| *** portdirect_ is now known as portdirect_away | 00:00 | |
| kfox1111 | so, just to be clear. | 00:01 |
|---|---|---|
| v1k0d3n | sbezverk_ secrets. | 00:01 |
| kfox1111 | going through the rules sdake posted. | 00:01 |
| kfox1111 | I think the concern is this: | 00:01 |
| kfox1111 | "They can of course represent the needs of their project team or their company, but in case of conflicts of interest, they should be ready to put those needs aside and make the best call for OpenStack as a whole." | 00:01 |
| v1k0d3n | how are secrets used during the injection of configmap data? | 00:01 |
| kfox1111 | I believe I've done that. If not, how so. | 00:01 |
| sbezverk_ | v1k0d3n: sorry, kubernetes secrets? | 00:02 |
| v1k0d3n | how does a user bring their own secrets to the helm charts, for importing into the configmap? | 00:02 |
| sbezverk_ | v1k0d3n: I must have missed the context, you do not approve how we use them? | 00:02 |
| v1k0d3n | sbezverk_ you said give you an example to srwilkers...i'm giving an example | 00:03 |
| v1k0d3n | sorry | 00:03 |
| v1k0d3n | sbezverk_: i'm not sure why everyone is getting offended. | 00:03 |
| v1k0d3n | i'm bringing up concerns that things are moving quickly and there are some misses. | 00:03 |
| sdake | how kubernetes secrets are handled is covered in the spec | 00:03 |
| sdake | specifically they are external, not internal | 00:03 |
| sbezverk_ | v1k0d3n: got it.. I do not think we use secrets as a part of config map | 00:03 |
| v1k0d3n | then i was told that i'm not contributing...fair enough. sbezverk_ you said yourself that i have not talked to cores, but i have. | 00:04 |
| sbezverk_ | basically we mount them inside of a template | 00:04 |
| sdake | v1k0d3n pretty sure your contributing | 00:04 |
| *** sayantan_ has quit IRC | 00:04 | |
| v1k0d3n | kfox1111: have i not asked if we're attempting to keep sync? | 00:04 |
| sdake | pretty sure the dev env was a huge huge contribution | 00:04 |
| v1k0d3n | and i get told "you're free to join" which i do... | 00:04 |
| v1k0d3n | sbezverk_: yes...secrets can be imported in. that's a powerful part. | 00:04 |
| v1k0d3n | user can bring their own, or generate | 00:04 |
| v1k0d3n | i think everyone so busy, that we're not looking at the other projects that are similar. | 00:05 |
| sbezverk_ | v1k0d3n: that is what I like :-) you put forward a technical issue which is fun to ger resolve :-) | 00:05 |
| v1k0d3n | and the other projects are asying that the goals don't align. i would like them to have less parity. | 00:05 |
| kfox1111 | v1k0d3n: I appologized for misunderstanding your level of contribution. | 00:05 |
| v1k0d3n | sbezverk_: you're one of my favorite people. you take recommendations. | 00:05 |
| v1k0d3n | but i fear that we're missing what is already done out there. | 00:06 |
| v1k0d3n | :( | 00:06 |
| v1k0d3n | oh i haven't contributed code. | 00:06 |
| v1k0d3n | :) | 00:06 |
| v1k0d3n | not here. it's not that i don't WANT to, or can't | 00:06 |
| kfox1111 | v1k0d3n: but you have contributed people, and communication, which is just as important. | 00:06 |
| v1k0d3n | kfox1111: this goes to the point i made earlier.... | 00:06 |
| v1k0d3n | my corp objective... | 00:06 |
| kfox1111 | thats a contribution in my book. | 00:06 |
| v1k0d3n | that's in a PoC | 00:06 |
| v1k0d3n | we're on the hook for that. not the community. | 00:06 |
| v1k0d3n | i said steal our stuff entirely. | 00:07 |
| kfox1111 | v1k0d3n: not to minimize that, but community is far more important then poc. | 00:07 |
| v1k0d3n | srwilkers: is supposed to merge...i think he's trying, but confused how they all fit | 00:07 |
| sdake | v1k0d3n in openstack we dont steal, copying is a totaly normal thing in openstack | 00:07 |
| *** portdirect_away is now known as portdirect_ | 00:07 | |
| kfox1111 | v1k0d3n: community first. code second. | 00:07 |
| sdake | in magnum for example, I copied ironic's implementation to rough out the initital code base | 00:07 |
| v1k0d3n | because they're so different currently. too much parity. | 00:07 |
| v1k0d3n | one think that maybe you guys are using later(?) livenessProbe: and readinessProbe: ? | 00:08 |
| sdake | ok gotta feed my child before she strangles me after being in the road rage incident | 00:08 |
| *** bachp has quit IRC | 00:08 | |
| kfox1111 | v1k0d3n: that only gets ironed out, if one party chucks their code, or the devs start talking I think. :/ | 00:08 |
| v1k0d3n | i didn't see them currently. that's what flagged me on the differences, then i started digging in ddeeper. | 00:08 |
| sdake | v1k0d3n thats in the spec | 00:08 |
| v1k0d3n | soryr i have to read back...i'm missing all of your comments and have to read back | 00:08 |
| kfox1111 | v1k0d3n: yeah. we need more checks. we havent focused on that. | 00:09 |
| kfox1111 | each project has focuesed on different things. | 00:09 |
| kfox1111 | so there are some one does better, some that the other does better, some things we both do the same, and some things we do differently. | 00:09 |
| v1k0d3n | steal i mean take...it's MEANT 100% for kolla-kubernetes | 00:09 |
| v1k0d3n | we were waiting for direction on operators, entrypoint, init, etc etc etc | 00:09 |
| v1k0d3n | that took a while | 00:09 |
| srwilkers | ill be back shortly. | 00:10 |
| *** srwilkers is now known as srwilkers_brb | 00:10 | |
| v1k0d3n | if you notice, that's when srwilkers paused for a bit and helped elsewhere. | 00:10 |
| kfox1111 | v1k0d3n: there hasn't been a good track record in open source communities for copmanies doing code drops and importing them into other projects. | 00:10 |
| kfox1111 | it tends to work much better when its being contributed to directly. | 00:10 |
| sdake | ya code dumps always fail | 00:10 |
| sdake | it results in lack of engagement | 00:10 |
| v1k0d3n | kfox1111: there also has been this level of communication. i have been so extremely clear about everything. | 00:10 |
| sdake | hence nobody to maintain it | 00:11 |
| kfox1111 | we need to grow a healthy community. | 00:11 |
| sdake | hence no community | 00:11 |
| v1k0d3n | we want to drop that repo | 00:11 |
| v1k0d3n | and again...we were asked to do a PoC in the first place | 00:11 |
| kfox1111 | so, can we get the folks working on aic-helm to try and take the templates there, convert them to use the microservices and contribute them back? | 00:11 |
| sdake | v1k0d3n by whom? | 00:11 |
| kfox1111 | then both parties get what they want? | 00:11 |
| v1k0d3n | (by kolla in barcelona) | 00:11 |
| v1k0d3n | PTL | 00:11 |
| sdake | ok, well ptl doesnt rule the universe | 00:12 |
| kfox1111 | v1k0d3n: thats a company request. | 00:12 |
| v1k0d3n | kfox... | 00:12 |
| v1k0d3n | have you looked at them?!?!?! | 00:12 |
| kfox1111 | and the ptl doesn't control what people contribute. | 00:12 |
| v1k0d3n | they ARE microserviced out | 00:12 |
| sdake | the ptl works with teh core team to make things happen :) | 00:12 |
| kfox1111 | v1k0d3n: yeah. I have. show in: https://github.com/att-comdev/aic-helm/tree/master/glance/templates | 00:12 |
| kfox1111 | where glance-api can be launched all by itself as a microservice. | 00:13 |
| kfox1111 | I thinik we may be disagreeing on terminology? | 00:13 |
| v1k0d3n | kfox1111: the one you pic is the new one that got added that we didn't get a chance to go over... | 00:13 |
| v1k0d3n | it's being refactored | 00:13 |
| *** mgiles has quit IRC | 00:13 | |
| v1k0d3n | for this reason | 00:13 |
| kfox1111 | v1k0d3n: can you point me at an example then? I can't find one. | 00:14 |
| kfox1111 | maybe a different branch? | 00:14 |
| *** srwilkers_brb is now known as srwilkers | 00:15 | |
| v1k0d3n | kfox1111: https://github.com/att-comdev/aic-helm/blob/37afa516ca396981e71e8ebcfa43020788a0779d/glance/values.yaml | 00:16 |
| v1k0d3n | images???? | 00:16 |
| sdake | bb-45mins | 00:16 |
| kfox1111 | thats not a microservice. thats a service. you cant do just one of those things, but hae to orchestrate the deployment of the entire set of glance services all at once. | 00:16 |
| kfox1111 | v1k0d3n: here's what I'm talking about. | 00:17 |
| *** sdake has quit IRC | 00:17 | |
| kfox1111 | say you take: | 00:17 |
| kfox1111 | https://github.com/att-comdev/aic-helm/tree/37afa516ca396981e71e8ebcfa43020788a0779d/glance/templates | 00:17 |
| v1k0d3n | wait, so a micoservice is not a container...? separated processes? | 00:17 |
| sbezverk_ | v1k0d3n: in the approach you push, is there a way to re-run glance db-manage as an individual task? | 00:17 |
| v1k0d3n | a microservice is a small atomic unit of a config file (helm chart)? | 00:17 |
| portdirect_ | sbezverk_: yes via the k8s api | 00:17 |
| v1k0d3n | this is the misunderstanding here.... | 00:17 |
| kfox1111 | glance registry, glance api tamples, and replace them with a glance-registry subpackage for the kolla-kubernetes microservcie, and the glance-api microservice. | 00:18 |
| jascott1 | that bypasses helm release right | 00:18 |
| v1k0d3n | microservice in my world...is an individual container for a specified purpose. | 00:18 |
| kfox1111 | v1k0d3n: microserviee is a package per k8s object. | 00:18 |
| v1k0d3n | not a small config unit. | 00:18 |
| *** jgriffith_AutoAw is now known as jgriffith | 00:18 | |
| v1k0d3n | ah.... | 00:19 |
| v1k0d3n | ok...so is a pod a microservice unit? | 00:19 |
| kfox1111 | one package per k8s object time. (deployment,daemonset,petset,etc) | 00:19 |
| sbezverk_ | portdirect_: it sounds kind of generic, could you be more specific | 00:19 |
| sbezverk_ | ? | 00:19 |
| kfox1111 | basicaly what you 'kubectl create -f <thingy>' | 00:19 |
| kfox1111 | you can then thake the microservices and bundle them into service level helm packages for orchestration, like your doing. | 00:20 |
| v1k0d3n | right | 00:20 |
| v1k0d3n | but in my <thing> | 00:20 |
| kfox1111 | its basicallyu just a bit of helm rapping around the embeded templates. | 00:20 |
| v1k0d3n | i have multiple contianers... | 00:20 |
| kfox1111 | there are multipole containers per pod. | 00:21 |
| kfox1111 | there are multiple pods per daemonset/deployment/etc | 00:21 |
| kfox1111 | there are one k8s object per microservie package. | 00:21 |
| kfox1111 | multiple microservices per service package. | 00:21 |
| portdirect_ | sbezverk_: yeah from my perspective (unless I'm missing somthing about helm) it simple puts the tempates out there - so in the case of the jobs we can once it's deployed then use the k8s api to update it - in the same way you would not install a .deb every time you wanted to initialise a user - but I was answering very much a specific tecnical point you raised - nothing more :/ | 00:21 |
| kfox1111 | v1k0d3n: does help explain the kolla-kubernetes terminology? I think we've been using different definitions and therefore talking past each other. | 00:22 |
| v1k0d3n | to me, there is only kubernetes terminology. | 00:23 |
| v1k0d3n | and helm | 00:23 |
| kfox1111 | v1k0d3n: there is no 'microservice' definition in k8s as far as I know. | 00:23 |
| kfox1111 | or helm. | 00:23 |
| kfox1111 | if they have a clear defiition you could point me to, I'll adopt it. | 00:24 |
| kfox1111 | a lot of problems end up when people define things differently. :/ | 00:25 |
