |
| *** dims has quit IRC | 11:49 | |
| *** shardy has joined #kolla | 11:49 | |
| *** rhallisey has joined #kolla | 11:53 | |
| *** dims has joined #kolla | 11:53 | |
| *** dims_ has joined #kolla | 11:56 | |
| *** dims has quit IRC | 11:58 | |
| *** dims_ has quit IRC | 12:04 | |
| *** dims has joined #kolla | 12:04 | |
| *** dims_ has joined #kolla | 12:08 | |
| *** dims has quit IRC | 12:10 | |
| *** tfukushima has joined #kolla | 12:11 | |
| *** dims has joined #kolla | 12:12 | |
| *** dims_ has quit IRC | 12:14 | |
| *** tfukushima has quit IRC | 12:16 | |
| *** dims_ has joined #kolla | 12:18 | |
| *** dims has quit IRC | 12:18 | |
| *** dims has joined #kolla | 12:19 | |
| *** dwalsh has joined #kolla | 12:20 | |
| openstackgerrit | Sam Yaple proposed stackforge/kolla: Remove Ansible abstraction for Nova https://review.openstack.org/211445 | 12:22 |
| *** dims_ has quit IRC | 12:23 | |
| *** dims_ has joined #kolla | 12:26 | |
| *** dims has quit IRC | 12:26 | |
| *** shardy_ has joined #kolla | 12:28 | |
| *** shardy has quit IRC | 12:29 | |
| *** dims has joined #kolla | 12:30 | |
| *** dims_ has quit IRC | 12:31 | |
| *** shardy_ has quit IRC | 12:33 | |
| *** shardy has joined #kolla | 12:34 | |
| *** dims has quit IRC | 12:35 | |
| *** dims has joined #kolla | 12:35 | |
| *** dims_ has joined #kolla | 12:38 | |
| *** dims has quit IRC | 12:40 | |
| *** cfarquhar has quit IRC | 12:42 | |
| *** dims has joined #kolla | 12:45 | |
| *** dims_ has quit IRC | 12:45 | |
| *** dims has quit IRC | 12:55 | |
| *** dwalsh has quit IRC | 12:55 | |
| *** klint has quit IRC | 12:55 | |
| *** tfukushima has joined #kolla | 12:56 | |
| *** dims has joined #kolla | 12:58 | |
| *** sdake has joined #kolla | 13:00 | |
| sdake | morning | 13:00 |
| akwasnie | hi:) | 13:04 |
| sdake | hey akwasnie! | 13:04 |
| *** britthouser has joined #kolla | 13:05 | |
| *** dims has quit IRC | 13:06 | |
| *** dims has joined #kolla | 13:09 | |
| *** dims has quit IRC | 13:14 | |
| *** dims has joined #kolla | 13:14 | |
| *** sdake_ has joined #kolla | 13:15 | |
| openstackgerrit | Alicja Kwasniewska proposed stackforge/kolla: Add Dockerfile template for designate https://review.openstack.org/213703 | 13:18 |
| *** sdake has quit IRC | 13:19 | |
| *** sdake_ has quit IRC | 13:19 | |
| *** dims has quit IRC | 13:21 | |
| pbourke | morning | 13:23 |
| *** dims has joined #kolla | 13:27 | |
| *** absubram has quit IRC | 13:29 | |
| *** britthou_ has joined #kolla | 13:36 | |
| *** dims has quit IRC | 13:36 | |
| *** stvnoyes has joined #kolla | 13:36 | |
| *** stvnoyes has left #kolla | 13:37 | |
| *** britthouser has quit IRC | 13:39 | |
| *** banix has joined #kolla | 13:41 | |
| *** diogogmt has joined #kolla | 13:41 | |
| *** dims has joined #kolla | 13:42 | |
| *** dims has quit IRC | 13:48 | |
| *** sdake has joined #kolla | 13:49 | |
| *** sdake has quit IRC | 13:49 | |
| *** sdake has joined #kolla | 13:49 | |
| *** dims has joined #kolla | 13:54 | |
| *** inc0 has quit IRC | 13:54 | |
| *** diogogmt has quit IRC | 13:54 | |
| *** dims has quit IRC | 13:59 | |
| *** dims has joined #kolla | 14:04 | |
| *** dwalsh has joined #kolla | 14:05 | |
| *** sdake has quit IRC | 14:10 | |
| *** dims has quit IRC | 14:11 | |
| *** dims has joined #kolla | 14:16 | |
| *** sdake has joined #kolla | 14:17 | |
| openstackgerrit | Steven Dake proposed stackforge/kolla: Remove compose from documenation https://review.openstack.org/213401 | 14:21 |
| sdake | jpeeler/rhallisey can you guys review that plz | 14:21 |
| sdake | or pbourke if your around | 14:21 |
| sdake | the review queue needs some love in general | 14:21 |
| pbourke | sdake: looking. I skipped reviews this morning as gerrit was acting up | 14:21 |
| *** dwalsh has quit IRC | 14:22 | |
| *** dims has quit IRC | 14:22 | |
| sdake | if we can get through the gate-from-source blueprint reviews that would help | 14:22 |
| sdake | they are mostly 1 liners | 14:22 |
| sdake | gerrit seems to be workingwell now | 14:24 |
| sdake | they probably restarted it | 14:24 |
| sdake | rhallisey what other kinds of cinder drivers can there be besdies lvm? | 14:29 |
| sdake | only ceph? | 14:29 |
| rhallisey | I figure those would be the 2 we'd support | 14:30 |
| rhallisey | there's more though | 14:30 |
| rhallisey | not sure what else | 14:30 |
| rhallisey | ec2 | 14:30 |
