| *** mgiles has quit IRC | 00:05 | |
| *** ayoung has quit IRC | 00:05 | |
| *** schwicht has joined #openstack-kolla | 00:05 | |
| portdirect | sbezverk_/srwilkers_: and slack comes trhough! the solution {{"{{"}} service_auth {{"}}"}} - have fun while you templates your templates... | 00:09 |
|---|---|---|
| sbezverk_ | :-) | 00:11 |
| srwilkers_ | Templateception | 00:11 |
| jascott1 | wow that is ugleeee | 00:13 |
| jascott1 | makes sense at least | 00:14 |
| sbezverk_ | when you look at it long enough it does make sense | 00:14 |
| *** schwicht has quit IRC | 00:15 | |
| *** sayantan_ has quit IRC | 00:16 | |
| *** schwicht has joined #openstack-kolla | 00:24 | |
| sdake | sbezverk_ looks like greek to me :) | 00:24 |
| sbezverk_ | sdake: cool, when you master gotpl, you can also say you mastered greek ;-) | 00:25 |
| sdake | greek is a dead language ;-) | 00:26 |
| sbezverk_ | sdake: really, what language then greek poeple speak ? | 00:27 |
| sdake | hrm | 00:27 |
| sdake | maybei ts not ded | 00:27 |
| sdake | maybe thats latin | 00:27 |
| sdake | dunno sounded good coming out :) | 00:27 |
| sdake | i'm not always right ;) | 00:27 |
| *** schwicht has quit IRC | 00:28 | |
| *** adrian_otto has joined #openstack-kolla | 00:28 | |
| sbezverk_ | sdake: yeah, latin despite of being the root of bunch of languages is note spoken, unles you reads doctor's prescription ;-) | 00:29 |
| *** Pavo has quit IRC | 00:29 | |
| *** Pavo has joined #openstack-kolla | 00:30 | |
| sdake | or need cloud tech :) | 00:30 |
| *** harlowja_ has quit IRC | 00:32 | |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Helm common user creation jobs https://review.openstack.org/408864 | 00:35 |
| jascott1 | https://www.youtube.com/watch?v=XbI-fDzUJXI | 00:37 |
| *** adrian_otto has quit IRC | 00:37 | |
| openstackgerrit | sean mooney proposed openstack/kolla: adds support for ovs with dpdk netdev datapath https://review.openstack.org/342354 | 00:39 |
| *** adrian_otto has joined #openstack-kolla | 00:41 | |
| *** portdirect has left #openstack-kolla | 00:42 | |
| *** portdirect has joined #openstack-kolla | 00:42 | |
| *** yingjun has joined #openstack-kolla | 00:43 | |
| openstackgerrit | Zeyu Zhu proposed openstack/kolla: Remove useless script from repo https://review.openstack.org/408493 | 00:44 |
| openstackgerrit | sean mooney proposed openstack/kolla: adds docs for ovs with dpdk https://review.openstack.org/408866 | 00:48 |
| portdirect | sbezverk_: ping? | 00:48 |
| sbezverk_ | portdirect: pong | 00:49 |
| *** jgriffith is now known as jgriffith_AutoAw | 00:49 | |
| *** duonghq has joined #openstack-kolla | 00:50 | |
| sbezverk_ | do you think the way kfox1111 uses openstack command will work in your case as well? | 00:51 |
| duonghq | morning guys | 00:51 |
| *** harlowja has joined #openstack-kolla | 00:51 | |
| sbezverk_ | duonghq: hello | 00:51 |
| duonghq | sbezverk_, what are you talking about? | 00:52 |
| portdirect | potenitially - it's certainly nicer in many ways (and how I've done it in the past) but just wanted to see if you knew the anser to this: does ansible fail if the user already exists? or update the password if it does? | 00:52 |
| portdirect | if it does, I'd prefer to keep it ansible | 00:53 |
| sbezverk_ | portdirect: I do not think it fails, but frankly not sure | 00:53 |
| portdirect | but otherwise will refactor round openstack-cli | 00:53 |
| sbezverk_ | portdirect: I mean it is again about consistency, not sure if it is good to use multiple appraoches.. will be confusing for people | 00:54 |
| portdirect | all good +2 to that - was just doing a 1-1 without thinking about potential improvements :) | 00:55 |
| portdirect | I'll let it run through the gate - just to check that what I have is working - and the refactor | 00:56 |
| *** zhubingbing has joined #openstack-kolla | 00:59 | |
| *** ayoung has joined #openstack-kolla | 00:59 | |
| sdake | portdirect haven't made it to fedex yet | 00:59 |
| portdirect | sbezverk_: also now passes with and without haproxy (with the exption of random kubeadm failures...) https://review.openstack.org/#/c/408810/, will rejeck but ready for review | 00:59 |
| sdake | hopfully this weekend | 00:59 |
| *** mattmceuen has quit IRC | 01:00 | |
| sbezverk_ | portdirect: I would do too 1:1 but what I like about kfox1111 is that he often thinks how to improve and forcing us to do the same :-) | 01:01 |
| *** jascott1 has quit IRC | 01:02 | |
| *** jascott1 has joined #openstack-kolla | 01:04 | |
| *** awiddersheim has quit IRC | 01:04 | |
| *** sayantan_ has joined #openstack-kolla | 01:07 | |
| openstackgerrit | sean mooney proposed openstack/kolla-ansible: adds ovs-dpdkctl tool https://review.openstack.org/408871 | 01:10 |
| openstackgerrit | sean mooney proposed openstack/kolla-ansible: extends ansible roles to deploy ovsdpdk https://review.openstack.org/408872 | 01:10 |
| *** zhurong has joined #openstack-kolla | 01:13 | |
| *** adrian_otto has quit IRC | 01:15 | |
| *** tonanhngo_ has quit IRC | 01:19 | |
| *** jrist has quit IRC | 01:21 | |
| *** awiddersheim has joined #openstack-kolla | 01:22 | |
| kfox1111 | hi | 01:25 |
| *** tovin07_ has joined #openstack-kolla | 01:25 | |
| portdirect | hey dude | 01:25 |
| kfox1111 | sorry. I kind of disapeared today. have some vacation time I have to burn by the end of the year or loose. | 01:26 |
| openstackgerrit | sean mooney proposed openstack/kolla-ansible: This change allow setting kernel commandline args https://review.openstack.org/408876 | 01:27 |
| portdirect | take it while you can - the thing I hate most about working for myself is that my boss never gives me any time off :) | 01:28 |
| kfox1111 | :) | 01:28 |
| kfox1111 | its inconvenient timing though. :/ I usually like to maximaize my time at work in december. its always so quiet. lets of uninterrupted time to get stuff done. | 01:29 |
| *** saneax is now known as saneax-_-|AFK | 01:30 | |
| *** tonanhngo has joined #openstack-kolla | 01:33 | |
| *** jrist has joined #openstack-kolla | 01:34 | |
| *** tonanhngo has quit IRC | 01:37 | |
| *** Marcellin__ has quit IRC | 01:39 | |
| portdirect | if anyone free to look at a squint at this would be appricated: https://review.openstack.org/#/c/408836/ | 01:39 |
| kfox1111 | will do. | 01:41 |
| kfox1111 | https://review.openstack.org/#/c/408836/2 is really close | 01:41 |
| *** sacharya has joined #openstack-kolla | 01:43 | |
| *** elynn has joined #openstack-kolla | 01:43 | |
| kfox1111 | https://review.openstack.org/#/c/408810/3 reviewed too | 01:44 |
| portdirect | doh! | 01:44 |
| portdirect | cheers :) | 01:45 |
| kfox1111 | and https://review.openstack.org/#/c/407788/ | 01:47 |
| *** sacharya has quit IRC | 01:47 | |
| *** elynn has quit IRC | 01:48 | |
| *** elynn has joined #openstack-kolla | 01:50 | |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Pull kube-system images as part of CNI setup https://review.openstack.org/407788 | 01:50 |
| kfox1111 | darn... nslookup nor dig is in kolla-toolbox already. :/ | 01:51 |
| portdirect | kfox1111: busybox? | 01:52 |
| portdirect | though pulling another container is a bit yuck | 01:52 |
| kfox1111 | nope. | 01:52 |
| kfox1111 | python -c 'import socket; print socket.gethostbyname("google.com")' should do the trick... | 01:54 |
| portdirect | kfox1111: quay.io/coreos/flannel:v0.6.1 has nslookup as well | 01:55 |
| kfox1111 | that would work too. | 01:55 |
| *** sayanta__ has joined #openstack-kolla | 02:01 | |
| kfox1111 | remember the node target trick off hand? | 02:02 |
| kfox1111 | its a node label, right? | 02:03 |
| kfox1111 | kubernetes.io/hostname: minikube | 02:03 |
| kfox1111 | yeah... | 02:03 |
| *** sayantan_ has quit IRC | 02:04 | |
| kfox1111 | doh... | 02:05 |
| kfox1111 | wrote it as a helm package... | 02:05 |
| kfox1111 | but need the test to ensure you can helm.... | 02:06 |
| kfox1111 | grr.. | 02:06 |
| kfox1111 | :) | 02:06 |
| portdirect | lol :) your mind works faster than my keyboard entry | 02:06 |
| portdirect | which good :) | 02:07 |
| portdirect | *is | 02:07 |
| kfox1111 | hmm... this is much more ugly without having access to the vars we define in helm... | 02:10 |
| kfox1111 | I think I solved this somewhere though... | 02:10 |
| kfox1111 | oh. setup-resolv-conf. | 02:11 |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Add DB creation and initial migration templates (keystone) https://review.openstack.org/408836 | 02:11 |
| portdirect | why not just nslookup 'kubernetes' ? | 02:11 |
| kfox1111 | portdirect: review looks good to me. workflowing. | 02:12 |
| kfox1111 | portdirect: the issue is before that. | 02:13 |
| kfox1111 | need to know what container to run. | 02:13 |
| Jeffrey4l | awiddersheim, around? | 02:13 |
| portdirect | I'll leave you to it - backseat driving is never fun :) | 02:13 |
| kfox1111 | I did that this way in setup-resolv-conf.sh like this: | 02:13 |
| kfox1111 | TOOLBOX=$(kolla-kubernetes resource-template create bootstrap neutron-create-db -o json | jq -r '.spec.template.spec.containers[0].image') | 02:13 |
| kfox1111 | but can't pull that trick with helm, | 02:14 |
| kfox1111 | as it requires you to have a working tiller before you can even render a template. :/ | 02:14 |
| kfox1111 | so this will fail as we go pure helm too. :/ | 02:14 |
| portdirect | oh - yeah, I wouldnt do the dns/cni check with helm at all - as we only really want to init that after we have a working env? | 02:15 |
| portdirect | looking that trick of yours is a pain :( | 02:15 |
| portdirect | the thing I've found most frustrating about working with helm is that it wont render a template at all without talking to tiller... | 02:16 |
| kfox1111 | yeah. thats really a pain. :/ | 02:16 |
| kfox1111 | maybe I punt in this ps. | 02:17 |
| kfox1111 | do it the same as setup-resolv-conf and then solve both later. | 02:17 |
| portdirect | kinda understand it but would be nice to have --dry-run-offline | 02:17 |
| kfox1111 | heh. | 02:18 |
| kfox1111 | I think I started test-dns once before... I've got a test-dns.yml file laying around.. | 02:18 |
| portdirect | yeah, you had it in the original docs somewhere i think? | 02:18 |
| *** elynn_ has joined #openstack-kolla | 02:19 | |
| kfox1111 | oh. yeah. that could be. | 02:19 |
| *** elynn has quit IRC | 02:20 | |
| portdirect | gonna ask a q to whihc i think the answer is gonna be no, but if you dont ask you never get... | 02:20 |
| portdirect | can we see logs from zuul as a job runs? | 02:21 |
| kfox1111 | yup. :) | 02:21 |
| portdirect | oh - see dont ask dont get | 02:21 |
| kfox1111 | but depending on your support, sometimes. | 02:21 |
| kfox1111 | if you go to zuul, and click on a non link in the job. | 02:21 |
| kfox1111 | it slides down into a full list of jobs. | 02:22 |
| kfox1111 | the running ones have a link. | 02:22 |
| kfox1111 | should be telnet://xx.xx.xx.xx:yyyyyy | 02:22 |
| kfox1111 | the no part is, some of the clouds in infra are ipv6 only, so you will see ipv6 addreses there instead. | 02:22 |
| kfox1111 | so if you don't have ipv6 support, you can't see those jobs run live. :/ | 02:22 |
| kfox1111 | I've still got like 0 ipv6 support. :/ | 02:22 |
| kfox1111 | most of the time though, at least one of the jobs gets a v4 address. | 02:23 |
| portdirect | sweet! cheers dude | 02:25 |
| kfox1111 | :) | 02:25 |
| portdirect | now I can watch my failure's unfold in real time :P | 02:26 |
| kfox1111 | hehe | 02:26 |
| portdirect | can you workflow this: so we can get your stuff in, https://review.openstack.org/#/c/407788/ | 02:28 |
| portdirect | (shamlessly tries to get commit count up) | 02:28 |
| kfox1111 | done | 02:29 |
| *** Pavo has quit IRC | 02:29 | |
| jascott1 | portdirect how am I supposed to invoke setup_dev_env.sh? | 02:31 |
| awiddersheim | Jeffrey4l: hey whats up | 02:31 |
| Jeffrey4l | awiddersheim, i saw the talk you and sdake . | 02:31 |
| openstackgerrit | Merged openstack/kolla-kubernetes: Pull kube-system images as part of CNI setup https://review.openstack.org/407788 | 02:31 |
| awiddersheim | ah right | 02:31 |
| Jeffrey4l | some patch is loss during repo split. | 02:32 |
| awiddersheim | yse | 02:32 |
| awiddersheim | yes | 02:32 |
| portdirect | jascott1: ah i see your in the world of half complete stuff | 02:32 |
| Jeffrey4l | there are patchset working on this. | 02:32 |
| jascott1 | ja | 02:32 |
| portdirect | 2 sec | 02:32 |
| Jeffrey4l | https://review.openstack.org/398563 | 02:32 |
| Jeffrey4l | but it is too big, and hard to backport. i am think backport one by one, may need a list to track this. | 02:33 |
| portdirect | jascott1: right you kinda want to work from patch set 6.5... | 02:33 |
| awiddersheim | ah okay | 02:33 |
| awiddersheim | yeah I see the change | 02:33 |
| portdirect | jascott1: you running halcyon on a linux host? | 02:34 |
| jascott1 | ya | 02:34 |
| jascott1 | using vagrant halcyon | 02:34 |
| awiddersheim | Jeffrey4l: yeah that is tough one but these definetly should get back in somehow | 02:34 |
| Jeffrey4l | thanks for point this out :) | 02:34 |
| jascott1 | vbox | 02:34 |
| awiddersheim | sure np | 02:34 |
| Jeffrey4l | yep | 02:34 |
| *** Pavo has joined #openstack-kolla | 02:34 | |
| portdirect | jascott1: ok if you get that ps and replace the junk etc/kolla-kubernetes/kolla-kubernetes.yml, etc/kolla-kubernetes/service_resources.yml with good ones | 02:35 |
| portdirect | jascott1: then we can get going. | 02:35 |
| jascott1 | ok thanks | 02:36 |
| portdirect | jascott1: first run the ./get-k8s-creds.sh from halcyon get your client side set up | 02:36 |
| jascott1 | right | 02:36 |
| portdirect | jascott1: then build the dockerfile i dumped at the root of kolla-k8s | 02:36 |
| jascott1 | oh thats new | 02:36 |
| portdirect | jascott1: docker build . --tag 'kolla/k8s-devstack:latest' | 02:37 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Have gate wait for dns ready https://review.openstack.org/408907 | 02:37 |
| portdirect | jascott1: https://review.openstack.org/#/c/405720/7/Dockerfile, L55 has the cmd to launch it | 02:38 |
| portdirect | jascott1: I'm gonna make al this run on mac os using vagrant/docker for mac when i get a chance | 02:38 |
| jascott1 | cool man thanks for the great work | 02:39 |
| portdirect | jascott1: only 45% there atm | 02:39 |
| jascott1 | im going to follow your reciipe and see what happens ;) | 02:39 |
| jascott1 | ps 6 or 5? | 02:39 |
| portdirect | ps 7 | 02:39 |
| jascott1 | oh ok cool thought you said get an earlier one | 02:39 |
| portdirect | nah just dump the dodgy files in etc in that last ps and I think you'll be good to go | 02:40 |
| jascott1 | wewt thx | 02:41 |
| openstackgerrit | Merged openstack/kolla-kubernetes: Add DB creation and initial migration templates (keystone) https://review.openstack.org/408836 | 02:41 |
| kfox1111 | portdirect: why abandon this one: https://review.openstack.org/#/c/408165/9 | 02:41 |
| portdirect | gimmie a holler once you have the docker image built and we can go through getting it up and running (at leat untillit all blows up horribly) | 02:41 |
| kfox1111 | I kind of liked it. | 02:41 |
| kfox1111 | just alittle on the fence on where to stick the value in all_values. | 02:42 |
| portdirect | could resurect it - just people have started working on service apis so thought renaming somthing in the common lib at thist atge was gonna cause more pain than good | 02:43 |
| kfox1111 | how many are in flight? | 02:44 |
| portdirect | think srwilkers has cinder inprogress and somtone else (jascott1) is working on cinder. | 02:44 |
| jascott1 | i am not working on cinder afaik :) | 02:45 |
| portdirect | whoops are you working on nova? or am I just imagining things (highly possible) | 02:45 |
| jascott1 | no they are claimed | 02:45 |
| jascott1 | i was going to help qin w horizon and hopefully help knock that out and then move on | 02:46 |
| kfox1111 | I see nova. | 02:46 |
| kfox1111 | looking for other things now. | 02:46 |
| *** schwicht has joined #openstack-kolla | 02:46 | |
| portdirect | think its just nova thats been pushed to gerrit - so shall i bring it back from the dead? | 02:47 |
| kfox1111 | thats the only one I see that would be effected. | 02:47 |
| kfox1111 | yeah. if we can fix up the vars thing real quick, we can get it merged quickly. | 02:47 |
| kfox1111 | how about we do something like done in this ps: | 02:48 |
| kfox1111 | https://review.openstack.org/#/c/408263 | 02:48 |
| jascott1 | who is on nova? | 02:48 |
| jascott1 | (unassigned)nova: TODO | 02:48 |
| sbezverk_ | kfox1111: hey man. do you mind to check mariadb, it is all ready | 02:49 |
| portdirect | mgiles has a ps in | 02:49 |
| *** absubram has joined #openstack-kolla | 02:49 | |
| kfox1111 | we make a common-api-python-deployment section in all_vars, | 02:49 |
| kfox1111 | and then tag it into the packages using it in helm_packages.py? | 02:49 |
| *** yuanying has quit IRC | 02:50 | |
| kfox1111 | sbezverk_: sure. | 02:50 |
| sbezverk_ | kfox1111: https://review.openstack.org/#/c/408160 | 02:50 |
| portdirect | kfox1111: looks good | 02:50 |
| kfox1111 | sbezverk_: trying to catch up on all the reviews. a lot going on. :) | 02:50 |
| sbezverk_ | portdirect: so are you going with openstack cli or ansible ;-) | 02:50 |
| portdirect | sbezverk_: will go with cli - but spining a lot of plates right now - I've not got it in with 90 mins beat me up | 02:52 |
| portdirect | sbezverk_: or just send hel-ming the merciless over :) | 02:52 |
| *** absubram_ has joined #openstack-kolla | 02:53 | |
| *** absubram has quit IRC | 02:53 | |
| *** absubram_ is now known as absubram | 02:53 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 02:54 |
| *** jascott1 has quit IRC | 02:55 | |
| *** harlowja has quit IRC | 02:55 | |
| kfox1111 | sbezverk_: should we come up with another name then servicename-job? | 02:58 |
| kfox1111 | its not very descriptive, and there may need to be other servicename-jobs. | 02:59 |
| kfox1111 | maybe servicename-element-init or something? | 02:59 |
| *** jascott1 has joined #openstack-kolla | 03:00 | |
| kfox1111 | or servicename-init-element? | 03:02 |
| kfox1111 | the latter is a bit more consistent with the other jobs. | 03:03 |
| portdirect | kfox1111: ping | 03:07 |
| *** schwicht has quit IRC | 03:07 | |
| portdirect | you want to make hapoxy options live in a section like common_create_keystone_admin? or happy to have it in common? I'd personally slightly pefer it in common - but dont really have stong views - just would like to leave helm_prebuild as mimila as possible | 03:11 |
| portdirect | *minimal | 03:11 |
| kfox1111 | a new section. | 03:14 |
| kfox1111 | named after the macro? | 03:14 |
| *** srwilkers__ has joined #openstack-kolla | 03:16 | |
| portdirect | no probs - 5 mins and i'll throw it at zuul | 03:16 |
| *** kbyrne has quit IRC | 03:17 | |
| *** srwilkers has quit IRC | 03:17 | |
| *** srwilkers__ is now known as srwilkers | 03:18 | |
| srwilkers | evening | 03:18 |
| *** kbyrne has joined #openstack-kolla | 03:18 | |
| *** srwilker- has joined #openstack-kolla | 03:18 | |
| *** srwilkers_ has left #openstack-kolla | 03:19 | |
| portdirect | evening srwilkers | 03:21 |
| portdirect | kfox1111: gonna call it pod-http-termination? | 03:22 |
| portdirect | seems the most flexible? | 03:22 |
| *** dave-mccowan has quit IRC | 03:25 | |
| *** slagle has joined #openstack-kolla | 03:29 | |
| *** jrist has quit IRC | 03:30 | |
| *** stevebaker has joined #openstack-kolla | 03:30 | |
| *** jrist has joined #openstack-kolla | 03:33 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible: [wip] Optimiztion reconfiguration for keystone https://review.openstack.org/408933 | 03:33 |
| kfox1111 | portdirect: that could work, but there may be other options from that common-deployment we might want to easily slide in? | 03:35 |
| kfox1111 | like, when we add ssl termination support to the haproxy? | 03:36 |
| *** sayanta__ has quit IRC | 03:36 | |
| kfox1111 | I guess we can cross that bridge when we get to it too though. | 03:37 |
| portdirect | yeah thats they I thought 'http-termination' made sense - as it provides a common point to add in functions related to that sort of thing for both haproxy and apache when we get there | 03:38 |
| kfox1111 | well, I don't think we want the haproxy option to show up in the apache based ones. | 03:38 |
| kfox1111 | just trying to figure this out... not exactly sure how to do it. | 03:39 |
| kfox1111 | we could just call it embeded-haproxy too. and make another one for ssl termination later when we support it. | 03:40 |
| * kfox1111 shrugs | 03:40 | |
| kfox1111 | probably doesn't matter too much. its easily changable if we pick wrong. | 03:40 |
| *** fguillot has joined #openstack-kolla | 03:44 | |
| *** kbyrne has quit IRC | 03:44 | |
| *** sacharya has joined #openstack-kolla | 03:44 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Have gate wait for dns ready https://review.openstack.org/408907 | 03:44 |
| *** kbyrne has joined #openstack-kolla | 03:48 | |
| *** sacharya has quit IRC | 03:48 | |
| openstackgerrit | dupengfei proposed openstack/kolla: Create a configure/parameter making no-pull base image, when it was loaded. And certainly, the default value is 'pull', to make the default building process stay the same. https://review.openstack.org/408940 | 03:48 |
| *** yuanying has joined #openstack-kolla | 03:50 | |
| *** yuanying has quit IRC | 03:50 | |
| *** yuanying has joined #openstack-kolla | 03:51 | |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Change _common_deployment to _common_api_python_deployment https://review.openstack.org/408165 | 03:52 |
| *** tovin07 has quit IRC | 03:55 | |
| *** tovin07 has joined #openstack-kolla | 03:56 | |
| kfox1111 | portdirect: a few lines to remove, then should be mergable. | 03:56 |
| portdirect | ah | 03:58 |
| portdirect | re all_values: https://github.com/openstack/kolla-kubernetes/blob/master/helm/all_values.yaml#L38 | 03:58 |
| portdirect | is there a way to update/rebase the commit more cleanly to take into acount the changes that have happened in the mean time while a ps has been under review? | 03:59 |
| kfox1111 | oh. um... | 03:59 |
| portdirect | beacuse it made be go wtf as well. - looks ok locally but not so pretty in gerrit? | 04:00 |
| kfox1111 | here's what I usually would do in this case. | 04:01 |
| kfox1111 | checkout master | 04:01 |
| kfox1111 | remove the old branch (if you use that). | 04:01 |
| kfox1111 | do a format patch pull | 04:01 |
| kfox1111 | git fetch https://git.openstack.org/openstack/kolla-kubernetes refs/changes/65/408165/10 && git format-patch -1 --stdout FETCH_HEAD > /tmp/foo.patch | 04:01 |
| kfox1111 | vi /tmp/foo.patch | 04:01 |
| kfox1111 | pull cut out the unwanted chunks out of all_values.yaml | 04:02 |
| kfox1111 | then do an: git am /tmp/foo.patch | 04:02 |
| kfox1111 | (oh. and make a branch before the git am if you do branches) | 04:02 |
| kfox1111 | there may be other ways to do that, but thats the easiest way I've found to fix up the all_vars merge conflicts. | 04:03 |
| portdirect | yeah - I need to really lean up my local working tree - its got to the stage where i have branch "scratch[1-somerandomnumber]' :) | 04:04 |
| kfox1111 | hehe. yeah. mine's getting pretty long too. :) | 04:04 |