| v1k0d3n | i gotta run guys. sorry for causing a stir, unfortunately this isn't really going anywhere for me, and i have too many concerns. it's 7:30 where i'm at on a friday. | 00:26 |
| sbezverk_ | portdirect_: I see, I just looked and glance temaplates and they are all in the same template folder, in my mind helm deploys all temapltes it finds in this folder, hence my question how it is done individually | 00:26 |
| v1k0d3n | if anyone needs to reach me, you guys know where to find me. | 00:26 |
| kfox1111 | sbezverk_: right. as far as I can understand, it just loads them all. | 00:27 |
| v1k0d3n | sbezverk_: https://github.com/att-comdev/aic-helm/blob/master/ceph/.gitignore#L1 | 00:27 |
| v1k0d3n | https://github.com/att-comdev/aic-helm/blob/6994418899d0309e12904b55d01aea5de03430c8/common/Makefile#L2 | 00:27 |
| v1k0d3n | https://github.com/att-comdev/aic-helm/blob/eb51f5b67efc05e2750ae07f31c060e66118a3ea/ceph/templates/secrets.yaml#L9-L49 | 00:28 |
| v1k0d3n | https://github.com/att-comdev/aic-helm/blob/eb51f5b67efc05e2750ae07f31c060e66118a3ea/bootstrap/templates/secrets.yaml#L9 | 00:28 |
| v1k0d3n | finally: https://github.com/att-comdev/aic-helm/blob/6994418899d0309e12904b55d01aea5de03430c8/common/utils/secret-generator/README.md#generate-raw-fsid-can-be-used-for-other-functions | 00:29 |
| kfox1111 | v1k0d3n: if your concern is primarily about being able to start over in helm/charts. thats fine with me. but you gota get the commuity to vote for it. | 00:29 |
| *** Pavo has quit IRC | 00:29 | |
| kfox1111 | and I'll abstain from voting, as I have a conflict of interest timeine wise. | 00:29 |
| kfox1111 | timeline | 00:29 |
| kfox1111 | that should clear up the issue. | 00:30 |
| kfox1111 | same with dropping mariadb microservice in exchange for upstream kubernetes/mariadb. | 00:30 |
| kfox1111 | get a vote, and I'll abstain. | 00:31 |
| v1k0d3n | i'll say what i said to you privately. i don't want to disrupt kolla-kubernetes again. | 00:32 |
| v1k0d3n | just fwiw...we talked about helm from the beginning, remember? | 00:32 |
| v1k0d3n | you said in the future, remember? | 00:32 |
| v1k0d3n | but you had a goal then. | 00:32 |
| kfox1111 | v1k0d3n: I want whats best for the community. | 00:32 |
| v1k0d3n | you needed something. | 00:32 |
| v1k0d3n | what gets coded in usually stays. | 00:33 |
| kfox1111 | v1k0d3n: we're working on helm. | 00:33 |
| v1k0d3n | ok | 00:33 |
| portdirect_ | sbezverk_: np - it would be nice if kubectl provided a command to rerun a job rather than editing the desired number of completeions in the spec as in it's current form it is a bit hacky. | 00:33 |
| *** cliles has quit IRC | 00:33 | |
| kfox1111 | when we talked last about helm, I said I didn't have time to work on helm. I can only say what I have time tocontribute. | 00:33 |
| kfox1111 | I was planning on looking ot helm more in the future. | 00:34 |
| *** Pavo has joined #openstack-kolla | 00:34 | |
| v1k0d3n | really need to go. today has been disruptive enough for me. :-/ | 00:34 |
| sbezverk_ | v1k0d3n: can we look at it from different angle. you want to use helm all the way, right? what we do now with microservices most likely will not prevent helm global orchestration, why you do not consider running global helm orchestration on microserviced chart temapltes? | 00:36 |
| *** jgriffith is now known as jgriffith_AutoAw | 00:37 | |
| kfox1111 | sbezverk_: +1 | 00:37 |
| *** ntpttr has quit IRC | 00:37 | |
| *** schwicht has joined #openstack-kolla | 00:37 | |
| *** jascott1_ has joined #openstack-kolla | 00:39 | |
| v1k0d3n | hey, i don't want to change direction | 00:39 |
| sbezverk_ | if there are some issue with current microservices let's discuss it and fix as a technical issues and not Trump/Clinton debate like | 00:39 |
| v1k0d3n | we're going around in circles | 00:39 |
| v1k0d3n | i want to remove myself from the loop. i've made my case, and others have politely chimed in. | 00:39 |
| v1k0d3n | i'm being the strong one, and it makes me a bit uncomfortable. | 00:40 |
| v1k0d3n | this is the beauty of open source. | 00:40 |
| *** jascott1 has quit IRC | 00:42 | |
| *** pece has quit IRC | 00:43 | |
| openstackgerrit | Merged openstack/kolla-kubernetes: Have gate wait for dns ready https://review.openstack.org/408907 | 00:43 |
| sbezverk_ | v1k0d3n: we are all human and do not like waste time, so far I see, I am offered to forget about last two month and switch to something different and new, I am sorry I do not take it with smile | 00:45 |
| sbezverk_ | I really really want to make sure that new is really worth such action | 00:46 |
| v1k0d3n | this is what i've been doing on behalf of kolla sbezverk_ working with other cloud native members to find out what is needed to merge projects | 00:47 |
| v1k0d3n | it would be nice to see openstack have a project dedicated to this mession. | 00:48 |
| v1k0d3n | it is what it needs for clear interop with kubernetes. | 00:48 |
| v1k0d3n | not just on the tenant-side, but also with kubernetes being first - class. | 00:49 |
| *** n0isyn0i1e has joined #openstack-kolla | 00:49 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 00:49 |
| kfox1111 | v1k0d3n: I think we're all in agreement on that. | 00:50 |
| kfox1111 | v1k0d3n: and I do thank you for continueing to try and bring the teams together. its a great thing. | 00:51 |
| kfox1111 | its just hard. and we're trying. | 00:51 |
| v1k0d3n | how so? | 00:51 |
| v1k0d3n | meaning, engaging other groups outside of openstack? | 00:52 |
| kfox1111 | bringing together different teams. | 00:52 |
| *** n0isyn0ise has quit IRC | 00:52 | |
| kfox1111 | with different ideas on how to get to the end goal. | 00:52 |
| kfox1111 | both teams agree on end goal I think. just not sure on the road to get there. | 00:53 |
| v1k0d3n | from stackanetes, fuel ccp, sap, pulling in these open source projects here? | 00:53 |
| kfox1111 | that is hard to work out. but worth trying. | 00:53 |
| v1k0d3n | dims: you still there? | 00:53 |
| v1k0d3n | i know you've been peeking here and there :) | 00:54 |
| v1k0d3n | if i'm the only one who feels this way, i will completely back down. | 00:55 |
| Pavo | anyone might know what this error means? http://paste.openstack.org/show/592000/ | 00:55 |
| v1k0d3n | so far this has just been you and me going through this. | 00:55 |
| *** sdake has joined #openstack-kolla | 00:56 | |
| sdake | back | 00:56 |
| v1k0d3n | kolla needs to do what it needs to do | 00:56 |
| v1k0d3n | this is why we brought this up in barcelona | 00:57 |
| v1k0d3n | i understand cores couldn't make it. but man, why have those working sessions with the PTL then? | 00:57 |
| Pavo | anyone might what this error means? http://paste.openstack.org/show/592000/ | 00:58 |
| *** tonanhngo has quit IRC | 00:59 | |
| kfox1111 | v1k0d3n: thats the thing about opensource governence. "One Contributor, One Vote" | 00:59 |
| jascott1_ | Pavo I would find the task thats failing and check it out | 00:59 |
| kfox1111 | the ptl helps try and lead the cats, but the cats get the final say. | 00:59 |
| v1k0d3n | point taken kfox1111 | 00:59 |
| v1k0d3n | noted. but we had people all in agreement there fwiw. | 01:00 |
| kfox1111 | v1k0d3n: https://github.com/openstack/governance/blob/master/reference/principles.rst is whate we all follow. | 01:00 |
| sdake | ya i think of a ptl with my 5 years of experience doing it s a "Project Team Facilitator" | 01:00 |
| sdake | PTL is not a dictator | 01:00 |
| kfox1111 | sdake: and a great job of it too. :) | 01:00 |
| jascott1_ | cat whisperer | 01:01 |
| sdake | openstack doesn't work under a dictator model | 01:01 |
| sdake | kfox1111 atm i'm retired as a ptl ;) | 01:01 |
| *** v1k0d3n is now known as severion | 01:01 | |
| srwilkers | jascott1_, hahaha | 01:01 |
| *** severion has quit IRC | 01:01 | |
| *** v1k0d3n has joined #openstack-kolla | 01:01 | |
| sdake | no BDFL model in openstack, unlike kernel.org | 01:02 |
| *** v1k0d3n is now known as severion | 01:02 | |
| Pavo | jascott1_ I have and found nothing that would indicate why it fails then continues on fine without issues | 01:02 |
| *** severion has quit IRC | 01:02 | |
| jascott1_ | Pavo maybe the service is up but check is failing incorrectly? | 01:03 |
| *** severion has joined #openstack-kolla | 01:03 | |
| Pavo | it only does it during a deployment | 01:03 |
| Pavo | and after I check and everything is running fine | 01:03 |
| Pavo | so have no clue why it errors and then works | 01:04 |
| jascott1_ | ya you kinda just described ansible ;) | 01:04 |
| srwilkers | could just be retrying until the mariadb containers up and responds | 01:04 |
| jascott1_ | pavo tried debug flag? | 01:06 |
| Pavo | ok will do | 01:06 |
| Pavo | if I have deployed and I want to add another compute node, do I just redeploy or will reconfigure work? | 01:08 |
| srwilkers | Pavo, im assuming thats to be expected with mariadb in kolla-ansible. the task to start mariadb is run with 'wait-for:' with retries, so it's expected to likely need to wait for a response depending on environment, etc | 01:10 |
| srwilkers | https://github.com/openstack/kolla-ansible/blob/master/ansible/roles/mariadb/tasks/start.yml#L21-L31 | 01:11 |
| Pavo | yeah I figured that srwilkers | 01:15 |
| Pavo | thanks | 01:15 |
| kfox1111 | sdake: yeah. but still, thnaks for all your help as ptl, and guidence afterwards. | 01:15 |
| srwilkers | cheers Pavo :) | 01:16 |
| srwilkers | unfortunately im not sure about reconfigure Pavo :( | 01:16 |
| Pavo | ok, guess I'll just have to try and findout | 01:16 |
| sdake | dinner time | 01:17 |
| sbezverk_ | sdake: que aproveche señor :-) | 01:20 |
| *** jgriffith_AutoAw is now known as jgriffith | 01:45 | |
| kfox1111 | sbezverk_: sbezverk_ the dns test either didn't work, or the issue is unrelated: http://logs.openstack.org/63/408263/12/check/gate-kolla-kubernetes-deploy-centos-binary-ceph-nv/dc431ce/logs/pods/kolla-neutron-create-keystone-endpoint-public-zlgts-main.txt | 01:50 |
| *** portdirect_ is now known as portdirect_away | 01:53 | |
| *** jgriffith is now known as jgriffith_AutoAw | 01:56 | |
| *** srwilkers_ has quit IRC | 01:58 | |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible: Move aodh precheck into its own role https://review.openstack.org/404389 | 02:00 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 02:03 |
| *** sayantan_ has joined #openstack-kolla | 02:11 | |
| *** awiddersheim has quit IRC | 02:16 | |
| *** sayantan_ has quit IRC | 02:21 | |
| *** sayantan_ has joined #openstack-kolla | 02:21 | |
| *** tonanhngo has joined #openstack-kolla | 02:23 | |
| *** mdnadeem has joined #openstack-kolla | 02:25 | |
| *** Pavo has quit IRC | 02:29 | |
| *** Pavo has joined #openstack-kolla | 02:30 | |
| sbezverk_ | kfox1111: this log looks a bit wierd | 02:32 |
| sbezverk_ | kfox1111: is there any way to attach to VM when it runs the gate to be able to troubleshoot this condition? | 02:33 |
| *** mdnadeem has quit IRC | 02:34 | |
| *** unicell1 has joined #openstack-kolla | 02:35 | |
| *** unicell has quit IRC | 02:35 | |
| *** mkoderer has quit IRC | 02:36 | |
| *** jtriley has joined #openstack-kolla | 02:44 | |
| *** tonanhngo has quit IRC | 02:48 | |
| openstackgerrit | Merged openstack/kolla-ansible: Always check enable_* variables first https://review.openstack.org/408529 | 03:01 |
