| openstackgerrit | Ryan Hallisey proposed a change to stackforge/kolla: Initial setup for nova-controller https://review.openstack.org/126305 | 00:07 |
|---|---|---|
| *** ChuckC has joined #tripleo | 00:07 | |
| *** chuckC_ has joined #tripleo | 00:09 | |
| larsks | radez_g0n3: around? | 00:17 |
| *** xuhaiwei has joined #tripleo | 00:38 | |
| *** dsneddon_ has quit IRC | 00:47 | |
| *** dsneddon_ has joined #tripleo | 00:48 | |
| SpamapS | larsks: did you have a chance to take a second look at os-cloud-config ? | 00:57 |
| *** nosnos has joined #tripleo | 01:01 | |
| openstackgerrit | Lars Kellogg-Stedman proposed a change to stackforge/kolla: renamed keystone services https://review.openstack.org/126453 | 01:03 |
| openstackgerrit | Lars Kellogg-Stedman proposed a change to stackforge/kolla: remove "master" from mariadb k8s configs https://review.openstack.org/126454 | 01:05 |
| larsks | SpamapS: I am looking at it right now. So, for example, I just want to initialize keystone with the identity service/endpoints and an admin user/tenant. I'm not sure how to do this. What would that look like? | 01:07 |
| SpamapS | larsks: init-keystone | 01:08 |
| *** dsneddon_ has quit IRC | 01:08 | |
| larsks | SpamapS: It sounds sort of like init-keystone will do that, but it also seems to want to ssh into something to work, and I just want it to talk to the keystone api. | 01:08 |
| *** yamahata has joined #tripleo | 01:08 | |
| *** dsneddon_ has joined #tripleo | 01:09 | |
| SpamapS | larsks: here's where we do _all of openstack_ in devtest http://git.openstack.org/cgit/openstack/tripleo-incubator/tree/scripts/devtest_overcloud.sh#n468 | 01:09 |
| larsks | It's not clear how to provide specific values for admin/internal/public endpoints. There's --public option, but nothing for admin/internal endpoints. | 01:09 |
| SpamapS | larsks: the SSH is disablable | 01:09 |
| larsks | SpamapS: How is it disabled? The help output does not provide any hints. | 01:09 |
| SpamapS | larsks: gah.. looks like it has not been added to the CLI. It is an option to os_cloud_config.keystone.initialize pki_setup=False | 01:10 |
| larsks | My take is that right now, this tool has been designed for a specific set of circumstances that isn't appropriate for what we're trying to do inside our containers right now. | 01:12 |
| larsks | And I'm going to say "right now" again, because apparently at this time of night I lack variety when I write. | 01:12 |
| SpamapS | larsks: no, this tool has been designed for general use as an openstack bootstrap tool. | 01:13 |
| larsks | That is not my impression. I guess we will just differ on that for now. | 01:14 |
| SpamapS | http://git.openstack.org/cgit/openstack/os-cloud-config/tree/README.rst#n7 | 01:14 |
| larsks | I have seen that several timses :) | 01:15 |
| SpamapS | larsks: It reads quite generically to me. :-P | 01:15 |
| openstackgerrit | A change was merged to openstack/diskimage-builder: Save extended attributes when creating tar https://review.openstack.org/125428 | 01:16 |
| *** yamahata_ has joined #tripleo | 01:16 | |
| SpamapS | larsks: I'm not sure what more I can do to convince you that this tool is intended to do exactly what you want to do. :-P | 01:20 |
| larsks | SpamapS: Take one of the start scripts I have written and make it use this tool. | 01:20 |
| larsks | ...with equivalent or reduced complexity. | 01:20 |
| larsks | That would totally convince me, and I would abandon all alternatives on the spot. | 01:21 |
| openstackgerrit | A change was merged to openstack/diskimage-builder: Move dpkg manifest creation to finalise https://review.openstack.org/124765 | 01:21 |
| SpamapS | larsks: need to land the --no-pki-setup option and I think it will be one line. :-P | 01:21 |
| larsks | I think you will also need to land the --no-ssh option. | 01:21 |
| SpamapS | same difference | 01:22 |
| larsks | Ah, okay. | 01:22 |
| larsks | I didn't realize that A == B there. | 01:22 |