| kfox1111 | I may just go bankrupt and recheckout a fresh one. :) | 04:04 |
| kfox1111 | everything should be checked in already. | 04:04 |
| portdirect | yeah been there a few times :) | 04:04 |
| kfox1111 | well, in review at least. | 04:05 |
| kfox1111 | it would be nice if you could just get a revew number. | 04:08 |
| kfox1111 | so you could get a new review number, thne make a branch after it, then do create your ps. | 04:08 |
| kfox1111 | would be easier to find. | 04:09 |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Change _common_deployment to _common_api_python_deployment https://review.openstack.org/408165 | 04:09 |
| portdirect | kfox1111: awesome, thanks for the git-tips :) | 04:10 |
| kfox1111 | ps looks good now. will wf as soon as it tests out. | 04:11 |
| kfox1111 | portdirect: np. :) | 04:11 |
| kfox1111 | the format-patch trick seems to work best with rebases where there are multiple deps involved too, as it just pulls off the most recent patch. | 04:12 |
| portdirect | yeah - im gonna quickly chuck that in a script before I forget | 04:13 |
| kfox1111 | cool. | 04:14 |
| *** fguillot has quit IRC | 04:15 | |
| *** devataa has joined #openstack-kolla | 04:15 | |
| *** prameswar has joined #openstack-kolla | 04:17 | |
| kfox1111 | I gota head out. talk to you later. | 04:17 |
| portdirect | jascott1: you getting on ok with that dev-env? | 04:17 |
| *** schwicht has joined #openstack-kolla | 04:19 | |
| *** elynn_ has quit IRC | 04:23 | |
| *** schwicht has quit IRC | 04:23 | |
| *** yuanying has quit IRC | 04:23 | |
| *** yuanying has joined #openstack-kolla | 04:24 | |
| *** sacharya has joined #openstack-kolla | 04:24 | |
| *** yuanying has quit IRC | 04:24 | |
| *** yuanying has joined #openstack-kolla | 04:25 | |
| srwilkers | gunna call it a night and sleep on this patch set. see you all in the morning | 04:25 |
| *** yuanying has quit IRC | 04:25 | |
| *** Pavo has quit IRC | 04:29 | |
| *** srwilkers has quit IRC | 04:30 | |
| *** tovin07 has quit IRC | 04:31 | |
| *** tovin07 has joined #openstack-kolla | 04:32 | |
| *** Pavo has joined #openstack-kolla | 04:34 | |
| *** kbyrne has quit IRC | 04:37 | |
| *** kbyrne has joined #openstack-kolla | 04:44 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Hel-ming mariadb components https://review.openstack.org/408160 | 04:51 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Hel-ming mariadb components https://review.openstack.org/408160 | 04:55 |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Helm common user creation jobs https://review.openstack.org/408864 | 04:55 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Hel-ming mariadb components https://review.openstack.org/408160 | 04:58 |
| *** mdnadeem has joined #openstack-kolla | 05:01 | |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Helm common user creation jobs https://review.openstack.org/408864 | 05:01 |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Change _common_deployment to _common_api_python_deployment https://review.openstack.org/408165 | 05:16 |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Update common deployment for python apis https://review.openstack.org/408165 | 05:21 |
| *** sdake has quit IRC | 05:24 | |
| *** yuanying has joined #openstack-kolla | 05:26 | |
| *** prameswar has quit IRC | 05:29 | |
| *** schwicht has joined #openstack-kolla | 05:30 | |
| *** schwicht has quit IRC | 05:34 | |
| *** yuanying has quit IRC | 05:35 | |
| *** yuanying has joined #openstack-kolla | 05:36 | |
| *** tonanhngo has joined #openstack-kolla | 05:42 | |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Helm Keystone K8s Services https://review.openstack.org/408963 | 05:42 |
| *** saneax-_-|AFK is now known as saneax | 05:46 | |
| *** yuanying has quit IRC | 05:49 | |
| *** elynn_ has joined #openstack-kolla | 05:52 | |
| *** HyperJohnGraham has joined #openstack-kolla | 05:53 | |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Helm Keystone K8s Services https://review.openstack.org/408963 | 05:55 |
| *** elynn__ has joined #openstack-kolla | 05:56 | |
| *** elynn_ has quit IRC | 05:56 | |
| *** kristian__ has joined #openstack-kolla | 06:01 | |
| *** schwicht has joined #openstack-kolla | 06:02 | |
| openstackgerrit | Li Yingjun proposed openstack/kolla-kubernetes: Fix different repo issue for documentation https://review.openstack.org/405054 | 06:05 |
| *** kristian__ has quit IRC | 06:05 | |
| *** schwicht has quit IRC | 06:07 | |
| *** prameswar has joined #openstack-kolla | 06:10 | |
| *** yuanying has joined #openstack-kolla | 06:28 | |
| *** sdake has joined #openstack-kolla | 06:29 | |
| *** Pavo has quit IRC | 06:29 | |
| sdake | sup peeps | 06:32 |
| *** Pavo has joined #openstack-kolla | 06:34 | |
| openstackgerrit | Qin Wang (qwang) proposed openstack/kolla-kubernetes: WIP - Hel-ming horizon components https://review.openstack.org/408975 | 06:35 |
| *** dupengfei has joined #openstack-kolla | 06:36 | |
| duonghq | sup sdake | 06:49 |
| *** kristian__ has joined #openstack-kolla | 06:50 | |
| *** msimonin has joined #openstack-kolla | 06:51 | |
| *** msimonin has quit IRC | 06:52 | |
| *** yuanying_ has joined #openstack-kolla | 06:55 | |
| *** kristian__ has quit IRC | 06:55 | |
| openstackgerrit | jianyi proposed openstack/kolla: Install solum dashboard into horizon image https://review.openstack.org/408981 | 06:57 |
| *** zhurong has quit IRC | 06:57 | |
| *** zhurong has joined #openstack-kolla | 06:58 | |
| *** yuanying has quit IRC | 06:58 | |
| *** mkoderer has joined #openstack-kolla | 06:58 | |
| *** sdake has quit IRC | 06:59 | |
| openstackgerrit | jianyi proposed openstack/kolla: Install solum dashboard into horizon image https://review.openstack.org/408981 | 06:59 |
| openstackgerrit | caoyuan proposed openstack/kolla: Install mistral dashboard into horizon image https://review.openstack.org/406726 | 07:00 |
| *** schwicht has joined #openstack-kolla | 07:03 | |
| *** prameswar has quit IRC | 07:06 | |
| *** msimonin has joined #openstack-kolla | 07:07 | |
| *** sacharya has quit IRC | 07:07 | |
| *** schwicht has quit IRC | 07:07 | |
| *** sacharya has joined #openstack-kolla | 07:08 | |
| *** kristian__ has joined #openstack-kolla | 07:11 | |
| *** sacharya has quit IRC | 07:13 | |
| openstackgerrit | jianyi proposed openstack/kolla-ansible: Enable solum dashboard when enable_solum is yes https://review.openstack.org/408985 | 07:13 |
| *** yuanying_ has quit IRC | 07:14 | |
| *** kristian__ has quit IRC | 07:15 | |
| *** yatin has quit IRC | 07:21 | |
| openstackgerrit | dupengfei proposed openstack/kolla: Not to pull the base image when building. https://review.openstack.org/408940 | 07:25 |
| *** yatin has joined #openstack-kolla | 07:32 | |
| openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible: Ansible-ize OpenStack Designate https://review.openstack.org/408465 | 07:35 |
| *** msimonin has quit IRC | 07:38 | |
| *** msimonin has joined #openstack-kolla | 07:38 | |
| *** kristian__ has joined #openstack-kolla | 07:41 | |
| *** kristian__ has quit IRC | 07:45 | |
| *** tonanhngo has quit IRC | 07:49 | |
| *** kristian__ has joined #openstack-kolla | 07:51 | |
| *** msimonin has quit IRC | 07:53 | |
| *** kristian__ has quit IRC | 07:56 | |
| openstackgerrit | zhubingbing proposed openstack/kolla: Install tacker-horizon into horizon image https://review.openstack.org/407810 | 07:58 |
| *** b_bezak has joined #openstack-kolla | 08:00 | |
| *** sacharya has joined #openstack-kolla | 08:00 | |
| *** tonanhngo has joined #openstack-kolla | 08:00 | |
| *** msimonin has joined #openstack-kolla | 08:02 | |
| *** fragatin_ has quit IRC | 08:03 | |
| *** msimonin has quit IRC | 08:04 | |
| *** tonanhngo has quit IRC | 08:05 | |
| *** sacharya has quit IRC | 08:05 | |
| *** fragatina has joined #openstack-kolla | 08:08 | |
| *** pcaruana has joined #openstack-kolla | 08:16 | |
| *** jascott1 has quit IRC | 08:23 | |
| *** fragatina has quit IRC | 08:27 | |
| *** Pavo has quit IRC | 08:29 | |
| *** sdake has joined #openstack-kolla | 08:34 | |
| sdake | sup duonghq | 08:34 |
| *** Pavo has joined #openstack-kolla | 08:34 | |
| *** HyperJohnGraham has quit IRC | 08:36 | |
| duonghq | sdake, still awake? | 08:36 |
| sdake | yup | 08:36 |
| *** matrohon has joined #openstack-kolla | 08:37 | |
| *** Jeffrey4l has quit IRC | 08:38 | |
| *** jascott1 has joined #openstack-kolla | 08:47 | |
| *** Serlex has joined #openstack-kolla | 08:53 | |
| *** egonzalez90 has joined #openstack-kolla | 08:54 | |
| *** schwicht has joined #openstack-kolla | 09:04 | |
| *** athomas has joined #openstack-kolla | 09:08 | |
| *** schwicht has quit IRC | 09:09 | |
| *** papacz has joined #openstack-kolla | 09:10 | |
| *** strigazi_AFK is now known as strigazi | 09:10 | |
| *** sdake_ has joined #openstack-kolla | 09:10 | |
| *** sdake_ has quit IRC | 09:10 | |
| *** sdake_ has joined #openstack-kolla | 09:10 | |
| *** sdake has quit IRC | 09:13 | |
| *** kristian__ has joined #openstack-kolla | 09:25 | |
| *** elynn__ has quit IRC | 09:29 | |
| *** yatin has quit IRC | 09:30 | |
| *** kristian__ has quit IRC | 09:30 | |
| zhubingbing | ;) | 09:34 |
| *** jascott1 has quit IRC | 09:40 | |
| *** yatin has joined #openstack-kolla | 09:42 | |
| duonghq | sup zhubingbing | 09:43 |
| *** kristian__ has joined #openstack-kolla | 09:45 | |
| *** zhurong has quit IRC | 09:46 | |
| *** Serlex has quit IRC | 09:46 | |
| *** kristian__ has quit IRC | 09:49 | |
| *** yingjun has quit IRC | 09:52 | |
| *** yingjun has joined #openstack-kolla | 09:53 | |
| openstackgerrit | zhubingbing proposed openstack/kolla: change "configurationso" to "configuration so" https://review.openstack.org/409051 | 09:54 |
| *** yingjun has quit IRC | 09:57 | |
| *** zhubingbing has quit IRC | 10:03 | |
| *** Serlex has joined #openstack-kolla | 10:13 | |
| *** yatin has quit IRC | 10:14 | |
| *** duonghq has quit IRC | 10:15 | |
| *** tovin07_ has quit IRC | 10:19 | |
| *** tovin07 has quit IRC | 10:22 | |
| *** yatin has joined #openstack-kolla | 10:25 | |
| *** prameswar has joined #openstack-kolla | 10:29 | |
| *** Pavo has quit IRC | 10:29 | |
| *** kristian__ has joined #openstack-kolla | 10:30 | |
| openstackgerrit | Eduardo Gonzalez proposed openstack/kolla: WIP: fix zaqar container https://review.openstack.org/407757 | 10:30 |
| *** Pavo has joined #openstack-kolla | 10:34 | |
| *** kristian__ has quit IRC | 10:34 | |
| *** jascott1 has joined #openstack-kolla | 10:40 | |
| *** kristian__ has joined #openstack-kolla | 10:43 | |
| *** schwicht has joined #openstack-kolla | 10:45 | |
| *** jascott1 has quit IRC | 10:45 | |
| *** Jeffrey4l has joined #openstack-kolla | 10:47 | |
| *** kristian__ has quit IRC | 10:47 | |
| openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible: Zaqar ansible role https://review.openstack.org/407760 | 10:49 |
| jmccarthy | Hmm I just noticed looks like there is no merge_configs capability for horizon, is that on purpose ? | 10:49 |
| openstackgerrit | Eduardo Gonzalez proposed openstack/kolla: Fix zaqar container https://review.openstack.org/407757 | 10:49 |
| *** schwicht has quit IRC | 10:50 | |
| openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible: Zaqar ansible role https://review.openstack.org/407760 | 10:56 |
| egonzalez90 | Jeffrey4l: thoughs about this? https://review.openstack.org/#/c/408493/3 | 11:09 |
| Jeffrey4l | egonzalez90, this file should be useless. | 11:09 |
| egonzalez90 | i cannot find where the file is/was used | 11:10 |
| Jeffrey4l | iirc, it is introduced for neutron namespace issue. but we have mount shared solution. | 11:10 |
| Jeffrey4l | need sync with sdake_ or SamYaple | 11:11 |
| egonzalez90 | https://review.openstack.org/#/c/272715/ seems to be removed here | 11:11 |
| Jeffrey4l | yes. so it should be a history file which is forgot to removed. | 11:12 |
| *** yingjun has joined #openstack-kolla | 11:15 | |
| *** schwicht has joined #openstack-kolla | 11:15 | |
| *** schwicht has quit IRC | 11:20 | |
| *** mdnadeem has quit IRC | 11:22 | |
| *** yingjun has quit IRC | 11:25 | |
| *** yingjun has joined #openstack-kolla | 11:25 | |
| openstackgerrit | Paul Bourke (pbourke) proposed openstack/kolla: Remove source install type for novnc / spice https://review.openstack.org/408599 | 11:26 |
| *** schwicht has joined #openstack-kolla | 11:26 | |
| openstackgerrit | Paul Bourke (pbourke) proposed openstack/kolla: Remove source install type for novnc / spice https://review.openstack.org/408599 | 11:28 |
| *** yingjun has quit IRC | 11:29 | |
| *** schwicht has quit IRC | 11:31 | |
| *** david-lyle_ has joined #openstack-kolla | 11:35 | |
| openstackgerrit | Merged openstack/kolla-ansible: Add *.egg-info into .gitignore file https://review.openstack.org/407312 | 11:36 |
| *** msimonin has joined #openstack-kolla | 11:37 | |
| *** david-lyle has quit IRC | 11:37 | |
| *** athomas has quit IRC | 11:38 | |
| *** msimonin has quit IRC | 11:41 | |
| openstackgerrit | Paul Bourke (pbourke) proposed openstack/kolla: Remove source install type for novnc / spice https://review.openstack.org/408599 | 11:44 |
| *** athomas has joined #openstack-kolla | 11:45 | |
| *** msimonin has joined #openstack-kolla | 11:45 | |
| *** msimonin has quit IRC | 11:49 | |
| *** mgiles has joined #openstack-kolla | 11:49 | |
| openstackgerrit | Merged openstack/kolla-ansible: Add missing per service Keystone config augments https://review.openstack.org/406224 | 11:50 |
| mliima | morning guys | 11:50 |
| openstackgerrit | Merged openstack/kolla-ansible: Change 'docker' to 'Docker' in docs https://review.openstack.org/408636 | 11:50 |
| *** absubram has quit IRC | 11:53 | |
| openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible: Zaqar ansible role https://review.openstack.org/407760 | 12:00 |
| egonzalez90 | morning mliima | 12:01 |
| *** schwicht has joined #openstack-kolla | 12:01 | |
| *** tonanhngo has joined #openstack-kolla | 12:02 | |
| *** jascott1 has joined #openstack-kolla | 12:03 | |
| *** Bico_Fino has joined #openstack-kolla | 12:03 | |
| *** tonanhngo has quit IRC | 12:03 | |
| *** williamc_ has quit IRC | 12:05 | |
| *** williamcaban has joined #openstack-kolla | 12:05 | |
| *** schwicht has quit IRC | 12:06 | |
| *** jascott1 has quit IRC | 12:08 | |
| *** dave-mccowan has joined #openstack-kolla | 12:09 | |
| *** kristian__ has joined #openstack-kolla | 12:09 | |
| *** kristian__ has quit IRC | 12:13 | |
| openstackgerrit | Merged openstack/kolla: Add monitoring group as a children of baremetal group https://review.openstack.org/408741 | 12:15 |
| openstackgerrit | Merged openstack/kolla: change "configurationso" to "configuration so" https://review.openstack.org/409051 | 12:16 |
| *** Bico_Fino has quit IRC | 12:17 | |
| *** kristian__ has joined #openstack-kolla | 12:17 | |
| *** fragatina has joined #openstack-kolla | 12:20 | |
| mliima | exit | 12:28 |
| *** Pavo has quit IRC | 12:29 | |
| *** rhallisey has joined #openstack-kolla | 12:33 | |
| *** Pavo has joined #openstack-kolla | 12:34 | |
| *** schwicht has joined #openstack-kolla | 12:40 | |
| *** msimonin has joined #openstack-kolla | 12:45 | |
| *** yingjun has joined #openstack-kolla | 12:46 | |
| *** msimonin has quit IRC | 12:50 | |
| *** papacz has quit IRC | 12:51 | |
| *** peterjenkins_ has quit IRC | 12:59 | |
| *** peterjenkins_ has joined #openstack-kolla | 12:59 | |
| *** Guest57 has joined #openstack-kolla | 13:03 | |
| *** elynn__ has joined #openstack-kolla | 13:04 | |
| *** jascott1 has joined #openstack-kolla | 13:05 | |
| *** pcaruana has quit IRC | 13:05 | |
| *** zhangshuai has quit IRC | 13:06 | |
| *** pcaruana has joined #openstack-kolla | 13:08 | |
| *** jascott1 has quit IRC | 13:10 | |
| *** Guest57 has quit IRC | 13:26 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Hel-ming mariadb components https://review.openstack.org/408160 | 13:38 |
| *** elynn__ has quit IRC | 13:40 | |
| *** msimonin has joined #openstack-kolla | 13:47 | |
| *** duonghq has joined #openstack-kolla | 13:49 | |
| *** msimonin has quit IRC | 13:51 | |
| *** Guest57 has joined #openstack-kolla | 13:55 | |
| *** fguillot has joined #openstack-kolla | 13:57 | |
| *** narasimha_SV has joined #openstack-kolla | 13:57 | |
| *** kristia__ has joined #openstack-kolla | 14:01 | |
| *** kristian__ has quit IRC | 14:01 | |
| *** kristia__ has quit IRC | 14:02 | |
| *** mgkwill has quit IRC | 14:03 | |
| *** mgkwill has joined #openstack-kolla | 14:03 | |
| *** kristian__ has joined #openstack-kolla | 14:03 | |
| narasimha_SV | host metrics are not getting collected even by gnocchi contaoiners though the metric details are specified in gnocchi-resources.yaml file in ceilometer container | 14:05 |
| *** gfidente has joined #openstack-kolla | 14:05 | |
| *** gfidente has joined #openstack-kolla | 14:05 | |
| *** kristian__ has quit IRC | 14:07 | |
| *** kristian__ has joined #openstack-kolla | 14:07 | |
| *** schwicht has quit IRC | 14:09 | |
| *** Guest57 has quit IRC | 14:09 | |
| *** duonghq has quit IRC | 14:14 | |
| *** papacz has joined #openstack-kolla | 14:17 | |
| *** yingjun has quit IRC | 14:22 | |
| *** yingjun has joined #openstack-kolla | 14:22 | |
| *** smekel_ has quit IRC | 14:23 | |
| *** smekel_ has joined #openstack-kolla | 14:23 | |
| openstackgerrit | Merged openstack/kolla: Remove useless script from repo https://review.openstack.org/408493 | 14:25 |
| openstackgerrit | Merged openstack/kolla: Add qemu-img into glance api image https://review.openstack.org/408129 | 14:25 |
| *** duonghq has joined #openstack-kolla | 14:27 | |
| *** kristian__ has quit IRC | 14:27 | |
| *** kristian__ has joined #openstack-kolla | 14:28 | |
| *** Pavo has quit IRC | 14:29 | |
| *** yingjun has quit IRC | 14:30 | |
| *** duonghq has left #openstack-kolla | 14:31 | |
| *** kristian__ has quit IRC | 14:32 | |
| *** kristian__ has joined #openstack-kolla | 14:32 | |
| *** tovin07 has joined #openstack-kolla | 14:33 | |
| *** schwicht has joined #openstack-kolla | 14:33 | |
| *** Pavo has joined #openstack-kolla | 14:34 | |
| *** goldyfruit has joined #openstack-kolla | 14:36 | |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible: Source the openrc file automatically when running tools/init-runonce https://review.openstack.org/409179 | 14:37 |
| *** duonghq has joined #openstack-kolla | 14:40 | |
| *** schwicht has quit IRC | 14:42 | |
| *** pcaruana has quit IRC | 14:42 | |
| *** tovin07 has quit IRC | 14:43 | |
| *** tovin07 has joined #openstack-kolla | 14:44 | |
| openstackgerrit | Merged openstack/kolla: Minor documentation fixes https://review.openstack.org/406534 | 14:44 |
| *** schwicht has joined #openstack-kolla | 14:45 | |
| *** kristian__ has quit IRC | 14:46 | |
| *** kristian__ has joined #openstack-kolla | 14:46 | |
| *** msimonin has joined #openstack-kolla | 14:47 | |
| *** schwicht has quit IRC | 14:50 | |
| *** jtriley has joined #openstack-kolla | 14:50 | |
| *** schwicht has joined #openstack-kolla | 14:51 | |
| *** macsz has quit IRC | 14:51 | |
| *** msimonin has quit IRC | 14:51 | |
| *** saneax is now known as saneax-_-|AFK | 14:52 | |
| *** macsz has joined #openstack-kolla | 14:52 | |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible: Source the openrc file automatically when running tools/init-runonce https://review.openstack.org/409179 | 14:52 |
| *** srwilkers has joined #openstack-kolla | 14:53 | |
| srwilkers | good morning | 14:54 |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible: Source the openrc file automatically when running tools/init-runonce https://review.openstack.org/409179 | 14:55 |
| *** portdirect_away has quit IRC | 14:55 | |
| *** ediardo has quit IRC | 14:56 | |
| *** portdirect_away has joined #openstack-kolla | 14:56 | |
| *** ediardo has joined #openstack-kolla | 14:56 | |
| portdirect_away | morning srwilkers o/ | 14:56 |
| *** pcaruana has joined #openstack-kolla | 14:56 | |
| *** portdirect_away is now known as portdirect_ | 14:57 | |
| srwilkers | hey man o/ | 14:58 |
| sbezverk_ | good morning gents | 14:58 |
| srwilkers | hey sbezverk_ :) | 15:01 |
| *** sdake_ has quit IRC | 15:01 | |
| *** tonanhngo has joined #openstack-kolla | 15:02 | |
| portdirect_ | sup sbezverk - that mariadb work is looking awesome - thanks for getting a common svc in | 15:02 |
| *** tonanhngo has quit IRC | 15:02 | |
| *** duonghq has quit IRC | 15:04 | |
| srwilkers | yeah, looks great. waiting for it to merge in so i can leverage it for cinder. had a few too many stouts last night, and things all of a sudden made much more sense. tends to happen quite often | 15:05 |
| portdirect_ | srwilkers: and as always xkcd has an explanation for this : https://xkcd.com/323/ | 15:06 |
| srwilkers | portdirect_, yes :D | 15:07 |
| sbezverk_ | srwilkers: portdirect_: thanks guys for reminding me, I was too lazy to do it right away ;-) | 15:07 |
| srwilkers | it happens :) | 15:08 |
| *** jascott1 has joined #openstack-kolla | 15:08 | |
| *** schwicht has quit IRC | 15:09 | |
| sbezverk_ | portdirect_: I saw keystone bits got merged, congrats!! it is a big block.. | 15:09 |
| srwilkers | \o/ | 15:09 |
| portdirect_ | sbezverk_: only bits... I'm waiting on your svc stuff coming in and we redid the common_deployment to make haproxy optional and specific to python services - the api will drop soonas i've finished a apache_api_deployment lib | 15:11 |
| *** jascott1 has quit IRC | 15:12 | |
| portdirect_ | though its amazing how much stuff everyones pulled together in the last few days - I think next week should see most services over to helm microservices? | 15:13 |
| sbezverk_ | portdirect_: microservices is great, but there are other things which frankly bugs me more.. | 15:15 |
| srwilkers | yeah, this week has been very productive i think | 15:15 |
| *** HyperJohnGraham has joined #openstack-kolla | 15:15 | |
| sbezverk_ | portdirect_: things like getting configuration generated so we could build config maps based upon | 15:15 |
| *** qwang has quit IRC | 15:16 | |
| portdirect_ | oh - I know the hard work is yet to begin :/ | 15:16 |
| *** qwang has joined #openstack-kolla | 15:16 | |
| portdirect_ | sbezverk_: yeah the config issue is the one that troubles me most as well | 15:16 |
| portdirect_ | We will need to think abou that quite a bit - to make sure we produce somthing that works for the whole community - but I'm sure we can get there | 15:18 |
| *** tovin07 has quit IRC | 15:20 | |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible: Source the openrc file automatically when running tools/init-runonce https://review.openstack.org/409179 | 15:20 |
| sbezverk_ | portdirect_: atm I do not see solution other than continue use ansible for this part.. | 15:23 |
| *** clsacramento has quit IRC | 15:24 | |
| sbezverk_ | but then you need 3 repos to get going.. which is kind of ugly.. | 15:24 |
| *** lamt has joined #openstack-kolla | 15:25 | |
| portdirect_ | sbezverk_: this is very true for now - and i think will need to continue for a bit. I must confess I am no where near as familar with kolla-ansible as I should be (so what I'm saying here could just be flat out wrong/inapropriate - I'm thinking out loud) | 15:30 |
| portdirect_ | but it would be nice to see kolla-ansible more clearly split between deployment and genconfig | 15:31 |