| *** bmace has joined #kolla | 14:38 | |
| *** britthou_ has quit IRC | 14:39 | |
| *** britthouser has joined #kolla | 14:39 | |
| openstackgerrit | Merged stackforge/kolla: Add Dockerfile template for swift https://review.openstack.org/213669 | 14:47 |
| *** diogogmt has joined #kolla | 14:48 | |
| *** dims has joined #kolla | 14:48 | |
| *** vbel has joined #kolla | 14:57 | |
| vbel | good morning/evening | 14:58 |
| bmace | good morning vbel :) | 14:58 |
| *** shardy_ has joined #kolla | 14:59 | |
| *** achanda has joined #kolla | 14:59 | |
| openstackgerrit | Merged stackforge/kolla: Fix openvswitch template https://review.openstack.org/213488 | 15:00 |
| *** shardy has quit IRC | 15:01 | |
| *** shardy_ has quit IRC | 15:04 | |
| *** achanda has quit IRC | 15:04 | |
| *** shardy has joined #kolla | 15:05 | |
| *** blahRus has joined #kolla | 15:10 | |
| *** jmccarthy has quit IRC | 15:15 | |
| *** jmccarthy has joined #kolla | 15:20 | |
| vbel | anyone could get centos-source-nova-compute running recently? | 15:25 |
| rhallisey | building from source is breaking for me atm | 15:26 |
| vbel | looks like nova-compute is missing libvirt-python package. | 15:27 |
| vbel | in binary distros it is requirement of openstack-nova-compute package | 15:27 |
| *** stevem has quit IRC | 15:35 | |
| *** smeisner has joined #kolla | 15:35 | |
| *** diogogmt_ has joined #kolla | 15:35 | |
| *** diogogmt has quit IRC | 15:37 | |
| *** diogogmt_ is now known as diogogmt | 15:37 | |
| *** vbel has quit IRC | 15:51 | |
| *** pbourke has quit IRC | 16:10 | |
| *** pbourke has joined #kolla | 16:11 | |
| pbourke | bmace: ping | 16:11 |
| *** unicell1 has quit IRC | 16:15 | |
| *** akwasnie has quit IRC | 16:23 | |
| sdake | hey guys just got back from pt whats hanging | 16:24 |
| sdake | vbel i neve rrun from source | 16:26 |
| sdake | vbel i am likely thinking it doesn't work well | 16:26 |
| sdake | rhallisey re cinder driver | 16:26 |
| sdake | i'd propose we remove the cinder specifc lvm thing | 16:27 |
| sdake | and if enabled_ceph is set, then cinder would just use that | 16:27 |
| sdake | i don't want config options unless absolutely mandatory to run the system | 16:27 |
| sdake | cinder specific stroage driver option | 16:27 |
| sdake | and if there are only two storage options lvm or ceph, ifcep his not set we know its lvm | 16:28 |
| sdake | i dont want to template anything in that isn't mandatory | 16:28 |
| sdake | ceph will be a mandatory templating | 16:28 |
| *** daneyon has quit IRC | 16:30 | |
| rhallisey | we should have local storage support though | 16:30 |
| *** daneyon has joined #kolla | 16:30 | |
| sdake | you just said lvm is the only option? | 16:31 |
| sdake | isn't that local storage or d oyou mean something else | 16:31 |
| rhallisey | ya what we have | 16:32 |
| rhallisey | just not iscsi | 16:32 |
| sdake | ok, so lets start over :) | 16:32 |
| sdake | there are going to be only 2 storage modes we will ever implement, lvm and ceph? | 16:33 |
| rhallisey | yea | 16:33 |
| rhallisey | I think just those 2 is good | 16:33 |
| sdake | if enabled_ceph==yes confiigure condier for ceph else configure for lvm | 16:33 |
| rhallisey | coorect | 16:33 |
| sdake | that allows us to remove the cinder_driver option | 16:33 |
| sdake | if someone wants a custom cinder driver they are on thei own with regards to config | 16:34 |
| sdake | ok so would you agree to a course of action of me filing a bug to remove that config option cinder_driver or whatever it is | 16:34 |
| sdake | (and you or coolsvap fixing it) | 16:35 |
| sdake | i hsould have caught it in review but idn't think of it until yesterday | 16:35 |
| *** tfukushima has quit IRC | 16:40 | |
| *** tfukushima has joined #kolla | 16:41 | |
| *** unicell has joined #kolla | 16:45 | |
| rhallisey | sdake, are you referring to -> {% if cinder_volume_driver == "lvm" %} | 16:48 |
| rhallisey | in the config | 16:49 |
| sdake | YES | 16:49 |
| sdake | just default to lvm | 16:50 |
| *** unicell has quit IRC | 16:50 | |
| sdake | and when ceph is implemented make a conditional around ceph_enabled | 16:50 |
| rhallisey | {% elif cinder_volume_driver == "ceph" %} | 16:51 |
| rhallisey | it's already there | 16:51 |
| *** mfalatic has joined #kolla | 16:51 | |
| *** mfalatic_ has joined #kolla | 16:52 | |
| sdake | i know I want to remove it | 16:52 |