| *** HyperJohnGraham has quit IRC | 03:02 | |
| *** sdake has quit IRC | 03:10 | |
| *** rhallisey has quit IRC | 03:15 | |
| *** imcsk8_ has joined #openstack-kolla | 03:17 | |
| *** ayoung has quit IRC | 03:19 | |
| openstackgerrit | Mark Giles proposed openstack/kolla-kubernetes: WIP - Helm nova services https://review.openstack.org/408829 | 03:25 |
| *** kollabot has quit IRC | 03:32 | |
| *** kollabot4 has joined #openstack-kolla | 03:32 | |
| *** kollabot4 has quit IRC | 03:33 | |
| *** kollabot has joined #openstack-kolla | 03:35 | |
| *** harlowja_ has quit IRC | 03:36 | |
| *** schwicht has quit IRC | 03:38 | |
| *** duonghq has joined #openstack-kolla | 04:01 | |
| duonghq | morning guys | 04:01 |
| srwilkers | morning duonghq | 04:01 |
| duonghq | anybody got ES cluster crash and lost data? | 04:05 |
| duonghq | srwilkers, iirc, you live in US? | 04:06 |
| srwilkers | duonghq, yep :) CST | 04:06 |
| duonghq | got it | 04:07 |
| *** mdnadeem has joined #openstack-kolla | 04:20 | |
| *** Pavo has quit IRC | 04:29 | |
| *** Pavo has joined #openstack-kolla | 04:34 | |
| *** yatin has quit IRC | 04:38 | |
| *** tonanhngo has joined #openstack-kolla | 04:49 | |
| *** tonanhngo has quit IRC | 04:50 | |
| *** yatin has joined #openstack-kolla | 04:50 | |
| *** zhubingbing has joined #openstack-kolla | 04:54 | |
| *** sdake has joined #openstack-kolla | 04:56 | |
| sdake | hey peeps | 04:56 |
| sdake | portdirect you about | 04:56 |
| duonghq | sup sdake | 04:56 |
| sdake | sup dounghq | 04:56 |
| sdake | duonghq :) | 04:56 |
| duonghq | :) | 04:56 |
| srwilkers | sup sdake | 04:56 |
| sdake | sup srwilkers | 04:56 |
| *** lamt has quit IRC | 04:57 | |
| *** portdirect_away is now known as portdirect_ | 04:58 | |
| * portdirect_ reboots | 04:59 | |
| portdirect_ | sup | 04:59 |
| srwilkers | :p | 05:01 |
| *** hogepodge has quit IRC | 05:06 | |
| *** duonghq has quit IRC | 05:10 | |
| *** severion has quit IRC | 05:12 | |
| *** v1k0d3n has joined #openstack-kolla | 05:12 | |
| v1k0d3n | srwilker-: you still up man? | 05:13 |
| v1k0d3n | lol | 05:13 |
| *** mdnadeem has quit IRC | 05:13 | |
| srwilkers | lol yeah, im here | 05:13 |
| v1k0d3n | wow...you're a machine srwilker- | 05:13 |
| srwilkers | can always sleep next week | 05:14 |
| *** jgriffith_AutoAw is now known as jgriffith | 05:28 | |
| sbezverk_ | duonghq: ping | 05:31 |
| *** unicell has joined #openstack-kolla | 05:34 | |
| *** unicell1 has quit IRC | 05:34 | |
| *** duonghq has joined #openstack-kolla | 05:35 | |
| *** portdirect_ is now known as portdirect_away | 05:40 | |
| *** jgriffith is now known as jgriffith_AutoAw | 05:41 | |
| *** yatin is now known as yatin_on_leave | 05:43 | |
| *** sdake_ has joined #openstack-kolla | 05:49 | |
| sdake_ | rbergeron are you around | 05:51 |
| *** jtriley has quit IRC | 05:51 | |
| sdake_ | rbergeron snakes on a plane? | 05:52 |
| *** sdake has quit IRC | 05:53 | |
| sdake_ | rogue one incoming | 05:56 |
| sdake_ | night folks | 05:56 |
| srwilkers | later sdake_ | 05:56 |
| sdake_ | night srwilkers | 05:56 |
| *** sdake_ has quit IRC | 05:57 | |
| *** saneax-_-|AFK is now known as saneax | 06:20 | |
| *** duonghq has quit IRC | 06:27 | |
| *** Pavo has quit IRC | 06:29 | |
| *** zhubingbing has quit IRC | 06:34 | |
| *** yatin_on_leave has quit IRC | 06:34 | |
| *** Pavo has joined #openstack-kolla | 06:34 | |
| *** zhubingbing has joined #openstack-kolla | 06:34 | |
| *** yatin_on_leave has joined #openstack-kolla | 06:47 | |
| *** adrian_otto has joined #openstack-kolla | 06:48 | |
| *** yatin_on_leave has quit IRC | 06:58 | |
| *** saneax is now known as saneax-_-|AFK | 07:01 | |
| *** saneax-_-|AFK is now known as saneax | 07:02 | |
| *** prameswar has joined #openstack-kolla | 07:04 | |
| *** adrian_otto has quit IRC | 07:07 | |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible: Move aodh precheck into its own role https://review.openstack.org/404389 | 07:09 |
| *** adrian_otto has joined #openstack-kolla | 07:10 | |
| *** yatin_on_leave has joined #openstack-kolla | 07:11 | |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible: Move manila precheck into its own role https://review.openstack.org/404338 | 07:15 |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible: Move manila precheck into its own role https://review.openstack.org/404338 | 07:17 |
| *** yatin_on_leave has quit IRC | 07:18 | |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible: Move watcher precheck into its own role https://review.openstack.org/404468 | 07:23 |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible: Move barbican precheck into its own role https://review.openstack.org/404465 | 07:28 |
| *** yatin_on_leave has joined #openstack-kolla | 07:30 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible: Enable tempest test in gate https://review.openstack.org/402122 | 07:32 |
| openstackgerrit | Merged openstack/kolla: Install watcher dashboard into horizon image https://review.openstack.org/406607 | 07:42 |
| *** sayantan_ has quit IRC | 07:47 | |
| *** sgordon has quit IRC | 07:51 | |
| *** sgordon has joined #openstack-kolla | 07:54 | |
| *** tonanhngo has joined #openstack-kolla | 08:01 | |
| *** tonanhngo has quit IRC | 08:03 | |
| *** adrian_otto has quit IRC | 08:10 | |
| *** prameswar has quit IRC | 08:16 | |
| *** rstarmer has quit IRC | 08:16 | |
| *** yee37911 has joined #openstack-kolla | 08:19 | |
| *** yee379 has quit IRC | 08:21 | |
| *** portdirect_ has joined #openstack-kolla | 08:21 | |
| *** tonyb has quit IRC | 08:22 | |
| *** portdirect_away has quit IRC | 08:23 | |
| *** tonyb has joined #openstack-kolla | 08:24 | |
| *** Pavo has quit IRC | 08:29 | |
| *** Pavo has joined #openstack-kolla | 08:34 | |
| *** msimonin has joined #openstack-kolla | 08:52 | |
| *** saneax is now known as saneax-_-|AFK | 08:52 | |
| *** tonyb has quit IRC | 08:53 | |
| *** tonyb has joined #openstack-kolla | 08:55 | |
| *** msimonin has quit IRC | 09:05 | |
| *** cmceniry has quit IRC | 09:05 | |
| *** cmceniry has joined #openstack-kolla | 09:10 | |
| openstackgerrit | Merged openstack/kolla: Install mistral dashboard into horizon image https://review.openstack.org/406726 | 09:19 |
| openstackgerrit | caoyuan proposed openstack/kolla: Remove the unnecessary "openstack-zaqar-ui" https://review.openstack.org/407576 | 09:25 |
| *** msimonin has joined #openstack-kolla | 09:27 | |
| *** msimonin has quit IRC | 09:31 | |
| *** Jeffrey4l has joined #openstack-kolla | 09:32 | |
| *** cmceniry has quit IRC | 09:32 | |
| *** cmceniry has joined #openstack-kolla | 09:33 | |
| *** sean-k-m1 has joined #openstack-kolla | 09:46 | |
| *** sean-k-mooney has quit IRC | 09:49 | |
| *** zhubingbing has quit IRC | 10:00 | |
| *** yee37911 has quit IRC | 10:22 | |
| *** kollabot has quit IRC | 10:23 | |
| *** kollabot has joined #openstack-kolla | 10:23 | |
| *** yee379 has joined #openstack-kolla | 10:23 | |
| *** Pavo has quit IRC | 10:29 | |
| *** Pavo has joined #openstack-kolla | 10:34 | |
| *** klindgren_ has joined #openstack-kolla | 10:37 | |
| *** klindgren__ has quit IRC | 10:39 | |
| *** n0isyn0i1e is now known as n0isyn0ise | 11:42 | |
| *** tonanhngo has joined #openstack-kolla | 12:02 | |
| *** tonanhngo has quit IRC | 12:04 | |
| *** msimonin has joined #openstack-kolla | 12:06 | |
| *** msimonin has quit IRC | 12:06 | |
| *** fragatina has quit IRC | 12:12 | |
| *** fragatina has joined #openstack-kolla | 12:13 | |
| *** zhubingbing has joined #openstack-kolla | 12:14 | |
| *** awiddersheim has joined #openstack-kolla | 12:16 | |
| *** fragatina has quit IRC | 12:19 | |
| *** fragatina has joined #openstack-kolla | 12:27 | |
| *** Pavo has quit IRC | 12:29 | |
| *** Pavo has joined #openstack-kolla | 12:34 | |
| openstackgerrit | Wei Cao proposed openstack/kolla-ansible: Add freezer ansible role https://review.openstack.org/402224 | 12:36 |
| openstackgerrit | Wei Cao proposed openstack/kolla-ansible: Add freezer ansible role https://review.openstack.org/402224 | 12:38 |
| *** sgordon has quit IRC | 12:45 | |
| *** sgordon has joined #openstack-kolla | 12:45 | |
| *** cinerama has quit IRC | 12:51 | |
| *** slagle has quit IRC | 12:51 | |
| *** portdirect has quit IRC | 12:51 | |
| *** kklimonda has quit IRC | 12:51 | |
| *** Marx314 has quit IRC | 12:51 | |
| *** wznoinsk has quit IRC | 12:51 | |
| *** esmiurium has quit IRC | 12:51 | |
| *** clayton has quit IRC | 12:51 | |
| *** cinerama has joined #openstack-kolla | 12:51 | |
| *** slagle has joined #openstack-kolla | 12:51 | |
| *** portdirect has joined #openstack-kolla | 12:51 | |
| *** kklimonda has joined #openstack-kolla | 12:51 | |
| *** Marx314 has joined #openstack-kolla | 12:51 | |
| *** wznoinsk has joined #openstack-kolla | 12:51 | |
| *** esmiurium has joined #openstack-kolla | 12:51 | |
| *** clayton has joined #openstack-kolla | 12:51 | |
| *** ipsecguy has quit IRC | 13:18 | |
| *** yatin_on_leave has quit IRC | 13:35 | |
| *** yatin_on_leave has joined #openstack-kolla | 13:49 | |
| openstackgerrit | zhubingbing proposed openstack/kolla-ansible: add panko role https://review.openstack.org/400122 | 13:52 |
| kfox1111 | sbezverk_: unfortunately not. | 13:57 |
| kfox1111 | I did find an issue with the blocker patch. it doesn't fail the job when it fails. so thats part of the issue. | 13:58 |
| *** dave-mccowan has joined #openstack-kolla | 13:59 | |
| kfox1111 | once that's resolved it should at least fail faster. | 14:00 |
| kfox1111 | sdake: you around? | 14:01 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 14:09 |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible: Enable watcher dashboard when enable_watcher is yes https://review.openstack.org/406608 | 14:13 |
| openstackgerrit | caoyuan proposed openstack/kolla-ansible: Enable mistral dashboard when enable_mistral is yes https://review.openstack.org/406728 | 14:14 |
| *** jascott1_ has quit IRC | 14:20 | |
| *** Pavo has quit IRC | 14:29 | |
| *** Pavo has joined #openstack-kolla | 14:34 | |
| openstackgerrit | caoyuan proposed openstack/kolla: Install solum dashboard into horizon image https://review.openstack.org/408981 | 14:37 |
| *** sdake has joined #openstack-kolla | 14:44 | |
| kfox1111 | sdake: morning | 14:45 |
| sdake | morning kfox1111 | 14:45 |
| srwilkers | morning kfox1111 sdake | 14:57 |
| kfox1111 | srwilkers: morning | 15:01 |
| *** jgriffith_AutoAw is now known as jgriffith | 15:01 | |
| *** jgriffith is now known as jgriffith_AutoAw | 15:02 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 15:02 |
| *** jgriffith_AutoAw is now known as jgriffith | 15:02 | |
| *** tonanhngo has joined #openstack-kolla | 15:03 | |
| sbezverk_ | morning gents | 15:03 |
| *** jtriley has joined #openstack-kolla | 15:05 | |
| *** tonanhngo has quit IRC | 15:05 | |
| *** jascott1 has joined #openstack-kolla | 15:06 | |
| *** fragatina has quit IRC | 15:12 | |
| *** fragatina has joined #openstack-kolla | 15:13 | |
| *** jascott1_ has joined #openstack-kolla | 15:21 | |
| kfox1111 | sbezverk_: morning. :) | 15:21 |
| *** adrian_otto has joined #openstack-kolla | 15:25 | |
| *** jascott1_ has quit IRC | 15:25 | |
| *** sp__ has joined #openstack-kolla | 15:26 | |