| *** srwilkers has quit IRC | 15:31 | |
| kfox1111 | morning. | 15:31 |
| kfox1111 | yeah. config will be tricky. | 15:32 |
| portdirect_ | as i think it would be great to hava a common config solution for both deployment models - also can genconfig just generate aconfig for a particlaur service (or set of services?) I've only ever used it to run the whole lot out? | 15:32 |
| portdirect_ | morning kfox :) | 15:32 |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible: Source the openrc file automatically when running tools/init-runonce https://review.openstack.org/409179 | 15:33 |
| kfox1111 | I'm kind of wondering if we can rework the config bits in such a way, that it can be rendered by either jinja2 or gotl. | 15:33 |
| kfox1111 | then we could make microservice packages out of them. | 15:33 |
| kfox1111 | or, | 15:34 |
| kfox1111 | we could make an operator like thing out of genconfig, and produce a microservice for the operator. | 15:34 |
| kfox1111 | so, if a use rjust wanted to deploy, they would need helm packages and containers. each could potentially be prebuilt/pretested from somewhere, so no need for 3 repos to checkout. | 15:35 |
| kfox1111 | portdirect_: found one more bug in the optional haproxy code. | 15:38 |
| kfox1111 | glad I slept on it. | 15:38 |
| portdirect_ | reargdless of how we go, i think your last point is spot on - idealy a user would just want to set up a k8s cluster, and the download some sorta config bundle (being vague intentonally) and not need to do much more than that | 15:38 |
| portdirect_ | kfox1111: dang! | 15:38 |
| portdirect_ | comment? | 15:38 |
| kfox1111 | commented | 15:39 |
| kfox1111 | portdirect_: yeah. another option is bundle the genconfgi bits right in a deployment/devleopment container. so user runs that, then does a genconfig. no checkout needed. | 15:40 |
| portdirect_ | well it would after 2 days :D - will fix asap :) | 15:40 |
| sbezverk_ | kfox1111: morning, can you check mariadb need to merge it | 15:40 |
| kfox1111 | portdirect_: yeah. true that. :) | 15:40 |
| kfox1111 | sbezverk_: yup. that was next on my list. | 15:40 |
| sbezverk_ | kfox1111: thanks | 15:40 |
| *** narasimha_SV has quit IRC | 15:42 | |
| kfox1111 | sbezverk_: np. :) | 15:42 |
| openstackgerrit | Pete Birley proposed openstack/kolla-kubernetes: Update common deployment for python apis https://review.openstack.org/408165 | 15:44 |
| *** srwilkers has joined #openstack-kolla | 15:46 | |
| *** krtaylor has joined #openstack-kolla | 15:47 | |
| srwilkers | of course the wifi kicks me off when im away :( | 15:47 |
| *** msimonin has joined #openstack-kolla | 15:48 | |
| mliima | awiddersheim, ping | 15:48 |
| awiddersheim | sup | 15:49 |
| *** kristian__ has quit IRC | 15:49 | |
| mliima | are you busy? awiddersheim | 15:49 |
| awiddersheim | I can spare a few minuets | 15:49 |
| awiddersheim | whats up? | 15:49 |
| mliima | i need some help with cinder + nfs | 15:49 |
| awiddersheim | ok | 15:49 |
| mliima | http://paste.openstack.org/show/591927/ awiddersheim | 15:50 |
| mliima | that is my error now | 15:50 |
| kfox1111 | sbezverk_: looks good to me. workflowed. | 15:50 |
| mliima | however, when i ran showmount -e ip from my host | 15:50 |
| mliima | it works | 15:50 |
| kfox1111 | portdirect_: also looks good. | 15:50 |
| sbezverk_ | kfox1111: thanks I will do some changes on rabbitmq as well to use new common things | 15:51 |
| kfox1111 | sbezverk_: cool. thanks. | 15:52 |
| awiddersheim | hmmm | 15:52 |
| mliima | awiddersheim, my cinder.conf http://paste.openstack.org/show/591958/ | 15:52 |
| awiddersheim | did you run the showmount from inside the container? | 15:53 |
| mliima | yes | 15:53 |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible: Source the openrc file automatically when running tools/init-runonce https://review.openstack.org/409179 | 15:53 |
| mliima | cinder-volume | 15:53 |
| mliima | and i get that error | 15:53 |
| mliima | whatever the ip, i get the same error | 15:53 |
| openstackgerrit | Merged openstack/kolla-kubernetes: Hel-ming mariadb components https://review.openstack.org/408160 | 15:53 |
| mliima | localhost, 172.24.53.1 | 15:53 |
| mliima | ... | 15:53 |
| awiddersheim | but not from your host? | 15:53 |
| mliima | yes | 15:54 |
| *** pece has joined #openstack-kolla | 15:54 | |
| awiddersheim | just in the container? | 15:54 |
| mliima | just in container | 15:54 |
| *** Guest57 has joined #openstack-kolla | 15:54 | |
| mliima | from host awiddersheim http://paste.openstack.org/show/591959/ | 15:54 |
| *** msimonin has quit IRC | 15:55 | |
| awiddersheim | makes me think the container is missing some RPC things or something | 15:55 |
| awiddersheim | like required packages | 15:55 |
| mliima | hmm | 15:55 |
| awiddersheim | I'm using a totally different driver | 15:55 |
| awiddersheim | netapp | 15:55 |
| mliima | ok awiddersheim, do you know which required packages are missing? | 15:56 |
| awiddersheim | no and that is just a guess | 15:56 |
| awiddersheim | what are you using for base? | 15:56 |
| mliima | sorry, i don't understand | 15:57 |
| mliima | :/ | 15:57 |
| mliima | base? | 15:57 |
| mliima | base image to build the container? | 15:57 |
| awiddersheim | yeah | 15:57 |
| mliima | ubuntu | 15:57 |
| awiddersheim | hrm we've been using centos mainly | 15:58 |
| awiddersheim | googling the clnt_create: RPC: Unknown host error seems to point to some DNS problems but you are using an IP | 15:58 |
| mliima | yes :/ | 15:58 |
| mliima | ok awiddersheim, thank you for support | 15:59 |
| mliima | :) | 15:59 |
| awiddersheim | you aer using NFSv3 or 4? | 15:59 |
| awiddersheim | if using v3 then things might not work | 16:00 |
| mliima | 4 | 16:00 |
| mliima | well | 16:00 |
| mliima | i think | 16:00 |
| mliima | should be 4 | 16:00 |
| awiddersheim | teh RPC stuff makes me thing v3 but might want to check | 16:01 |
| awiddersheim | unfortunately between using a different base for container and differenet NFS drivers not much | 16:01 |
| awiddersheim | I can do on my end to help | 16:01 |
| kfox1111 | portdirect_: sorry. one more... the mariadb one merged so yours has a mergeconflict now. :/ | 16:01 |
| portdirect_ | np, expected that... -will fix up now we can take advantage of sbezverk's work :) | 16:02 |
| portdirect_ | will also have keystone svc ready in a few mins as well - might as well move the lowhanging fruit quickly | 16:03 |
| kfox1111 | yours should go quick, so I'll rebase the neutron endpoint one on yours, to prevent another merge conflict. | 16:04 |
| *** Guest57 has quit IRC | 16:04 | |
| mliima | awiddersheim, http://paste.openstack.org/show/591961/ | 16:05 |
| mliima | .. | 16:05 |
| *** gfidente has quit IRC | 16:06 | |
| *** egonzalez90 has quit IRC | 16:08 | |
| *** tonanhngo has joined #openstack-kolla | 16:09 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Have gate wait for dns ready https://review.openstack.org/408907 | 16:10 |
| *** sdake has joined #openstack-kolla | 16:13 | |
| *** klindgren__ has joined #openstack-kolla | 16:15 | |
| *** klindgren_ has quit IRC | 16:17 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: PS modifies Rabbitmq templates to use new common elements https://review.openstack.org/409214 | 16:18 |
| *** matrohon has quit IRC | 16:21 | |
| sdake | morning folks | 16:22 |
| srwilkers | hey sdake :) | 16:22 |
| sdake | sup srwilkers | 16:22 |
| mliima | morning sdake | 16:23 |
| *** sayantan_ has joined #openstack-kolla | 16:23 | |
| *** sayantan_ has quit IRC | 16:24 | |
| *** sayantan_ has joined #openstack-kolla | 16:25 | |
| sbezverk_ | morning sdake | 16:26 |
| *** adrian_otto has joined #openstack-kolla | 16:27 | |
| kfox1111 | sdake: morning. | 16:27 |
| kfox1111 | sdake: how close is a 3.0.2? | 16:27 |
| kfox1111 | cause 3.0.1 has a bad openstack client in the kolla toolbox. :/ | 16:28 |
| *** Pavo has quit IRC | 16:29 | |
| *** jascott1 has joined #openstack-kolla | 16:29 | |
| *** mattmceuen has joined #openstack-kolla | 16:30 | |
| *** Pavo has joined #openstack-kolla | 16:34 | |
| sbezverk_ | kfox1111: have you done concatination of a gotpl variable with a string inside of {{}} ? | 16:35 |
| openstackgerrit | Merged openstack/kolla: Fix zaqar ui https://review.openstack.org/407065 | 16:36 |
| *** jascott1 has quit IRC | 16:36 | |
| kfox1111 | with a new var. | 16:36 |
| kfox1111 | theres a function for concat. | 16:36 |
| kfox1111 | or printf works too. | 16:36 |
| kfox1111 | there's a second form of printf too. | 16:37 |
| kfox1111 | {{ $foo | printf "%s-bar" }} | 16:37 |
| kfox1111 | if you pass it via pipeline, the pipeline shows up to printf as an appended last argument. | 16:38 |
| sbezverk_ | kfox1111: got it thanks forgot about printf | 16:39 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: PS modifies Rabbitmq templates to use new common elements https://review.openstack.org/409214 | 16:41 |
| *** lrensing has joined #openstack-kolla | 16:44 | |
| kfox1111 | sbezverk_: looks good. | 16:45 |
| kfox1111 | will workflow when the gate passes. | 16:45 |
| sbezverk_ | kfox1111: sounds good, thank you | 16:46 |
| kfox1111 | np. :) | 16:46 |
| kfox1111 | thanks for working on it. :) | 16:46 |
| *** msimonin has joined #openstack-kolla | 16:51 | |
| *** msimonin has quit IRC | 16:55 | |
| *** schwicht has joined #openstack-kolla | 16:56 | |
| *** sp__ has joined #openstack-kolla | 16:59 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: PS modifies Rabbitmq templates to use new common elements https://review.openstack.org/409214 | 16:59 |
| openstackgerrit | Merged openstack/kolla-kubernetes: Update common deployment for python apis https://review.openstack.org/408165 | 17:00 |
| kfox1111 | ah... think I figured out how trove's artefacting to tarballs.openstack.org. | 17:01 |
| kfox1111 | we should be able to do the same... :) | 17:01 |
| *** fragatina has quit IRC | 17:02 | |
| *** fragatina has joined #openstack-kolla | 17:03 | |
| *** pcaruana has quit IRC | 17:03 | |
| sayantan_ | Hi guys | 17:06 |
| kfox1111 | hi | 17:06 |
| sayantan_ | the kolla-build.conf file is generated by kolla right? Correct me if I am wrong | 17:07 |
| sayantan_ | Getting a little confused since I see "genconfig Generate configuration files for enabled OpenStack services" in kolla-ansible script | 17:10 |
| *** devataa has quit IRC | 17:10 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 17:11 |
| portdirect_ | sayantan_: just knocked this up talking to someone esle - whihc generates a kolla-build.conf | 17:11 |
| srwilkers | kfox1111, portdirect_: left some comments regarding the common_manage_db bits in the keystone review because i thought it was a lot to put here. just wanted some clarification on something as im working through the cinder bits | 17:11 |
| portdirect_ | sayantan_: may be of help? https://gist.github.com/intlabs/dc8b14c09afab8132b18c3a9605b2483 | 17:12 |
| sayantan_ | Thanks portdirect_ let me check the link | 17:12 |
| kfox1111 | srwilkers: which review? The workflow -1'ed one? | 17:13 |
| sayantan_ | yes. It should be done from kolla. I will open a patch to remove the genconfig option from kolla-ansible. Can be a little misleading | 17:13 |
| portdirect_ | srwilkers: you got link? | 17:14 |
| srwilkers | yeah sorry kfox: https://review.openstack.org/#/c/407780/28 | 17:14 |
| srwilkers | im sure its my newbie showing, but just wanted some clarification that im probably overthinking something | 17:14 |
| kfox1111 | srwilkers: ah. sorry. I was ignoring it as it was getting reworked. I'll have a look. | 17:14 |
| *** Marcellin__ has joined #openstack-kolla | 17:14 | |
| srwilkers | i just posted them a couple of minutes ago kfox1111, no worries :) | 17:15 |
| portdirect_ | srwilkers: this is the one that got merged: https://review.openstack.org/#/c/408836/3/helm/kolla-common/templates/_common_manage_db_job.yaml | 17:15 |
| portdirect_ | but your questions are still relivant as its just the same | 17:15 |
| srwilkers | oh thats right. that did get split off. should get in a habit of closing tabs out more frequently | 17:16 |
| *** sdake_ has joined #openstack-kolla | 17:18 | |
| *** schwicht has quit IRC | 17:18 | |
| portdirect_ | np :) | 17:19 |
| portdirect_ | I just buy more ram, much easier than closing tabs | 17:20 |
| portdirect_ | though perhaps thats why i'm always broke :) | 17:20 |
| *** sdake has quit IRC | 17:21 | |
| kfox1111 | portdirect_: I kind of do the same. found some interesting browser bugs that way. ;) | 17:25 |
| *** b_bezak has quit IRC | 17:25 | |
| *** jascott1 has joined #openstack-kolla | 17:25 | |
| *** b_bezak has joined #openstack-kolla | 17:26 | |
| portdirect_ | I've on occasion restarted my computer to stop soundcloud playing - as finding it was much harder than wheres wally (waldo in the states I think?) | 17:28 |
| *** jascott1 has quit IRC | 17:28 | |
| *** jascott1 has joined #openstack-kolla | 17:29 | |
| *** Guest57 has joined #openstack-kolla | 17:30 | |
| sbezverk_ | kfox1111: updated rabbit came backk all green ;-) | 17:30 |
| *** b_bezak has quit IRC | 17:30 | |
| kfox1111 | cool. | 17:30 |
| kfox1111 | workflowed | 17:31 |
| sbezverk_ | kfox1111: I guess the rabbit hole was recently painted ;-) | 17:31 |
| *** schwicht has joined #openstack-kolla | 17:31 | |
| srwilkers | thanks for looking at that kfox1111 portdirect_ | 17:31 |
| kfox1111 | srwilkers: np. thanks for raising the issue. | 17:31 |
| openstackgerrit | Merged openstack/kolla-kubernetes: PS modifies Rabbitmq templates to use new common elements https://review.openstack.org/409214 | 17:35 |
| *** schwicht has quit IRC | 17:36 | |
| mgiles | sbezverk_ I have a question on the common_svc you added if you have a moment | 17:39 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Have gate wait for dns ready https://review.openstack.org/408907 | 17:40 |
| *** HyperJohnGraham has quit IRC | 17:42 | |
| *** strigazi is now known as strigazi_AFK | 17:43 | |
| *** Jeffrey4l has quit IRC | 17:45 | |
| *** athomas has quit IRC | 17:48 | |
| *** srwilkers is now known as srwilkers_lunch | 17:48 | |
| *** jemcevoy has joined #openstack-kolla | 17:48 | |
| sbezverk_ | mgiles: sure, how I can helm you? | 17:49 |
| mgiles | sbezverk_ trying to use the common_svc in the nova stuff and running into a couple of issues | 17:49 |
| mgiles | both having to do with the fact that this uses element_name for all values | 17:49 |
| mgiles | First, for nova-api, the service selector and port name differ (nova and nova-api respectively) | 17:50 |
| mgiles | I'm not sure if the service selector name really matters? Is that ever used to look it up? | 17:50 |
| mgiles | I'm not sure if you know the answer, but just wondering if these should always be the same or not | 17:50 |
| mgiles | since they didn't used to be | 17:51 |
| jemcevoy | Anyone know how to make kolla-build work without crashing a CentOS7 VM? I have had panics due to Null pointers in yum and also "see kernel:BUG: soft lockup - CPU#1 stuck for 23s! [yum:10359]" | 17:51 |
| kfox1111 | service_name/service_key shoudl come from values. should default to kolla_controller for everything but nova-compute. | 17:51 |
| *** sdake_ has quit IRC | 17:51 | |
| mgiles | sbezverk_ :) Just caught the pun there... | 17:51 |
| jemcevoy | I am building 3.0.1 | 17:51 |
| kfox1111 | should be able to override the default service_key: kolla_compute on the nova-compute microservice in all_values.yaml | 17:51 |
| *** msimonin has joined #openstack-kolla | 17:52 | |
| mgiles | kfox1111 if that's directed at me, I think we're talking about different things | 17:53 |
| portdirect_ | jemcevoy: that's unlikly to directly be a kolla-build issue - but a docker one, as it pushed the daeon pterry hard with connurent builds | 17:53 |
| portdirect_ | jemcevoy: try --threads=1 and see if resolves it | 17:53 |
| mgiles | This is the service selector listed inside the service description (for looking up the service I think) | 17:54 |
| kfox1111 | mgiles: yeah. ok. let me try rereading. | 17:54 |
| mgiles | for nova-api svc, used to be: | 17:54 |
| mgiles | selector: | 17:54 |
| mgiles | service: nova | 17:54 |
| mgiles | type: api | 17:54 |
| srwilkers_lunch | mgiles, i think the same issue arises with cinder too | 17:54 |
| sbezverk_ | mgiles: for nova-api you do not need to use element_name to name the actual process | 17:54 |
| sbezverk_ | in this case element_name: would be just "nova" | 17:55 |
| sbezverk_ | or I am missing something here | 17:55 |
| mgiles | but if I set element_name to nova, then the port name also ends up being just "nova" instead of "nova-api" | 17:55 |
| kfox1111 | sbezverk_: yeah. I don't think there is any reason to have multiple elements of nova-api. | 17:55 |
| openstackgerrit | Qin Wang (qwang) proposed openstack/kolla-kubernetes: WIP - Hel-ming horizon components https://review.openstack.org/408975 | 17:55 |
| kfox1111 | mgiles: I think serviceName should just be nova-api in that case? | 17:55 |
| mgiles | yep, I can do that. Just inconsistent with what we had before | 17:56 |
| *** msimonin has quit IRC | 17:56 | |
| mgiles | The second issue I have is that the nova-api-svc chart produces two services from two templates (nova-api and nova-metadata). | 17:57 |
| mgiles | Obviously, I can override the element_name for the second, but kind of defeats the purpose of having a consistent name passed at runtime | 17:57 |
| *** Serlex has quit IRC | 17:57 | |
| *** schwicht has joined #openstack-kolla | 17:57 | |
| portdirect_ | nova-metadata is a special snowflake - I'd be tempted to have it use it's own template | 17:57 |
| kfox1111 | mgiles: I think we don't mess with element names at all with nova. as there should only be one of them? | 17:58 |
| jemcevoy | portdirect_, So I should just add --threads=1 to my command line like "kolla-build --registry kolla4echo.scm.penguincomputing.com:4000 --push --threads=1" | 17:58 |
| kfox1111 | portdirect_: yeah. was on the fence on that one when I did generic-service origionally. | 17:58 |
| portdirect_ | jemcevoy: I'd try that | 17:59 |
| jemcevoy | I have a 4 core KVM VM with 8GB RAM and btrfs filesystem on a raw 128GB volume | 18:00 |
| mgiles | kfox1111, not sure what you're proposing. element name for everything is nova, instead of each of the microservice names? That's going to mean very little could use the common templates, so I'm guessing I'm misunderstanding | 18:00 |
| kfox1111 | ah... I think I see the issue. | 18:00 |
| sbezverk_ | mgiles: if you want a custom service name you can always build one on top of element_name | 18:00 |
| kfox1111 | .element_name is used in the _common_svc macro. | 18:00 |
| openstackgerrit | Qin Wang (qwang) proposed openstack/kolla-kubernetes: WIP - Hel-ming horizon components https://review.openstack.org/408975 | 18:00 |
| kfox1111 | but it's a . var, so in nova-api-svc, you can just hard code it to nova-api? | 18:00 |
| kfox1111 | {{ with $env := dict .... "element_name" "nova-api" ... }} ? | 18:01 |
| mgiles | yes, I could, but it means it couldn't be overridden at runtime | 18:01 |
| *** jemcevoy has left #openstack-kolla | 18:01 | |
| kfox1111 | mgiles: yes. but not sure it needs to be. | 18:01 |
| mgiles | which I thought was the purpose of using the passed in element_name for stuff | 18:01 |
| *** jmcevoy has joined #openstack-kolla | 18:01 | |
| kfox1111 | mgiles: element_name is for multi-instance. | 18:01 |
| sbezverk_ | svc1 := printf "$s-custom" .Values.element_name | 18:02 |
| kfox1111 | so, nova-compute for example. if you need different sets of nova-compute for different host aggregates. | 18:02 |
| sbezverk_ | and then pass to common_svc svc1 as element name | 18:02 |
| kfox1111 | but I think there can only be one set of nova-api in a region? | 18:02 |
| mgiles | sbezverk_ Yeah, like that idea to just append api to the element_name | 18:02 |
| sbezverk_ | mgiles: yep, you prepare the name you want the way you want it and pass to common_svc | 18:03 |
| *** pbourke has quit IRC | 18:03 | |
| kfox1111 | so, why support configuring a second nova-api set if it is impossible to use? | 18:03 |
| mgiles | kfox1111 That makes sense to me - I didn't know if that was possible | 18:03 |
| *** pbourke has joined #openstack-kolla | 18:03 | |
| sbezverk_ | kfox1111: who knwos what next version of openstack will throw at us ;-) | 18:03 |
| portdirect_ | jemcevoy: if be tempted to use upstream docker, and must confess i have no experience with btrfs backend, my build systems run centos with the lts upstream kernel and overlay2 backend and I've not had issues - but have seen similar things to you in the past with the loopback backend o low spec vm's and concurrent builds. | 18:04 |
| mgiles | Thanks for the input, guys. I think I have enough to go on and I can propose some stuff in a ps and we can all tear it apart there :) | 18:04 |
| kfox1111 | sbezverk_: true. but its been cosistently that way since the begining. | 18:04 |
| sbezverk_ | kfox1111: true | 18:05 |
| kfox1111 | multiple elements make the most sence to me for mariadb/rabbit/memcached/nova-compute*/neutron-agents*(for dvr) | 18:05 |
| *** TxGirlGeek has joined #openstack-kolla | 18:05 | |
| kfox1111 | and maybe horizon to do safe upgrades. | 18:05 |
| kfox1111 | not sure if any others need multiple instances in the same k8s namespace. | 18:06 |
| kfox1111 | oh. maybe cinder-volume too. | 18:06 |
| portdirect_ | think cells may change the above re nova? | 18:06 |
| * portdirect_ trolls kfox | 18:06 | |
| kfox1111 | portdirect_: yeah. though last I looked cells were asically just a whole nother region basically. | 18:07 |
| kfox1111 | so we may want to do those with a namespace still? | 18:07 |
| portdirect_ | ^^ +2 to that *when* we get there | 18:08 |
| *** schwicht has quit IRC | 18:08 | |
| kfox1111 | unfortunatly, I've never deployed cells, so I lack some knowlege there. :/ | 18:08 |
| portdirect_ | in kolla a cell would be most analagous to k8s fereation i think | 18:08 |
| kfox1111 | neutron doesn't have cell support yet though, so havent' seen where they could help too much with scaling. :/ | 18:09 |
| kfox1111 | yeah. cells could be loaded into different k8s instances. | 18:09 |
| kfox1111 | though I think k8s scales better these days then nova. | 18:10 |
| portdirect_ | lets get to that rabbit warren further down the road :) | 18:10 |
| kfox1111 | +1 | 18:11 |
| * portdirect_ tiries to put the lid back on pandoras box | 18:11 | |
| portdirect_ | :) | 18:11 |
| kfox1111 | yeah. I think just supporting one nova-api per namespace is probably just fine for now. | 18:11 |
| kfox1111 | we'll evaluate cells when one of us has time to figure out how cells actually work. | 18:12 |
| kfox1111 | cells and tricircle (which I think is trying to do cells for neutron) | 18:12 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Have gate wait for dns ready https://review.openstack.org/408907 | 18:14 |
| *** jgriffith_AutoAw is now known as jgriffith | 18:14 | |
| kfox1111 | k. so the dns-test passed the gate... so hopefully that will fix the race. | 18:15 |
| *** ipsecguy has joined #openstack-kolla | 18:16 | |
| kfox1111 | portdirect_: do you know much about docker export? | 18:16 |
| portdirect_ | a bit | 18:17 |
| portdirect_ | sup? | 18:17 |
| *** msimonin has joined #openstack-kolla | 18:17 | |