| sdake | I want to remove the config option cinder_volume_driver | 16:53 |
| sdake | the les config options our software has the beetter it will be for operators | 16:53 |
| sdake | oh sorry didn't dsee the ceph in teh conditional | 16:53 |
| *** unicell has joined #kolla | 16:53 | |
| sdake | can you just make that ceph_enables == yes? | 16:54 |
| rhallisey | so is it ok? if lvm else ceph | 16:54 |
| sdake | ceph_enabled == yes? | 16:54 |
| sdake | if ceph_enabled == yes use ceph else use lvm | 16:54 |
| sdake | is the logic i'd like | 16:54 |
| rhallisey | ya I can abstract out to ansible | 16:54 |
| rhallisey | fine with that. Easier to track | 16:55 |
| sdake | just add a ceph_enabled=n to the config options list | 16:55 |
| rhallisey | k | 16:55 |
| *** mfalatic has quit IRC | 16:56 | |
| *** bmace has quit IRC | 17:01 | |
| sdake | rhallisey if you want to build from source you have to pull my patchset | 17:03 |
| rhallisey | sdake, which one? | 17:03 |
| sdake | or atleast cherrypick this patch https://review.openstack.org/#/c/213472/18 | 17:03 |
| rhallisey | kk | 17:03 |
| openstackgerrit | Steven Dake proposed stackforge/kolla: Make gnocchi-base build from source https://review.openstack.org/213522 | 17:06 |
| openstackgerrit | Steven Dake proposed stackforge/kolla: Make heat build from source in the templates https://review.openstack.org/213523 | 17:06 |
| openstackgerrit | Steven Dake proposed stackforge/kolla: Neutron base does not need python-neutron-client package https://review.openstack.org/213539 | 17:06 |
| openstackgerrit | Steven Dake proposed stackforge/kolla: Just a test patch to test building with templates https://review.openstack.org/213538 | 17:06 |
| openstackgerrit | Steven Dake proposed stackforge/kolla: Make cinder build from source https://review.openstack.org/213524 | 17:06 |
| openstackgerrit | Steven Dake proposed stackforge/kolla: Make openvswitch build from source https://review.openstack.org/213525 | 17:06 |
| openstackgerrit | Steven Dake proposed stackforge/kolla: Fix gate source building https://review.openstack.org/213472 | 17:06 |
| openstackgerrit | XiaBing Yao proposed stackforge/kolla: update config-external.sh in galera,keystone,rabbitmq,swift https://review.openstack.org/213787 | 17:09 |
| sdake | pbourke around? | 17:18 |
| rhallisey | cool that fixed it | 17:19 |
| sdake | well give it a +2 then ! :) | 17:19 |
| sdake | i've got like 20 patches outstanding | 17:19 |
| sdake | really hard to keep a tidy workspace with that much work outstanding | 17:19 |
| sdake | the rest in that chain are all 1 liners as ewll | 17:20 |
| sdake | brb relogging | 17:20 |
| *** sdake_ has joined #kolla | 17:21 | |
| *** athomas has quit IRC | 17:21 | |
| *** jasonsb has quit IRC | 17:23 | |
| *** jasonsb has joined #kolla | 17:24 | |
| *** sdake has quit IRC | 17:24 | |
| *** achanda has joined #kolla | 17:27 | |
| *** jasonsb has quit IRC | 17:28 | |
| *** sdake has joined #kolla | 17:31 | |
| *** sdake_ has quit IRC | 17:35 | |
| sdake | rhallisey https://bugs.launchpad.net/kolla/+bug/1485704 | 17:39 |
| openstack | Launchpad bug 1485704 in kolla "remove cinder_volume_driver config option from globals.yml" [Critical,Confirmed] - Assigned to Ryan Hallisey (rthall14) | 17:39 |
| sdake | anytime before we release is fine | 17:39 |
| rhallisey | roger that | 17:39 |
| sdake | its marked critical because it must be fixed | 17:40 |
| sdake | config file is in lockdown | 17:40 |
| sdake | samyaple you around atm? | 17:40 |
| openstackgerrit | Steven Dake proposed stackforge/kolla: Move enable_cinder to the correct location https://review.openstack.org/213798 | 17:45 |
| rhallisey | sdake, I suggest the same thing on the review for cinder ^ | 17:46 |
| rhallisey | sam did not agree | 17:46 |
| sdake | because? | 17:47 |
| sdake | i dont want defaults.yml changing for every new service | 17:47 |
| sdake | that is rediculous | 17:47 |
| sdake | lets have another go at the discussion | 17:48 |
| sdake | or whatever the main config file is | 17:48 |
| rhallisey | https://review.openstack.org/#/c/208868/7/etc/kolla/globals.yml | 17:48 |
| rhallisey | wait no that's not it | 17:49 |
| rhallisey | oh nvm ya it it is | 17:49 |
| rhallisey | that's the suggestion and response | 17:49 |
| sdake | i think you saying it should be in defaults.yml was not in queestion | 17:49 |