| larsks | So, here are the things I am looking for in particular: | 01:23 |
| StevenK | We want to stop init-keystone doing pki_setup anyway | 01:23 |
| StevenK | Now that greghaynes' is back, I can shake him until a plan falls out | 01:23 |
| larsks | - idempotent creation of users/tenants/roles/user-role-memberships | 01:23 |
| SpamapS | StevenK: thats what I thought but it's been a while since I participated so wasn't sure if maybe it stuck around because "reasons" | 01:23 |
| larsks | - idempotent creation of services/endpoints | 01:23 |
| larsks | - the ability to *replace* endpoints in a single operation | 01:23 |
| *** sdake has joined #tripleo | 01:24 | |
| StevenK | Huh, replace is a new one | 01:24 |
| StevenK | larsks: Why do you want to replace endpoints, out of interest? | 01:25 |
| larsks | StevenK: Without real "external ip" support in kubernets, I think we may need that for being able to swap out the public endpoint on container reboot. But maybe not! | 01:25 |
| larsks | We have not quite gotten to that point. | 01:25 |
| larsks | If that is an issue, don't worry about that one right now. | 01:26 |
| SpamapS | larsks: isn't that what hautoproxy is going to do? | 01:26 |
| larsks | No. That primarily handles *internal* communication between pods...so that glance can find keystone, and nova can find glance, etc, even when the pods are relocated to different minions. | 01:26 |
| larsks | The external ip problem is more difficult: what do you use as the external ip? | 01:27 |
| StevenK | larsks: There's a register-endpoints script in os-cloud-config that may deal with most of your concerns, init-keystone is only concerned with performing stuff that keystone requires | 01:27 |
| *** takadayuiko has joined #tripleo | 01:27 | |
| larsks | Sure. Is there a create-users script there, too? | 01:27 |
| larsks | Hmmm, it doesn't look like it. | 01:27 |
| StevenK | larsks: No, I believe register-endpoints does that | 01:27 |
| SpamapS | larsks: in the heat-only TripleO we use a keepalived controlled VIP | 01:28 |
| SpamapS | larsks: which is allocated via Neutron | 01:28 |
| larsks | SpamapS: we could totally do something like that, although I am trying to take advantage of what kubernetes provides wherever possible. | 01:28 |
| StevenK | larsks: If you want a script to give it a list of Keystone users, and it uses the Keystone API to make sure they exist, then I'd be delighted to help you work on it | 01:28 |
| larsks | And then we could pass in the vip into our pods via an environment variable. | 01:29 |
| SpamapS | larsks: does Kubernetes do something similar? | 01:29 |
| larsks | Kuberenetes doesn't do anythign for external ips right now. There is an issue open on that with a lot of discussion. | 01:29 |
| larsks | StevenK: I already have a script that does everything I want :). SpamapS is trying to convince me that os-cloud-config will do those things, too. | 01:29 |
| StevenK | It probably does most of them | 01:29 |
| StevenK | And shared code is excellent | 01:30 |
| larsks | SpamapS: https://github.com/GoogleCloudPlatform/kubernetes/issues/1161 | 01:30 |
| StevenK | SpamapS, larsks: Let me push up a patch for --no-pki-setup | 01:30 |
| larsks | StevenK: This is what I put together that currently handles my use cases: https://github.com/larsks/crux | 01:30 |
| SpamapS | larsks: the question is, would your script be generically useful to non-kolla users or is the environment special enough to warrant a special tool? | 01:30 |
| larsks | SpamapS: Oh, it would totally be useful elsewhere. | 01:30 |
| larsks | This is code that originated when I first started deploying OpenStack at my last job. | 01:31 |
| SpamapS | larsks: so we can entertain the opposite idea: replace os-cloud-config with yours. :) | 01:31 |
| larsks | SpamapS: Well, my use case is primarily "provisioning keystone objects" | 01:31 |
| larsks | Not "bootstrapping openstack" | 01:31 |
| larsks | So I would not recommend that direction at all! | 01:32 |