| kfox1111 | I think I figured out how peple are artefacting things in the gate. | 18:17 |
| kfox1111 | so I might be able to cache kolla container builds. | 18:17 |
| kfox1111 | it would be nice if they didn't flatten out though. is there a way to export while still keeping the layers in tact? | 18:18 |
| sbezverk_ | kfox1111: about dns, I think it would be better to resolve kubernetes.default.svc, in this case we know for sure that internal name resolution works | 18:18 |
| portdirect_ | not that I know of | 18:18 |
| portdirect_ | you also loose some metadata, though not everything like docker save (used? - maty be better now) does | 18:19 |
| kfox1111 | sbezverk_: most of the cases of failure I've seen have been external dns. | 18:19 |
| kfox1111 | I could add a kubernetes one too though. | 18:19 |
| kfox1111 | check both cases. | 18:20 |
| *** ipsecguy_ has quit IRC | 18:20 | |
| kfox1111 | sbezverk_: is it kubernetes.default.svc or kubernetes.default? | 18:21 |
| portdirect_ | i agree with sbezverk on having an internal dns check, but have also seen quite a fex where external has gone down :/ | 18:21 |
| portdirect_ | though i have some suspisions that docker bub may also be to blame on occasion | 18:21 |
| portdirect_ | *hub | 18:21 |
| kfox1111 | portdirect_: yeah. I've seen the hub break it. hence the precacheing / docker export question. :) | 18:22 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Have gate wait for dns ready https://review.openstack.org/408907 | 18:23 |
| *** prameswar has quit IRC | 18:23 | |
| kfox1111 | we should cache: 1. kolla containers. 2. kube containers. 3. kube tools like kubeadm/kubectl/helm binary. 4. kolla-kube packages | 18:24 |
| kfox1111 | then have gate jobs that test/refresh them. | 18:24 |
| kfox1111 | oh. and prebuilt genconfig. :) | 18:24 |
| kfox1111 | should be able to then shave off 10's of minutes off the gate job runtime. | 18:24 |
| kfox1111 | while at the same time making it more stable. | 18:24 |
| portdirect_ | I'm not so sure about that :/ has there been any progress on a registry in infra - my experience with export/importing container has ususally resulted in more hassle than it's solved | 18:25 |
| sbezverk_ | kfox1111: please add an internal one too, then we cover all bases | 18:25 |
| portdirect_ | as docker intionally makes it hard to do it without wrping th image in some way | 18:25 |
| kfox1111 | sbezverk_: the most recent update has that. | 18:26 |
| portdirect_ | also can we get the pull_images back in? | 18:26 |
| kfox1111 | portdirect_: thats the alternative I think. maybe createing docker repo container with the containers preloaded. | 18:26 |
| kfox1111 | portdirect_: I never removed it. | 18:26 |
| kfox1111 | not sure why your asking about that. | 18:26 |
| *** prameswar has joined #openstack-kolla | 18:27 | |
| portdirect_ | ah - your working from an old setup_canal from before we merged it: https://github.com/openstack/kolla-kubernetes/blob/master/tests/bin/setup_canal.sh#L16 | 18:27 |
| sbezverk_ | kfox1111: what about using kolla's docker registry, it should be up soon? | 18:28 |
| kfox1111 | portdirect_: ah. yeah. | 18:28 |
| kfox1111 | sbezverk_: do you have a review link? | 18:28 |
| kfox1111 | either way though, all the other things should be cached too. | 18:28 |
| *** schwicht has joined #openstack-kolla | 18:29 | |
| *** Pavo has quit IRC | 18:29 | |
| sbezverk_ | kfox1111: need to check with Jeffrey4l | 18:29 |
| sbezverk_ | for latest update on registry | 18:29 |
| qwang | a quick q: what's the difference between include "common_containers" . and include "common_containers" $env ? | 18:30 |
| kfox1111 | qwang: '.' is a magic variable in gotl that is kind of a var namespace. | 18:32 |
| *** harlowja has joined #openstack-kolla | 18:32 | |
| kfox1111 | when you include another macro, | 18:32 |
| kfox1111 | you specify as the second argument what the definition of '.' is to that macro's perspective. | 18:33 |
| kfox1111 | include "x" . | 18:33 |
| kfox1111 | passes the . environment from the current template directly to the included macro. | 18:33 |
| kfox1111 | include "x" $env uses the $env variavble as the '.' vars in the macro. | 18:33 |
| v1k0d3n | hey guys...have a general question. | 18:34 |
| kfox1111 | you can also consider these two thigns eqiuv: | 18:34 |
| *** Pavo has joined #openstack-kolla | 18:34 | |
| *** harlowja_ has joined #openstack-kolla | 18:34 | |
| kfox1111 | {{ import "foo" $env }} | 18:34 |
| kfox1111 | and | 18:34 |
| *** harlowja has quit IRC | 18:35 | |
| jascott1 | qwang its a common concept of context in template systems | 18:35 |
| kfox1111 | {{- with . := $env }} | 18:35 |
| kfox1111 | {{ import "foo" . }} | 18:35 |
| kfox1111 | {{- end | 18:35 |
| v1k0d3n | is kolla-kubernetes a new project...looking at things from a fresh perspective (i.e. greenfield) or is this really intended to be the same project entirely? | 18:35 |
| *** schwicht has quit IRC | 18:35 | |
| kfox1111 | v1k0d3n: some of both. | 18:36 |
| qwang | thank you kfox1111 and jascott1. | 18:36 |
| kfox1111 | kubernetes enables some things to be done differently, and its a good idea to do them differently. | 18:36 |
| v1k0d3n | the way the repo is shaping up, i'm really comfused. it looks like some of the same templating that makes groups like stackanetes and CCP tack a step back away from Kolla. | 18:36 |
| kfox1111 | some of the learning on how kolla-ansible did things shouldn't be lost though. | 18:37 |
| v1k0d3n | i really thought this was a greenfield project, which was the point of breaking up the repos. | 18:37 |
| kfox1111 | v1k0d3n: how so? | 18:37 |
| v1k0d3n | so i want to put all these services in a helm repo....local for development, or i want to deliver these services to the kubernetes team so it can gain wider adoption wiht helm users. how would you propose this? | 18:38 |
| kfox1111 | been trying to write things in such a way those groups could reuse the helm packages. | 18:38 |
| v1k0d3n | one of the main issues that Openstack is facing is a interop question. where does openstack fit in with kubernetes. | 18:39 |
| kfox1111 | we depend on a stock kubernetees and a stock helm. | 18:39 |
| kfox1111 | then we provide regular helm packages. | 18:39 |
| v1k0d3n | what do you mean by stock, so i don't misunderstand. | 18:39 |
| v1k0d3n | ? | 18:39 |
| kfox1111 | users can then assemble an openstack using them. | 18:39 |
| kfox1111 | stock = unpatched | 18:40 |
| kfox1111 | we don't need like a custom thingy like openshift does to k8s. | 18:40 |
| v1k0d3n | so what about providing cinder, glance, nova as https://github.com/kubernetes/charts | 18:40 |
| kfox1111 | I think we can do that eventually maybe, | 18:40 |
| v1k0d3n | that should be a goal from the onset | 18:41 |
| kfox1111 | kubernetes/chart's build system is.... insufficient currently. :/ | 18:41 |
| v1k0d3n | wider adoption and it addresses interop immediately. | 18:41 |
| v1k0d3n | no it's not. | 18:41 |
| kfox1111 | we're going to have many dozens of packages. | 18:41 |
| v1k0d3n | we've been doing just fine working right with it. | 18:41 |
| kfox1111 | kubernetes/charts has no way to build that many packages currently. | 18:41 |
| v1k0d3n | right... | 18:42 |
| kfox1111 | I need a way to cut a release of dozens of packages with a few minutes of effort or less. | 18:42 |
| v1k0d3n | so take a play form openstack-ansible | 18:42 |
| v1k0d3n | provide the core openstack services. everything else can be bolted right on. | 18:42 |
| kfox1111 | thats what we're doing. | 18:42 |
| kfox1111 | still, we're talking about dozens of of packages. | 18:42 |
| kfox1111 | and I want to be able to rerelease whenever a dep changes. | 18:43 |
| kfox1111 | you can't just checkout kubernetets/charts and do a "build all" | 18:43 |
| v1k0d3n | we're working with deis and kubernetes for the end goal, i'm assuming? | 18:43 |
| kfox1111 | they will get there eventually. but doesn't yet. | 18:43 |
| v1k0d3n | i.e. = interop | 18:44 |
| kfox1111 | v1k0d3n: there's a lot of stuff to do. thats one task yes. | 18:44 |
| kfox1111 | but, for now, just trying to get our stuff to work in a helm world. once that's done, | 18:44 |
| v1k0d3n | can we invite them in to help? | 18:44 |
| v1k0d3n | so kfox1111 we're doing the same thing. | 18:44 |
| kfox1111 | we can see about getting the kubernetes/charts repo build upgraded. | 18:44 |
| kfox1111 | v1k0d3n: sure. :) | 18:45 |
| v1k0d3n | there's too much parity. | 18:45 |
| v1k0d3n | i'm concerned. | 18:45 |
| kfox1111 | v1k0d3n: its not clear to me, exactly what features we needed out of the build system. | 18:45 |
| kfox1111 | for example, we need to figure out how to do version tagging still. | 18:45 |
| kfox1111 | and that may change things drasticially. | 18:46 |
| v1k0d3n | it looks like there's a larger effort to provide what kolla is today, keeping all that intact for kolla-kube, when kolla-kube should be a greenfield deployment strategy. | 18:46 |
| kfox1111 | so I didn't want to try and push upstream for features until we were shure we knew what we needed them to look like. | 18:46 |
| v1k0d3n | but the structure is different. | 18:46 |
| kfox1111 | the structure is a bit different, because w hat helm provides is insufficient. :/ | 18:47 |
| kfox1111 | but we're not sure exactly teh feature ask yet. | 18:47 |
| kfox1111 | but we will work with upstream once we are sure exactly what we need to ask for. | 18:47 |
| v1k0d3n | so the SAP folks...have been been inviting them in? | 18:48 |
| kfox1111 | like, I think we need a way to include vars from a subpackage. | 18:48 |
| kfox1111 | v1k0d3n: sure. everyone's welcome | 18:48 |
| v1k0d3n | ok... | 18:48 |
| v1k0d3n | this is what i mean. | 18:48 |
| v1k0d3n | their solution is minor gaps, but they are able to do what you are looking for. | 18:49 |
| sbezverk_ | v1k0d3n: we are public IRC anybody who has interest and willing to contribute can pick up a task from action and do it.. | 18:49 |
| jascott1 | sorry to jump in here, i went to SAP repo and they were very much 'this is for us' and I dont see a problem with that | 18:49 |
| kfox1111 | v1k0d3n: last I looked there was one big difference in aproaches. | 18:49 |
| v1k0d3n | sbezverk_: i agree, but you are not running kubernetes on openstack. you are running openstack on kubernetes, so the interop discussion goes both ways. | 18:49 |
| sbezverk_ | I was following this dialogue and had a sense that we were blamed for something, I could not grasp what exactly | 18:49 |
| v1k0d3n | not "we are here if they want to join". | 18:49 |
| v1k0d3n | just saying... | 18:49 |
| jascott1 | not sayin they dont take PRs but they are very driven to implement their way it seems | 18:49 |
| kfox1111 | v1k0d3n: one of using helm for orchestration and one that sperates orchestration outside of the base micropackages. | 18:49 |
| kfox1111 | which does make the requirements of the build system quite different. | 18:50 |
| *** papacz has quit IRC | 18:50 | |
| kfox1111 | kolla-kubernetes prime archetectural goal right now is flexibility in deployments, to ensure anyone can reuse the kolla-kubernetes building blocks. | 18:50 |
| kfox1111 | sap/stackanetes havent' been designed that way. | 18:50 |
| v1k0d3n | sbezverk_: not blamed...it just looks more brownfield adhering to what kolla is, as opposed to greenfield and how helm/kubernetes makes things easier. | 18:50 |
| v1k0d3n | kfox1111: kolla is opinionated | 18:51 |
| v1k0d3n | i don't want kolla-kube to be opinionated. | 18:51 |
| kfox1111 | v1k0d3n: I agree with you. we need closer ties with k8s. but we're very very new project. we're al stilll learning. | 18:51 |
| v1k0d3n | this is where vendors take advantage. | 18:51 |
| v1k0d3n | we want to throw everything into kolla-kube. | 18:51 |
| kfox1111 | v1k0d3n: I don't want an opinionated kolla-kube either. trying very very hard to keep it unopinionated. | 18:51 |
| v1k0d3n | fwiw, i think we are starting to show signs of that. | 18:52 |
| v1k0d3n | kfox1111: then why do the repos not look more similar to this... | 18:52 |
| kfox1111 | v1k0d3n: please don't get me wrong. I want to know of all signs of opinionatedness. so we can fix them. | 18:52 |
| v1k0d3n | then why don't the repos look more like this: https://github.com/sapcc/openstack-helm | 18:52 |
| jascott1 | v1k0d3n what part of the approach is a problem? | 18:52 |
| v1k0d3n | or this: https://github.com/att-comdev/aic-helm | 18:53 |
| kfox1111 | vincent_vdk: because that looks quite opinionated to me. | 18:53 |
| v1k0d3n | or this: https://github.com/kubernetes/charts/tree/master/stable | 18:53 |
| kfox1111 | v1k0d3n: I mean. it looks quite opinionated to me. | 18:53 |
| kfox1111 | v1k0d3n: for example: https://github.com/sapcc/openstack-helm/tree/master/neutron/templates | 18:53 |
| kfox1111 | it forces helm to be your orchestration engine. | 18:53 |
| v1k0d3n | that is yes... | 18:53 |
| v1k0d3n | and https://github.com/att-comdev/aic-helm removes that | 18:54 |
| kfox1111 | it tries to launch all the things in there up. | 18:54 |
| v1k0d3n | we were asked by inc0 to do a PoC | 18:54 |
| v1k0d3n | we're doing what we were asked. | 18:54 |
| v1k0d3n | we gather folks who have done this before. | 18:54 |
| kfox1111 | v1k0d3n: yeah. just saying. you asked why not do it like openstack-helm. I'm saying, becuase we're trying to be unopinionated. | 18:54 |
| v1k0d3n | and the goal was to share, but they are shaping up differently. | 18:54 |
| kfox1111 | aic-helm looks less opionated. | 18:54 |
| v1k0d3n | very much less. | 18:54 |
| kfox1111 | v1k0d3n: thats one nof the problems with having teams not work together. | 18:55 |
| v1k0d3n | and uses Probe | 18:55 |
| kfox1111 | v1k0d3n: if you want a converged solution, its best to have the teams work on the same thing. :/ | 18:55 |
| v1k0d3n | and entrypoint (we talked about entrypoint) | 18:55 |
| kfox1111 | otherwise, they tend to naturally diverge. :/ | 18:55 |
| *** haplo37_ has quit IRC | 18:55 | |
| sbezverk_ | v1k0d3n: I think approach taken was not correct, asking some people to do PoC and the rest sit and wait does not make too much sense.. | 18:55 |
| v1k0d3n | i told folks to take | 18:56 |
| kfox1111 | v1k0d3n: look at https://github.com/att-comdev/aic-helm/tree/master/glance/templates | 18:56 |
| openstackgerrit | Qin Wang (qwang) proposed openstack/kolla-kubernetes: WIP - Hel-ming horizon components https://review.openstack.org/408975 | 18:56 |
| kfox1111 | it also looks opinionated to me. | 18:56 |
| *** haplo37_ has joined #openstack-kolla | 18:56 | |
| portdirect_ | kfox1111, though to be ballanced we are becoming opionated in how we package our work... | 18:56 |
| v1k0d3n | kfox1111: ...explain this for me..... | 18:56 |
| kfox1111 | portdirect_: I agree there, I'd rather be temporarily opinionated on how we package, so that we provide the most unopionated deployment options. | 18:57 |
| v1k0d3n | portdirect_: ^^ yes, exactly | 18:57 |
| v1k0d3n | temp opnionated is going to cause problems. | 18:57 |
| kfox1111 | portdirect_: the build system is easily replacable in the future though. so once we get helm to the point it natively can do the things we need, then we can go stock again. | 18:57 |
| v1k0d3n | greenfield.... | 18:57 |
| kfox1111 | v1k0d3n: this is greenfield. | 18:58 |
| v1k0d3n | this is part of why i'm bring this up. | 18:58 |
| *** gagehugo has joined #openstack-kolla | 18:58 | |
| v1k0d3n | the point is to remove opinion from the onset | 18:58 |
| v1k0d3n | otherwise it loses it's value | 18:58 |
| kfox1111 | v1k0d3n: we're in violent agreement on the goal. :) | 18:58 |
| openstackgerrit | Mark Giles proposed openstack/kolla-kubernetes: WIP - Helm nova services https://review.openstack.org/408829 | 18:58 |
| kfox1111 | v1k0d3n: I think we're disagree'ing on what the requirements of the proejct are? | 18:58 |
| kfox1111 | here's where I'm at: | 18:59 |
| *** haplo37_ has quit IRC | 18:59 | |
| *** haplo37_ has joined #openstack-kolla | 18:59 | |
| kfox1111 | 1. kolla-kubernetes provides a set of basic set of helm package building blocks that can be assembled by a user to stand up an openstack. these correspond closely to the openstack provided components (neutron-l3-agent, nova-api, etc) | 19:00 |
| kfox1111 | very fign grained. | 19:00 |
| v1k0d3n | kfox1111: we are constantly trying to remove option from the deployment PoC we're working on. case in point: https://github.com/att-comdev/aic-helm/pull/14 | 19:00 |
| kfox1111 | This allows other projects to reuse those base packages the same way they currenty can reuse kolla-containers. | 19:00 |
| kfox1111 | v1k0d3n: +1 for that. | 19:01 |
| kfox1111 | v1k0d3n: I'm saying though, our archetecture removes the opinion of orchestration out too. something your not doing. that makes it look much different. | 19:01 |
| v1k0d3n | ok kfox1111 | 19:02 |
| kfox1111 | v1k0d3n: the aic-helm code base for example, wouldn't work so well with say a user that needed to use chef to do their orchestration. | 19:02 |
| v1k0d3n | i have nothing else to say. | 19:02 |
| kfox1111 | kolla-kubernetes has been tooling in such a way as that could be done. have chef generate the configmaps, upload them, then use helm to launch the microservice packages. | 19:03 |
| kfox1111 | but, we also want to support your use case. using helm to orchestrate. | 19:03 |
| kfox1111 | so, like, have a neutron package that luanches all the neutron components. | 19:03 |
| v1k0d3n | have chef generate configmaps?! | 19:03 |
| kfox1111 | those can be built with the microservice model. they just become subcharts of the neutron chart. | 19:03 |
| portdirect_ | kfox1111: that last stament conerns me, I'm not sure why would want to use anything other thank k8s native tooling for kolla-kubernetes? | 19:03 |
| v1k0d3n | dude | 19:03 |
| v1k0d3n | we're not using chef | 19:04 |
| v1k0d3n | period | 19:04 |
| kfox1111 | v1k0d3n: your not. and thats cool. | 19:04 |
| v1k0d3n | helm generates the configmaps | 19:04 |
| v1k0d3n | i think there's a massive gap here. | 19:04 |
| kfox1111 | but "you must use k8s for absolutely everyting" is opinionated. | 19:04 |
| v1k0d3n | and unfortunately, there's too much parity from the start | 19:04 |
| kfox1111 | the reason kolla-kubernetes looks the way it does, is its unopinionated. | 19:04 |
| kfox1111 | you can build opinionated workflows on top of it. | 19:05 |
| portdirect_ | the whole point of k8s is a clean slate is is not? i thought that kolla-k8s was about using k8s? I must confess I'm a bit confused by this as well? | 19:05 |
| kfox1111 | if you take the aic-helm neutron example, | 19:05 |
| v1k0d3n | kfox1111: by nature...kolla-kubernetes makes kubernetes first class dude. | 19:05 |
| kfox1111 | if you replace the service templates, whith kolla-kubernetes microservice dependenceis, it should work. | 19:05 |
| *** srwilkers_lunch is now known as srwilkers_ | 19:05 | |
| v1k0d3n | in the future...do you really think that vms aren't going to just be volume claims? | 19:05 |
| *** srwilkers_ is now known as srwilkers | 19:05 | |
| v1k0d3n | vms, serverless, claims...everything is changing. | 19:06 |
| portdirect_ | v1k0d3n: exactly^ | 19:06 |
| kfox1111 | v1k0d3n: I agree that k8s is going to eat the world. | 19:06 |
| kfox1111 | but, | 19:06 |
| v1k0d3n | no but | 19:06 |
| v1k0d3n | use openstack for what it's for infra | 19:06 |
| v1k0d3n | make it great again | 19:06 |
| v1k0d3n | :) | 19:06 |
| *** g3ek has quit IRC | 19:06 | |
| kfox1111 | the reason so many openstack on k8s projects are springing up is too much opinionated ness. | 19:06 |
| v1k0d3n | (slipped that one in there for comic relief) | 19:06 |
| kfox1111 | one tool wont work for another person's use case. | 19:06 |
| srwilkers | i see what you did there v1k0d3n | 19:06 |
| kfox1111 | so the wheel keeps getting reinvented, voer and over again. | 19:07 |
| kfox1111 | but, the vast majority of solutions reuse kolla's containers. | 19:07 |
| kfox1111 | why? because they are pretty well unopionated. | 19:07 |
| kfox1111 | kolla-kuberneets is striving for the same thing at the lower level. | 19:07 |
| v1k0d3n | kfox1111: ...no...we're picking a dynamic single source of truth for our infra-environment. | 19:07 |
| v1k0d3n | kubernetes | 19:08 |
| kfox1111 | provide helm packages that are unopinionated, so that they c an be used either directly, | 19:08 |
| v1k0d3n | which is what i thought this effort was about. | 19:08 |
| v1k0d3n | the others that you mention...are doing the same | 19:08 |
| kfox1111 | or can be built into opinionated tools that are easier to use, but more specific to a use case. | 19:08 |
| v1k0d3n | infra is needed right now | 19:08 |
| v1k0d3n | so kubernetes makes infra stronger. that's the hope right? | 19:08 |
| v1k0d3n | so what's with all the additional "stuff"? | 19:08 |
| kfox1111 | yup. and kolla-kubernetes primary focu right now is on infra. trying to get common micro packages going. | 19:08 |
| kfox1111 | which additional stuff? | 19:08 |
| v1k0d3n | well, that just opens the door for every vendor to say that's why we're different | 19:09 |
| v1k0d3n | i.e. chef | 19:09 |
| *** schwicht has joined #openstack-kolla | 19:09 | |
| v1k0d3n | i.e. kolla-build as it currently stands. | 19:09 |
| v1k0d3n | i.e. ansible | 19:09 |
| *** g3ek has joined #openstack-kolla | 19:09 | |
| kfox1111 | if we don't alow chef to do their own configmaps, and reuse the other components, | 19:09 |
| kfox1111 | then they will just rebuild it all. | 19:09 |
| kfox1111 | then the community fractures further. | 19:10 |
| kfox1111 | we're trying to make it as inclusive as possible to enable sharing/reuse. | 19:10 |
| kfox1111 | being unopionated. | 19:10 |
| v1k0d3n | unfortunately, there's a lot of opinion | 19:10 |
| v1k0d3n | this is far from opnionated. | 19:10 |
| *** sp__ has quit IRC | 19:10 | |
| v1k0d3n | i.e. operators | 19:10 |
| kfox1111 | v1k0d3n: we have no operators at present. | 19:11 |
| v1k0d3n | i.e. jinja2 | 19:11 |
| kfox1111 | and they will be optional, on top of the microservices. | 19:11 |
| kfox1111 | jinja2 is going away. | 19:11 |
| kfox1111 | its historical. | 19:11 |
| portdirect_ | kfox1111: I think the poin her that v1k0d3n is making needs to be addressed anfd though about properly - what he is saying is that if we dont produce *source* what can be resused or even userstood outside of our comminity then the fatureing is enviable | 19:11 |
| v1k0d3n | ok. well, i brought up my concerns y'all. | 19:11 |
| v1k0d3n | i'd said my part, as others have in the past as well. | 19:12 |
| kfox1111 | portdirect_: I agree with that. | 19:12 |
| kfox1111 | v1k0d3n: I'm trying very hard to understand your concerns. I don't want you to feel excluded. | 19:12 |
| v1k0d3n | kfox1111: this NEEDS to be successful. | 19:12 |
| v1k0d3n | we are BACKING this is with many more resources soon. | 19:13 |
| kfox1111 | I'm also trying to deal with bootstraping a project, and migrating a lot of existing code. | 19:13 |
| kfox1111 | things can't just happen overnight. | 19:13 |
| v1k0d3n | but i need to make sure we're on the same page. | 19:13 |
| kfox1111 | agreed. | 19:13 |
| v1k0d3n | kfox1111: actually...you can take your time with it | 19:13 |
| v1k0d3n | do it right :D | 19:13 |
| *** david-lyle_ is now known as david-lyle | 19:13 | |
| v1k0d3n | get more consensus, take more ideas. | 19:14 |