| sdake | its the enabled by default in sam's comment | 17:49 |
| *** inc0 has joined #kolla | 17:50 | |
| sdake | hey inc0 | 17:50 |
| inc0 | afternoon | 17:50 |
| rhallisey | iirc I think there was more discussion on irc.. | 17:51 |
| rhallisey | will have to bring it up again just to clarify since I agree with the patch | 17:51 |
| sdake | well +2 it then :) | 17:52 |
| * sdake wtbs lots of reviews from our core reviewers today | 17:52 | |
| sdake | mandre if your around pls vote on coolsvap's core proposal | 17:53 |
| sdake | or let me know your abstaining so i can pull the trigger :) | 17:53 |
| *** dims_ has joined #kolla | 17:53 | |
| rhallisey | sdake, he starts next week | 17:55 |
| inc0 | jpeeler, around? | 17:56 |
| sdake | nice what is he going to be working on? | 17:56 |
| *** dims has quit IRC | 17:57 | |
| inc0 | sdake, do you know status of ironic container? | 17:57 |
| sdake | not sure if its started or not | 17:58 |
| sdake | it says started in launchpad | 17:58 |
| sdake | it is just a conductor and api so should be 30 minute job to make the container | 17:58 |
| inc0 | yeah, would be cool to setup functional undercloud with ansible tho | 17:59 |
| inc0 | might be good use case | 17:59 |
| inc0 | for tripleo | 17:59 |
| inc0 | instead of all this devtest bullshit | 17:59 |
| inc0 | anyway, I just came to say hello | 18:00 |
| inc0 | now I'm going to have evening glass of scotch and chill | 18:01 |
| inc0 | cya all tomorrow | 18:01 |
| *** inc0 has quit IRC | 18:01 | |
| * rhallisey struggles with devtest | 18:02 | |
| sdake | rhallisey is https://bugs.launchpad.net/kolla/+bug/1460129 finished? | 18:02 |
| openstack | Launchpad bug 1460129 in kolla "Cinder needs to be broken out into thin containers" [Medium,Confirmed] - Assigned to Ryan Hallisey (rthall14) | 18:02 |
| rhallisey | not yet since tgt is still running along side cinder-volumes | 18:03 |
| sdake | but as is, cinder definately needs docker 1.7? | 18:04 |
| sdake | if looking at the deployment today | 18:04 |
| sdake | if its enabled | 18:04 |
| sdake | neutron are not in thin containers so that rgument is moot | 18:04 |
| *** tfukushima has quit IRC | 18:07 | |
| rhallisey | sdake, oh I thought they've been broken out | 18:08 |
| rhallisey | cinder will need 1.7 | 18:08 |
| rhallisey | at least for local storage | 18:08 |
| rhallisey | not sure about ceph yet | 18:08 |
| *** jasonsb has joined #kolla | 18:11 | |
| openstackgerrit | Merged stackforge/kolla: Add Dockerfile template for designate https://review.openstack.org/213703 | 18:14 |
| openstackgerrit | Merged stackforge/kolla: Fix heat docker templates https://review.openstack.org/213547 | 18:16 |
| openstackgerrit | XiaBing Yao proposed stackforge/kolla: update config-external.sh in galera,keystone,rabbitmq,swift https://review.openstack.org/213787 | 18:18 |
| mandre | sdake: i'm abstaining on coolsvap's core proposal, i feel i don't have the required background to vote | 18:20 |
| mandre | been away a lot lately | 18:20 |
| mandre | btw, I'm back to work on 25th | 18:21 |
| mandre | i'm totally trusting other cores' judgment, I'm sure coolsvap will do excellent job as core | 18:25 |
| sdake | mandre our process requires an email ack | 18:26 |
| sdake | is that not possible? | 18:26 |
| sdake | if not, I can ack in your stead linking to irc chat | 18:26 |
| mandre | nah, I can do it | 18:27 |
| mandre | give me a few minutes | 18:27 |
| sdake | rhallisey cinder will need, or does need? | 18:30 |
| sdake | rhallisey difference being will need = in the future, does need = needs today | 18:31 |
| sdake | this needs an ack pls https://review.openstack.org/#/c/213472/ | 18:31 |
| sdake | mandre if you can review quickly ;-) | 18:32 |
| openstackgerrit | Steven Dake proposed stackforge/kolla: Remove compose from documenation https://review.openstack.org/213401 | 18:34 |
| mandre | sdake: sure thing, once I put the kids to bed | 18:35 |
| sdake | mandre if you could spend 20 mins on the queue that would be fantastic :) | 18:35 |
| sdake | our reviews are really lagging with you mia :( | 18:35 |
| sdake | once the kids are asleep ;) | 18:35 |
| sdake | rhallisey this is an easy one https://review.openstack.org/#/c/213524/ ;-) | 18:37 |
| rhallisey | ya that must've been an accidental addition | 18:38 |
| *** britthou_ has joined #kolla | 18:43 | |