| SpamapS | why? the act of bootstrapping is to create the initial keystone objects necessary for normal use. | 01:32 |
| larsks | Well, I notice that os-cloud-config has bits for neutron and baremetal/ironic stuff, too, which is why I said that. | 01:33 |
| SpamapS | as well as neutron objects, ironic objects, and any others that deployers need to seed into their cloud before users can use the cloud | 01:33 |
| SpamapS | It's for the same reason. The API's are not useful to non-admins without these objects. | 01:34 |
| larsks | Yup. | 01:36 |
| SpamapS | larsks: really what should be done is keystoneclient / openstack client should be made idempotent. | 01:40 |
| SpamapS | Or at least have an idempotent option | 01:40 |
| larsks | SpamapS: e.g., like ovs has the --may-exist and --if-exists options. Yes, that would also be a good solution. | 01:41 |
| SpamapS | yeah I really appreciate how ovs adapts to whichever way you need it to work | 01:42 |
| SpamapS | StevenK: hey have you tried a fresh 'tox -epy27' on os-cloud-config lately? | 01:44 |
| SpamapS | StevenK: I'm getting weird distutils problems. | 01:44 |
| StevenK | SpamapS: Huh | 01:45 |
| SpamapS | distutils.errors.DistutilsError: Could not find suitable distribution for Requirement.parse('pbr') | 01:45 |
| StevenK | Heh. Running tox -epy27 while you have that venv activated does not end well | 01:47 |
| *** ccrouch has joined #tripleo | 01:48 | |
| *** ccrouch has quit IRC | 01:48 | |
| *** ccrouch has joined #tripleo | 01:48 | |
| SpamapS | definitely weird | 01:50 |
| SpamapS | trying to install in the venv myself produced equally confusing things | 01:50 |
| SpamapS | also.. wtf cffi.. :-P | 01:50 |
| SpamapS | distutils.errors.DistutilsError: Could not find suitable distribution for Requirement.parse('cffi>=0.8') | 01:51 |
| SpamapS | seems like one of those "you ahve the wrong version of something I"m not even mentioning" pbr-ish problems | 01:52 |
| * SpamapS goes into family mode whilst investigating | 01:52 | |
| tchaypo | StevenK: speaking of the gregory | 01:54 |
| tchaypo | he has a change up to switch us to using generate-ssl-cert | 01:54 |
| StevenK | tchaypo: Oh? Link me? | 01:56 |
| tchaypo | hard from gertty | 01:57 |
| tchaypo | but it's 126343 | 01:57 |
| tchaypo | also it's a completel fail at present as it can't find the command | 01:57 |
| tchaypo | but you know, just minor details... | 01:57 |
| *** andreaf has quit IRC | 01:57 | |
| *** andreaf has joined #tripleo | 01:58 | |
| StevenK | Oh, it probably needs a new os-cloud-config | 02:01 |
| StevenK | tchaypo: And it will certainly also require the change I'm prepping | 02:01 |
| *** yamahata_ has quit IRC | 02:03 | |
| *** yamahata_ has joined #tripleo | 02:03 | |
| openstackgerrit | Steve Kowalik proposed a change to openstack/os-cloud-config: Teach init-keystone about PKI setup https://review.openstack.org/126456 | 02:04 |
| tchaypo | heh | 02:05 |
| tchaypo | friday in the office? | 02:06 |
| StevenK | tchaypo: Sounds like a good plan | 02:06 |
| StevenK | tchaypo: I have commented on greghaynes' review, thanks for the pointer | 02:06 |
| *** pcrews has quit IRC | 02:19 | |
| *** Yanyanhu has joined #tripleo | 02:24 | |
| sdake | larsks around? | 02:27 |
| larsks | Briefly! :) | 02:27 |
| sdake | say w emay need to put crux under gating | 02:27 |
| sdake | if we do that, it would be ideal if the license was asl2.0 | 02:27 |
| sdake | any objections to changing the license? | 02:27 |
| *** zigo has quit IRC | 02:27 | |
| larsks | sdake: https://github.com/larsks/crux/commit/ad95cbfd78114c59d4151e042e1b285fda57693b | 02:27 |
| sdake | cool | 02:28 |
| sdake | I'm not sure it matters yet | 02:28 |
| *** cwolferh_ has joined #tripleo | 02:28 | |
| sdake | but we may need to get that into stackforge if its a long term solution | 02:28 |
| larsks | No worries. | 02:28 |
| sdake | lets give it a mo or two and see what happens with cloud config | 02:28 |
| *** Yanyanhu has quit IRC | 02:29 | |