| kfox1111 | this is interesting: http://logs.openstack.org/63/408263/15/check/gate-kolla-kubernetes-deploy-ubuntu-binary-ceph-nv/c4db305/logs/pods/kube-system-kube-dns-654381707-dd954-kube-dns.txt | 15:28 |
| sbezverk_ | kfox1111: is that from health container that comes with dns? | 15:29 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 15:29 |
| *** jtriley has quit IRC | 15:33 | |
| *** ipsecguy has joined #openstack-kolla | 15:35 | |
| *** portdirect_ has quit IRC | 15:37 | |
| *** ipsecguy has quit IRC | 15:37 | |
| *** ipsecguy has joined #openstack-kolla | 15:37 | |
| kfox1111 | sbezverk_: here's the whole logs: http://logs.openstack.org/63/408263/15/check/gate-kolla-kubernetes-deploy-ubuntu-binary-ceph-nv/c4db305/logs/pods/ | 15:38 |
| kfox1111 | the dnsmasq and helthz containers have nothing in their logs. | 15:39 |
| kfox1111 | sdake: you alive? | 15:43 |
| *** adrian_otto has quit IRC | 15:46 | |
| *** Jeffrey4l has quit IRC | 15:48 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 16:00 |
| *** ipsecguy has quit IRC | 16:02 | |
| *** ipsecguy has joined #openstack-kolla | 16:03 | |
| portdirect | afternoon o/ :) | 16:05 |
| kfox1111 | portdirect: afternoon. :) | 16:05 |
| *** Jeffrey4l has joined #openstack-kolla | 16:07 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: test the parent review minus a few bits. https://review.openstack.org/409455 | 16:07 |
| portdirect | hey man :) - just reseting the mainframe - will be back in 5 | 16:08 |
| srwilkers | hey portdirect o/ | 16:10 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 16:10 |
| sdake | kfox1111 i am alive | 16:13 |
| sdake | portdirect alive? | 16:14 |
| *** jgriffith is now known as jgriffith_AutoAw | 16:14 | |
| kfox1111 | sdake: ran into an issue on the hub. | 16:15 |
| kfox1111 | kolla-toolbox 3.0.1 has a bad openstack client in it. | 16:15 |
| kfox1111 | is there a 3.0.2 soon, or should we try and update the 3.0.1 one with a fresher version? | 16:16 |
| sdake | kfox1111 no idea i don't maintain the hub | 16:16 |
| sdake | not sure who does | 16:16 |
| sdake | kfox1111 can you see your pm please ;) | 16:16 |
| *** Kdecherf has quit IRC | 16:17 | |
| *** Kdecherf has joined #openstack-kolla | 16:18 | |
| *** jgriffith_AutoAw is now known as jgriffith | 16:21 | |
| *** jascott1_ has joined #openstack-kolla | 16:22 | |
| kfox1111 | sdake: ping | 16:25 |
| *** jascott1_ has quit IRC | 16:27 | |
| *** Pavo has quit IRC | 16:29 | |
| *** jgriffith is now known as jgriffith_AutoAw | 16:34 | |
| *** Pavo has joined #openstack-kolla | 16:34 | |
| *** ipsecguy has quit IRC | 16:35 | |
| sdake | srwilkers pingola | 16:35 |
| *** ipsecguy has joined #openstack-kolla | 16:35 | |
| srwilkers | hey sdake :) | 16:38 |
| sdake | srwilkers see pm plz | 16:39 |
| *** Serlex has joined #openstack-kolla | 16:40 | |
| portdirect | config-map rollout poc: https://github.com/kargakis/configmap-rollout | 16:44 |
| SamYaple | w00t | 16:44 |
| *** jascott1_ has joined #openstack-kolla | 16:44 | |
| SamYaple | got the ansible docker module converted to salt in a rough manner | 16:44 |
| portdirect | :) | 16:44 |
| SamYaple | it will eventually be mostly a rewrite to take advantage of salt things like __context__ | 16:45 |
| SamYaple | but for now, I am starting kolla containers with salt | 16:45 |
| SamYaple | im pretty excited | 16:46 |
| SamYaple | its looking clean | 16:46 |
| SamYaple | and _quick_ | 16:46 |
| srwilkers | thanks for that link portdirect :) | 16:47 |
| SamYaple | 2 seconds to run a container "upgrade" (remove/recreate) for memcache container | 16:47 |
| *** srwilkers is now known as srwilkers_away | 16:47 | |
| SamYaple | not a fan of the output from salt though. but eh | 16:48 |
| *** jascott1_ has quit IRC | 16:49 | |
| *** Serlex has quit IRC | 16:51 | |
| *** msimonin has joined #openstack-kolla | 17:05 | |
| *** msimonin has quit IRC | 17:06 | |
| *** msimonin has joined #openstack-kolla | 17:07 | |
| kfox1111 | oh. interesting... | 17:09 |
| *** msimonin has quit IRC | 17:09 | |
| kfox1111 | https://review.openstack.org/#/c/409455/ is still running... | 17:09 |
| kfox1111 | so, maybe not libvirt/nova-compute related. | 17:09 |
| kfox1111 | I would have really guessed those where the issue. | 17:10 |
| kfox1111 | hmm.. what else could it be... | 17:10 |
| *** msimonin has joined #openstack-kolla | 17:12 | |
| *** msimonin has quit IRC | 17:13 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: test the parent review minus a few bits. https://review.openstack.org/409455 | 17:15 |
| *** jtriley has joined #openstack-kolla | 17:18 | |
| *** rstarmer has joined #openstack-kolla | 17:21 | |
| *** dave-mccowan has quit IRC | 17:23 | |
| *** rstarmer has quit IRC | 17:26 | |
| *** rstarmer has joined #openstack-kolla | 17:26 | |
| *** rstarmer has quit IRC | 17:26 | |
| *** rstarmer has joined #openstack-kolla | 17:27 | |
| *** rstarmer has quit IRC | 17:27 | |
| *** rstarmer has joined #openstack-kolla | 17:28 | |
| *** rstarmer has quit IRC | 17:28 | |
| *** rstarmer has joined #openstack-kolla | 17:28 | |
| *** rstarmer has quit IRC | 17:29 | |
| *** rstarmer has joined #openstack-kolla | 17:29 | |
| *** rstarmer has quit IRC | 17:30 | |
| *** sdake has quit IRC | 17:31 | |
| *** sdake has joined #openstack-kolla | 17:33 | |
| sdake | folks - here is the repo reorganization for kolla-kubernetes and kolla-ansible's benefit to unify the default config: https://review.openstack.org/#/c/399147/ | 17:38 |
| *** sp__ has quit IRC | 17:40 | |
| *** jgriffith_AutoAw is now known as jgriffith | 17:41 | |
| SamYaple | sdake: i review that, that seems like a very bad idea | 17:50 |
| SamYaple | those configs are very much ansible configs | 17:50 |
| kfox1111 | sdake: thanks for the link. | 17:51 |
| kfox1111 | SamYaple: I'm less convinced. they use vars defined by ansible, but arn't super ansible specific. | 17:51 |
| SamYaple | kfox1111: lol half of them say ansible | 17:52 |
| SamYaple | hostvars[inventory_hostname]['ansible_'..... | 17:52 |
| kfox1111 | I think with just a little bit of reworking, the ansible specific logic can be put in ansible based var defines, and then the templates themselves can be used for whatever. | 17:52 |
| kfox1111 | SamYaple: yeah. and if that was replaced with a more generic var name, that should work. | 17:53 |
| kfox1111 | and then have foo = hostvars[inventory_hostname]['ansible_'..... | 17:53 |
| SamYaple | i dont know. that seems not very useful. having all kolla-* projects have to work in changes that might be needed for only kolla-ansible or kolla-k8s defeats the point of spliting the repos | 17:53 |
| kfox1111 | we've conditionaled around most of them already for kolla-kubernetes anyway. | 17:53 |
| SamYaple | this also assume jinja2 templating | 17:54 |
| kfox1111 | this would make it so that kolla kubernetes wouldn't need to depend on kolla-ansible. | 17:54 |
| kfox1111 | yeah. jinja2 templating it a thing. though I don't think too much of jinja2 specific features are in there. | 17:54 |
| SamYaple | which is just silly. copy the configs out. don't add in kolla-k8s stuff into kolla-ansible configs | 17:55 |
| kfox1111 | it could pretty easily be sed'ed into a helm compatable format too I think. | 17:55 |
| SamYaple | trying to make a generic set of configs that exist in a repo that builds containers is very strange to me | 17:55 |
| kfox1111 | s/{$/ {{ ./ etc. | 17:56 |
| SamYaple | shoehorning is a term that might fit | 17:56 |
| kfox1111 | well, its called kolla, not kolla-containers. | 17:56 |
| kfox1111 | if it was called kolla-containers I'd agree. :) | 17:56 |
| SamYaple | the whole reason for the split was to remove the deployment stuff from kolla main repo | 17:56 |
| kfox1111 | (and it might be nice to have that anyway. but I know folks don't want yet another repo split) | 17:56 |
| SamYaple | configs are definetely deployment | 17:56 |
| kfox1111 | we disagree there. | 17:56 |
| kfox1111 | I see there bing 4 pieces to the puzzle. | 17:57 |
| kfox1111 | config generation, orchestration, deployment, and software. | 17:57 |
| kfox1111 | ansible is currently doing 3 of the 4. | 17:57 |
| kfox1111 | and sofware is the containers in kolla. | 17:57 |
| SamYaple | ok. fair enough. i think thats correct then | 17:58 |
| SamYaple | i suppose i dont care so much as long as its not going to be a requirement for other kolla-* projects to use the configs from kolla | 17:58 |
| kfox1111 | yeah. there to use if you want to. but not required. | 17:58 |
| SamYaple | because that is going to lead to alot of headache for kolla-k8s and kolla-ansible | 17:58 |
| SamYaple | i assure you | 17:58 |
| kfox1111 | thats why I like splitting it up into those 4 catagories. | 17:58 |
| kfox1111 | so in kolla-kubernetes, you can use the kolla provided genconfig stuff, or not. | 17:59 |
| kfox1111 | if you want to say, hand roll your own configs. | 17:59 |
| SamYaple | its going to suck for kolla-k8s when ansible changes a config for something that works for ansible, but not for k8s | 17:59 |
| kfox1111 | SamYaple: thats why gating is important. | 17:59 |
| kfox1111 | config has to be gated against those things that use it. | 17:59 |
| SamYaple | gating is important, and it is never 100% | 17:59 |
| kfox1111 | so ansible won't be able to break kolla-kubernetes without someone knowing. | 17:59 |
| kfox1111 | and visa versa. | 18:00 |
| SamYaple | im thinking about for loops over inventories that subtly break that the gate just wont catch | 18:00 |
| kfox1111 | but right now, kolla-ansible owns genconfig and it breaks kolla-kubernetes. :/ | 18:00 |
| SamYaple | i would bet my salary that is going to happen | 18:00 |
| SamYaple | so move genconfig into kolla-k8s as it should be | 18:00 |
| SamYaple | this is a very silly change | 18:00 |
| kfox1111 | sharing code has advantages and disadvantages. | 18:01 |
| SamYaple | im not convinved this has advantages or is even less work | 18:01 |
| kfox1111 | sharing is usuaaly better I think when there aren't a lot of devs to deal with fork maintenence. | 18:01 |
| kfox1111 | its more costly in dev time to have 2 seperate copies. | 18:02 |
| kfox1111 | dvr support for exampl,e requires two sets of changes then. | 18:02 |
| SamYaple | maybe. until the first time ansible breaks k8s subtly | 18:02 |
| kfox1111 | thats the risk. or the other way around. | 18:02 |
| SamYaple | and then its far more work | 18:02 |
| kfox1111 | but thats why tests are so important. anything, not just config stuff can break things. | 18:03 |
| SamYaple | besides, these were always supposed to be bare-bone configs | 18:03 |
| SamYaple | where user supplied any overrides or other thigsn. so it shouldnt take any dev time to enable dvr | 18:03 |
| *** pbourke has quit IRC | 18:03 | |
| SamYaple | thats user config stuff where then need to consult neutron on how dvr works | 18:03 |
| *** pbourke has joined #openstack-kolla | 18:03 | |
| kfox1111 | there nees to be logic though that configures certain overrides per host role. so its not quite so easy. :/ | 18:04 |
| kfox1111 | its not just "enable_dvr=true" :/ | 18:04 |
| SamYaple | but that exists though | 18:04 |
| kfox1111 | only because the genconfig was patched to support it. | 18:05 |
| kfox1111 | and that code woiuld be nice to share. | 18:05 |
| SamYaple | no no no, this is an ansible specifc thing | 18:05 |