| kfox1111 | yup. trying to do that. | 19:14 |
| v1k0d3n | there are projects out there in development for a long time. | 19:14 |
| srwilkers | doing this right, first is key. i think transitioning this in the future will end up being painful | 19:14 |
| v1k0d3n | this is something, that for interop sake...should be done correctly, not quickly. | 19:14 |
| kfox1111 | and please do the same for me. I totally get you thing chef is a bad idea/dead end. I actulaly think that too. but, I'm trying to be incluseive. | 19:14 |
| v1k0d3n | interactively sure...but with all voices heard | 19:14 |
| kfox1111 | I'm not going to write chef suporting code myself, but I'm not going to block it either. | 19:14 |
| v1k0d3n | so CoreOS came in...said some things. | 19:15 |
| v1k0d3n | SAP...they've been here, came to meetings in barcelona. | 19:15 |
| v1k0d3n | are they contributing? | 19:15 |
| kfox1111 | v1k0d3n: there are two types of listening. | 19:15 |
| v1k0d3n | helping out. | 19:15 |
| v1k0d3n | ? | 19:15 |
| kfox1111 | "we do it this way, you should too". | 19:15 |
| v1k0d3n | ^^ was getting to it. | 19:15 |
| v1k0d3n | kfox1111: .... for just a moment... | 19:15 |
| kfox1111 | they may have a good reason to do it. | 19:15 |
| v1k0d3n | think about what they do... | 19:15 |
| v1k0d3n | could they possibly have something to share? | 19:16 |
| kfox1111 | v1k0d3n: want to have a call? this is getting to ocomplicated for irc. | 19:16 |
| v1k0d3n | i'm just saying...when people reach out, they are looking to share something. | 19:16 |
| v1k0d3n | idea's etc. | 19:16 |
| kfox1111 | agreed. and we're trying to be as acomidating as possible. | 19:17 |
| kfox1111 | but, realize there's a flip side. | 19:17 |
| kfox1111 | sometimes the stuff atempted to be contributed in is tainted with opinions that wont fly well in an unopionated style project. | 19:17 |
| kfox1111 | its not good or bad. its a good fit for an opionated project. but not nessisarily a good for an unopionated project. | 19:18 |
| kfox1111 | we should listen to ALL ideas. | 19:18 |
| kfox1111 | in both directions. | 19:18 |
| kfox1111 | a new person coming in might not realize why the other poject did things in a very different looking way. | 19:18 |
| v1k0d3n | kfox1111: tell me the cloud native tooling you're using vs. standard tools for kolla-kubernetes? | 19:18 |
| kfox1111 | v1k0d3n: there are two options for fully cloud native orchestration for example. | 19:19 |
| kfox1111 | helm and operators. | 19:19 |
| kfox1111 | we're going to support both use cases. | 19:19 |
| v1k0d3n | you said chef. | 19:19 |
| kfox1111 | its unclear which cloud native tool is best. | 19:19 |
| kfox1111 | v1k0d3n: you aske for cloud native options. helm and operators are there. | 19:19 |
| kfox1111 | I'm sahying, if chef wanted to do things, they won't be blocked. | 19:19 |
| kfox1111 | they are a 3rd option. | 19:20 |
| kfox1111 | or puppet, or heat. | 19:20 |
| v1k0d3n | lost us | 19:20 |
| kfox1111 | kolla's not going to write that code. | 19:20 |
| v1k0d3n | right there... | 19:20 |
| v1k0d3n | i have to think about this for a while. | 19:20 |
| kfox1111 | but kolla-kubernetes's building blocks shouild be reusable by them. | 19:20 |
| portdirect_ | why write helm pagages if we dont use helm to deploy? | 19:21 |
| srwilkers | ^ | 19:21 |
| kfox1111 | there are several use cases in the layer cake. layed out by the spec. | 19:21 |
| v1k0d3n | i see where you're coming from, but there are some fundamental differences and misunderstandings i had about the direction for kolla-kubernetes. | 19:21 |
| kfox1111 | helm provides a way to package services, to do templating, and orchestration. | 19:21 |
| kfox1111 | 3 different use cases. | 19:21 |
| v1k0d3n | kfox1111: why 3 different USE CASE!? | 19:21 |
| kfox1111 | the microservices layer explicitly excludes orchestration. so it uses helm for the first 2. packaging and templating. | 19:22 |
| kfox1111 | the service layer can add the 3rd option and add the microservice packages as subpackages. | 19:22 |
| kfox1111 | therefore using helm for all 3. | 19:22 |
| v1k0d3n | if you are working on kolla-KUBERNETES...why not use all standard kubernetes tools only. | 19:22 |
| kfox1111 | v1k0d3n: becauwe that's very opionated. | 19:22 |
| kfox1111 | v1k0d3n: I get you believe helm orchestration is the only way forward. but other cloud native folkx don't all agree with you. | 19:22 |
| kfox1111 | some think operators are a better solution to orchestration. | 19:23 |
| kfox1111 | so, even if we just talk about purely cloud native, | 19:23 |
| kfox1111 | orcehstration needs to be seperatble to be unopionated. | 19:23 |
| kfox1111 | so thats what we did. | 19:23 |
| kfox1111 | microservices are unopionated with reguards to orchestration. | 19:23 |
| *** jgriffith is now known as jgriffith_AutoAw | 19:24 | |
| kfox1111 | they can be assembled into helm packages though that are helm orchestrated. | 19:24 |
| kfox1111 | meeting your desire. | 19:24 |
| kfox1111 | while not blocking others clout native desires. | 19:24 |
| v1k0d3n | kfox1111: i've made it so abundantly clear...we are abusing the operator in kolla-kubernetes. | 19:24 |
| v1k0d3n | kfox1111: cloud native folks...who? | 19:24 |
| kfox1111 | coreos beleives in operators. | 19:25 |
| v1k0d3n | HA!!! | 19:25 |
| v1k0d3n | wrong | 19:25 |
| v1k0d3n | operators replace operator functions. | 19:25 |
| v1k0d3n | they are not intended to be this magical clue that deploys things for you. | 19:25 |
| v1k0d3n | i have had meetings with the coreos team about this. | 19:26 |
| v1k0d3n | so we're getting conflicting information. | 19:26 |
| kfox1111 | v1k0d3n: I have significant doubts that helm can orchestrate an openstack live major rolling upgrade. | 19:26 |
| kfox1111 | that coudl be. | 19:26 |
| v1k0d3n | kfox1111: could it be that we're learning about helm's capabilities? | 19:26 |
| kfox1111 | but they aren't contributing. :/ how are we supposed to kn[Cow what they think? | 19:26 |
| v1k0d3n | this is really hard to keep up with afteral. | 19:26 |
| kfox1111 | agreed. | 19:27 |
| v1k0d3n | "they" who is they. | 19:27 |
| kfox1111 | we're all still learning. | 19:27 |
| rhallisey | don't need to use operators the way they intended. ThirdPartyResource is what is big there | 19:27 |
| kfox1111 | and this is still a very new project. we gota code something, and then decide how wrong it is. | 19:27 |
| kfox1111 | we don't have enough konwlege to produce it perfect out of the gate. | 19:27 |
| v1k0d3n | kfox1111: i know you weren't there...but this is why we were asked to do a poc from the PTL | 19:27 |
| kfox1111 | v1k0d3n: honestly, we're a ways away from having any operator code checked in. microservices are the goal right now. | 19:28 |
| kfox1111 | once they re in, I fully expect someone to do a helm service package next. | 19:28 |
| v1k0d3n | and that PoC could totally be in kolla...i'll have it merged today... | 19:28 |
| v1k0d3n | however... | 19:28 |
| kfox1111 | you could be proven right at that point, and operators never show up. | 19:28 |
| v1k0d3n | when we came back from barcelona....discussions happened for weeks about operators, entrypoint. | 19:28 |
| kfox1111 | a poc from whom? | 19:28 |
| v1k0d3n | all the things that were discussed came unraveled from barcelona | 19:29 |
| *** msimonin has left #openstack-kolla | 19:29 | |
| kfox1111 | still think we should have a call and talk... this is tricky. | 19:29 |
| v1k0d3n | we were asked to start a poc | 19:29 |
| kfox1111 | everyone's got a poc. :/ tons of them. :/ | 19:29 |
| kfox1111 | I wasn't at barcelona either. :/ | 19:29 |
| v1k0d3n | yeah, some of them looks oddly alike kfox1111 | 19:29 |
| v1k0d3n | i wished you were there. it is what it was | 19:29 |
| kfox1111 | I only kjnow what happend in agreements afterwards. | 19:29 |
| sbezverk_ | v1k0d3n: that was not a fair request, because that request cause this split we are seing now.. | 19:30 |
| v1k0d3n | sbezverk_: we started working as operators were being discussed repeatedly. | 19:30 |
| kfox1111 | v1k0d3n: there are a lot of poc's too that are totally unusable outside of a poc. :/ | 19:30 |
| v1k0d3n | i don't like the split either. | 19:30 |
| kfox1111 | aic-helm came a long way after the lessons f rom the other poc. | 19:30 |
| v1k0d3n | and hearing that there are 3 options for kolla-kube | 19:30 |
| v1k0d3n | i mean WHAT!? | 19:30 |
| kfox1111 | and aic-helm and kolla-kubernetes share a lot more then you think. | 19:30 |
| v1k0d3n | operators or helm, or this or that. | 19:30 |
| v1k0d3n | so incredibly messy. | 19:31 |
| v1k0d3n | those are all separate efforts | 19:31 |
| kfox1111 | v1k0d3n: why doesn't nova/neutron/glance/cinder/keystone all just merge into one binary? | 19:31 |
| kfox1111 | it would be easier for everyone. | 19:31 |
| kfox1111 | they are split now in a lot of ways, due to needing to be unopionated in deployment. | 19:31 |
| kfox1111 | at our basest level, we're following what openstack itself is doing. | 19:32 |
| v1k0d3n | look. this is my concern with kolla. i brought it up. | 19:32 |
| kfox1111 | one k8s object/container per openstack process. following openstacks lead. this leads to flexaaaability. | 19:32 |
| v1k0d3n | i don't want to keep going round and about. | 19:32 |
| kfox1111 | on top of that, can be built the layer that makes it easy to deploy. | 19:32 |
| kfox1111 | so you just have t odeal with "give me a compute kit" | 19:33 |
| kfox1111 | or, "give me a neutron" | 19:33 |
| kfox1111 | I think everyone has a place to contribute in this model. | 19:33 |
| v1k0d3n | your point is made. i'll stand down. | 19:33 |
| sbezverk_ | v1k0d3n: how many from kolla-kubernetes core team got involved in your PoC? | 19:33 |
| jmcevoy | portdirect_: After a hard reboot of the stuck VM docker would not start due to corruption in the /var/lib/docker directory. I ran the nuke script on the directory and recreated the registry and so far so good with threads=1 but the build is very slow... | 19:33 |
| v1k0d3n | sbezverk_: not one person has even asked about it (core) | 19:33 |
| jmcevoy | how many threads are used by default? | 19:33 |
| portdirect_ | jmcevoy: 8 | 19:34 |
| v1k0d3n | except for inc0 because he wants to see merged. sdake wanted us to merge in... | 19:34 |
| kfox1111 | v1k0d3n: opensource is not just a bout a license. its about governenece. | 19:34 |
| v1k0d3n | but according to what kfox1111 is saying...it doesn't fit in kolla-kube | 19:34 |
| v1k0d3n | and i agee | 19:34 |
| kfox1111 | v1k0d3n: if you have an opensource license, but no governence in place, its hard to get contributions. | 19:34 |
| v1k0d3n | we have zero place for chef, ansible etc. | 19:34 |
| kfox1111 | kolla-kubernetes is trying to do things under the openstack governence model , to allow contributers to more easily contribute. | 19:34 |
| v1k0d3n | we're only going with a cloud native approach | 19:34 |
| kfox1111 | v1k0d3n: I'm glad your going fully cloud native. thats a great place to be. :) | 19:35 |
| kfox1111 | I konw some folks at my org that won't be able to make the same commitment right away thoguh. :/ | 19:35 |
| *** jmccarthy has quit IRC | 19:35 | |
| portdirect_ | jmcevoy: sounds like you have a docker related issue there | 19:35 |
| v1k0d3n | kfox1111: that's where openstack should be in order to solve the interop issue. | 19:35 |
| v1k0d3n | there's extreme potential in kolla | 19:35 |
| v1k0d3n | but as soon as you slap that "kubernetes" name on it...many people will assume at first it is cloud native. | 19:36 |
| *** jmccarthy has joined #openstack-kolla | 19:36 | |
| kfox1111 | right. so lets get everyone that wants to contribute under the kolla-kubernetes umbrealla, | 19:36 |
| v1k0d3n | and by very nature of your description, it is not intended to be. | 19:36 |
| kfox1111 | all agree on what needs they have, and we'll try nad implement something that meets everyones needs. | 19:36 |
| sbezverk_ | v1k0d3n: the problem was that since you did not include any core who could express concenrs we are having this discussion now | 19:37 |
| jmcevoy | portdirect_: yes docker... https://github.com/docker/docker/issues/19758 I am running the dockerproject.org version of docker | 19:37 |
| kfox1111 | v1k0d3n: we're not going to be shipping non cloud native stuff in kolla-kubernetes. | 19:37 |
| *** haplo37 has quit IRC | 19:37 | |
| kfox1111 | you are confusing not blocking things from using kolla-kubernetes primiatives with non cloud native tooling, from kolla-kubernetes actively suporting the use case with chef recipies. | 19:38 |
| v1k0d3n | sbezverk_: what do you mean? | 19:38 |
| v1k0d3n | not sure i follow | 19:38 |
| v1k0d3n | kfox1111: WHY ARE WE TALKING ABBOUT CHEF?! | 19:39 |
| kfox1111 | v1k0d3n: I think he's saying, if aim-helm or the other poc's had kolla cores involved, it would be less divergent. | 19:39 |
| portdirect_ | jmcevoy: i've had ok results with the overlay back end - may be worth a shot | 19:39 |
| v1k0d3n | openstack kubernetes | 19:39 |
| sbezverk_ | v1k0d3n: you knew there was a team working on kolla-kube, right? so expect people sitting and waiting for results of your poc is kind of strange | 19:39 |
| kfox1111 | because some of the core ideas here would hve contibuted back the other way. | 19:39 |
| v1k0d3n | and ansible because unfortunately it's inherited from kolla | 19:39 |
| v1k0d3n | i'm not sure i follow | 19:39 |
| sbezverk_ | v1k0d3n: instead you should have actively involved people | 19:40 |
| *** jgriffith_AutoAw is now known as jgriffith | 19:40 | |
| kfox1111 | v1k0d3n: been actively trying to get rid of ansible. there's even a ps up here: | 19:40 |
| v1k0d3n | sbezverk_: i have actively involved. | 19:40 |
| kfox1111 | https://review.openstack.org/#/c/408263/ | 19:40 |
| v1k0d3n | thought that's what you meant. | 19:40 |
| v1k0d3n | i have reached out a lot. | 19:40 |
| v1k0d3n | just the other day i asked if y'all had seen it or what you thought... | 19:40 |
| v1k0d3n | can anyone please back me up here??? | 19:40 |
| sbezverk_ | v1k0d3n: for sure I remember, but where are review requests? | 19:41 |
| v1k0d3n | i'm kind of alone...working for all of you. | 19:41 |
| v1k0d3n | :) | 19:41 |
| kfox1111 | can we make this a call? I think there's a long of emotions being a ssumed that are really hard to convey over irc. | 19:41 |
| kfox1111 | s/long/lot/ | 19:41 |
| jmcevoy | portdirect_: What do you use for storage to back the VM? I am using a 1TB 7200 RPM /dev/sdb1 with a raw LVM volume... When I have seen Soft lockups in the past it has been I/O related... Maybe more spindles | 19:41 |
| kfox1111 | v1k0d3n: +1. thank you for your help trying to bridge communities. I think thats great. | 19:41 |
| *** haplo37 has joined #openstack-kolla | 19:41 | |
| *** schwicht has quit IRC | 19:41 | |
| v1k0d3n | kfox1111: it's unfortunately not working | 19:42 |
| kfox1111 | but, long term, to get the communities closer together, we need to start working on a common code base. | 19:42 |
| kfox1111 | we can't just have one party work on their tree, and another in another tree, | 19:42 |
| kfox1111 | and never talk. | 19:42 |
| v1k0d3n | kfox1111: you're confusing "continuing" with joining the communities. | 19:42 |
| v1k0d3n | that is not the case. | 19:42 |
| portdirect_ | jmcevoy: yeah man - i think your going to need a lot more than that to handle 8 build threads in a vm - I build on bare metal with a pci ssd... | 19:42 |
| v1k0d3n | listening to peoples concerns...that's what brings it closer. | 19:42 |
| kfox1111 | openstack can't work on aic-helm as its not under openstack governence. | 19:43 |
| v1k0d3n | i'm not asking you to work on it. | 19:43 |
| srwilkers | so kfox1111, if i put a patch set in to start bringing it inside kolla-k8s, can we work on reviewing it and seeing where that path takes us? | 19:43 |
| jmcevoy | portdirect_: There in lies the difference! | 19:43 |
| kfox1111 | and we're not seeing many people show up here to talk about their ideas. | 19:43 |
| v1k0d3n | i'm asking you to steal it....take it...low hanging fruit | 19:43 |
| v1k0d3n | try it | 19:43 |
| *** openstack has joined #openstack-kolla | 19:47 | |
| *** Marcellin__ has quit IRC | 19:49 | |
| kfox1111 | v1k0d3n: what we need to do is grow a kolla-kubernetes project, that can meet the needs of att&t, sap, coreos, pnnl, and everone else. | 19:49 |
| v1k0d3n | kfox1111: you have to understand that from what i see though...these are all outsiders currently. | 19:50 |
| kfox1111 | that I don't believe can be done by just taking one of the existing poc's, and just importing it making it kolla-kuberenes. that isn't how openstack projects have usually worked. | 19:50 |
| v1k0d3n | there's an opportunity to bring them in, but in bringing up concerns they get squashed... | 19:50 |
| v1k0d3n | and leave | 19:50 |
| kfox1111 | now that there is alot of knowlege, we need to import ideas from all of the poc's as well as the new use cases coming up and make something brand new. | 19:51 |
| v1k0d3n | and then end up doing their own thing | 19:51 |
| v1k0d3n | so as a core... | 19:51 |
| kfox1111 | and thats whats been happening in kolla-kubernetes trunk. | 19:51 |
| v1k0d3n | you have to find out why that is. | 19:51 |
| *** mattmceuen is now known as mattmceuen_away | 19:51 | |
| v1k0d3n | and that may mean that you need to open up some. | 19:51 |
| v1k0d3n | and do something that may be different | 19:51 |
| kfox1111 | v1k0d3n: I understand. its a verhy hard balance to strike in the opensource community. | 19:51 |
| v1k0d3n | and in the end.... | 19:51 |
| v1k0d3n | who knows | 19:51 |
| v1k0d3n | they may be right. | 19:51 |
| kfox1111 | how do you balance, a user coming in that has a great new idea, | 19:51 |
| v1k0d3n | because you're losing more as time goes one. | 19:52 |
| kfox1111 | from one that comes in and has a great idea for their restricted use case, that doesn't work in the more general project? | 19:52 |
| v1k0d3n | *on | 19:52 |
| v1k0d3n | personally? find out what their thoughts are on orchestration.... | 19:52 |
| *** schwicht has quit IRC | 19:52 | |
| v1k0d3n | all of them are going to say...no chef, no ansible, no puppet. | 19:52 |
| v1k0d3n | that may be you. | 19:52 |
| kfox1111 | v1k0d3n: I totally support their view of orchestration.l | 19:52 |
| v1k0d3n | but it's not them | 19:52 |
| v1k0d3n | they can add that on | 19:52 |
| kfox1111 | I will actively encurage that code to land in tree. | 19:52 |
| v1k0d3n | cloud native apps | 19:52 |
| *** fragatina has joined #openstack-kolla | 19:53 | |
| v1k0d3n | there are cloud native apps | 19:53 |
| v1k0d3n | helm. not perfect | 19:53 |
| v1k0d3n | CoreOS supports KPM btw....that;s the competition to helm | 19:53 |
| v1k0d3n | because they like jasonet | 19:53 |
| v1k0d3n | rather than go templates... | 19:53 |
| v1k0d3n | however...they realize that helm is the path forward and they dont want to support a one off | 19:53 |
| v1k0d3n | so they want to merge that in | 19:53 |
| kfox1111 | v1k0d3n: there is a place in tree for helm orchestrated packaging. kolla-kubernetes/helm/services. | 19:54 |
| *** mliima has quit IRC | 19:54 | |
| v1k0d3n | possibly a jsonet combo? who knows...but that;s for them. | 19:54 |
| kfox1111 | thats the perfect place where stuff like aic-helm to go. | 19:54 |
| v1k0d3n | kfox1111: greenfield | 19:54 |
| v1k0d3n | get rid of the need for any orchestration | 19:54 |
| v1k0d3n | make it easy | 19:54 |
| v1k0d3n | simple | 19:54 |
| v1k0d3n | i don;t even want to hear about ansible | 19:54 |
| kfox1111 | v1k0d3n: getting reid of the need for orchestration, impossible. | 19:54 |
| v1k0d3n | i love ansible...but still. | 19:54 |
| v1k0d3n | this is a openstack/kubernetes project | 19:55 |
| *** fragatina has quit IRC | 19:55 | |
| kfox1111 | v1k0d3n: your trying to define it a way as saying it doesn't happen becuase helm does it. | 19:55 |
| v1k0d3n | i brought up helm | 19:55 |
| v1k0d3n | i brought up operators (althoguh i'm sorry i did0 | 19:55 |
| v1k0d3n | kfox1111: i'm saying that we've figured out how to use helm for what you're looking for | 19:55 |
| v1k0d3n | there are differences | 19:55 |
| v1k0d3n | so has SAP | 19:55 |
| kfox1111 | v1k0d3n: your trying to convince me we should go full helm without doing microservices? | 19:55 |
| kfox1111 | I'm not sure what you are tryign to say. | 19:56 |
| v1k0d3n | kfox1111: what is a microservice? | 19:56 |
| kfox1111 | v1k0d3n: I disagree there. no poc I've seen actually does a nova live upgrade from mitaka to newton. so how do you know its complete? | 19:56 |
| kfox1111 | v1k0d3n: this is where we have the disconnect here. and an example of my point. | 19:56 |
| v1k0d3n | so that Glance chart you pulled up... | 19:56 |
| v1k0d3n | is this not microserviced? | 19:57 |
| v1k0d3n | https://github.com/att-comdev/aic-helm/blob/master/glance/requirements.yaml | 19:57 |
| kfox1111 | we have come up with an abstraction to allow all use cases. | 19:57 |
| kfox1111 | you came in, said, your way is the best wayk, | 19:57 |
| kfox1111 | without understanding our way. | 19:57 |
| *** sdake has joined #openstack-kolla | 19:57 | |
| kfox1111 | lets talk about it. | 19:57 |
| v1k0d3n | "we have come up with" | 19:57 |
| kfox1111 | thjen we can see what way is best going forward. | 19:57 |
| v1k0d3n | i want to figure out what kubernetes teams have come up wtih | 19:57 |
| kfox1111 | v1k0d3n: the kolla-kubernetes community, of which anyone can join | 19:57 |
| v1k0d3n | right... | 19:58 |
| v1k0d3n | i have a phone number...anyone can call | 19:58 |
| v1k0d3n | but do they know i'm going to give them $100 if they do? | 19:58 |
| v1k0d3n | how do they know what "i'm" working on | 19:58 |
| kfox1111 | v1k0d3n: https://github.com/openstack/kolla-kubernetes/tree/master/helm/microservice/neutron-l3-agent | 19:58 |
| v1k0d3n | goes both ways | 19:58 |
| kfox1111 | there's a microservice. | 19:58 |
| v1k0d3n | right. different way of deploying those | 19:59 |
| v1k0d3n | we're using the same images as you :) | 19:59 |
| kfox1111 | a small package that does one thing, and maps to one openstack thing. | 19:59 |
| v1k0d3n | so is SAP | 19:59 |
| v1k0d3n | so is stackanetes | 19:59 |
| v1k0d3n | so CCP | 19:59 |
| kfox1111 | those microservices can then be wrapped up into a greater service that does a whole openstack project like thing. | 19:59 |
| kfox1111 | say, "openstack-neutron" | 19:59 |
| v1k0d3n | right | 19:59 |
| kfox1111 | depends on neutron-l3-agent, neutron-server, neutron-openvswitch-agent, etc. | 20:00 |
| *** schwicht has joined #openstack-kolla | 20:00 | |
| kfox1111 | when done that way, tis exactly the same as aic-helm does, except you can launch the sub packages through other ways then just helm install the service package. | 20:00 |
| kfox1111 | supporting other folks use case. | 20:00 |
| kfox1111 | so I don't think they are conflicting ways of doing things, | 20:01 |
| v1k0d3n | look man, i'm only bringing up because it helps us all if kolla-kube is successful. | 20:01 |
| v1k0d3n | many people are looking at this project to be the bridge between opentsack and cloud native. | 20:01 |
| kfox1111 | yup. and I apreciate the discussion. | 20:01 |
| kfox1111 | and we totally need to provide a solution that is cloud native. | 20:01 |