| *** zigo has joined #tripleo | 02:29 | |
| larsks | Sure. I was talking with SpamapS about os-cloud-config earlier this evening. I am still dubious, but there are some patches landing soon that will at least drop the 'ssh-to-remote-host' default behavior. | 02:29 |
| larsks | Until then, I think crux will at least allow us to move forward with our existing containers in a sane fashion. | 02:31 |
| sdake | ya we aren't looking for perfect this early | 02:32 |
| sdake | just looking for works :) | 02:32 |
| *** Yanyanhu has joined #tripleo | 02:42 | |
| sdake | time for diablo 3 with the rugrats | 02:47 |
| sdake | ttyl | 02:47 |
| *** Yanyanhu has quit IRC | 02:51 | |
| openstackgerrit | Ian Wienand proposed a change to openstack/diskimage-builder: Standarise tracing for scripts https://review.openstack.org/119023 | 02:53 |
| *** noslzzp has quit IRC | 02:59 | |
| *** sdake_ has quit IRC | 02:59 | |
| *** untriaged-bot has joined #tripleo | 03:01 | |
| untriaged-bot | No untriaged bugs so far! \o/ | 03:01 |
| *** untriaged-bot has quit IRC | 03:01 | |
| *** sdake has quit IRC | 03:01 | |
| *** ramishra has joined #tripleo | 03:09 | |
| *** nosnos has quit IRC | 03:24 | |
| *** ramishra has quit IRC | 03:31 | |
| *** ramishra has joined #tripleo | 03:31 | |
| *** ramishra has quit IRC | 03:33 | |
| *** ramishra has joined #tripleo | 03:34 | |
| *** spzala has quit IRC | 03:34 | |
| *** cwolferh_ has quit IRC | 03:38 | |
| *** rushiagr_away is now known as rushiagr | 03:38 | |
| *** rushiagr is now known as rushiagr_away | 04:09 | |
| *** nosnos has joined #tripleo | 04:10 | |
| *** pensu has joined #tripleo | 04:11 | |
| openstackgerrit | A change was merged to stackforge/kolla: use "crux" for creating users/endpoints https://review.openstack.org/126067 | 04:13 |
| *** yamahata has quit IRC | 04:13 | |
| *** yamahata_ has quit IRC | 04:13 | |
| *** yamahata_ has joined #tripleo | 04:13 | |
| *** yamahata_ has quit IRC | 04:14 | |
| *** yamahata has joined #tripleo | 04:14 | |
| *** yamahata_ has joined #tripleo | 04:15 | |
| openstackgerrit | A change was merged to stackforge/kolla: added hautoproxy auto-configuring haproxy https://review.openstack.org/126373 | 04:18 |
| *** pensu has quit IRC | 04:19 | |
| *** dshulyak_ has joined #tripleo | 04:26 | |
| *** dshulyak_ has quit IRC | 04:26 | |
| *** tzumainn has quit IRC | 04:50 | |
| StevenK | tchaypo: Did you want to put up an appointment for Friday? | 05:07 |
| *** lazy_prince2 has joined #tripleo | 05:10 | |
| *** pradeep has joined #tripleo | 05:10 | |
| *** lazy_prince2 has quit IRC | 05:12 | |
| openstackgerrit | Steve Kowalik proposed a change to openstack/os-cloud-config: Add support for creating baremetal flavors https://review.openstack.org/115872 | 05:18 |
| openstackgerrit | Steve Kowalik proposed a change to openstack/os-cloud-config: Add support for injecting extra-specs into flavors https://review.openstack.org/123345 | 05:19 |
| *** loki184 has joined #tripleo | 05:21 | |
| *** rushiagr_away is now known as rushiagr | 05:23 | |
| tchaypo | StevenK: sure | 05:24 |
| loki184 | please review https://review.openstack.org/#/c/106727/ | 05:26 |
| *** cwolferh_ has joined #tripleo | 05:36 | |
| *** nosnos has quit IRC | 05:37 | |
| *** nosnos has joined #tripleo | 05:38 | |
| *** amitpp has joined #tripleo | 05:40 | |
| *** amitpp has quit IRC | 05:41 | |
| *** nosnos has quit IRC | 05:42 | |
| *** nosnos has joined #tripleo | 05:43 | |
| *** dshulyak_ has joined #tripleo | 05:59 | |
| *** pensu has joined #tripleo | 06:03 | |
| tchaypo | StevenK: i can find a 4 perosn room that's free | 06:03 |
| StevenK | tchaypo: In SNO02, or 01? | 06:05 |
| tchaypo | 02 | 06:07 |
| StevenK | tchaypo: 5.13 looks free all day Friday and it's 6 | 06:07 |
| StevenK | But round table, so it may be a squeeze | 06:07 |
| StevenK | tchaypo: 6.12 is also free | 06:08 |
| tchaypo | okay, done | 06:08 |
| StevenK | tchaypo: Accepted | 06:09 |
| tchaypo | cool. | 06:10 |