| SamYaple | it looks up a file on the host in question and can merge configs | 18:05 |
| SamYaple | k8s not doing that unless they are deploying with ansible | 18:05 |
| SamYaple | which to be honest, i dont know if you are | 18:05 |
| kfox1111 | why is merging config's a ansible specific feature? | 18:05 |
| SamYaple | the way its implemented in kolla-ansible | 18:06 |
| kfox1111 | today it was implemented in a way that only works with ansible, which is a pain. :/ | 18:06 |
| kfox1111 | I'd like the genconfig to a little more generically handle config merging. | 18:06 |
| SamYaple | this is a change that is going to cause more headaches than solutions in my opinion. no to mention dirty kollas git history with unrelated things. | 18:07 |
| SamYaple | but if its not a requirement and only going to be used in those projects, i guess its not a big deal | 18:07 |
| SamYaple | making this a requirement for future kolla-* projects i am 100% against | 18:07 |
| kfox1111 | kollas rhistroy is alrady pretty durty I think due to the repo split. :/ | 18:08 |
| kfox1111 | I'm with you. its optional to use genconfig. | 18:08 |
| kfox1111 | I want it optional in kolla-kubernetes too. | 18:08 |
| kfox1111 | you can use genconfig to do your config for kolla-kubernets, or some other way if you want. | 18:08 |
| SamYaple | esh what a mess this is going to be for people | 18:09 |
| kfox1111 | heh. yeah. it would be easier if we just stuck to one technology. | 18:09 |
| SamYaple | no forget that, thats what the split was about | 18:10 |
| kfox1111 | but for now, supporting both kolla-ansible and kolla-kubernetes makes a lot of sense. | 18:10 |
| SamYaple | so now we can use different technologies that have no dependancies on each other | 18:10 |
| kfox1111 | maybe it makes sense to suport it forever. I dont konw. | 18:10 |
| kfox1111 | right. | 18:10 |
| SamYaple | but this is introducing a dependacy | 18:10 |
| kfox1111 | optional dependency is ok. | 18:11 |
| SamYaple | its not option for kolla-ansible | 18:11 |
| kfox1111 | kolla-ansible already depends on kolla containers. | 18:11 |
| SamYaple | not the repo though | 18:11 |
| SamYaple | and the containeres a a closed box | 18:11 |
| SamYaple | an api almost | 18:11 |
| SamYaple | this is certainly not ansible best practice either | 18:11 |
| SamYaple | template files belong in the roles | 18:11 |
| *** jgriffith is now known as jgriffith_AutoAw | 18:12 | |
| * kfox1111 shrugs | 18:12 | |
| kfox1111 | best practes are guidelines. | 18:12 |
| kfox1111 | should be followed most of the time. | 18:12 |
| kfox1111 | but sometimes exceptions are good when it actualy benifits a project. | 18:12 |
| kfox1111 | the benifits need to outweigh the drawbacks. | 18:13 |
| SamYaple | we are just disagreeing on benefits here. i see complication and problems | 18:13 |
| kfox1111 | yup. difference in opinion. thats ok. | 18:14 |
| SamYaple | just for my own clarity, what is the issue with kolla-k8s having thier own configs? | 18:14 |
| kfox1111 | thats why the community votes on things. to try and get a common direction rather then one opinion driving too much. | 18:14 |
| kfox1111 | SamYaple: maintenence. | 18:15 |
| SamYaple | you think all the work and complication surrounding this will be less dev time than maintaining configs? | 18:15 |
| kfox1111 | we have very few devs, and if we have to maintain a whole set of configs, its going to be more work then it solves trying to share in my opinion. | 18:15 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: test the parent review minus a few bits. https://review.openstack.org/409455 | 18:15 |
| SamYaple | wow. no. i would strongly disagree there. maintaining these configs takes almost no time at all | 18:16 |
| SamYaple | i think the liberty -> mitaka config refresh i did took 2 hours | 18:16 |
| kfox1111 | across all projects? | 18:17 |
| SamYaple | yes | 18:17 |
| kfox1111 | it will also cause more schism where kolla-kubernetes or kolla-ansible gets a patch to add support for a new project and the other one doesn't. | 18:17 |
| kfox1111 | the more work to do, the less likely that it will get done on the other side. | 18:17 |
| SamYaple | i mean thats a strange assumption that ansible and kubernetes projects will always support the same things | 18:18 |
| SamYaple | weve had container images for projects ansible didnt support since the bengingin | 18:18 |
| kfox1111 | true. just trying to say, we should maek the bar as low as possible. | 18:18 |
| kfox1111 | true. true. | 18:19 |
| SamYaple | i agree there, i think this is devs deving and overcomplicating though. making the bar much higher | 18:19 |
| SamYaple | it certainly makes the complexity of the project higher | 18:19 |
| kfox1111 | maybe. | 18:19 |
| kfox1111 | I think it may actually clean things up in the end. but thats juts my opinion. | 18:19 |
| sbezverk_ | kfox1111: do you recall what rpm name is for hq bits? | 18:20 |
| kfox1111 | it will be more clear reading throug hthe templates that dont have hostvars[ansible].... stuff in it. | 18:20 |
| kfox1111 | sbezverk_: hq? | 18:20 |
| SamYaple | kfox1111: i agree. we can have different opinions. since we agree on my biggest point (this isn't a *requirement*) im perfectly fine | 18:20 |
| sbezverk_ | json tool | 18:20 |
| kfox1111 | SamYaple: yup. :) | 18:20 |
| kfox1111 | sbezverk_: oh. jq you mean? | 18:21 |
| kfox1111 | sbezverk_: the rpm package is 'jq' | 18:21 |
| kfox1111 | I think it comes from epel. | 18:21 |
| sbezverk_ | kfox1111: that the one, no wonder I could not find it :-) | 18:21 |
| *** zhubingbing has quit IRC | 18:27 | |
| kfox1111 | weird... | 18:28 |
| kfox1111 | the ubuntu gate job is failing often on the openstackclient fix... | 18:28 |
| kfox1111 | while the others arn't. | 18:28 |
| kfox1111 | http://logs.openstack.org/63/408263/18/check/gate-kolla-kubernetes-deploy-ubuntu-binary-ceph-nv/4022901/ | 18:29 |
| *** Pavo has quit IRC | 18:29 | |
| sdake | SamYaple inc0 proposed the spec, i was providing it for people that asked for it | 18:29 |
| kfox1111 | oh. wierd. | 18:30 |
| kfox1111 | socket.gaierror: [Errno -2] Name or service not known | 18:30 |
| kfox1111 | 2016-12-10 16:16:11.950219 | + '[' 0 -eq 0 ']' | 18:30 |
| kfox1111 | that should have trhown an error. | 18:30 |
| *** Jeffrey4l_ has joined #openstack-kolla | 18:30 | |
| *** Jeffrey4l has quit IRC | 18:32 | |
| *** Pavo has joined #openstack-kolla | 18:34 | |
| *** jtriley has quit IRC | 18:35 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 18:39 |
| portdirect | kfox1111: re, http://logs.openstack.org/63/408263/18/check/gate-kolla-kubernetes-deploy-ubuntu-binary-ceph-nv/4022901/logs/pods/kolla-neutron-create-keystone-endpoint-internal-j2yz2-main.txt | 18:44 |
| portdirect | are we on the wrong track here? pip's freaking out over: "requests.packages.urllib3.connection.VerifiedHTTPSConnection" | 18:45 |
| portdirect | is there a proxy infront of some vm's in zuul? | 18:45 |
| *** jascott1_ has joined #openstack-kolla | 18:46 | |
| kfox1111 | portdirect: it looks like my dns test still failed to block on that review... | 18:50 |
| kfox1111 | still tryign to fix that part. | 18:50 |
| kfox1111 | portdirect: see: http://logs.openstack.org/63/408263/18/check/gate-kolla-kubernetes-deploy-ubuntu-binary-ceph-nv/4022901/console.html#_2016-12-10_16_16_11_950208 | 18:51 |
| kfox1111 | looks very weird. | 18:51 |
| *** jascott1_ has quit IRC | 18:52 | |
| portdirect | yeah - i think that was a brain squeltch re the proxy | 18:52 |
| portdirect | (ok it WAS a brain squeltch) | 18:52 |
| portdirect | but wieird to get that far before cropping up again? | 18:54 |
| *** jgriffith_AutoAw is now known as jgriffith | 18:55 | |
| kfox1111 | it should have not passed that [ $? -eq 0 ] | 18:56 |
| kfox1111 | not sure how it managed to do that. | 18:56 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: test the parent review minus a few bits. https://review.openstack.org/409455 | 18:57 |
| *** jascott1_ has joined #openstack-kolla | 19:13 | |
| kfox1111 | portdirect: we were kind of talking about it before, but lost track of it. if there's not a way to export a docker container with layers intact, think there's a way to build a docker-registry preloaded with containers? | 19:15 |
| portdirect | kfox1111: possibly, never tried but (and bear with me, 'cause this is gonna sound ugly, and just off the top of my head): | 19:16 |
| portdirect | 1. registry in container | 19:17 |
| portdirect | 2. load images into it | 19:17 |
| portdirect | 3 export that | 19:17 |
| portdirect | 4. profit? | 19:17 |
| portdirect | may be worth looking at what fuel used to do, as i think it did somthing like what yur thinking about | 19:18 |
| kfox1111 | yeah. thats kind of what I was thinking. | 19:20 |
| kfox1111 | k. I'll see what I can dig up. thanks for the tip. | 19:21 |
| *** jgriffith is now known as jgriffith_AutoAw | 19:26 | |
| kfox1111 | portdirect: yeah, I'm kind of using the neutron endpoint review to continue to debug the dns issue. | 19:26 |
| portdirect | figured :) | 19:27 |
| kfox1111 | as it seems to tickle it frequently, due to the pip upgrade workaround in it. | 19:27 |
| kfox1111 | its really weird how much its hitting ubuntu relative to centos. | 19:28 |
| portdirect | lol, so it turns out that trusting me is a bad move | 19:28 |
| portdirect | docker save does what you want i think | 19:28 |
| portdirect | just tested it and it's leaving the layers etc intact | 19:28 |
| kfox1111 | can you do multiple saves? | 19:29 |
| portdirect | 2 sec and I'll try | 19:29 |
| *** awiddersheim has quit IRC | 19:29 | |
| kfox1111 | like, what i'm trying to do is cache the downloads for the gate. | 19:29 |
| kfox1111 | but if it ends up blowing up the number of layers, the amount of download might cancel out any benifits from the cache. | 19:29 |
| kfox1111 | the man page has IMAGE [IMAGE...] | 19:31 |
| kfox1111 | so maybe that would work. | 19:31 |
| portdirect | https://paste.fedoraproject.org/503480/81398260/raw/ | 19:31 |
| kfox1111 | looks good. :) | 19:32 |
| portdirect | see if you can spot the craig wright like mistake :) but I can assure you it was unintentional :) | 19:32 |
| kfox1111 | hmm.. they don't seem to share any layers though? | 19:33 |
| portdirect | thats cause the 2 images i chose dont | 19:33 |
| portdirect | will check with some that do | 19:34 |
| kfox1111 | k. thanks. :) | 19:34 |
| *** jmccarthy has quit IRC | 19:36 | |
| *** msimonin has joined #openstack-kolla | 19:36 | |
| *** jmccarthy has joined #openstack-kolla | 19:37 | |
| kfox1111 | hmm..... | 19:38 |
| kfox1111 | so the nova pause seems to be db related maybe... | 19:38 |
| portdirect | kfox1111: seems good: https://paste.fedoraproject.org/503498/13987511/raw/ | 19:39 |
| portdirect | still not 100% sold on it - but worth a shot | 19:40 |
| kfox1111 | portdirect: yeah. looks good to me. thanks for testing. :) | 19:40 |
| kfox1111 | should be able to precache things pretty easily then I think. | 19:41 |
| kfox1111 | even with trunk. | 19:41 |
| kfox1111 | just gota make a simple job that docker-save's a tarball periodically, and for trunk, does a build and does the export. | 19:41 |
| *** awiddersheim has joined #openstack-kolla | 19:43 | |