| sdake | rhallisey which must've been accidental? | 18:43 |
| rhallisey | the line you deleted | 18:44 |
| *** britthouser has quit IRC | 18:45 | |
| sdake | it wasn't accidental i deleted it | 18:49 |
| sdake | the container won't build with it | 18:49 |
| sdake | i htink it was just never tested | 18:49 |
| sdake | i am fixing that now | 18:49 |
| sdake | soon we will have 8 build gates | 18:49 |
| sdake | this needs merging or our documentation wont work | 18:50 |
| sdake | https://review.openstack.org/#/c/213398/ | 18:50 |
| Slower | sdake: http://git.openstack.org/cgit/openstack/puppet-neutron/tree/manifests/params.pp | 18:51 |
| Slower | SamYaple: ^ | 18:51 |
| Slower | so this is the craziness we were talking about with multiple config files | 18:51 |
| openstackgerrit | Steven Dake proposed stackforge/kolla: Templatize Murano container https://review.openstack.org/213558 | 18:52 |
| rhallisey | ya a more flexible config support would make dealing with that easier | 18:53 |
| rhallisey | s/config/externalconfig | 18:53 |
| sdake | slower we are aware of the problem and have a solution involving a json file | 18:53 |
| Slower | sdake: ah interesting | 18:53 |
| rhallisey | sdake, there's a bp for that right? | 18:53 |
| sdake | there is not | 18:53 |
| sdake | there should probably be, but it would be low priority | 18:54 |
| sdake | if you want to file one | 18:54 |
| sdake | it requires big changes all ove the code base and its a minor improvement | 18:54 |
| Slower | sdake: so basically an env var with json defining the configs? | 18:54 |
| sdake | i'm satisified with adding to config-external for the short term | 18:54 |
| sdake | config-external.sh would copy a json file | 18:54 |
| Slower | ah ok | 18:54 |
| sdake | or set_configs would read the json and do the copying | 18:55 |
| sdake | and we could remove config-external entirely | 18:55 |
| sdake | rhallisey can you make a blueprint pls and link to channel | 18:55 |
| sdake | i would highy prefer yaml but it introduces a dependency on the hot | 18:55 |
| sdake | PyYAML | 18:55 |
| sdake | as a result, we are going with json, as sucky as it is | 18:56 |
| Slower | I think json is fine | 18:57 |
| sdake | we dont use env variables for passing data any longer | 18:57 |
| sdake | except for special conditoins like bootstrapping | 18:57 |
| Slower | yeah that's probably a good call | 18:57 |
| sdake | it just makes two config systems, which is maddening to deal with :) | 18:57 |
| sdake | I dont know if that work will land for liberty or not | 18:57 |
| sdake | it definately wont land for liberty-3 | 18:58 |
| sdake | our launchpad tracker is overflowing with work for liberty | 18:58 |
| sdake | slower we need critical/high to land for liberty-3 https://launchpad.net/kolla/+milestone/liberty-3 | 18:59 |
| rhallisey | sdake, re cinder discussion above. I accidentally added the line you correctly deleted it | 19:05 |
| sdake | roger | 19:05 |
| sdake | what about th docker 1.7 thing | 19:05 |
| rhallisey | cinder will need 1.7 | 19:06 |
| sdake | is it needed *NOW* to run cinder as is in teh source base | 19:06 |
| sdake | you say will - meaning future tense | 19:06 |
| sdake | do you mean does? | 19:06 |
| rhallisey | ya right now | 19:06 |
| sdake | ok cinder does need 1.7 then ;) | 19:06 |
| rhallisey | for local storage you need /dev properly mounter | 19:06 |
| rhallisey | mounted | 19:06 |
| rhallisey | which is only supported 1.7+ | 19:06 |
| rhallisey | sorry for lack of clarity | 19:06 |
| rhallisey | too many conversations at once | 19:07 |
| sdake | ya i jump around alot | 19:07 |
| sdake | thats how I think too ;) | 19:07 |
| rhallisey | uncontrolled add today | 19:07 |
| rhallisey | sdake, so I'm not quite able to visualize this json file | 19:09 |
| rhallisey | what would be in it? List of configs? | 19:09 |
| sdake | rhallisey can you look at https://review.openstack.org/#/c/213401/ | 19:10 |
| sdake | so i dont know json format | 19:10 |
| sdake | but i'll put it in yaml so you can get an idea | 19:11 |
| sdake | files: | 19:11 |
| sdake | project_name: | 19:11 |
| sdake | let me start over | 19:12 |
| sdake | files: | 19:12 |
| sdake | keystone: | 19:12 |
| sdake | let me start over | 19:12 |
| sdake | files: | 19:12 |
| sdake | - source: /var/lib/kolla/configs/keystone/keystone.conf | 19:13 |