| StevenK | tchaypo: We tend to book 8s, remember? :-P | 06:10 |
| tchaypo | Yes | 06:12 |
| tchaypo | but outlook is painful and I don't want to go back | 06:13 |
| *** jml has quit IRC | 06:14 | |
| *** rdopieralski has joined #tripleo | 06:15 | |
| *** rdopieralski has quit IRC | 06:15 | |
| *** jml has joined #tripleo | 06:15 | |
| *** rdopieralski has joined #tripleo | 06:15 | |
| *** mrunge has joined #tripleo | 06:31 | |
| *** jprovazn has joined #tripleo | 06:31 | |
| tchaypo | loki184: love your branch name :) | 06:32 |
| loki184 | tchaypo: lol | 06:33 |
| greghaynes | StevenK: oh, did I break everything? | 06:34 |
| greghaynes | StevenK: oh, so as-is generate-keystone-pki essentially calls the other script | 06:36 |
| greghaynes | my thinking was to move over to the other script, then deprecate generate-keystone-pki | 06:36 |
| greghaynes | as for the os-c-c needs to be released, agreed | 06:36 |
| tchaypo | loki184: Reviewed. | 06:37 |
| tchaypo | loki184: I feel a little bit guilty because I really think most of the things I've mentioned are fairly minor | 06:37 |
| tchaypo | greghaynes: OH HAI | 06:37 |
| greghaynes | tchaypo: HAI | 06:37 |
| * greghaynes returns | 06:37 | |
| tchaypo | how was the wilderness? | 06:39 |
| tchaypo | were there tumbleweeds? | 06:39 |
| greghaynes | hah, I was sort of in the antithesis of wilderness: Tokyo | 06:40 |
| greghaynes | definitely no tombleweeds | 06:40 |
| greghaynes | er, tumble | 06:40 |
| loki184 | tchaypo: ty will check it out :) | 06:40 |
| StevenK | loki184: I've also reviewed it | 06:40 |
| StevenK | greghaynes: I was talking about the plan to deprecate pki_setup in init-keystone | 06:41 |
| loki184 | stevenk: ty will resubmit with changes :) | 06:41 |
| StevenK | Since the sooner we can stop sshing into the keystone endpoint to run keystone-manage, the better | 06:41 |
| greghaynes | StevenK: totally, is that needed for my change though? | 06:42 |
| greghaynes | (this is one of those patches that took months to get merged so having a hard time recalling) | 06:42 |
| StevenK | greghaynes: It's needed for your incubator patch which hasn't merged yet | 06:44 |
| greghaynes | oh, huh, how are we not currently doing keystone pki init in two places already? | 06:45 |
| GheRivero | morning all | 06:49 |
| *** jtomasek has joined #tripleo | 06:53 | |
| *** gfidente has joined #tripleo | 06:53 | |
| greghaynes | StevenK: so, this is kind of hitting my boundary of keystone knowledge. AFAICT my change isnt an actual functional differenct for the commands we are running | 06:54 |
| greghaynes | StevenK: seems like currently we both create the ssl pki on jump host, copy it in via heat, and then ssh in and run pki_setup | 06:55 |
| *** andreaf has quit IRC | 06:59 | |
| tchaypo | morning ghe | 07:00 |
| greghaynes | StevenK: Looks like "keystone-manage pki_setup" fails if you run it when certs are already present, so this might be what is currently going on | 07:07 |
| gfidente | isn't it time for the meeting? | 07:09 |
| *** jcoufal has joined #tripleo | 07:09 | |
| gfidente | oh no it's in 12 hours :P | 07:09 |
| GheRivero | I though it was in 13 :) | 07:11 |
| GheRivero | no, 12 is right. This week is in Tuesday night for me :/ | 07:12 |
| tchaypo | ah, the early morning meetings! woo! | 07:19 |
| tchaypo | thanks to DST kicking over here, they're now at 6am for me instead of 5am | 07:20 |
| tchaypo | *dances* | 07:20 |
| *** ifarkas has joined #tripleo | 07:21 | |
| GheRivero | DST already? in Europe the 26th | 07:26 |
| *** ramishra has quit IRC | 07:26 | |
| *** ramishra has joined #tripleo | 07:32 | |
| StevenK | GheRivero: Yeah, there is like two or three weeks difference when Europe, the US and we change. | 07:46 |
| *** jistr has joined #tripleo | 07:46 | |
| *** lazy_prince has quit IRC | 07:48 | |
| *** ramishra has quit IRC | 07:49 | |
| *** lazy_prince has joined #tripleo | 07:49 | |
| *** dsneddon_ has quit IRC | 07:49 | |