| v1k0d3n | kfox1111: don't misunderstand me | 20:01 |
| v1k0d3n | this is a concern being brought up | 20:02 |
| kfox1111 | yeah. | 20:02 |
| v1k0d3n | if i wanted to just have a discussion, i would bring up something far less controversial than talk about peoples work | 20:02 |
| v1k0d3n | i respect the amount of time and energy goes into it. | 20:02 |
| kfox1111 | and I'm committed to that. so its a concern that's based gut feeling and not discussion I think? | 20:02 |
| v1k0d3n | and passions | 20:02 |
| openstackgerrit | Merged openstack/kolla: Add ability to output the things built/failed/not build https://review.openstack.org/395273 | 20:03 |
| v1k0d3n | gut feeling? no i was looking through the code and differences with other projects. | 20:03 |
| v1k0d3n | not sure how i feel about that statement. | 20:03 |
| v1k0d3n | i gotta run man. i seriously don't want to put my self in front of this anymore. | 20:03 |
| openstackgerrit | Merged openstack/kolla: Fix octavia images https://review.openstack.org/404645 | 20:03 |
| kfox1111 | v1k0d3n: yeah, but you didn't know what microservices are. | 20:03 |
| kfox1111 | just because something doesn't look like another doesn't mean that one thing is bad. | 20:04 |
| kfox1111 | your discounting our ideas I think because they don't look like other folks ideas. | 20:04 |
| kfox1111 | cloud native is all very very new. | 20:04 |
| kfox1111 | so room for a lot of new ideas to come in, and eventually it will all settile out and things will look very consistant, | 20:05 |
| kfox1111 | when all the new, good ideas propigate around everywhere. | 20:05 |
| v1k0d3n | kfox1111: dont assume | 20:05 |
| v1k0d3n | i know exactly what a microservice is.... | 20:05 |
| v1k0d3n | i was making sure we're on the same page | 20:05 |
| kfox1111 | ah. ok. fair enough. | 20:05 |
| v1k0d3n | :-/ | 20:06 |
| kfox1111 | sorry I assumed. | 20:06 |
| kfox1111 | like I said, this is getting probbaly emotiaonall ycharged, because of lack of unserstandig via irc, | 20:06 |
| kfox1111 | not relaying tone well. | 20:06 |
| kfox1111 | I mean no offence. | 20:06 |
| *** jtriley has quit IRC | 20:07 | |
| v1k0d3n | it's ok. | 20:08 |
| v1k0d3n | i just assume we're all really sharp people | 20:08 |
| kfox1111 | v1k0d3n: I think aso a lot of this stems from lack of formal requirements. | 20:08 |
| v1k0d3n | especially the ones obviously contributing to something related ;) | 20:08 |
| kfox1111 | we really should write taht down, as all coding decisiions stem from there. | 20:08 |
| v1k0d3n | kfox1111: OMG YES | 20:08 |
| kfox1111 | yeah. | 20:08 |
| v1k0d3n | i don't think formality is clearly stated | 20:08 |
| v1k0d3n | like the docs unfortunately. | 20:09 |
| kfox1111 | and some of the weirdness in kolla-kuberentes directlyl stems from my view of the requirements. | 20:09 |
| kfox1111 | like being able to easily support plugging in ci/cd's and other such things. | 20:09 |
| v1k0d3n | lot of work. everyone excited. but missing pieces that are critical. | 20:09 |
| kfox1111 | yeah. | 20:09 |
| v1k0d3n | kfox1111: i have to be honestly with you... | 20:09 |
| v1k0d3n | you're view brings orchestration back into the fold, and you're learning helm. | 20:10 |
| v1k0d3n | i don't mean to be offensive. | 20:10 |
| kfox1111 | I don't quite follow, and I take no offence. | 20:10 |
| kfox1111 | you saying, my view on orchestration is a problem? | 20:10 |
| v1k0d3n | if we're talking in context of tools outside of kubernetes, yes. | 20:11 |
| v1k0d3n | what kubernetes doesn't natively bring as far as organization, variables into config data, etc...helm does. | 20:11 |
| kfox1111 | my view on orchestration is orchestration is deeply religious, just like config management system is deeply religious. | 20:11 |
| v1k0d3n | it's not perfect...but not for the things that i think you've expressed. | 20:11 |
| kfox1111 | I believe kolla-kubernetes sould take a batteries included aproach, | 20:11 |
| kfox1111 | and provide a solution. cloud native. | 20:11 |
| v1k0d3n | right. kfox1111... greenfield | 20:12 |
| kfox1111 | but shouldn't blokc others. or else they will form competetors to kolla-kubernetes. | 20:12 |
| v1k0d3n | drop what you think is important regarding orchestration today. | 20:12 |
| kfox1111 | that are mostly the same except for orchestration. | 20:12 |
| v1k0d3n | because we're not using any of those | 20:12 |
| v1k0d3n | and we get everything we need. | 20:12 |
| v1k0d3n | and yes...microservices. | 20:12 |
| kfox1111 | microservices and helm orchestrated services can be done today with our current path. whats the problem? | 20:13 |
| v1k0d3n | why add what's not needed for a greenfield project? | 20:13 |
| v1k0d3n | we missed each other i think.... | 20:13 |
| kfox1111 | you just not liking me say 'chef' or 'ansible' out loud? cuase I have no intention of writing that code. | 20:13 |
| v1k0d3n | let's just let this rest for today | 20:13 |
| portdirect_ | kfox1111: please consider that v1k0d3n (and to a much smaller extend myself) both run projects parrell to kolla-k8s that we are trying to wind up | 20:13 |
| portdirect_ | kfox1111: and fold into here - we need to work in a way that enables that | 20:14 |
| kfox1111 | portdirect_: +1. I do understand that. I'm very glad your contrubting. and support you. | 20:14 |
| kfox1111 | I'm trying to figure out how we can get v1k0d3n to be able to do the same. | 20:14 |
| *** jrich523 has quit IRC | 20:15 | |
| kfox1111 | I think the main difference is this. you came in and we've talked a lot, you've looked at some of the way we've done things, and contributed some of the way you've done things. | 20:15 |
| kfox1111 | v1k0d3n's been in a harder place. he has team members that are contributing to a poc of their own, | 20:16 |
| kfox1111 | and they aren't coming here to learn our way, and bring in their ideas here, so everyhing ends up under one roof. | 20:16 |
| openstackgerrit | Merged openstack/kolla: Consistently use a exception vs sys.exit https://review.openstack.org/334690 | 20:16 |
| v1k0d3n | kfox1111: no...i pulled srwilkers off the PoC to support you and your effort | 20:16 |
| openstackgerrit | Qin Wang (qwang) proposed openstack/kolla-kubernetes: WIP - Hel-ming horizon components https://review.openstack.org/408975 | 20:16 |
| *** mattmceuen_away is now known as mattmceuen | 20:16 | |
| * srwilkers waves | 20:16 | |
| v1k0d3n | and btw...he's also leading in reviews | 20:16 |
| v1k0d3n | so i'd say we're trying to pull weight | 20:16 |
| kfox1111 | oh. sorry. I forgot about where you came from srwilkers. my appologies. | 20:16 |
| v1k0d3n | and be supportive of this effort | 20:16 |
| v1k0d3n | btw...also lrensing is on our team | 20:17 |
| kfox1111 | srwilkers has been doing good work too. | 20:17 |
| v1k0d3n | another triage assistant | 20:17 |
| kfox1111 | ok. cool. then I really do appologise for misunderstanding. | 20:17 |
| kfox1111 | I have been trying to listen to them. (and everyone else). | 20:17 |
| v1k0d3n | if you want to see what we do overall...do a search for our folks. we try to practice what we preach. | 20:17 |
| kfox1111 | srwilkers: lrensing: do you feel like I've squelched your ideas? | 20:17 |
| v1k0d3n | we're on the openstack board | 20:18 |
| kfox1111 | I hope not. | 20:18 |
| v1k0d3n | we're on CNCF | 20:18 |
| v1k0d3n | we're on OCI | 20:18 |
| v1k0d3n | kfox1111: | 20:19 |
| v1k0d3n | https://www.openstack.org/foundation/board-of-directors/ | 20:19 |
| v1k0d3n | Toby Ford | 20:19 |
| v1k0d3n | yes...we are actually | 20:19 |
| kfox1111 | v1k0d3n: cool. :) | 20:19 |
| v1k0d3n | oh ...your comment was for someone else | 20:19 |
| v1k0d3n | sorry! | 20:19 |
| v1k0d3n | lol | 20:19 |
| v1k0d3n | :) actually chuckled a bit | 20:20 |
| kfox1111 | v1k0d3n: oh. yeah. no, I think what you folks are doing is great. :) | 20:20 |
| v1k0d3n | so just to be clear... | 20:20 |
| v1k0d3n | kolla-kubernetes isn't just some kolla 2.0 where you can bring your own orchestrator and deploy kubernetes manifests...am i correct? | 20:21 |
| v1k0d3n | :) | 20:21 |
| kfox1111 | let me attempt to be clear. | 20:21 |
| sbezverk_ | v1k0d3n: could you explain why that would be bad? | 20:22 |
| kfox1111 | in kolla, there are two peices. orchestration: kolla-ansible and software: kolla-containers. | 20:22 |
| kfox1111 | in kolla-kubernetes, I'm attempating to do the same thing. orchestration: cloud native tooling. and software: helm based microservice packages. | 20:22 |
| v1k0d3n | sbezverk_: fair enough...is that clearly stated in the spec? | 20:23 |
| kfox1111 | in kolla, you can use the kolla containers without using the orchestrator. | 20:23 |
| v1k0d3n | (i probably missed it). | 20:23 |
| kfox1111 | the same should be true I believe in kolla-kubernetes. | 20:23 |
| v1k0d3n | kfox1111: isnt that what kolla-ansible is??? | 20:23 |
| portdirect_ | kfox1111: that point i dont follow?^^ | 20:23 |
| kfox1111 | portdirect_: ok, let me explain further. | 20:24 |
| v1k0d3n | this is why this should be an entirely different project. | 20:24 |
| portdirect_ | is kolla-kubernetes not the orchestration of kolla kontainers with kubernetes tooling? | 20:24 |
| v1k0d3n | i'm with adrian_otto on this one. | 20:24 |
| kfox1111 | if I want to use an orchestration system other then kolla-ansible, I can. by just reusing the containers. | 20:24 |
| v1k0d3n | kfox1111: the point is that the container is the artifact. | 20:24 |
| kfox1111 | in the kubernetes stack, there is a seperation between orchestration and deployemnt. | 20:24 |
| v1k0d3n | kolla ansible orchestrates that artifact logically | 20:25 |
| kfox1111 | orchestration being "I need to run these pods together, whith these services to connect them, etc" | 20:25 |
| v1k0d3n | in barcelona it was discussed in depth that helm/kubernetes would be used. | 20:25 |
| kfox1111 | and deployment "k8s, here's a description of a neutron-l3-agent. spawn pods as described" | 20:25 |
| v1k0d3n | not some open-ended, bring your own to the show. | 20:25 |
| kfox1111 | v1k0d3n: excactly the container is an artefect. a kolla-kubernetes microservice is a nother type of artefact imo. | 20:25 |
| v1k0d3n | this is why the projects should be seprate, with separate goals, and i feel separate missions | 20:26 |
| v1k0d3n | no. | 20:26 |
| v1k0d3n | kubernetes is the orchestration scheduler/api | 20:26 |
| kfox1111 | and then a second type of artefact is an orchestration artefact. | 20:26 |
| kfox1111 | helm/neutron being such a thing. | 20:26 |
| v1k0d3n | helm is the part that takes user-defined data and places that correctly into configmaps. | 20:27 |
| v1k0d3n | see this is where the disconnect is | 20:27 |
| kfox1111 | v1k0d3n: no, its only partially so. if you read their docs they say they don't want to be in the orchestration buisness. | 20:27 |
| v1k0d3n | when you go all in on helm...what else do you need | 20:27 |
| v1k0d3n | this mission...this path...is everything | 20:27 |
| kfox1111 | in fact, helm does a lot more orchestration then k8s does. | 20:27 |
| kfox1111 | v1k0d3n: thats the issue though. you assume everyone's all in on helm. | 20:27 |
| kfox1111 | v1k0d3n: I'm saying, for microservices, we all are agreed to be all in. | 20:27 |
| kfox1111 | v1k0d3n: for orchestration though, not everyone's all in. | 20:27 |
| v1k0d3n | kfox1111: sorry...more clearly (i stated it correctly) kubernetes is the scheduler/API | 20:28 |
| kfox1111 | v1k0d3n: right. k8s is scheduling and watchdogging. | 20:28 |
| v1k0d3n | we need to let each other type and finish | 20:28 |
| v1k0d3n | :) | 20:28 |
| v1k0d3n | this would help | 20:28 |
| kfox1111 | k. | 20:28 |
| *** schwicht has quit IRC | 20:28 | |
| kfox1111 | I'll stop for a bit. go ahead. | 20:28 |
| v1k0d3n | the thoughts are passing, and i'm not sure we're taking time to process. there are misses. and to be in the same place is important to everyone who is on this channel. | 20:29 |
| *** Pavo has quit IRC | 20:29 | |
| kfox1111 | +1 | 20:29 |
| v1k0d3n | i was at all the meetings in barcelona. | 20:30 |
| kfox1111 | sorry. I wasn't there. I can't be held to know what happened. :/ | 20:30 |
| v1k0d3n | we never discussed this. i'm not saying that direction doesn't change. but where is this clearly stated? | 20:30 |
| v1k0d3n | it's ok | 20:30 |
| kfox1111 | we tried to in the spec. we then voted on it as agreeing. | 20:31 |
| v1k0d3n | but this may be part of a bigger misunderstanding too just tbh. | 20:31 |
| kfox1111 | yeah. could be. | 20:31 |
| v1k0d3n | so the spec wasn't completed, was there a disagreement, what was the issue with clearly identifying this path? | 20:31 |
| v1k0d3n | sorry. | 20:31 |
| v1k0d3n | what if i have members on vacation. | 20:32 |
| v1k0d3n | ? | 20:32 |
| v1k0d3n | etc. this is where my heads at... | 20:32 |
| *** prameswar has quit IRC | 20:32 | |
| kfox1111 | let me see if I can find it. it was voted on and agreed to. | 20:32 |
| openstackgerrit | sayantani proposed openstack/kolla-ansible: Remove genconfig command from kolla-ansible script https://review.openstack.org/409325 | 20:32 |
| *** TxGirlGeek has quit IRC | 20:33 | |
| v1k0d3n | as you're trying to find it, i just want to express my concern with kolla and make it vividly clear to you as a core member... | 20:33 |
| v1k0d3n | kolla is broadly focused. | 20:33 |
| kfox1111 | https://review.openstack.org/#/c/392257/ | 20:33 |
| v1k0d3n | my hope was that it would narrow with this new "kubernetes" approach. | 20:33 |
| *** TxGirlGeek has joined #openstack-kolla | 20:33 | |
| v1k0d3n | this is also turning into an extremely broad project. unlike any project in openstack. | 20:34 |
| *** Pavo has joined #openstack-kolla | 20:34 | |
| v1k0d3n | i would much prefer it to have laser focus. | 20:34 |
| v1k0d3n | not open focus, or made for all | 20:34 |
| v1k0d3n | and what i mean (for clarity) | 20:34 |
| v1k0d3n | taking another project.... | 20:34 |
| v1k0d3n | openstack ansible. | 20:34 |
| kfox1111 | I don't see kolla-kubernetes microservices being more or less broad then kolla containers. | 20:35 |
| v1k0d3n | they made a laser focused choice (good, bad, different), for going with ansible. | 20:35 |
| v1k0d3n | using galaxy roles for additions. | 20:35 |
| kfox1111 | I hear what your saying. | 20:36 |
| v1k0d3n | and in barcelona | 20:36 |
| v1k0d3n | during design summit | 20:36 |
| portdirect_ | and i thought we were going to do the same here with k8s - the spec does not discuss any other platforms | 20:36 |
| v1k0d3n | (i know you weren't there...i know, and i'm sorry) | 20:36 |
| *** sayantan_ has quit IRC | 20:36 | |
| kfox1111 | portdirect_: it says there is a layerd apoach. which is what microservices are about. | 20:36 |
| v1k0d3n | but cores, and community developers, some who are reviewing a lot of your code right now... | 20:37 |
| sbezverk_ | v1k0d3n: I think you are trying to call incorrectly the flexibility we aim to achieve with microservices as broad scope.. | 20:37 |
| v1k0d3n | all talked about helm and kubernetes. | 20:37 |
| v1k0d3n | sbezverk_: the "flexibility" is changing your helm deployments | 20:37 |
| v1k0d3n | for considerations of things that were discussed later. | 20:37 |
| kfox1111 | I feel you seem to be hung up assuming just because I'm not shutting the door to someone doing some traditional config management system that I'm actively encuraging it and am wasting time on it? | 20:38 |
| v1k0d3n | namely how to add orchestration opinions into the mix. | 20:38 |
| sbezverk_ | v1k0d3n: not sure it is true, we use helm to deploy microservices | 20:38 |
| kfox1111 | I don't believe we're spending much time actively working on use cases for not cloud native orchestration. | 20:38 |
| v1k0d3n | kfox1111: said that there are limitations to helm | 20:38 |
| v1k0d3n | are these limitations documented in these community? | 20:39 |
| v1k0d3n | are we being good community citizens outside of openstack? | 20:39 |
| * adrian_otto looks at scrollback | 20:39 | |
| kfox1111 | v1k0d3n: I do plan on writing those up and submitting them as issues to helm, and amybe even proposing fixes. | 20:39 |
| kfox1111 | v1k0d3n: just havent' had time yet. I'm under a deadline to get something kolla-kubernetes usable by the end of the month. | 20:40 |
| v1k0d3n | kfox1111: why are we waiting? | 20:40 |
| v1k0d3n | can i help submit these for you? | 20:40 |
| kfox1111 | so focus is on getting microservices to work. then can work on getting the build system in shape. | 20:40 |
| kfox1111 | v1k0d3n: would love that. but the issue is finding time enough to tell you, or tell upstream. it will take me about as much time to figure out the right wording either way I'm thinkin? | 20:41 |
| *** TxGirlGeek has quit IRC | 20:41 | |
| v1k0d3n | but to be fair...you're having issues with helm, and continuing on because you have an orchestration tool that you use, correct? | 20:41 |
| kfox1111 | and if I don't word it right, then it becuomes pushback, "don't do it that way" kinds of rhitns. | 20:41 |
| kfox1111 | things | 20:41 |
| kfox1111 | no. the issues are not orchestration related at all. | 20:42 |
| kfox1111 | but building related. | 20:42 |
| *** TxGirlGeek has joined #openstack-kolla | 20:42 | |
| kfox1111 | ultimately, I want to be able to load kolla-kubernetes into a ci/cd system, | 20:42 |
| kfox1111 | and have it build new kolla containers whenever a security update comes out, | 20:42 |
| kfox1111 | and rebuild all helm packages that are touched by that container. | 20:42 |
| kfox1111 | so for example, there's no way of bumping the default image tag or helm package version from the cli. | 20:43 |
| *** dupengfei has quit IRC | 20:43 | |
| kfox1111 | you have to edit Chart.yaml or values.yaml by hand. | 20:43 |
| kfox1111 | or automate their generation, | 20:43 |
| kfox1111 | buth then our not 100% pure checkout and run helm and it works. | 20:43 |
| kfox1111 | does that make sense? | 20:44 |
| v1k0d3n | i can understand and relate to all of that. we have similar goals. of course. | 20:44 |
| *** TxGirlGeek has quit IRC | 20:45 | |
| kfox1111 | once helm grows that functionality, I'm be extatic from removing the hack from kolla-kubernete's build. | 20:45 |
| v1k0d3n | wait, sorry... | 20:45 |
| v1k0d3n | you're missing me again. | 20:45 |
| *** TxGirlGeek has joined #openstack-kolla | 20:45 | |
| v1k0d3n | is this a helm issue to solve? | 20:45 |
| kfox1111 | I'm also happy to help push that functionality upstream. but for now, am time constrained. | 20:45 |
| kfox1111 | I believe so. | 20:45 |
| kfox1111 | so, for example, | 20:45 |
| v1k0d3n | did they state that this is their issue to solve? | 20:45 |
| kfox1111 | I want to be able to "helm package foo --version 1.3.2-1" | 20:45 |
| kfox1111 | rather then edit the chart and put 1.3.2-1 in the yaml. | 20:46 |
| kfox1111 | no. and that has to be discussed too. | 20:46 |
| v1k0d3n | kfox1111: have you seen lachies gator app? | 20:46 |
| kfox1111 | but without either a helm_prebuild like thing we're doing, or the feature in helm, | 20:46 |
| v1k0d3n | this covers helm chart lifecycle. | 20:46 |
| v1k0d3n | and using jenkins | 20:46 |
| v1k0d3n | pipelines to be exact | 20:46 |
| kfox1111 | then the stuff can't go into kubernetes/charts. | 20:46 |
| kfox1111 | v1k0d3n: I have not. got a link? | 20:47 |
| v1k0d3n | kfox1111: https://github.com/lachie83/croc-hunter | 20:47 |
| v1k0d3n | this repo has everything | 20:47 |
| v1k0d3n | however... | 20:47 |
| v1k0d3n | this is outside of helm | 20:47 |
| v1k0d3n | and i feel that this is outside of kubernetes | 20:48 |
| * adrian_otto read the scrollback. Wonders if you need my input here. | 20:48 | |
| v1k0d3n | and outside of kolla | 20:48 |
| kfox1111 | adrian_otto: always apreciate input. :) | 20:49 |
| kfox1111 | v1k0d3n: yeah. the tricky balancing act is how much to do fully cloud native, and when to step outside. like jenkins. | 20:50 |
| kfox1111 | jenins could just rewrite the Chart.aml for exmple. | 20:50 |
| kfox1111 | but I also see a caese a human might want to bump versions across all the packages too. | 20:51 |
| adrian_otto | kfox1111: I do share a general concern that Kolla is at risk for having a focus that's too wide | 20:51 |
| kfox1111 | like when we're a bout to cut a release. | 20:51 |
| *** TxGirlGeek has quit IRC | 20:51 | |
| kfox1111 | adrian_otto: I agree there. | 20:51 |
| *** TxGirlGeek has joined #openstack-kolla | 20:51 | |
| adrian_otto | we had that problem in Magnum, and addressed it by starting new projects to focus on the new areas of ambition | 20:51 |
| adrian_otto | or at least one new project | 20:51 |
| kfox1111 | I think there are two aproaches kolla can take. | 20:51 |
| adrian_otto | and I remember your participation in that decision making process, kfox1111 | 20:52 |
| kfox1111 | 1. be a complete openstack distro in its own right. | 20:52 |
| kfox1111 | 2. be a provider of components that openstnack distro's pull from in their distro. | 20:52 |
| kfox1111 | adrian_otto: talking about zun? | 20:52 |
| adrian_otto | yep | 20:52 |
| kfox1111 | I think 2 is the thing that will make kolla-kubernetes stick around. and #1 is a thing that is a batteries included thing. | 20:53 |
| kfox1111 | adrian_otto: yah. I think that turned out pretty well in the end. | 20:53 |
| adrian_otto | se see a lot of potential value in 2. | 20:53 |
| adrian_otto | s/se/I/ | 20:53 |
| v1k0d3n | kfox1111 and adrian_otto i would really like to see a cloud-native approach towards providing openstack components. | 20:53 |
| kfox1111 | adrian_otto: oter then the, yet another service to deply. :/ But I think kolla-kubernetes can be a solution to that problem. | 20:53 |
| *** mattmceuen is now known as mattmceuen_away | 20:53 | |
| adrian_otto | I'm concerned about the #1 approach, as many community stakeholders have very different ideas about how to do that. | 20:53 |
| v1k0d3n | i think this is key for interop with large contributing projects like kubernetes as well. | 20:54 |
| kfox1111 | adrian_otto: yeah. I think as #2, kolla-kubernetes microservices should be consumable from tripleo for exapmle, so they get the power of k8s, and they have room to do what they do best. | 20:54 |
| kfox1111 | I don't think kolla-kubernetes will have the resources to ever do a gui as well as fuel for example. | 20:54 |
| kfox1111 | but as #2, fuel gui could be wrapped around kolla-kubernetes deployments. and everyone wins. | 20:55 |
| adrian_otto | does kolla-kubernets share a mission statement with Kolla? | 20:55 |
| v1k0d3n | kfox1111: fuel is no more. | 20:55 |
| kfox1111 | adrian_otto: think so, but the mission statement could be also taken a few different ways I think. | 20:55 |
| kfox1111 | my focus though is on #2 like things. | 20:56 |
| adrian_otto | that's a red flag, to think about. | 20:56 |
| *** TxGirlGeek has quit IRC | 20:56 | |
| kfox1111 | v1k0d3n: yeah. but those devs wroking on the ui might want to conitue that part. | 20:56 |
| kfox1111 | adrian_otto: which part do you object to? | 20:56 |
| kfox1111 | (just trying to understand) | 20:56 |
| adrian_otto | in the absense of a clear mission, we have conversations like this one. | 20:56 |
| kfox1111 | adrian_otto: ah. | 20:57 |
| kfox1111 | agreed. | 20:57 |
| adrian_otto | so we have an opportunity to get better alignment | 20:57 |
| v1k0d3n | adrian_otto: kfox1111 yes...mission | 20:57 |