| sdake | - dest: /etc/keystone/keystone.conf | 19:13 |
| sdake | - owner: keystone | 19:13 |
| sdake | - perm: 064 | 19:14 |
| sdake | that needs to be the entry in the list | 19:14 |
| sdake | and thee needs to be a bunch of entries | 19:14 |
| sdake | my yaml foo isn't great either - i usually edit other people's yamls :) | 19:14 |
| rhallisey | same | 19:14 |
| sdake | basically 4 parameters over and over in a dictionary | 19:15 |
| rhallisey | ok so we just add based on whats needed | 19:15 |
| rhallisey | gotcha | 19:15 |
| sdake | you add to tripleo | 19:15 |
| sdake | we add to ansible | 19:15 |
| rhallisey | set_config will interpret | 19:15 |
| sdake | thatsthe idea | 19:15 |
| rhallisey | heat will read for tripleo | 19:16 |
| rhallisey | ok will outline this | 19:16 |
| openstackgerrit | Merged stackforge/kolla: Add Dockerfile template for keepalived https://review.openstack.org/213479 | 19:16 |
| sdake | it has to be in json because of ansible | 19:17 |
| sdake | and because of the dependency problem | 19:17 |
| sdake | the dependency problem is on the genreation side | 19:17 |
| sdake | not on the decoding side | 19:17 |
| sdake | i dont care if PyYaml is in a container image | 19:18 |
| sdake | but i do care if its required on the host | 19:18 |
| rhallisey | right | 19:18 |
| rhallisey | no worries | 19:18 |
| sdake | this would make hosts like coreos or aotmic not work since they area read only filesystems and proably dont include pymal | 19:18 |
| rhallisey | ya json is fine | 19:20 |
| sdake | this needs a review to finish up the build from source gating bluleprint | 19:23 |
| sdake | https://review.openstack.org/#/c/213525/ | 19:23 |
| rhallisey | https://blueprints.launchpad.net/kolla/+spec/replace-config-external | 19:24 |
| sdake | all I have to say is centos binary and source gating work with this patch stream with our automated tests : https://review.openstack.org/#/c/213539/ | 19:24 |
| rhallisey | sdake, add further comments to that if needed | 19:25 |
| openstackgerrit | Jeff Peeler proposed stackforge/kolla: Change default install type to binary https://review.openstack.org/213831 | 19:35 |
| *** tfukushima has joined #kolla | 19:43 | |
| openstackgerrit | Merged stackforge/kolla: Fix gate source building https://review.openstack.org/213472 | 19:44 |
| openstackgerrit | Merged stackforge/kolla: Make gnocchi-base build from source https://review.openstack.org/213522 | 19:44 |
| openstackgerrit | Merged stackforge/kolla: Make heat build from source in the templates https://review.openstack.org/213523 | 19:45 |
| sdake | i've made an rc1 milestone in launchpad | 19:46 |
| sdake | https://launchpad.net/kolla/+milestone/liberty-rc1 | 19:46 |
| sdake | it includes all the things not yet started | 19:46 |
| openstackgerrit | Merged stackforge/kolla: Make cinder build from source https://review.openstack.org/213524 | 19:46 |
| sdake | the one problem is "mongodb" | 19:46 |
| *** bmace has joined #kolla | 19:46 | |
| openstackgerrit | Ryan Hallisey proposed stackforge/kolla: Change cinder_volume_driver config option to enable_ceph https://review.openstack.org/213840 | 19:46 |
| sdake | which is a dependency of ceilometer in liberty-3 | 19:47 |
| sdake | deadline for rc1 is sept 25th | 19:47 |
| rhallisey | I'm still testing ceph.. | 19:47 |
| rhallisey | since I have no idea what I'm doing taking some time | 19:47 |
| sdake | rhallisey i think you have too much work to do between now and 31st (deadling for liberty-3) to do mongodb | 19:48 |
| sdake | do you mind if i remove you from the assignee? | 19:48 |
| rhallisey | for ceph? | 19:48 |
| *** dims_ has quit IRC | 19:48 | |
| sdake | no for mongodb | 19:48 |
| sdake | your #1 job is ceph :) | 19:48 |
| rhallisey | just push to rc1 | 19:48 |
| sdake | pushceilometer to rc1? | 19:48 |
| rhallisey | unless someone else wants it | 19:48 |
| rhallisey | no mongo | 19:48 |
| rhallisey | isn't that what you did? | 19:48 |
| sdake | ceilometer depends on mongo | 19:48 |
| *** dims has joined #kolla | 19:48 | |
| rhallisey | ohh | 19:48 |
| sdake | i want to pull it back into liberty-3 and possibly get coolsvap to take it oon | 19:49 |
| rhallisey | sorry I missed your earlier comments :/ jumped in at the wrong time | 19:49 |
| rhallisey | ya just set it loose | 19:50 |
| rhallisey | I grabbed cause no one wanted it | 19:50 |
| sdake | i think nobody still wants it ;-) | 19:52 |