| GheRivero | the US one week later than Europe. We'll go nuts for the following meetings. | 07:50 |
| *** dsneddon_ has joined #tripleo | 07:51 | |
| *** lazy_prince2 has joined #tripleo | 07:51 | |
| StevenK | GheRivero: :-) | 07:51 |
| StevenK | GheRivero: And just think, all the Americans in Paris for the summit miss the clock change | 07:52 |
| *** lazy_prince is now known as killer_prince | 07:53 | |
| *** jang1 has joined #tripleo | 07:53 | |
| *** killer_prince is now known as lazy_prince | 07:56 | |
| *** ramishra has joined #tripleo | 07:57 | |
| *** derekh has joined #tripleo | 07:57 | |
| GheRivero | StevenK: :) It could be worst. I was in the UK during 6 months when we change to the Euro, so when I came back home didn't know how to use the new currency and missed the 3 months adaption period with the two currencies | 07:58 |
| *** lazy_prince2 has quit IRC | 07:58 | |
| StevenK | GheRivero: Haha! | 07:59 |
| StevenK | GheRivero: So you had this money in your wallet, and you couldn't spend it? | 08:00 |
| openstackgerrit | Chmouel Boudjnah proposed a change to stackforge/kolla: Validate JSON https://review.openstack.org/126484 | 08:03 |
| GheRivero | yeah. You can change it in the bank but not spend it. And the conversion was 1000pts to 6€ but I was already thinking in £ | 08:03 |
| openstackgerrit | xu-haiwei proposed a change to stackforge/kolla: Remove unnecessary comma to correct json format https://review.openstack.org/126487 | 08:07 |
| *** ramishra has quit IRC | 08:07 | |
| greghaynes | StevenK: Are you able to parse this py33 CI fail? http://logs.openstack.org/34/111334/4/check/gate-os-collect-config-python33/5443472/console.html#_2014-10-06_13_24_10_881 | 08:08 |
| StevenK | greghaynes: WAT | 08:09 |
| greghaynes | wat indeed | 08:10 |
| StevenK | greghaynes: I can recall the LC_ALL failures from my porting | 08:10 |
| *** lucasagomes has joined #tripleo | 08:11 | |
| *** openstackgerrit has quit IRC | 08:11 | |
| StevenK | greghaynes: I see FAIL: os_collect_config.tests.test_collect.TestCollect.test_main | 08:22 |
| StevenK | locally, but not the LC_ALL | 08:22 |
| *** nosnos has quit IRC | 08:23 | |
| yuanying | I also saw that fail, fixtures is broken at py33. | 08:24 |
| StevenK | Huh, yeah, it could be a fixtures failure | 08:25 |
| yuanying | https://bugs.launchpad.net/python-fixtures/+bug/1358085 | 08:25 |
| uvirtbot | Launchpad bug 1358085 in python-fixtures "FakePopen does not work in Python 3.3+" [High,Fix released] | 08:25 |
| StevenK | Oh, that. I fixed that. | 08:25 |
| StevenK | greghaynes: Rebase your branch? | 08:26 |
| *** athomas has joined #tripleo | 08:31 | |
| yuanying | I think that it has not been fixed yet. | 08:35 |
| StevenK | yuanying: The fixtures stuff? Yeah, it has. I wrote the bug fixes, and lifeless merged them and released it. | 08:36 |
| yuanying | hmm... | 08:37 |
| *** nosnos has joined #tripleo | 08:40 | |
| yuanying | StevenK: My patch is rebased master now, but tox -epy33 was failed because of fixtures broken issue. | 08:42 |
| *** pradeep has quit IRC | 08:44 | |
| yuanying | StevenK: I installed fixtures-0.3.17 in my environment, but os-collect-config tox -epy33 still failed at master. | 08:56 |
| *** jp_at_hp has joined #tripleo | 08:56 | |
| yuanying | fixtures-0.3.17 is latest source I found. | 08:58 |
| *** regebro has joined #tripleo | 08:58 | |
| *** untriaged-bot has joined #tripleo | 09:00 | |
| untriaged-bot | No untriaged bugs so far! \o/ | 09:00 |
| *** untriaged-bot has quit IRC | 09:00 | |
| yuanying | StevenK: I think it's another problem, because of difference of handling subprocess. | 09:08 |
| *** akrivoka has joined #tripleo | 09:18 | |
| *** dtantsur|afk is now known as dtantsur | 09:20 | |
| *** loki184 has quit IRC | 09:26 | |
| *** lazy_prince has quit IRC | 09:29 | |
| *** andreaf_ is now known as andreaf | 09:32 | |
| *** dsneddon_ has quit IRC | 09:32 | |
| *** dsneddon_ has joined #tripleo | 09:33 | |
| *** yamahata has quit IRC | 09:34 | |
| *** pradeep has joined #tripleo | 09:35 | |