| kfox1111 | so, maybe we need a mission statement specificaly for kolla-kubernetes? | 20:57 |
| SamYaple | kfox1111: i would recommend it | 20:57 |
| v1k0d3n | that part is huge | 20:57 |
| adrian_otto | my suggestion is to choose an approach that will leverage the talents and assets that already exist within our community and beyond | 20:57 |
| kfox1111 | +1. I like that idea. | 20:58 |
| adrian_otto | try our best not to duplicate efforts, but to combine an multiple | 20:58 |
| adrian_otto | multiply | 20:58 |
| * kfox1111 nods | 20:58 | |
| *** TxGirlGeek has joined #openstack-kolla | 20:59 | |
| adrian_otto | one key concern I have with kolla is that the same name gets attached to the "openstack distribution" and the "openstack service application containers" | 20:59 |
| adrian_otto | which means they share a common vision | 20:59 |
| v1k0d3n | kfox1111: i know we're going to disagree on this, but due to the nature of this being openstack, in containers, and scheduled by kubernetes.... | 20:59 |
| kfox1111 | adrian_otto: right. yeah. | 20:59 |
| v1k0d3n | i'd really like to see a cloud-native mission statement | 20:59 |
| adrian_otto | which makes the intent of the project unclear | 20:59 |
| v1k0d3n | that would be a big win for the openstack community | 20:59 |
| v1k0d3n | and could spark a lot of attention (positive). | 21:00 |
| kfox1111 | v1k0d3n: I agree, cloud native is a must support feature. | 21:00 |
| adrian_otto | I won't tell anyone *how* to solve this in implementation. But I will help you as much as I can with *what* to solve and the *why* we should. | 21:00 |
| v1k0d3n | and try to work towards that. | 21:00 |
| kfox1111 | adrian_otto: thanks. its very much appreciated. | 21:00 |
| *** schwicht has joined #openstack-kolla | 21:00 | |
| v1k0d3n | adrian_otto: this is really why i pinged you. :) | 21:01 |
| kfox1111 | v1k0d3n: I think your more concerned with use case #1 goals, and I'm with #2? | 21:01 |
| v1k0d3n | thanks | 21:01 |
| kfox1111 | just trying to understand. | 21:01 |
| v1k0d3n | actually a bit different kfox1111 | 21:01 |
| adrian_otto | so my first suggestion is to think about ways to laser focus kolla on being foundational components that are very useful to other projects in the ecosystem. | 21:01 |
| adrian_otto | and have another mission/name that refers to any vertical integration of those components. | 21:02 |
| kfox1111 | adrian_otto: +1 | 21:02 |
| adrian_otto | and have them tightly compliment each other. | 21:02 |
| *** msimonin1 has joined #openstack-kolla | 21:02 | |
| v1k0d3n | this is being advertised as openstack schedule via kubernetes, so kubernetes is first class in this case. | 21:02 |
| kfox1111 | adrian_otto: thats where I've been drawing the line. orcestration being the vertical integration, and microservices the components. | 21:02 |
| adrian_otto | but please resist the temptation to characterize Kolla as an openstack distro or an orchestration solution. | 21:03 |
| kfox1111 | adrian_otto: do you see that aproach being viable, or is something else missing? | 21:03 |
| *** TxGirlGeek has quit IRC | 21:03 | |
| adrian_otto | that might work. It sounds like you've thought about a way to carve it out | 21:03 |
| kfox1111 | adrian_otto: yeah. the spec covered it in the layer aporach. | 21:04 |
| v1k0d3n | so kfox1111 i would say 3. openstack via cloud native integration components and tooling. | 21:04 |
| kfox1111 | https://review.openstack.org/#/c/392257/ | 21:04 |
| kfox1111 | layer 2 being microservices. | 21:05 |
| adrian_otto | so I'd love to keep diving in more, but I have to step away for a lunch break | 21:07 |
| kfox1111 | so, ideally we would have a second repo for orcestration level things. which I've argued for, | 21:07 |
| kfox1111 | but others want just one repo to deal with. | 21:07 |
| adrian_otto | and my afternoon is pretty full. Happy to regroup with you again after the weekend if I can add value. | 21:07 |
| kfox1111 | adrian_otto: thanks for the help. :) | 21:07 |
| adrian_otto | :-) | 21:08 |
| v1k0d3n | not sure if that answered your question kfox1111 ? | 21:08 |
| *** adrian_otto has quit IRC | 21:08 | |
| kfox1111 | v1k0d3n: sorrym issed it for a sec. | 21:08 |
| kfox1111 | v1k0d3n: I think that is wrapped up in both use cases. | 21:08 |
| kfox1111 | I agree cloud native should touch as much of the components as possible in both use cases. | 21:09 |
| kfox1111 | the argectecture should be split though such that orchestration is seperable. | 21:09 |
| kfox1111 | I personally want to manually orchestrate everything on my own deployment. | 21:10 |
| kfox1111 | I need that flexability in my crazy environment. :( | 21:10 |
| kfox1111 | so I'll be useing the microservices but no service level packages, orchestrators, traditional config management engines, etc. | 21:11 |
| v1k0d3n | just to be clear, i don't totally agree with number 1, and number 1 would really cause friction with vendors who roll their own. | 21:11 |
| v1k0d3n | i'm still kind of mulling over | 21:11 |
| kfox1111 | v1k0d3n: I agree. but #1 has been in kolla's charter for a while, in a way. | 21:11 |
| v1k0d3n | but we're laser focused... | 21:11 |
| kfox1111 | at minimal, as a development platform. | 21:11 |
| v1k0d3n | yeah, but i just want to also be clear that i really wish this was a separate project entirely. | 21:11 |
| kfox1111 | but its something that kolla will have a hard time competing on. | 21:12 |
| v1k0d3n | because it's not bound entirely by kolla | 21:12 |
| kfox1111 | so I don't think it should get much focus. | 21:12 |
| v1k0d3n | *now | 21:12 |
| v1k0d3n | and this was one of my votes for separating from kolla | 21:12 |
| kfox1111 | yeah. get where your coming from. | 21:12 |
| kfox1111 | kolla's general mission statement thoguh covers the use case. | 21:12 |
| kfox1111 | so I can see it being under the umbrella project. | 21:13 |
| v1k0d3n | if i was perfectly honest...this is what i would want out of "the perfect" openstack project.... | 21:13 |
| kfox1111 | kolla-kubernetes does need its own mission statement though, to be clear what it is all about. | 21:13 |
| v1k0d3n | that just happens to look a little like kolla-kubernetes | 21:13 |
| v1k0d3n | (it does) | 21:13 |
| v1k0d3n | cloud native is first class in a kubernetes-kolla project | 21:14 |
| v1k0d3n | so kubernetes tools 1. | 21:14 |
| v1k0d3n | that includes helm | 21:14 |
| kfox1111 | +1 | 21:14 |
| v1k0d3n | promethus first class | 21:14 |
| *** TxGirlGeek has joined #openstack-kolla | 21:14 | |
| v1k0d3n | open tracing first class | 21:14 |
| kfox1111 | we're going through a huge amount of effort to retool somethign that was working before to helm. so I think the whole team's on board there. | 21:14 |
| *** mattmceuen_away is now known as mattmceuen | 21:15 | |
| v1k0d3n | and as the project is being developed...always try to accomplish the goals with those projects first | 21:15 |
| kfox1111 | yeah. those two techs can come in too. just need someone to work on them. I'm totally cool with them. | 21:15 |
| v1k0d3n | if there is an issue with that project, then ticket in our dependency (helm/kubernetes/etc). i.e. interoperability for our members. | 21:15 |
| kfox1111 | fist goal though is getting compute kit deploying cloud native. | 21:15 |
| kfox1111 | that doesn't have to include monitoring (promethius) | 21:16 |
| kfox1111 | once monitoring really comes in though, it should come in asap. | 21:16 |
| kfox1111 | part of this project though, | 21:16 |
| *** sayantan_ has joined #openstack-kolla | 21:17 | |
| v1k0d3n | oh i don't know...it could come whenever someone wants to add it. | 21:17 |
| v1k0d3n | but cores focus should yeah...be control plane. | 21:17 |
| kfox1111 | is those developers that contribue, get to decide what they want to work on. | 21:17 |
| v1k0d3n | but that microservice issue you're referring to.... | 21:17 |
| v1k0d3n | i'm not exactly what you're referring to...what doesn't work? | 21:17 |
| v1k0d3n | i don't know/care... | 21:17 |
| kfox1111 | if my boss is interested in me contributing collectd/influxdb/grafana before promethius, thats what I have to do. :/ | 21:18 |
| v1k0d3n | i'm just saying that needs to be an issue on github. | 21:18 |
| kfox1111 | ah. | 21:18 |
| v1k0d3n | pronto, because then someone could say... "it is intended to work this way...let me show you", or "that's an issue we're looking into" | 21:18 |
| kfox1111 | yeah, we should have blueprints for "all the things" | 21:18 |
| kfox1111 | ah. so, your concerned more with scaling out the devel team? | 21:19 |
| kfox1111 | new person wants to come in, not sure what to work on, | 21:19 |
| kfox1111 | we should be able to say "promethius" | 21:19 |
| v1k0d3n | well, in this case, we need really good citizens... | 21:19 |
| v1k0d3n | well, there's actually a prometheus chart out there :0 | 21:19 |
| v1k0d3n | what i'm saying is, the project should just be able to absorb that | 21:20 |
| v1k0d3n | and be flexible enough to just absorb things | 21:20 |
| kfox1111 | cool. so we just need a doc that says how to use it? | 21:20 |
| v1k0d3n | so the formatting is everything | 21:20 |
| srwilkers | agreed v1k0d3n. this goes well beyond just openstack. need to promote the idea that reaching out across the divide to other open source projects should be a matter of habit, not last resort | 21:20 |
| portdirect_ | from my perspective the issue thats worring me most is not the detail at this stage, its impossible to have lazer focus, without a clearly defined point that we are aiming for - and I'm not meaning the spec - but the vision of what we deliver. | 21:20 |
| kfox1111 | srwilkers: agreed | 21:20 |
| v1k0d3n | +1 | 21:20 |
| srwilkers | agreed portdirect_ | 21:21 |
| kfox1111 | srwilkers: portdirect_ v1k0d3n "the code is done not when there is nothing more to add, but nothing more to remove" | 21:21 |
| kfox1111 | that includes not ding things others can/are doing better. | 21:21 |
| kfox1111 | sometimes bad tradeoffs (temporarily) have to be made to bootstrap though. :/ | 21:21 |
| v1k0d3n | and kfox1111 what would be helpful for the group in general...is if we're using a cloud native approach, there are certain characteristics.... | 21:21 |
| kfox1111 | and can then be removed as the product matures. | 21:21 |
| v1k0d3n | for instance, the chart for prometheus should be drop in. | 21:21 |
| v1k0d3n | the chart for mariadb...drop in | 21:22 |
| v1k0d3n | if it's not...make that chart better | 21:22 |
| v1k0d3n | don't own it. | 21:22 |
| kfox1111 | v1k0d3n: I agree, but not sure how we can manage that on the tight timeline we're on. | 21:22 |
| portdirect_ | yup - we are at real danger of reinventing the wheel | 21:22 |
| v1k0d3n | we tell our people all the time...use upstream...save ourselves work. if it doesn't quite work and we can make it better...do it. | 21:22 |
| v1k0d3n | kfox1111: the timeline? | 21:22 |
| kfox1111 | short term, I thinik we do a mariadb chart in kolla, just to flesh out the working system, | 21:22 |
| v1k0d3n | why can't we get it right? | 21:22 |
| v1k0d3n | do PoC's..... | 21:23 |
| kfox1111 | then as soon as possible, contribute back to the main mariadb chart and drop our own. | 21:23 |
| kfox1111 | that kind of thing. | 21:23 |
| kfox1111 | v1k0d3n: I need to get a system up very early january. | 21:23 |
| v1k0d3n | we went full circle :( | 21:23 |
| v1k0d3n | for work or for kolla-kubernetes? | 21:23 |
| kfox1111 | kolla-kubernetes was actually useful for that purpoese before we started wo nthe helm route. | 21:23 |
| kfox1111 | I think we can finish the conversion to helm and get back to useable for it by the deadline though. | 21:24 |
| v1k0d3n | please tell me you're not hijacking the project because of an item at work. | 21:24 |
| v1k0d3n | :( | 21:24 |
| portdirect_ | dude - that statement terrifies me, we all have commercial concerns here | 21:24 |
| kfox1111 | v1k0d3n: its really easy to convert the existing owrking mariadb chart from jinja2 to helm | 21:24 |
| v1k0d3n | ^^ | 21:24 |
| v1k0d3n | exactly why we have a PoC | 21:24 |
| kfox1111 | its hard to integrate an external one. | 21:25 |
| v1k0d3n | so you totally feel me know :) | 21:25 |
| v1k0d3n | we didn't want to hijack as good citizens | 21:25 |
| v1k0d3n | we wanted to guide | 21:25 |
| kfox1111 | v1k0d3n: there's always tradeoffs. :/ | 21:25 |
| kfox1111 | I've contributed all the work I have at my bosses good graces. :/ | 21:25 |
| v1k0d3n | yeah, look where we're at...we're explaining our intent | 21:25 |
| kfox1111 | they have told me, if ican't get a usable test sytem up january, my time will be moved elsewhere. | 21:26 |
| v1k0d3n | to use upstream helm charts where we can. | 21:26 |
| v1k0d3n | not reinvite the wheel and own all the work | 21:26 |
| kfox1111 | will kolla be better or worse off if we convert an existing mariadb template from jinja2 to gotl, or | 21:26 |
| v1k0d3n | what's teh benefit of getting the jan date, if kolla-kubernetes has to change in the future? | 21:26 |
| kfox1111 | risk loosing a core? | 21:26 |
| kfox1111 | :/ | 21:26 |
| kfox1111 | we can iterate on the templates. bosses are cool with that. | 21:27 |
| kfox1111 | they just need something working to tell their bosses they have something working. | 21:27 |
| v1k0d3n | you're losing cores from others.... | 21:27 |
| v1k0d3n | our list above...do you remember? | 21:27 |
| kfox1111 | possibly. | 21:27 |
| v1k0d3n | this is depressing me. | 21:27 |
| * kfox1111 shurgs. | 21:27 | |
| kfox1111 | yeah. me too. :/ | 21:27 |
| v1k0d3n | we need to find alignment. | 21:27 |
| kfox1111 | I agree with your end goal. | 21:27 |
| kfox1111 | I want to get us there too. | 21:27 |
| v1k0d3n | our approach is everything to our end goal. | 21:28 |
| kfox1111 | the journay is never easy on something this tough. :/ | 21:28 |
| v1k0d3n | to be clear... | 21:28 |
| v1k0d3n | we are 100% community focused | 21:28 |
| kfox1111 | yup. and I think we are still. | 21:28 |
| kfox1111 | the community is made up of members. | 21:28 |
| v1k0d3n | not if there's a jan goal | 21:28 |
| v1k0d3n | :( | 21:28 |
| kfox1111 | and the community gives and takes to keep it a stable whole. | 21:28 |
| portdirect_ | i cannot dontate time to a project that is being directed/driven by once companies/empoyer and projects concerns - we need this to work for everyone. | 21:28 |
| kfox1111 | portdirect_: I fully agree. | 21:29 |
| kfox1111 | and I'd resign before I thought it came to that. | 21:29 |
| kfox1111 | I don' tthink we're there though? | 21:29 |
| v1k0d3n | this is why i'm bringing up the concerns | 21:30 |
| kfox1111 | I have tried to review things fairly and increase the quality of things, not just push stuff through to get somethign done. | 21:30 |
| v1k0d3n | realign possibly | 21:30 |
| v1k0d3n | trust me, everyone wants work done. | 21:31 |
| kfox1111 | I'm just stateing where my own issues lay. 100% open here. | 21:31 |
| kfox1111 | I want whats best for the project. | 21:31 |
| v1k0d3n | but i'm not sure if maintaining helm charts in this repo, that need to work a specific way is the right way to do this long term. | 21:31 |
| v1k0d3n | let's do this...let's rest on this. it's friday. | 21:32 |
| v1k0d3n | let's all think over the weekend, and regroup. | 21:32 |
| v1k0d3n | plan? | 21:33 |
| kfox1111 | v1k0d3n: long timer, i think ti would be really good to get all the charts into kubernetes/charts. | 21:33 |
| kfox1111 | term | 21:33 |
| kfox1111 | its more of a how, and when. | 21:33 |
| v1k0d3n | yeah, but it's better to just start that way. | 21:33 |
| v1k0d3n | this may make more sense now...? | 21:33 |
| v1k0d3n | greenfield | 21:33 |
| v1k0d3n | not later. it's a start. start right. | 21:34 |
| v1k0d3n | i know it's not a "start" for you... | 21:34 |
| v1k0d3n | and that's sucks. | 21:34 |
| kfox1111 | if we actualliy ddid start there, then I'd agree. | 21:34 |
| v1k0d3n | but it's a start for everyone else. | 21:34 |
| kfox1111 | but we started kolla-kuberenets in this repo a long time ago. | 21:34 |
| v1k0d3n | i know. and then we regrouped. | 21:34 |
| v1k0d3n | and i know...you weren't there | 21:34 |
| v1k0d3n | but there was consensus. | 21:34 |
| kfox1111 | we've regruped several times. | 21:34 |
| v1k0d3n | and it's warping back in a way | 21:34 |
| kfox1111 | :/ | 21:34 |
| v1k0d3n | i know. trust me...it hurts. | 21:35 |
| v1k0d3n | but there are folks out there making good cases for the parity with kolla-kubernetes | 21:35 |
| v1k0d3n | as a good citizen, i'm trying to break that down for more cores | 21:35 |
| kfox1111 | yeah. we want more cores. | 21:36 |
| * kfox1111 shrugs. | 21:36 | |
| v1k0d3n | you too dude | 21:37 |
| v1k0d3n | let me think on this. | 21:37 |
| kfox1111 | I don't see a way to get kolla-kubernetes to kubernetes/charts | 21:37 |
| v1k0d3n | in the end...everyone benefits | 21:37 |
| v1k0d3n | i don't even think there's a ton of work | 21:37 |
| v1k0d3n | but i need srwilkers help | 21:37 |
| v1k0d3n | probably lrensing | 21:37 |
| kfox1111 | and me have the level of involvment I'v had recently, in a short amount of time. :/ | 21:37 |
| v1k0d3n | to help you quite a bit | 21:37 |
| *** adrian_otto has joined #openstack-kolla | 21:37 | |
| v1k0d3n | and we need to hang out on the helm weekly meetings | 21:37 |
| v1k0d3n | and we probably need to hang out on some others :-/ | 21:38 |
| srwilkers | im slated for 40 hours a week kfox1111, but willing to work twice that if thats what it takes | 21:38 |
| v1k0d3n | you and i | 21:38 |
| kfox1111 | but I do thin kwe should work on getting it there asap. | 21:38 |
| v1k0d3n | i know portdirect_ is all in | 21:38 |
| srwilkers | kfox1111, thankfully this is 100% my focus in and out of work | 21:38 |
| kfox1111 | v1k0d3n: cool. that would help. | 21:38 |
| v1k0d3n | my plan is to get coreos to help too. | 21:39 |
| v1k0d3n | i need the repos to look....similar | 21:39 |
| kfox1111 | making kolla-kubernetes fly is 100% my focus in and out of work too. | 21:39 |
| v1k0d3n | also a couple more intel folks. | 21:39 |
| portdirect_ | yup - I've put a lot of eggs in this basket - and will do anything I can to make it work here while I can | 21:39 |
| v1k0d3n | so you actually have MORE help than you had before | 21:39 |
| v1k0d3n | i promise. | 21:39 |
| kfox1111 | and grow a healthy community around it. | 21:39 |
| v1k0d3n | now i have a clear message. | 21:39 |
| v1k0d3n | dude, we have some huge meetings about how much we want to throw at this...it's sick | 21:39 |
| v1k0d3n | :) | 21:39 |
| v1k0d3n | not for AT&T...for community | 21:40 |
| v1k0d3n | because we committed to that promise | 21:40 |
| v1k0d3n | in Austin | 21:40 |
| v1k0d3n | we're community first, AT&T second. it's a nice luxury to have. | 21:40 |
| * dims peeks | 21:40 | |
| * srwilkers waves at dims | 21:41 | |
| v1k0d3n | haha hey dims! | 21:41 |
| dims | hi srwilkers v1k0d3n | 21:41 |
| v1k0d3n | if i find a roomful of att folks in here, i'm going to laugh my butt off! | 21:41 |
| srwilkers | lrensing, mattmceuen srwilkers v1k0d3n lamt gagehugo: thats 6 of us at least | 21:42 |
| v1k0d3n | oh man | 21:42 |
| dims | v1k0d3n : i am still hoping against hope to get something going with fuel-ccp :( | 21:42 |
| v1k0d3n | lol | 21:42 |
| v1k0d3n | oh no!! | 21:42 |
| v1k0d3n | lol | 21:42 |
| dims | above my pay grade | 21:43 |
| * v1k0d3n things i don't talk about | 21:43 | |
| v1k0d3n | :) | 21:43 |
| dims | :) | 21:43 |
| mattmceuen | based on this exchange I'm trying to loop in a couple more folks. | 21:43 |
| portdirect_ | hey can anyone fix my phone? It's been really weird the last few days. | 21:44 |
| kfox1111 | dims: best of luck. if you can't, we're here to help. | 21:44 |
| v1k0d3n | oh boy. i need to start saying some truly inspirational things | 21:44 |
| dims | haha portdirect_ | 21:44 |
| dims | thanks kfox1111 | 21:44 |
| kfox1111 | I'm an operator trying to build a tool I'd want to use myself. | 21:45 |
| kfox1111 | just | 21:45 |
| v1k0d3n | kfox1111: you're surrounded by developers who can make it better. | 21:45 |
| v1k0d3n | :) | 21:46 |
| kfox1111 | so trying to minimize operator pain along the whole life cycle. | 21:46 |
| kfox1111 | +1 | 21:46 |
| srwilkers | thats what we're here for | 21:46 |
| v1k0d3n | so kfox1111 our whole intent is exactly what you're talking about. | 21:46 |
| v1k0d3n | you can even take our jenkins code which is going to come soon. | 21:47 |
| v1k0d3n | (for the helm bits) | 21:47 |
| v1k0d3n | but i assure you we want to have the helm charts as pluggable as possible | 21:47 |
| v1k0d3n | and autobuilds in the repos for security CVE's | 21:48 |
| v1k0d3n | and kfox1111 just fyi...i came from a security organization before this one. | 21:48 |
| kfox1111 | cool. thats goign to be important. | 21:48 |
| v1k0d3n | so this is a top concern of mine | 21:48 |
| kfox1111 | and "cloud native", we should be able to run the ci/cd inside k8s. too. :) | 21:48 |
| v1k0d3n | it also aligns with our core objectives...but hey. | 21:48 |
| v1k0d3n | yes, we're using that as a chart | 21:49 |
| kfox1111 | I think we're all pretty soidly in agreement on where we want to go. its mostly a how and a when. | 21:49 |
| kfox1111 | nice. | 21:49 |
| v1k0d3n | best thing about the chart that's provided is it already has the tools required for the CI for this kubernetes/helm piece | 21:49 |
| v1k0d3n | built right in at deploy time. | 21:49 |
| v1k0d3n | so you just have your jenkins code in a repo...done | 21:50 |
| v1k0d3n | kfox1111: the devil is in the details | 21:50 |
| v1k0d3n | at least for us | 21:50 |
| *** dave-mccowan has quit IRC | 21:50 | |
| v1k0d3n | just so you know (no pressure here) our decision is in your hands :) | 21:50 |
| v1k0d3n | lol | 21:50 |
| jmcevoy | portdirect_: Threads=1 looks like it worked... Are the errors at the end expected? https://salemherbals.com/Kolla/kolla301-Results.txt | 21:50 |
| kfox1111 | nice. | 21:50 |
| kfox1111 | v1k0d3n: the decision for what? | 21:51 |
| v1k0d3n | ^^ :) | 21:51 |
| *** britthouser has quit IRC | 21:51 | |
| v1k0d3n | well, we're really big on that reusability part | 21:51 |
| v1k0d3n | and not really owning the work | 21:51 |
| *** britthouser has joined #openstack-kolla | 21:51 | |
| v1k0d3n | and developer workflow | 21:52 |
| kfox1111 | I've lost a bit of context. are you talking about the CI tool? | 21:52 |
| *** goldyfruit has quit IRC | 21:52 | |
| v1k0d3n | and all automated (which goes to your point of operator ease) | 21:52 |
| kfox1111 | yeah. if you can point me at some repo's and stuff, I'd love to give it a look. | 21:53 |
| portdirect_ | jmcevoy: I'm not the best person on that front, but I'd expect similar results | 21:53 |
| kfox1111 | I think kolla-kubernetes really needs the ci stuff for building containers. | 21:54 |
| kfox1111 | (and its not specific to just kolla-kubernetes. many other projects too.) | 21:54 |
| kfox1111 | really its cloud native problem. | 21:54 |
| *** jrich523 has joined #openstack-kolla | 21:54 | |
| portdirect_ | jmcevoy: many services are not supported on diffeent platforms (eg source/binary) different base containers | 21:54 |
| kfox1111 | how to deal with the next openssl volnerability in a cloud native world. | 21:54 |