| *** jtriley has joined #openstack-kolla | 19:46 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: test the parent review minus a few bits. https://review.openstack.org/409455 | 19:48 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 19:51 |
| *** jtriley has quit IRC | 19:55 | |
| kfox1111 | portdirect: kind of lost track of your ps. | 19:55 |
| kfox1111 | anything you need help with? | 19:55 |
| portdirect | Think I'm ok on thekeystone side - that just needs a shake up to take into account the suff thats dropped since | 19:56 |
| portdirect | but if you want ceph, its all yours | 19:56 |
| portdirect | I'm trying to catch up on a few things i should have done last week for work - but was gonna hit em hard tomorrow so that on monday it should be good to merge | 19:58 |
| kfox1111 | stil working on neutron. was just looking around to help out inflight reviews if help was needed. | 19:58 |
| kfox1111 | ok. cool. | 19:58 |
| portdirect | np :) cheers dude | 19:58 |
| kfox1111 | do you konw if kubeadm supports specifying a repo to pull from? | 19:58 |
| kfox1111 | or an existing issue for it? | 19:58 |
| portdirect | not that I'm aware of | 19:59 |
| kfox1111 | k. I might raise one then. | 19:59 |
| kfox1111 | without that kind of option, it will be hard for some folks to deply in an enterprise. | 19:59 |
| kfox1111 | kind of similar to this though: https://github.com/kubernetes/kubeadm/issues/34 | 20:00 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: test the parent review minus a few bits. https://review.openstack.org/409455 | 20:10 |
| SamYaple | you know kfox1111, to make my case just a bit more on these configs I was just packaging rabbitmq into the kolla-salt thing im doing | 20:11 |
| SamYaple | and kolla-ansible doesnt use seperate vhosts per service | 20:11 |
| SamYaple | this means kolla-kubernetes can't either | 20:11 |
| SamYaple | nor could anything else using those configs | 20:12 |
| kfox1111 | SamYaple: that may be a good reason though to share? kolla-ansible probably should support vhosts. so would benifit from the effort too? | 20:12 |
| kfox1111 | kolla-kubernetes is in the same boat. | 20:12 |
| kfox1111 | we probably want to support vhosts too at some point. | 20:12 |
| SamYaple | aha! but thats a problem isnt it? thats not use something that can be easily added | 20:12 |
| SamYaple | migration and other things must occur | 20:12 |
| SamYaple | but... only for kolla-ansible | 20:13 |
| kfox1111 | it won't break existing things. its an additional feature. | 20:13 |
| SamYaple | you lose queues | 20:13 |
| kfox1111 | so should be able to add without breaking existing things? | 20:13 |
| SamYaple | its not an additional feature | 20:13 |
| kfox1111 | just default to the non vhost way? | 20:13 |
| SamYaple | everything has a vhost | 20:13 |
| SamYaple | always | 20:13 |
| SamYaple | in this case though its the same vhost accross all services | 20:13 |
| kfox1111 | k. just speaking generically. | 20:13 |
| SamYaple | it need not be, and there is some security in not having it that way | 20:14 |
| SamYaple | my point is, now kolla-k8s is bound to kolla-ansible with whether they should use vhosts or not | 20:14 |
| SamYaple | one can't make the decision without the other | 20:14 |
| kfox1111 | right. but if kolla-kubernetes contributed optional support for vhosts, then kolla-ansible's path to support is easier. | 20:15 |
| kfox1111 | as they won't have to do all the config bits. just the ansible side? | 20:15 |
| SamYaple | that makes the configs _way_ more complicated though | 20:15 |
| SamYaple | now you support vhost and not vhost | 20:15 |
| kfox1111 | possibly? you said, its always vhost. we just set the per service thing to the same default for kolla-ansible, and then make it per service overridable? | 20:16 |
| SamYaple | nah its rabbitmq. its never simple | 20:16 |
| portdirect | ^^ truth | 20:16 |
| portdirect | :) | 20:16 |
| SamYaple | but the point is, there is going to be if/else branches in all of that that make the configs nigh unreadable very quickly | 20:16 |
| * kfox1111 shrugs. kolla-kubernete's will probably solve it yet a different way. by one rabbit per service. | 20:16 | |
| SamYaple | just because you are trying to shoehorn the configs together | 20:16 |
| kfox1111 | SamYaple: yeah. aleays a risk. | 20:17 |
| SamYaple | one rabbitmq per service.. hmm | 20:17 |
| kfox1111 | just so far, it hasn't taken much to get kolla-kubernetes off the ground using kolla-ansibles genconfig. | 20:17 |
| SamYaple | thatll break openstack in some interesting ways | 20:17 |
| kfox1111 | so, so far, little risk of lots of conditionals. | 20:17 |
| kfox1111 | its already shared. | 20:17 |
| portdirect | yeah metering will become fun | 20:17 |
| kfox1111 | SamYaple: we run mujltiple rabbits all the time at work. it works ok. | 20:17 |
| kfox1111 | SamYaple: I know of other sites doing it too. | 20:18 |
| SamYaple | oh im not worried about the not working part | 20:18 |
| kfox1111 | SamYaple: cross servids don't need to talk between rabbits except notifications, | 20:18 |
| SamYaple | its the when rabbitmq split brains... but only for glance | 20:18 |
| SamYaple | that kind of thing | 20:18 |
| kfox1111 | and notifications can speak to yet another rabbit. | 20:18 |
| SamYaple | yea | 20:18 |
| kfox1111 | we have a dedicated rabbit cluster just for ceilometer. | 20:18 |
| kfox1111 | for scaling reasons. | 20:18 |
| portdirect | but being serious we need to think abou this issue and get all parites on board, Sam's raising some very good points - and as kolla-k8s becomes more developed we will start to place really heavy demands on a shared config | 20:18 |
| SamYaple | yep. certainly has a value add there | 20:18 |
| kfox1111 | portdirect: yup. I agree. it may make sense to split it at some point. just so far, it hasn't seemed different enough to justify carying a near identicle copy I think. | 20:19 |
| kfox1111 | but as we get further into the helm/servicepackages, maybe that changes. | 20:20 |
| kfox1111 | but splitting them seems premature to me. | 20:20 |
| kfox1111 | the same could be said of the kolla containers. we could make them kolla-kubernetes specific versions (I'm not advocating this at all) | 20:20 |
| portdirect | yeah - I'm not saying we should split now - but we should not create a complexity minefeild for ourselfs when we hit the wall | 20:21 |
| kfox1111 | but sharing has benifit until it becomes too much of a problem. | 20:21 |
| SamYaple | i disagree. yo ucant say that about the contaienrs | 20:21 |
| kfox1111 | portdirect: +1 | 20:21 |
| SamYaple | the containers have a single point of entry | 20:21 |
| kfox1111 | SamYaple: sure you can. | 20:21 |
| kfox1111 | the containers were designed for ansible deployment. they could in theory designed differently for k8s deployment. | 20:21 |
| SamYaple | they were not designed for ansible deployment | 20:21 |
| SamYaple | you are wrong | 20:21 |
| kfox1111 | but, close enough for now. (and hopefully forever) | 20:21 |
| SamYaple | they were infact designed to be deployment tool agnostic | 20:22 |
| SamYaple | i know, because thats waht i wrote them as | 20:22 |
| kfox1111 | sure. but stuff like kubernetes-entrypoint wern't things at the time. | 20:22 |
| kfox1111 | not saying you didn't do the right thing. I like the way the containers were built. | 20:23 |
| *** msimonin has quit IRC | 20:23 | |
| kfox1111 | just saying. yuou can make the share/not share argument about any part of the system. | 20:23 |
| SamYaple | eh, i dont agree with the kubernetes-entrypoint thing personally since the extend start stuff exists to filel that | 20:23 |
| SamYaple | you can entrypoint from the deployment tool itself | 20:23 |
| SamYaple | or so was the original design | 20:23 |
| SamYaple | its what kolla-mesos did | 20:23 |
| kfox1111 | SamYaple: I agree with you, but just saying. that kind of things will be argued by folks. | 20:24 |
| kfox1111 | SamYaple: we happened to find a solution with a beta level k8s feature to make it not need to change the images. | 20:24 |
| portdirect | Yeah - i think the containers are fine as they stand, the great thing about k8s it it lets us build round them (eg run ku8s-entrypoint) in an init container | 20:24 |
| kfox1111 | but if it weren't for that, the containers would have to be changed in an kubernetes-entrypoint world. | 20:24 |
| SamYaple | kfox1111: that solution exists, you need to have an extend start that you copy into the container to override the default one | 20:24 |
| SamYaple | that was the original purpose of this | 20:25 |
| kfox1111 | I think we're getting off the rails. | 20:25 |
| SamYaple | i agree | 20:25 |
| kfox1111 | :) | 20:25 |
| SamYaple | but my point going back was the containers are 100% meant to be generic without any deployment specific code | 20:25 |
| kfox1111 | yeah. and I think the genconfig should do the same. | 20:26 |
| kfox1111 | but thats just my opinion. | 20:26 |
| kfox1111 | and I admit I could be wrong. :) | 20:27 |
| SamYaple | personally, i think genconfig stuff should be its own projects. like as a whole | 20:27 |
| kfox1111 | so far, it has worked out prety well. | 20:27 |
| kfox1111 | SamYaple: that, I agree with too. | 20:27 |
| SamYaple | im not saying a shared genconfig is wrong, im saying genconfig in kolla main repo is wrong | 20:27 |
| kfox1111 | +1 | 20:27 |
| kfox1111 | a kolla-genconfig would be a better fit I think too. | 20:28 |
| kfox1111 | but there's been a lot of pushback around extra repos. :/ | 20:28 |
| SamYaple | that could be tailored away from jinja2 | 20:28 |
| kfox1111 | I don't quite understand why. | 20:28 |
| SamYaple | or abstracted rather | 20:28 |
| kfox1111 | shoudl we propose a kolla-genconfig repo then in the spec? | 20:28 |
| SamYaple | idk. i think that deserves a new spec | 20:29 |
| SamYaple | it would have to discuss how to do it generically | 20:29 |
| *** Pavo has quit IRC | 20:29 | |
| SamYaple | so the issues like vhost between projects could be solved | 20:29 |
| kfox1111 | most of it woudl be the same though, whether it was in kolla repo or kolla-genconfig repo? | 20:29 |
| *** portdirect is now known as portdirect_away | 20:30 | |
| kfox1111 | the work's the genericness. | 20:30 |
| kfox1111 | not as much, where it lives. | 20:30 |
| SamYaple | idk. a seperate repo seems better, but im still not convinced | 20:31 |
| SamYaple | certainly worth exploring | 20:31 |
| SamYaple | im much more open to that | 20:31 |
| kfox1111 | kk | 20:31 |
| kfox1111 | I'm all for doing it in seperate repos. cramming everything into too few repos is something that I think has caused kolla pain over the years. | 20:32 |
| SamYaple | if our build process worked a bit different (less/no layering) i think seperate repos for each service would work well | 20:32 |
| SamYaple | but our build process is very much integrated | 20:33 |
| kfox1111 | true | 20:33 |
| SamYaple | i like the idea of having a user supplied repo that builds an image along side the rest | 20:33 |
| SamYaple | but thats another subject for another time | 20:33 |
| kfox1111 | ah. yeah, that would be cool. | 20:33 |
| *** Pavo has joined #openstack-kolla | 20:34 | |
| SamYaple | ive been working on fleshing out some ideas to help the building | 20:34 |
| SamYaple | including building a repo_build container that runs pypi server to feed the containers | 20:34 |