| rhallisey | just leave it for me then I'll hopefully get to it | 19:54 |
| rhallisey | I'm still stuck on the cinder db issue | 19:54 |
| rhallisey | I have a running cinder elsewhere that I'm trying to attach to a ceph node | 19:55 |
| rhallisey | I think it's working | 19:55 |
| rhallisey | I don't know | 19:55 |
| rhallisey | I'm just not quite satisfied yet | 19:55 |
| rhallisey | I need to understand it more to make a doc at the very least | 19:56 |
| openstackgerrit | Ryan Hallisey proposed stackforge/kolla: Cinder bootstrap container is not initializing the DB https://review.openstack.org/213845 | 19:58 |
| rhallisey | issues I've found so far ^ | 19:58 |
| *** bizarrochristy has joined #kolla | 19:58 | |
| *** mfalatic_ has quit IRC | 19:58 | |
| *** mfalatic has joined #kolla | 19:58 | |
| openstackgerrit | Ryan Hallisey proposed stackforge/kolla: Change cinder_volume_driver config option to enable_ceph https://review.openstack.org/213840 | 20:02 |
| sdake | rhallisey i left two comments - one in partchset 1 and patchset 2 | 20:05 |
| rhallisey | wfm | 20:06 |
| sdake | rhallisey there is a comment for you in https://review.openstack.org/#/c/213623/ | 20:07 |
| * sdake works peopole like dogs! | 20:07 | |
| rhallisey | didn't even notice the line #'s | 20:08 |
| sdake | yup thats why i expalined it :) | 20:08 |
| rhallisey | well I did. But didn't notice that the last line wasn't given a new line # | 20:09 |
| sdake | jpeeler or mandre can we get an ack on https://review.openstack.org/#/c/213401/ | 20:09 |
| openstackgerrit | Ryan Hallisey proposed stackforge/kolla: Change cinder_volume_driver config option to enable_ceph https://review.openstack.org/213840 | 20:10 |
| sdake | rhallisey you missed my comment in patch set #1 | 20:10 |
| rhallisey | adding | 20:12 |
| sdake | nobrainer review: https://review.openstack.org/#/c/213620/2/docker_templates/memcached/Dockerfile.j2 | 20:12 |
| openstackgerrit | Ryan Hallisey proposed stackforge/kolla: Change cinder_volume_driver config option to enable_ceph https://review.openstack.org/213840 | 20:13 |
| sdake | ok folks, mandre rhallisey need revies on this one, its been through the ringer- i've fixed all the problems and tested the deploy several times with it in various different patch sets | 20:14 |
| sdake | https://review.openstack.org/#/c/212804/ | 20:14 |
| *** daneyon has quit IRC | 20:14 | |
| sdake | also rhallisey no brainer review here too https://review.openstack.org/#/c/213831/ | 20:15 |
| sdake | and then you can get bck to ceph work :) | 20:15 |
| sdake | TIA :) | 20:15 |
| openstackgerrit | Merged stackforge/kolla: Add mongodb to docker_templates https://review.openstack.org/213623 | 20:18 |
| openstackgerrit | Merged stackforge/kolla: Remove compose from documenation https://review.openstack.org/213401 | 20:19 |
| sdake | yay the docs patch merged finally ;) | 20:19 |
| openstackgerrit | Merged stackforge/kolla: Change default install type to binary https://review.openstack.org/213831 | 20:26 |
| *** tfukushima has quit IRC | 20:27 | |
| sdake | jpeeler when you have a moment to read https://review.openstack.org/#/c/213489/1/tools/build.py | 20:30 |
| sdake | what are your thoughts on somehow running build.py multiple times if it fails and collating the data into one data set? | 20:30 |
| openstackgerrit | Merged stackforge/kolla: Add memcached to docker_templates https://review.openstack.org/213620 | 20:30 |
| sdake | in test_build.py rather then adding a flag | 20:31 |
| jpeeler | i'd think people would be using build.py directly locally too, so although it may not be as easy a retry flag would be nice | 20:32 |
| sdake | can you add a comment in the review pleae | 20:32 |
| * jpeeler copy pastes | 20:32 | |
| jpeeler | it's sad that the mirrors are the reason we need this | 20:33 |
| sdake | it is terribly sad agree | 20:33 |
| *** jasonsb has quit IRC | 20:35 | |
| sdake | jpeeler if you could ack https://review.openstack.org/#/c/213525/ its holding up build from source gating working properly | 20:35 |
| *** jasonsb_ has joined #kolla | 20:35 | |
| sdake | queue isn't 3 pages long anymore ;) | 20:36 |
| *** tfukushima has joined #kolla | 20:40 | |
| sdake | rhallisey jpeeler mandre this needs reviewing love - would really like to finish the job on this change | 20:44 |