| sbezverk_ | since I need to split for a meeting, I want to express on the record that I 100% support microserivices approach taken atm I strongly believe we need to move in that direction. | 21:55 |
| portdirect_ | jmcevoy: all of the core services images are built - so you should be good to go with that :) | 21:55 |
| jmcevoy | portdirect_ It looks like way more than what I'll need built... Thanks so much for cutting the threads... Now to edit globals and install... ;-) | 21:57 |
| portdirect_ | jmcevoy: no problem :) I'm not that hot at the kolla-ansible side but there are plenty of knowledgeable folk here who sould be able to help | 21:58 |
| srwilkers | kfox1111: i know its been mentioned a few times, but i think it'll really help when the core team expands for this project in particular and its not just a handful of people. i feel thatll drive the project in a way thats beneficial for all parties, and it'll have the added benefit of handling the influx of work thats bound to come and not rely so heavily on you and the other few who are responsible for managing that | 22:00 |
| openstackgerrit | Mark Giles proposed openstack/kolla-kubernetes: WIP - Helm nova services https://review.openstack.org/408829 | 22:00 |
| srwilkers | i cant imagine what it's like needing to be the workflow guy | 22:00 |
| *** fragatina has joined #openstack-kolla | 22:01 | |
| kfox1111 | srwilkers: +1. its been mostly just me and sbezverk for a long time. I'm very happy to see others contributing. you all have been doing a great job. | 22:02 |
| kfox1111 | hoping to be able to promote some new cores soon. :) | 22:02 |
| srwilkers | kfox1111, i know my commits havent matched my reviews, but hopefully soon i can help you guys out :) | 22:03 |
| kfox1111 | srwilkers: its all good. :) | 22:03 |
| kfox1111 | reviews are quite valuable too. | 22:03 |
| srwilkers | portdirect_ has been busting his tail too. convinced hes a machine and im in the matrix some days | 22:03 |
| kfox1111 | we've had few of those too. | 22:03 |
| portdirect_ | machines? | 22:03 |
| kfox1111 | hehe. portdirect's awesome. :) | 22:03 |
| portdirect_ | I'm 100% organic | 22:03 |
| kfox1111 | jascott1 too. :) | 22:04 |
| kfox1111 | ok. I'm really bad a name, so, if I haven't mentioned your name yet, please dont take offence. :/ | 22:04 |
| kfox1111 | bad at names | 22:04 |
| kfox1111 | the team as grown quite a bit reacently | 22:04 |
| portdirect_ | but on a serious not, kfox you and sbezverk have been doing an amazing job considering the pressure and scrutinly this project is under - I dont envy what your doing for a minute | 22:04 |
| kfox1111 | and everyones doing great. | 22:04 |
| portdirect_ | *note | 22:05 |
| kfox1111 | yeah. its been a delicate balancing act trying to keep everyone hapy and continue to push the project to become healthy and mature. | 22:05 |
| v1k0d3n | "I think kolla-kubernetes really needs the ci stuff for building containers." we were just going to use kolla as is. | 22:06 |
| v1k0d3n | perfectly fine. | 22:06 |
| kfox1111 | yeah. thats good too. | 22:07 |
| kfox1111 | I'd like some automated way to build containers and check for security updates though. | 22:07 |
| v1k0d3n | well, you're talking about using kolla...container in kube to do the current kolla-build stuff, right? | 22:07 |
| kfox1111 | so if you have some thing thats close, I can probably contribute to it to get it hte rest of the way. :) | 22:07 |
| v1k0d3n | yeah, what you're asking for a really common. | 22:07 |
| kfox1111 | what I want is something like: | 22:07 |
| v1k0d3n | the harder stuff is upgrades, etc...but helm manages that too. | 22:08 |
| kfox1111 | * prcess running kubernetes does a kubectl get pods -o yaml ... greap for images | sort -u | 22:08 |
| v1k0d3n | and to be fair...we'd like a commit-based deployment of git repo build... | 22:08 |
| kfox1111 | run a k8s job per image to run an cve checker on it, and do a yum upgrade on it. | 22:08 |
| kfox1111 | if either one says there nees to be updates, | 22:08 |
| kfox1111 | kick off a rebuild of that container, and all deps. | 22:08 |
| openstackgerrit | Steve Wilkerson proposed openstack/kolla-kubernetes: Helm charts for adding Cinder https://review.openstack.org/409349 | 22:08 |
| v1k0d3n | so we have a patch to keystone, a developer (of ours) submits patch, commit, auto-build, pluck out for deployment. | 22:08 |
| kfox1111 | and updated helm packages pointing to them. | 22:08 |
| portdirect_ | it is not an area where we should be looking to reach out - the coreos guys must deal with this the whole time, and I think project atomics doing a few things in this area? | 22:09 |
| kfox1111 | yeah. git repo based is important too. | 22:09 |
| kfox1111 | but some things aren't tracked in git. | 22:09 |
| kfox1111 | like openssl package gets updated due to cve. | 22:09 |
| v1k0d3n | source. we're a development shop. we like code/source :) | 22:09 |
| kfox1111 | all containers containing it needs updated. | 22:09 |
| kfox1111 | do you build all of the base distro from source? | 22:09 |
| * portdirect_ bets the probably do, but suspects they will be pretty coy about it | 22:10 | |
| kfox1111 | portdirect_: /me shrugs. I think the issue of containers needing rebuild is a major general concern to all docker based systems. | 22:11 |
| v1k0d3n | ha! only so much | 22:11 |
| v1k0d3n | right now taking what kolla gives | 22:11 |
| v1k0d3n | ubuntu-source | 22:11 |
| v1k0d3n | and git | 22:11 |
| openstackgerrit | Steve Wilkerson proposed openstack/kolla-kubernetes: Helm charts for adding Cinder https://review.openstack.org/409349 | 22:11 |
| kfox1111 | v1k0d3n: yeah. thas my concern. some of the base deps need updating. so having a procedure to watch for them. | 22:11 |
| v1k0d3n | working with builds now as portdirect_ knows. helped me with a container...which i'm working with slightly now. | 22:12 |
| kfox1111 | the docker hub workflow doesn't handle it at all. :/ | 22:12 |
| *** msimonin1 has quit IRC | 22:12 | |
| v1k0d3n | ^^ agreed | 22:12 |
| sdake | wow | 22:12 |
| sdake | just experienced full on road rage | 22:12 |
| sdake | guy got out of his car | 22:12 |
| sdake | started approaching my vehlice | 22:13 |
| sdake | so i drove around him | 22:13 |
| kfox1111 | sdake: wow. :/ | 22:13 |
| sdake | he threw shit at my car | 22:13 |
| sdake | was flipping me of | 22:13 |
| sdake | etc | 22:13 |
| portdirect_ | what did you do to him? | 22:13 |
| sdake | i told him "keep following me, we are going to the police station to settle the conflict" | 22:13 |
| sdake | that settled the conflict ;) | 22:13 |
| srwilkers | sdake, damn dude | 22:13 |
| kfox1111 | scary. | 22:14 |
| v1k0d3n | sdake: we may have more than usual traffic on our channel...just saying sdake | 22:14 |
| v1k0d3n | glad you're safe! | 22:14 |
| sdake | damn - good thing he didnt 'hve a gun | 22:14 |
| srwilkers | understatement of the year goes tooooo v1k0d3n | 22:14 |
| sdake | or id be dead | 22:14 |
| sdake | gotta pick up son brb | 22:15 |
| portdirect_ | stay tight! | 22:15 |
| v1k0d3n | play some relaxing music on the way :) | 22:16 |
| *** mattmceuen is now known as mattmceuen_away | 22:16 | |
| srwilkers | something totally relaxing, like slayer maybe | 22:16 |
| jascott1 | slayer - the christmas album | 22:17 |
| srwilkers | its snowing blood! | 22:18 |
| jascott1 | lol | 22:18 |
| *** bmace has quit IRC | 22:20 | |
| *** bmace has joined #openstack-kolla | 22:20 | |
| *** fguillot has quit IRC | 22:23 | |
| sdake | all i remember about the guy is he was wearing a red sweatshirt | 22:25 |
| sdake | and was 18-25 | 22:25 |
| sdake | and had long hair | 22:25 |
| sdake | should have snapped a picture of his license plate | 22:25 |
| srwilkers | yeah, thats crazy man | 22:27 |
| srwilkers | on a lighter note, youve got plenty of scrollback if you havent made your way through it yet | 22:28 |
| *** Pavo has quit IRC | 22:29 | |
| srwilkers | heading home -- bbiab | 22:30 |
| *** srwilkers has quit IRC | 22:30 | |
| *** Pavo has joined #openstack-kolla | 22:30 | |
| *** srwilkers_ has joined #openstack-kolla | 22:31 | |
| *** Guest57 has quit IRC | 22:32 | |
| *** b_bezak has joined #openstack-kolla | 22:34 | |
| *** b_bezak has quit IRC | 22:35 | |
| *** b_bezak has joined #openstack-kolla | 22:35 | |
| *** b_bezak has quit IRC | 22:40 | |
| *** msimonin has joined #openstack-kolla | 22:41 | |
| *** unicell1 has quit IRC | 22:43 | |
| *** fragatina has quit IRC | 22:43 | |
| *** fragatina has joined #openstack-kolla | 22:44 | |
| *** fragatina has quit IRC | 22:45 | |
| *** unicell has joined #openstack-kolla | 22:45 | |
| *** fragatina has joined #openstack-kolla | 22:46 | |
| *** cinerama has quit IRC | 22:47 | |
| *** cinerama has joined #openstack-kolla | 22:48 | |
| *** HyperJohnGraham has joined #openstack-kolla | 22:52 | |
| *** TxGirlGeek has quit IRC | 22:56 | |
| kfox1111 | sbezverk_: https://review.openstack.org/#/c/408907/ is ready | 22:56 |
| kfox1111 | lets see it that fixes the gate network issue. | 22:56 |
| *** sdake has quit IRC | 22:59 | |
| *** msimonin has quit IRC | 23:00 | |
| *** sdake has joined #openstack-kolla | 23:01 | |
| jascott1 | hmm should I change memcached to memcached-dp? its lonely being the only one without a - in the name | 23:02 |
| openstackgerrit | Mark Giles proposed openstack/kolla-kubernetes: WIP - Helm nova services https://review.openstack.org/408829 | 23:07 |
| *** schwicht has quit IRC | 23:13 | |
| *** rstarmer has joined #openstack-kolla | 23:14 | |
| *** msimonin has joined #openstack-kolla | 23:17 | |
| *** msimonin has quit IRC | 23:17 | |
| *** msimonin has joined #openstack-kolla | 23:17 | |
| portdirect_ | jascott1: was it you who I was talking to about my docker dev WIP last night? (if so how did you get on?) | 23:18 |
| openstackgerrit | Joshua Harlow proposed openstack/kolla: Allow for externally managed configuration https://review.openstack.org/332590 | 23:22 |
| sdake | adrian_otto the mission statement is set in stone at this point | 23:22 |
| sdake | adrian_otto chisseled by the 500+ contributors to kolla nd its deliverlables | 23:22 |
| sdake | we hvae a very clear mission | 23:23 |
| v1k0d3n | sdake: so kolla-kubernetes and kolla and kolla-ansible are exact same mission? | 23:23 |
| sdake | kfox1111 no we dont need a mission statement for kolla-kubernetes - each project has one mission statement under which all deliverables deliver upon | 23:23 |
| v1k0d3n | nothing about kolla-kubernetes and cloud native or different approach for integration there? | 23:23 |
| sdake | v1k0d3n yup | 23:24 |
| sdake | the tc wants mission statements to be broad | 23:24 |
| sdake | and i understand the reasoning - it keeps us open to differnet options down the road | 23:25 |
| v1k0d3n | that's unfortunate. | 23:25 |
| sdake | the community defined the mission statement early on, and we have been executing it | 23:25 |
| v1k0d3n | i would like to have something more cloud natively focused. | 23:25 |
| sdake | that is an implementation detail and not something the tc would accept | 23:25 |
| sdake | our original mission statement was something ike " | 23:26 |
| v1k0d3n | and not tied directly to kolla. this is not exactly good news for us. | 23:26 |
| sdake | Provide docker containers and ansible playbooks to deploy openstack clouds" | 23:26 |
| *** srwilkers has joined #openstack-kolla | 23:26 | |
| srwilkers | evning | 23:26 |
| sdake | thats wht we started with | 23:26 |
| v1k0d3n | yeah, doesn't fit at all with kolla-kube | 23:26 |
| sdake | righ tthat doesn't fit | 23:26 |
| sdake | our mission statement now is : | 23:26 |
| kfox1111 | sdake: I think having a mission statement just for kolla-kubernetes to focus more on what its goals are would help people not misunderstand its goal. | 23:27 |
| kfox1111 | the kolla goal is a good one. but may be misunderstood in the kolla-kubernetes context. | 23:27 |
| kfox1111 | the idea would just be to clarify. | 23:27 |
| kfox1111 | maybe "mission statement" is the wrong word though. | 23:28 |
| sdake | https://github.com/openstack/governance/blob/master/reference/projects.yaml#L1861 | 23:28 |
| v1k0d3n | kfox1111: mission statement is typical wording though | 23:28 |
| kfox1111 | v1k0d3n: yeah. but in openstack governence I think it has more connotation then just the term. | 23:28 |
| *** takashi_sogabe has joined #openstack-kolla | 23:28 | |
| sdake | read the mission statement - that fits with kolla deliverable, kolla-ansible deliverlable, nd kolla-kubernetes deliverable | 23:29 |
| kfox1111 | so sdake's riht, in that a "mission statement" might not be the right thing. but still, a document that is esentially the same thing, but not for the tc might make sense. | 23:29 |
| kfox1111 | a statement for what kolla-kubernetes is about. | 23:29 |
| kfox1111 | or, maybe the goals you are setting out v1k0d3n are more inline with.... design goals or something. | 23:30 |
| srwilkers | statement of purpose | 23:30 |
| kfox1111 | I kind of started a document like that at one point. | 23:30 |
| sdake | statemnet of purpose = mission statement | 23:30 |
| kfox1111 | more of a, how are we planning on going about meeting the mission with a bit more detail? | 23:30 |
| *** schwicht has joined #openstack-kolla | 23:30 | |
| sdake | if you wnat to clarify what kolla-kubernetes deliverable in particular does, that is something we can do | 23:30 |
| sdake | that operates inside our mission statement | 23:31 |
| kfox1111 | kind of like: https://review.openstack.org/#/c/361186/ | 23:31 |
| kfox1111 | sdake: yeah. that. | 23:31 |
| kfox1111 | that review's fairly dated at this point though. | 23:31 |
| sdake | kfox1111 - ya that looks ok - those are implementation details that fit in with kolla's mission statement | 23:33 |
| kfox1111 | kind of a dcoument that lays out the roles of the game. what things we want to strive for while implementing. | 23:33 |
| kfox1111 | rules | 23:33 |
| *** adrian_otto has quit IRC | 23:33 | |
| *** lrensing has quit IRC | 23:33 | |
| v1k0d3n | your call guys. i'm not really sure of what to think of today. i need to sit on it over the weekend. there was a lot of ground covered, and we really need to find a way forward. | 23:33 |
| sdake | i only read one page of scrollback thus far | 23:33 |
| sdake | i'm not sure what other ground led to the desire to modify the mission statement | 23:34 |
| *** takashi_sogabe has quit IRC | 23:34 | |
| portdirect_ | yeah there's a lot to go over, but I think we can emerge stronger | 23:34 |
| sdake | as a core reviewer, I am -2 on modifying the mission statement | 23:34 |
| v1k0d3n | voting, rather than a spec to determine goals outside of what was discussed and agreed on at Barcelona. | 23:34 |
| *** takashi_sogabe has joined #openstack-kolla | 23:35 | |
| v1k0d3n | not having these items documented in a reviewable spec | 23:35 |
| v1k0d3n | and the closely coupled operations flow of kolla/kolla-kubernetes | 23:35 |
| v1k0d3n | and personal deadlines (sorry, but this bothers me) | 23:36 |
| v1k0d3n | it's shaping the project | 23:36 |
| sdake | not all cores are present at barcelona | 23:36 |
| v1k0d3n | fair enough | 23:36 |
| v1k0d3n | spec? | 23:36 |
| sdake | cores are the extended leadership team | 23:37 |
| sdake | specs are a pain in the ass to deal with | 23:37 |
| sdake | i could go read all the scrollback ify ou like | 23:37 |
| sdake | so i have full context | 23:37 |
| sdake | unfortunately I can't od that today as i had a crown - and I feel like absolute shit | 23:37 |
| v1k0d3n | sorry the spec process is painful? | 23:37 |
| kfox1111 | v1k0d3n: stuff is more binding when in spec form. | 23:38 |
| sdake | right - it requires a rollcall majority vote | 23:38 |
| sdake | we already have a spec | 23:38 |
| sdake | we went through the spec process to define what kolla-kubernetes implementation would look like | 23:38 |
| kfox1111 | v1k0d3n: let me speak to that one more time. | 23:38 |
| sdake | specs are not about defining mission statements | 23:38 |
| kfox1111 | v1k0d3n: we all have personal deadlines/goals. | 23:38 |
| v1k0d3n | kfox1111: our goals are outside of the project | 23:38 |
| kfox1111 | v1k0d3n: so do you. sdake, sbezverk. what matters is growing a diverse set of contributers, | 23:38 |
| sdake | i dont hae a personal deadline ) | 23:39 |
| kfox1111 | so that no one set of goals drives the project. | 23:39 |
| sdake | ;) | 23:39 |
| v1k0d3n | sdake: and kfox1111 neither do we | 23:39 |
| v1k0d3n | we're open source first.... | 23:39 |
| sdake | my deadline is ocata - not personal - that is openstack's deadline | 23:39 |
| sdake | the principles of openstack are here: | 23:39 |
| v1k0d3n | and it's alarming i have people reviewing kolla-kubernetes to meet a specific date requirement from a core. | 23:40 |
| v1k0d3n | ocata is a fair deadline...that's an openstack date | 23:41 |
| kfox1111 | v1k0d3n: Like I said, I have my own deadlines. it is relatet do if I can contribute or not, not if you can contribute or not, or what you contribute. | 23:41 |
| kfox1111 | everyone's bound in that way. | 23:41 |
| kfox1111 | if your boss comes in and says show someong tomorrow or else you can't contribute, your bound in the same way. | 23:42 |
| kfox1111 | this is a community. we're trying to help each other come up with something that will be good for eveyone and outlast hopefully us all. | 23:43 |
| kfox1111 | that is what I'm trying to do to. | 23:43 |
| sdake | https://github.com/openstack/governance/blob/master/reference/principles.rst | 23:43 |
| sdake | kfox1111 right on the money - the goal here is to create a deliverable that lives on far beyond our limited time on this planet | 23:44 |
| sdake | id encourage pople to rea the principles | 23:44 |
| kfox1111 | that takes give and take within the communtity to get to that point. | 23:45 |
| sdake | they are also chissled in stone - by the 5k+ ATCs | 23:45 |
| kfox1111 | here's the distinction. My boss has given me a deadline for how long I can contrube without being able to stand up a test system. that has bearing on kolla-kubernetes outside of the kolla-kubernetes community wanting my time. | 23:46 |
| sdake | v1k0d3n the reson the date of the 15th was set was covered on the mailing list | 23:46 |
| v1k0d3n | kfox1111: https://github.com/openstack/governance/blob/master/reference/principles.rst#openstack-first-project-team-second-company-third | 23:46 |
| portdirect_ | sdake: v1k0d3n is raising a there is some concern raised re: https://github.com/openstack/governance/blob/master/reference/principles.rst#openstack-first-project-team-second-company-third | 23:46 |
| kfox1111 | that has no bearing on | 23:46 |
| kfox1111 | I meant. | 23:46 |
| portdirect_ | whoops | 23:46 |
| v1k0d3n | same timing portdirect_ | 23:46 |
| v1k0d3n | :) | 23:46 |
| kfox1111 | yup. that rule is true. | 23:46 |
| v1k0d3n | trust me, i'm glad not to be the only one here. | 23:46 |
| kfox1111 | all I'm saying is where my time comes from. being open about it. | 23:47 |
| v1k0d3n | there's some disconnect here we can't seem to resolve. | 23:47 |
| kfox1111 | my company is not strangeholding the community saying what they want. | 23:47 |
| sdake | portdirect_ what is the concern with that particular principle portdirect? | 23:47 |
| v1k0d3n | i'm putting myself out there too much, guys. | 23:47 |
| kfox1111 | just saying where my time can go. | 23:47 |
| sdake | or v1k0d3n ? | 23:47 |
| v1k0d3n | if this is disruptive, or if we even have to discuss this for this long...something is wrong. | 23:47 |
| kfox1111 | the community is free to do what it needs to do. | 23:48 |
| v1k0d3n | kfox1111: right. i gotcha. | 23:48 |
| portdirect_ | sdake: there is some cenern that the timeframes we are under for commerical issues are guiding some of the dession making process | 23:48 |
| sdake | yup every9one works on their own objectives for their own interests, this is #3 | 23:48 |
| v1k0d3n | well, it's guiding goals within the project | 23:49 |
| sdake | portdirect where did that originte from? | 23:49 |
| *** msimonin has quit IRC | 23:49 | |
| sdake | i guess i need to read the full scrolback | 23:49 |
| portdirect_ | sdake: some commentry in the srollback | 23:49 |
| sdake | i'd encourage peopel to just trust the process | 23:49 |
| sdake | if you can't trust the process, I dont' know how else to convince ya :) | 23:50 |
| portdirect_ | sdake: i think it's a issue that we can resolve though and has been more of a communication issuse than a systemic one | 23:50 |
| sdake | smarter people then us chisselled the governance repo into stone | 23:50 |
| kfox1111 | sdake: I was trying to be open and let folks konw I really need a working prototype of microservices by early january. not as a way to force development the way I want, but just stating if we tried to be to ambitious with some things, my time to contribute might be cut short. | 23:50 |
| kfox1111 | I think it was taken to mean my company was controlling kolla-kubernetes. and that was not the intention. | 23:51 |
| v1k0d3n | sdake: i think that speaks volumes, but i'm not sure it's making me warm. | 23:51 |
| v1k0d3n | *feel warm | 23:51 |
| sdake | ok, well guys i've had a bad hair week | 23:51 |
| sdake | and i don't have the energy to help everyone understand this today | 23:51 |
| sdake | can we take it up monday? | 23:52 |
| kfox1111 | sdake: yup | 23:52 |
| sdake | i've been through the ringer on personal issues | 23:52 |
| sdake | including above mentioned road rage | 23:52 |
| sdake | and a crown on a front tooth | 23:52 |
| sdake | which is what i was doing today | 23:52 |
| *** jgriffith is now known as jgriffith_AutoAw | 23:52 | |
| srwilkers | not necessarily directly kfox1111. we all know you're vested in this project and you've given a lot, and we appreciate it. i think there's just some concern that the ability to do this in a cloud native way that benefits everyone is made a bit more difficult due to needing to show some tangible value on your end in terms of kolla-k8s being able to prove itself very soon | 23:52 |
| sdake | and my kids not cooperating with school | 23:53 |
| sdake | and i could go on | 23:53 |
| sdake | monday would be better for me to help people understand the process | 23:53 |
| portdirect_ | sdake: of course | 23:53 |
| sdake | in the meantime read the philsophy document and the other mission statements for other projects | 23:53 |
| sdake | there are other documents in the tc governance repository | 23:54 |
| sdake | the idea of the tc governance repo is to be super lightweight | 23:54 |
| sbezverk_ | srwilkers: could you give an example of something substencial wrt your concern, so I could undersatnd | 23:54 |
| sdake | yet provide enough guidance so folks don't have conflict :) | 23:54 |
| sdake | i'd run for the tc, but instead i'm running for the board of directors to solve more strategic issues wrt openstack | 23:55 |
| sdake | can't be in two places at once ;) | 23:55 |
| *** schwicht has quit IRC | 23:56 | |
| sdake | https://www.openstack.org/election/2017-individual-director-election/CandidateList | 23:56 |
| sdake | if selected by the openstack membership, i do intend to remain core reviewer for all kolla deliverabls | 23:57 |
| portdirect_ | yeah - you guys have a heck of a job trying to balance this stuff - on multiple levels | 23:57 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!