| *** Dafna has joined #tripleo | 09:40 | |
| *** GheRivero has quit IRC | 09:47 | |
| *** GheRivero has joined #tripleo | 09:47 | |
| *** pblaho has joined #tripleo | 09:48 | |
| *** loki184 has joined #tripleo | 09:48 | |
| *** pelix has joined #tripleo | 09:51 | |
| *** rlandy has joined #tripleo | 09:51 | |
| *** pradeep has quit IRC | 10:05 | |
| *** pradeep has joined #tripleo | 10:06 | |
| *** pradeep has quit IRC | 10:22 | |
| *** sdake has joined #tripleo | 10:32 | |
| *** sdake has quit IRC | 10:32 | |
| *** sdake has joined #tripleo | 10:32 | |
| *** lazy_prince has joined #tripleo | 10:34 | |
| *** lucasagomes_ has joined #tripleo | 10:48 | |
| *** jprovazn_ has joined #tripleo | 10:51 | |
| *** dsneddon_ has quit IRC | 10:54 | |
| *** jp_at_hp has quit IRC | 10:54 | |
| *** lucasagomes has quit IRC | 10:54 | |
| *** jprovazn has quit IRC | 10:54 | |
| *** yuanying has quit IRC | 10:54 | |
| *** wfoster has quit IRC | 10:54 | |
| *** mordred has quit IRC | 10:54 | |
| *** jp_at_hp has joined #tripleo | 10:57 | |
| *** loki184 has quit IRC | 11:02 | |
| *** dsneddon_ has joined #tripleo | 11:03 | |
| *** yuanying has joined #tripleo | 11:03 | |
| *** wfoster has joined #tripleo | 11:03 | |
| *** mordred has joined #tripleo | 11:03 | |
| *** hewbrocca has joined #tripleo | 11:08 | |
| *** rlandy_ has joined #tripleo | 11:13 | |
| *** weshay has quit IRC | 11:15 | |
| *** rlandy has quit IRC | 11:16 | |
| *** tzumainn has joined #tripleo | 11:18 | |
| *** lazy_prince has quit IRC | 11:23 | |
| *** lazy_prince has joined #tripleo | 11:23 | |
| *** sdake has quit IRC | 11:27 | |
| *** sdake has joined #tripleo | 11:28 | |
| *** lucasagomes_ is now known as lucasagomes | 11:28 | |
| *** xuhaiwei has quit IRC | 11:30 | |
| *** lazy_prince has quit IRC | 11:30 | |
| *** jistr has quit IRC | 11:32 | |
| *** yuanying has quit IRC | 11:32 | |
| *** yuanying has joined #tripleo | 11:33 | |
| *** takadayuiko has quit IRC | 11:35 | |
| *** takadayuiko has joined #tripleo | 11:37 | |
| *** yuanying has quit IRC | 11:37 | |
| *** pblaho has quit IRC | 11:38 | |
| *** lucasagomes has quit IRC | 11:44 | |
| *** noslzzp has joined #tripleo | 11:47 | |
| *** jml has quit IRC | 11:52 | |
| *** jistr has joined #tripleo | 11:52 | |
| *** weshay has joined #tripleo | 11:52 | |
| *** jistr is now known as jistr|english | 11:54 | |
| *** lucasagomes has joined #tripleo | 11:54 | |
| *** lucasagomes is now known as lucas-hungry | 11:54 | |
| *** TheJulia has joined #tripleo | 11:54 | |
| *** jml has joined #tripleo | 11:55 | |
| tchaypo | Ng: around? | 11:55 |
| tchaypo | https://fedorahosted.org/fpc/ticket/269 seems relevant to the discussion about static uids | 11:55 |
| Ng | tchaypo: hey, sup? | 11:55 |
| tchaypo | i feel like this is something distros have dealt with; i feel like we should review what they've done before we go and do things they have decided dont work for them | 11:56 |
| Ng | tchaypo: I think they've dealt with it by just picking the next available UID when useradd gets called | 11:57 |
| tchaypo | our problem is on a different scale of course but i feel like we'd do well to think about why/how this situation is different when we think about what to do differently | 11:58 |
| Ng | if we do go the static route, we should definitely look to see if they have reserved UID ranges for static allocations | 11:58 |
| tchaypo | the distros? | 11:58 |
| Ng | yeah | 11:59 |
| *** dprince has joined #tripleo | 11:59 | |
| tchaypo | https://fedoraproject.org/wiki/Packaging:UsersAndGroups seems to be the policy for fedora | 11:59 |
| Ng | I love the idea in comment 1 about a centralised authority that allocates static UIDs. Like IANA for daemons \o/ | 12:00 |
| tchaypo | its called ldap | 12:00 |
| Ng | no, eww ldap | 12:00 |
| *** pradeep has joined #tripleo | 12:01 | |
| tchaypo | if i want to have one, i have one, and im going to be oissed off if tripleo is ignoring it and trampling all over my assigned uids | 12:01 |