| SamYaple | so the containers dont ened build tools | 20:34 |
| kfox1111 | wow. is the gate really that empty? or is it just my view somehow? | 20:35 |
| *** sayantan_ has joined #openstack-kolla | 20:35 | |
| SamYaple | wow. its dead it seems | 20:35 |
| *** portdirect_away is now known as portdirect | 20:35 | |
| kfox1111 | yeah. I know some others want the thinner containers too. | 20:35 |
| portdirect | +2 | 20:35 |
| portdirect | :) | 20:35 |
| kfox1111 | portdirect: was it you that was working on super thin containers? | 20:35 |
| SamYaple | i actually want non-layered containers | 20:36 |
| portdirect | SamYaple, I was pretty much doing that for a bit | 20:36 |
| SamYaple | i know i pushed for the layers, and thats served us well, but i think we can get rid of them now | 20:36 |
| SamYaple | if we are clever | 20:36 |
| SamYaple | yea portdirect, the issue is build time mostly | 20:36 |
| SamYaple | but with a repo container, and some other magic we might get away with it | 20:36 |
| SamYaple | i really hate having gcc in the final image | 20:37 |
| SamYaple | alright. sleep time | 20:37 |
| portdirect | yep - was building using alpine as a base and some crazy scripts - but it put build time up a lot | 20:37 |
| portdirect | eg: https://github.com/portdirect/harbor/blob/latest/docker/openstack/openstack-keystone/openstack-keystone-alpine/Dockerfile | 20:37 |
| SamYaple | yea, when i first came to kolla the build time was like an hour | 20:37 |
| SamYaple | and that was with a fraction of the services we have now | 20:38 |
| SamYaple | of course it was also bash script building back in those days | 20:38 |
| SamYaple | we didnt get parallel until build.py | 20:38 |
| portdirect | (which I'm still using... thanks for your work on that :) ) | 20:38 |
| SamYaple | i actually wrote parallelism in a patchset for the bash script | 20:39 |
| kfox1111 | hmm... so after 3 minutes, kube-dns still wasn't working. | 20:39 |
| kfox1111 | and its happening fairly consistently on ubuntu right now.... | 20:39 |
| kfox1111 | how very strange. | 20:39 |
| SamYaple | it worked, but it was so awful and full of 'evals' that i said screw it and started the initial build.py script | 20:39 |
| SamYaple | and the community has just run with that thankfully | 20:39 |
| SamYaple | for real this time. sleep | 20:40 |
| SamYaple | thanks for the level-headed chat kfox1111 and portdirect | 20:40 |
| SamYaple | always up for those | 20:40 |
| kfox1111 | SamYaple: you too. :) | 20:40 |
| portdirect | laters man :) | 20:40 |
| kfox1111 | itsa always good to have conversaions when folks don't agree but can still be civil. :) | 20:41 |
| kfox1111 | I really try not to take disagreement personally. | 20:41 |
| kfox1111 | so.... why is dns failing... so strange... | 20:43 |
| kfox1111 | three minutes should be pleanty of time for things to settle.... | 20:43 |
| portdirect | its ubuntu only? | 20:43 |
| kfox1111 | at the moment, it seems that way. | 20:43 |
| kfox1111 | the last 10 runs or so has consistently landed that way. :/ | 20:43 |
| portdirect | can i run an exact copy of the gate locally? (I have hardware spare) | 20:44 |
| kfox1111 | will run it a bunch more times to see if Ican see any pattern. | 20:44 |
| kfox1111 | shoudl be able to. but never tested it. | 20:45 |
| portdirect | or just get the image that it uses | 20:45 |
| kfox1111 | I don't think setup_gate has anything too gate specific in it, | 20:45 |
| kfox1111 | other then assuming a complete fresh install is there on each run. | 20:45 |
| kfox1111 | I was going to tweak it to run in minikube at some point, but just havent had time. | 20:46 |
| kfox1111 | I know others are starting to play with the gate scripts in the vagrent scripts. | 20:46 |
| portdirect | yeah think think, jascott might be playing with that dicker container I started | 20:48 |
| portdirect | *docker | 20:49 |
| kfox1111 | ah. yeah. | 20:49 |
| portdirect | but wondering if its a CNI issue, perhaps? | 20:49 |
| kfox1111 | we should try and get that container merged and get a chart into kubernetes/charts. | 20:49 |
| kfox1111 | portdirect: yeah. could be. one of the last things left. as the job now makes it fail really early in the process. | 20:50 |
| portdirect | weird that its erratic - maybe worth trying flannel and seing if that makes a difference? | 20:51 |
| kfox1111 | it basically is now deploying canal, then emediately testing that dns eventually becomes responsive. | 20:51 |
| kfox1111 | yeah. could try a differentn one. but might be some work. as they embed addresss into the deaemonset that needs tweaking. | 20:52 |
| portdirect | weave would be the easiest one to swtich to - should only require a tiny tweak to those sed' | 20:53 |
| portdirect | (I'm using romana here - but i wouldnt for the gate, too heavy) | 20:54 |
| *** jgriffith_AutoAw is now known as jgriffith | 20:55 | |
| kfox1111 | interesting... | 20:55 |
| kfox1111 | so the logs do line up: | 20:55 |
| kfox1111 | failure to forward request "read udp 127.0.0.1:47510->127.0.0.1:53: i/o timeout" | 20:55 |
| kfox1111 | lots of those during the window where dns was tested and failed. | 20:55 |
| kfox1111 | but, here, the kicker, | 20:56 |
| kfox1111 | that is within a pod. | 20:56 |
| kfox1111 | so in theory networking should have any affect on it? | 20:56 |
| kfox1111 | seems like dnsmasq is not responding. | 20:57 |
| *** adrian_otto has joined #openstack-kolla | 20:57 | |
| kfox1111 | hmm.. the nova one really needs to be broken up. there seems to be muiltiple issues with it. | 21:03 |
| *** adrian_otto has quit IRC | 21:04 | |
| portdirect | yeah - could you make me a happy chap? and either expictly confirm that that rpm command can come out or file a bug so we can get it sorted - it makes me a sad panda :) | 21:05 |
| *** portdirect has left #openstack-kolla | 21:06 | |
| kfox1111 | it needs to be in for at least 2.0.2 suport. | 21:06 |
| *** portdirect has joined #openstack-kolla | 21:06 | |
| kfox1111 | if they cut a 2.0.3 I'd be happy, but not sure they will ever do that. :/ | 21:06 |
| kfox1111 | still, wouldn't hurt to file a bug. | 21:07 |
| kfox1111 | but they don't really support prebuilt containers on the hub. | 21:07 |
| portdirect | cool - i'll put a comment in for a '|| true' then? I know we are not doing set -e there but having an rpm command kinda exculdes ubuntu | 21:08 |
| kfox1111 | https://bugs.launchpad.net/kolla/+bug/1649009 | 21:08 |
| openstack | Launchpad bug 1649009 in kolla "centos 2.0.2 hub container qemu wrong." [Undecided,New] | 21:08 |
| kfox1111 | portdirect: without the -e, it will just always continue on. | 21:08 |
| kfox1111 | ubuntu or not. | 21:08 |
| kfox1111 | and this one too https://bugs.launchpad.net/kolla/+bug/1649010 | 21:09 |
| openstack | Launchpad bug 1649010 in kolla "kolla-toolbox 3.0.1 broken openstack cli" [Undecided,New] | 21:09 |
| kfox1111 | portdirect: but if you want to make the || true explicit, I'm cool with that. | 21:11 |
| kfox1111 | or wrap it in an if conditional. | 21:11 |
| * srwilkers_away reads scrollback | 21:11 | |
| kfox1111 | srwilkers_away: wb. :) | 21:11 |
| srwilkers_away | kfox1111, thanks man :) | 21:12 |
| *** srwilkers_away is now known as srwilkers | 21:12 | |
| portdirect | kfox1111: wicked - cheers :) added them to kolla-k8s as well so I can keep an eye on them | 21:18 |
| kfox1111 | hmm.. looks like the only thing I havent' switched out yet is the svc's... | 21:21 |
| kfox1111 | maybe thats the issue. | 21:21 |
| *** adrian_otto has joined #openstack-kolla | 21:22 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: test the parent review minus a few bits. https://review.openstack.org/409455 | 21:24 |
| kfox1111 | and another ubuntu only failure. | 21:26 |
| kfox1111 | last time was rackspace. this time osic. | 21:26 |
| *** clayton has quit IRC | 21:26 | |
| kfox1111 | so, very different networks. | 21:27 |
| portdirect | thinking about slim images - I need to get back on that soon as I get time - the savings are signifigant (eg harbor's glance api is 65mb) would be great to try and get a kolla image close to that size. | 21:27 |
| kfox1111 | +1 | 21:27 |
| kfox1111 | between that and a gate cache, we could get the gate to run significantly faster I think. | 21:28 |
| *** clayton has joined #openstack-kolla | 21:28 | |
| *** Jeffrey4l_ has quit IRC | 21:31 | |
| kfox1111 | weird.... | 21:33 |
| kfox1111 | how does minikube do dns.... | 21:33 |
| kfox1111 | there isn't the kube dns pod on it. | 21:33 |
| *** msimonin has joined #openstack-kolla | 21:36 | |
| *** Jeffrey4l_ has joined #openstack-kolla | 21:36 | |
| portdirect | never used minikube - apart from that day I barged into a kolla meeting and tried to do a code dump, seemed like the best way to say hello at the time :) | 21:36 |
| kfox1111 | just trrying to see if Ican quickly get a kube-dns log dump from a working one to compare against. | 21:36 |
| kfox1111 | hehe. yeah. :) | 21:37 |
| portdirect | i can get you one in 2 sec if thats any good? | 21:37 |
| kfox1111 | seems like that's the common way folks say hello in the opensource world. :) | 21:37 |
| kfox1111 | that would be awesome. :) | 21:37 |
| srwilkers | all caught up on scrollback. happy saturday | 21:37 |
| *** dave-mccowan has joined #openstack-kolla | 21:37 | |
| kfox1111 | srwilkers: yeah. far less discussion today | 21:38 |
| srwilkers | kfox1111, no kidding. no 6 hours worth of discussion to play catch up with | 21:38 |
| portdirect | :) | 21:38 |
| *** adrian_otto has quit IRC | 21:39 | |
| kfox1111 | srwilkers: yeah, I know right? yesterday was a very long discussion. :/ | 21:39 |
| portdirect | kfox1111: actually scrath my offer of help - ive only got k8s 1.5 atm - whihc i think is different | 21:40 |
| kfox1111 | srwilkers: its ok if people are trying hard to understand each other and not just talk past each other. talking to each other, while time consuming sometimes is better then not and just making assumptions. | 21:41 |
| kfox1111 | portdirect: ok. thanks anyway. | 21:41 |
| srwilkers | kfox1111, i agree 100% :) | 21:42 |
| portdirect | kfox1111: but dont think its of much use https://paste.fedoraproject.org/503729/14814061/raw/ | 21:42 |
| kfox1111 | portdirect: can you get me the dump of the other two ones too? | 21:43 |
| kfox1111 | dnsmask/health check? | 21:43 |
| portdirect | there is nothing to show | 21:43 |
| kfox1111 | blank? | 21:43 |
| portdirect | yeap | 21:43 |
| kfox1111 | ok. thats helpful. | 21:43 |
| kfox1111 | they are blank on the broken one too. | 21:44 |
| kfox1111 | so that validates that blank doesn't mean broken. | 21:44 |
| kfox1111 | thanks. :) | 21:44 |
| portdirect | np :) | 21:44 |
| kfox1111 | nice... | 21:45 |
| kfox1111 | this is looking very promicing: https://review.openstack.org/#/c/409455/7/tests/bin/ceph_workflow.sh | 21:45 |
| kfox1111 | one of the gates passed so far. | 21:46 |
| kfox1111 | second passed. | 21:48 |
| kfox1111 | this should be an easy fix. :) | 21:48 |
| kfox1111 | one more success. :) | 21:54 |
| kfox1111 | one more and its good. :) | 21:54 |
| *** portdirect is now known as portdirect_away | 22:00 | |
| *** msimonin has quit IRC | 22:04 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 22:05 |