| sdake | https://review.openstack.org/#/c/212804/ | 20:44 |
| mandre | on it :) | 20:45 |
| *** shardy has quit IRC | 20:50 | |
| sdake | look ma, the containers are all passing the gate: https://review.openstack.org/#/c/213845/ | 20:52 |
| openstackgerrit | Merged stackforge/kolla: Make openvswitch build from source https://review.openstack.org/213525 | 20:53 |
| openstackgerrit | Merged stackforge/kolla: Neutron base does not need python-neutron-client package https://review.openstack.org/213539 | 20:53 |
| openstackgerrit | Steven Dake proposed stackforge/kolla: Use one data container for all data volumes https://review.openstack.org/212804 | 21:01 |
| sdake | mandre for your viewing pleasure ;) | 21:02 |
| sdake | mandre re the newline, I did try copying over the old file and just making the changes i needed, same result | 21:05 |
| mandre | sdake, bad copy and paste, you need to remove the RUN instruction :) | 21:05 |
| mandre | sdake: have you tried with git reset -p? | 21:06 |
| mandre | not a big deal, though | 21:06 |
| sdake | which file has he bad copy and paste? | 21:07 |
| mandre | the ones you modified | 21:08 |
| mandre | :P | 21:08 |
| openstackgerrit | Steven Dake proposed stackforge/kolla: Use one data container for all data volumes https://review.openstack.org/212804 | 21:08 |
| mandre | 1 sec, I'll add comments to the review | 21:09 |
| *** banix has quit IRC | 21:12 | |
| *** britthou_ has quit IRC | 21:14 | |
| sdake | are you talking about the whitespace? | 21:14 |
| sdake | I just tested it and it no worky | 21:14 |
| sdake | so rebuilding imges, then if that works ill remove the workflow -1 | 21:14 |
| sdake | but comments still appreciated | 21:15 |
| * sdake wtb delorean mirrors | 21:16 | |
| mandre | nah, I don't care about the whitespace, i care about valid dockerfile | 21:21 |
| mandre | have you seen my comments about the extra RUN? | 21:21 |
| mandre | RUN is not a valid bash command, you'll get an error | 21:22 |
| mandre | git reset for the witespace, yeah... but not important, really | 21:25 |
| *** achanda has quit IRC | 21:35 | |
| *** unicell1 has joined #kolla | 21:38 | |
| *** unicell has quit IRC | 21:39 | |
| *** gfidente has quit IRC | 21:41 | |
| *** bizarrochristy has quit IRC | 21:43 | |
| *** bmace has quit IRC | 21:46 | |
| sdake | oh i see mandre | 21:50 |
| sdake | ok i'll fix that up | 21:50 |
| sdake | mandre sorry was away farrying children from school | 21:50 |
| *** banix has joined #kolla | 21:51 | |
| openstackgerrit | Merged stackforge/kolla: Fix volumes_from for the ansible playbooks https://review.openstack.org/213492 | 21:52 |
| openstackgerrit | Steven Dake proposed stackforge/kolla: Use one data container for all data volumes https://review.openstack.org/212804 | 21:54 |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed stackforge/kolla: Update Fedora magnum images to use fedora repo https://review.openstack.org/212972 | 22:02 |
| sdake | mandre that patch just tested as working | 22:03 |
| sdake | with a rebuild of images | 22:04 |
| sdake | mandre git reset -p prints out "No changes." | 22:05 |
| *** sdake_ has joined #kolla | 22:10 | |
| *** sdake has quit IRC | 22:13 | |
| *** achanda has joined #kolla | 22:16 | |
| *** achanda has quit IRC | 22:16 | |
| *** achanda has joined #kolla | 22:16 | |
| *** dims_ has joined #kolla | 22:25 | |
| *** dims has quit IRC | 22:28 | |
| *** banix has quit IRC | 22:38 | |
| *** banix has joined #kolla | 22:40 | |
| sdake_ | yo coolsvap|away | 22:45 |
| sdake_ | welcome to the core reviewer team :) | 22:45 |
| sdake_ | mandre https://review.openstack.org/#/c/212804/ could use a final review ;) | 22:46 |
| *** dims_ has quit IRC | 22:47 | |
| *** dims has joined #kolla | 22:47 | |
| sdake_ | coolsvap|away if you want to cut your chops on a hard review have a look at thatone ;) | 22:48 |
| *** dims has quit IRC | 22:51 | |
| *** sdake_ is now known as sdake | 23:01 | |
| sdake | samyaple is shade adding domain support in some way? | 23:11 |
| *** rhallisey has quit IRC | 23:11 | |
| *** sdake_ has joined #kolla | 23:23 | |
| *** sdake has quit IRC | 23:26 | |
| *** blahRus has quit IRC | 23:31 | |
| sdake_ | samyaple that method with my wireless worked! | 23:41 |
| *** britthouser has joined #kolla | 23:57 | |
| *** britthou_ has joined #kolla | 23:58 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!