| *** fragatina has quit IRC | 22:06 | |
| srwilkers | hey kfox1111: question. are we just hardcoding replicas in currently, similar to whats done for rabbitmq here: https://github.com/openstack/kolla-kubernetes/blob/master/helm/microservice/rabbitmq-pod/templates/rabbitmq-pod.yaml#L11 ? | 22:07 |
| *** ipsecguy_ has joined #openstack-kolla | 22:08 | |
| *** msimonin has joined #openstack-kolla | 22:10 | |
| *** msimonin has left #openstack-kolla | 22:10 | |
| *** rstarmer has joined #openstack-kolla | 22:11 | |
| *** ipsecguy has quit IRC | 22:12 | |
| kfox1111 | srwilkers: we're setting defaults in all_values. | 22:15 |
| kfox1111 | rabbit is somewhat unique, as things break badly if replica != 1. | 22:15 |
| kfox1111 | same with mariadb and memcached. | 22:15 |
| kfox1111 | the rest should be scalable I think. | 22:15 |
| srwilkers | cool. makes sense. thanks | 22:17 |
| *** mgiles has joined #openstack-kolla | 22:18 | |
| *** bmace has quit IRC | 22:20 | |
| *** bmace has joined #openstack-kolla | 22:20 | |
| openstackgerrit | Mark Giles proposed openstack/kolla-kubernetes: Helm creation of Nova services https://review.openstack.org/409473 | 22:23 |
| openstackgerrit | Mark Giles proposed openstack/kolla-kubernetes: Helm database creation for Nova https://review.openstack.org/409474 | 22:24 |
| *** Pavo has quit IRC | 22:29 | |
| openstackgerrit | Mark Giles proposed openstack/kolla-kubernetes: Helm database creation for Nova https://review.openstack.org/409474 | 22:32 |
| kfox1111 | hmm... so its not the dns resolving container either... forcing it to centos still fails... | 22:32 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 22:33 |
| *** Pavo has joined #openstack-kolla | 22:34 | |
| openstackgerrit | Steve Wilkerson proposed openstack/kolla-kubernetes: Helm charts for adding Cinder https://review.openstack.org/409349 | 22:34 |
| openstackgerrit | Steve Wilkerson proposed openstack/kolla-kubernetes: Helm charts for adding Cinder https://review.openstack.org/409349 | 22:39 |
| kfox1111 | srwilkers: once it passes the gate, I think its mergable. | 22:41 |
| kfox1111 | wana pull the wip off the commit message? | 22:41 |
| srwilkers | still need to get the keystone endpoints and such in, unless you want those to move in separately | 22:41 |
| kfox1111 | oh. still need the gate test update too. | 22:42 |
| kfox1111 | more smaller revies are probably better then one big one. | 22:42 |
| kfox1111 | the endpoints are kind of held up on the neutron-endpoints one. | 22:42 |
| mgiles | Are the gates working? I'm getting errors that appear completely unrelated to my change (see logs for https://review.openstack.org/#/c/409473/) | 22:42 |
| portdirect_away | mgiles: kolla dir in commit | 22:43 |
| kfox1111 | mgiles: that looks very strange. | 22:43 |
| kfox1111 | oh. yeah. that dir would break things. | 22:44 |
| mgiles | portdirect_away Hmmm. Not sure how that got there. thanks for noticing | 22:44 |
| portdirect_away | if your testing locally it's caught me out a few times | 22:44 |
| portdirect_away | also I'm here! | 22:44 |
| *** portdirect_away is now known as portdirect | 22:44 | |
| openstackgerrit | Steve Wilkerson proposed openstack/kolla-kubernetes: Helm charts for adding Cinder https://review.openstack.org/409349 | 22:45 |
| srwilkers | wb portdirect o/ | 22:45 |
| portdirect | hey dude! | 22:45 |
| openstackgerrit | Mark Giles proposed openstack/kolla-kubernetes: Helm creation of Nova services https://review.openstack.org/409473 | 22:46 |
| mgiles | portdirect_away Thanks - fixed | 22:47 |
| mgiles | or just portdirect now | 22:47 |
| kfox1111 | oh. very interesting... | 22:52 |
| kfox1111 | kubernetes resolves, but google does not. | 22:52 |
| *** dave-mccowan has quit IRC | 22:53 | |
| kfox1111 | I wonder if this is an ubuntu resolver issue. | 22:53 |
| kfox1111 | anyone have an ubuntu box handy? | 22:56 |
| mgiles | yep | 22:56 |
| kfox1111 | can you please cat /etc/resolv.conf? | 22:56 |
| mgiles | but I can't resolv google anyway because I'm behind a corp firewall | 22:56 |
| kfox1111 | I'm wondering if they stuck 127.0.0.1 in there. | 22:56 |
| mgiles | yes | 22:56 |
| mgiles | nameserver 127.0.0.1 | 22:57 |
| kfox1111 | thats probably the issue. :/ | 22:57 |
| mgiles | that's the only entry | 22:57 |
| kfox1111 | yeah. that would explain it. | 22:57 |
| kfox1111 | how to workaround that though.... | 22:57 |
| kfox1111 | hmm... | 22:57 |
| kfox1111 | seems like this: https://github.com/kubernetes/kubernetes/issues/19076 | 22:59 |
| sbezverk_ | good evening! anything interesting going on? | 23:01 |
| srwilkers | hey sbezverk_ :) | 23:02 |
| kfox1111 | hi sbezverk_ | 23:03 |
| portdirect | sbezverk_: hey! | 23:04 |
| sbezverk_ | wow everybody is working! that's the holiday spirit ;-) | 23:05 |
| kfox1111 | :) | 23:05 |
| sbezverk_ | kfox1111: qq do we have an image with jq and kube bits in? | 23:06 |
| kfox1111 | sbezverk_: the kolla-toolbox:3.0.1 has jq and crudini. | 23:09 |
| kfox1111 | I don't think there's a container yet with kubectl/helm in it. | 23:09 |
| kfox1111 | though portdirect's been working on a more general client container. | 23:09 |
| openstackgerrit | Mark Giles proposed openstack/kolla-kubernetes: Helm for nova-api https://review.openstack.org/409479 | 23:09 |
| kfox1111 | I don't have a review link handy though. portdirect do you? | 23:10 |
| portdirect | you mean the kolla-k8s-dev container? | 23:10 |
| kfox1111 | yeah. | 23:10 |
| portdirect | thats not really for running in a cluster but: | 23:10 |
| portdirect | https://review.openstack.org/#/c/405720/7/Dockerfile | 23:11 |
| portdirect | dont think it's what sbezverk_ needs :( | 23:11 |
| portdirect | also only has helm | 23:11 |
| kfox1111 | it would be good to add kubectl and jq too. | 23:12 |
| portdirect | oh I'm telling lies again | 23:12 |
| portdirect | it does have them | 23:13 |
| sbezverk_ | portdirect: it would really like to have kubectl in one of kolla toolboxes | 23:13 |
| portdirect | I could make one for you know if you want? would be in my docker namespace for now but loud put an ps for kolla-kube-toolbox? | 23:14 |
| portdirect | would just be kolla-toolbox butwith k8s specific things added | 23:15 |
| kfox1111 | sbezverk_: I guess I should ask, what use case do you ahve for it? | 23:16 |
| kfox1111 | is it for embedding in a pod, or for using directly? | 23:16 |
| sbezverk_ | I am putting together configmap monitor and need to run jq which is already there and kubectl to be able to check kube object revision number | 23:17 |
| kfox1111 | oh. for the sidecar thing. | 23:18 |
| kfox1111 | then, yeah. a more specific use container might be in order. | 23:18 |
| sbezverk_ | as soon as I get it running. I will post it just to get your opinion around it | 23:18 |
| kfox1111 | as the devel one might have stuff like a git checkout and other stuff that is heavier weight. | 23:19 |
| kfox1111 | sounds good. | 23:19 |
| portdirect | I'll have an image for you to *play* with in 5 mins | 23:19 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 23:20 |
| sbezverk_ | portdirect: thanks!!! | 23:23 |
| portdirect | sbezverk_: docker.io/port/centos-binary-kolla-kube-toolbox:3.0.1 <<--kolla toolbox with kubectl | 23:23 |
| kfox1111 | sbezverk_: shoudl that restart feature be added to kubernetes-entrypoint? | 23:23 |
| kfox1111 | it has most of the code already to contact kubernetes and do things when things change. | 23:24 |
| portdirect | not if the toolbox only runs as an init container? | 23:24 |
| kfox1111 | there could be a sidecar run with the same container. | 23:24 |
| kfox1111 | just thinking it may be easier to add the feature to kubernetes-entrypoint then to write the code from scratch. | 23:25 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 23:27 |
| portdirect | kfox1111: lets see what sbezverk_ comes up with - the protype I've seen ther people do it is 20 lines of bash including comments - so it could be very lightweight poentially | 23:27 |
| *** adrian_otto has joined #openstack-kolla | 23:27 | |
| kfox1111 | portdirect: does it busyloop, or long wait? | 23:28 |
| kfox1111 | I can see a busyloop implementation potentially being fairly easy to do. | 23:28 |
| kfox1111 | but would be pretty hard on the api on a larger cluster. | 23:29 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 23:30 |
| sbezverk_ | portdirect: no, just init container does not work | 23:30 |
| sbezverk_ | portdirect: side car check with an interval revision version | 23:31 |
| sbezverk_ | of configmap | 23:31 |
| portdirect | sbezverk_: we are in agreement about the init container not working :) | 23:31 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 23:32 |
| sbezverk_ | cool, let me finish then we can discuss | 23:32 |
| kfox1111 | weird... no resolvconf service in the newer ubuntu.... | 23:34 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 23:35 |
| openstackgerrit | Steve Wilkerson proposed openstack/kolla-kubernetes: Helm charts for adding Cinder https://review.openstack.org/409349 | 23:35 |
| *** adrian_otto has quit IRC | 23:35 | |
| portdirect | kfox1111: networkmanager afaik? | 23:36 |
| kfox1111 | but it is only 127.0.0.1... | 23:36 |
| sbezverk_ | kfox1111: I saw that on my 16.04 | 23:36 |
| kfox1111 | network manager is doing resolving on 16.04? | 23:36 |
| openstackgerrit | Mark Giles proposed openstack/kolla-kubernetes: Helm for nova-api https://review.openstack.org/409479 | 23:37 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 23:39 |
| sbezverk_ | portdirect: could you send me in pm the location of that image, I need to step out again, but when I get back I will try it.. | 23:41 |
| kfox1111 | arg... not finding many traces of network manager either. | 23:43 |
| kfox1111 | something is odd on this ubuntu box. :/ | 23:43 |
| portdirect | oh - it may be systemd-resolved.service ? | 23:46 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 23:47 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 23:48 |
| kfox1111 | not seeing that one either. | 23:49 |
| kfox1111 | http://logs.openstack.org/63/408263/29/check/gate-kolla-kubernetes-deploy-ubuntu-binary-ceph-nv/286f405/console.html | 23:50 |
| kfox1111 | has the service list. | 23:50 |
| kfox1111 | I wonder how custom this image is. | 23:50 |
| kfox1111 | not much by name resolv either. | 23:51 |
| kfox1111 | so weird... | 23:51 |
| kfox1111 | this is the fastest gate turnaround I've had in a long long time. :) | 23:51 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 23:51 |
| openstackgerrit | Mark Giles proposed openstack/kolla-kubernetes: Helm for nova-libvirt https://review.openstack.org/409481 | 23:53 |
| kfox1111 | there we go: | 23:53 |
| kfox1111 | 2016-12-10 23:53:08.942026 | tcp 0 0 127.0.0.1:53 0.0.0.0:* LISTEN 1011/unbound | 23:53 |
| kfox1111 | something called unbound | 23:53 |
| *** kmadac has joined #openstack-kolla | 23:55 | |
| *** kmadac has quit IRC | 23:58 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: Add neutron keystone service/endpoint creation jobs https://review.openstack.org/408263 | 23:59 |
| openstackgerrit | Steve Wilkerson proposed openstack/kolla-kubernetes: Helm charts for adding Cinder https://review.openstack.org/409349 | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!