| tchaypo | but | 12:01 |
| Ng | yet another controlplane service we have to care for, secure, make HA. Also one we're going to have to make available at build time | 12:01 |
| tchaypo | oh no | 12:01 |
| tchaypo | i dont want to suggest we fo that | 12:01 |
| Ng | but us assigning static UIDs wouldn't stop someone from using LDAP in their deployment | 12:02 |
| tchaypo | i just thunk we should provide an inout so people who want to pre-assign can | 12:02 |
| tchaypo | but maybe this is because i'm thinking at the wrong level. | 12:02 |
| *** regebro has quit IRC | 12:02 | |
| tchaypo | i have a mental image of using ldap to control access to the hardware for sysadmins, and already having GID1000 assigned to the sysadmins group, and then tripleo assigns it to ceilometer | 12:03 |
| *** yuanying has joined #tripleo | 12:03 | |
| tchaypo | maybe sysadmins having access to the nodes isnt a thing that happens in cloudland though? | 12:03 |
| Ng | if they're using ldap, they've already configured libnss to look at LDAP before /etc/passwd though | 12:04 |
| *** takadayuiko has left #tripleo | 12:04 | |
| tchaypo | perhaps. | 12:04 |
| tchaypo | perhaps they give local accounts preference for DR | 12:04 |
| tchaypo | or perhaps they use nsscache and write the ldap groups directly into /etc/passwd | 12:05 |
| tchaypo | i'm not arguing against static assignment, im arguing that the assignment should be configureable | 12:06 |
| Ng | Yeah, that is fine | 12:06 |
| Ng | this is probably something that belongs in a root element, and we can make it as flexible as we want | 12:07 |
| tchaypo | what's a root element? | 12:08 |
| tchaypo | something other things depend on? | 12:08 |
| Ng | one that contains root.d | 12:08 |
| Ng | there can only be one in a build, it's essentially the "chosoe your OS" | 12:08 |
| tchaypo | so we'd provide a tripleo-uids-map element, and other people can fork and override if they want? | 12:09 |
| tchaypo | or put it in the ubuntu element? | 12:10 |
| Ng | I think each element that needs a UID/GID should expose that fact, so the root element can collect the info and either just go with the default values the elements suggest, or apply some operator customisation | 12:10 |
| Ng | (and lint that the resulting IDs make sense) | 12:10 |
| tchaypo | so each element grows a uids.d directory maybe? | 12:11 |
| Ng | the simplest implementation I can think of, would be that the element contains a passwd file and a group file, each containing properly formatted lines. The simplest possible root.d mechanism would just cat those all together, but I suspect it would be better to parse them and feed the data to useradd/groupadd | 12:13 |
| tchaypo | how would operators override? | 12:13 |
| tchaypo | fork every element they want to twiddle? | 12:13 |
| *** pradeep has quit IRC | 12:14 | |
| tchaypo | .... That seems to sound a good deal more argumentative than I intended | 12:16 |
| TheJulia | Perhaps have the logic check to see if the user or group is already present | 12:16 |
| *** yamahata has joined #tripleo | 12:17 | |
| Ng | I think that would be overly burdensome. We could go the REPOREF style route and throw a million env vars at the problem, we could allow them to inject passwd/group file fragments that trump the element defaults, we could make sure that they're able to trump whatever we do with nsswitch.conf, we could allow them to disable us from doing anything because they | 12:17 |
| Ng | have it covered with some other libnss thing they're going to do. I don't know exactly :) | 12:17 |
| *** rhallisey has quit IRC | 12:17 | |
| Ng | yeah, any useradd/groupadd should always be guarded by a check for whether it already exists. If they trump us with their own mechanism, that's fine | 12:18 |
| Ng | we could allow for a single element to provide some kind of map and/or script that trumps whatever we were going to do | 12:18 |
| *** dsneddon_ has quit IRC | 12:18 | |
| sdake | morning | 12:19 |
| *** dsneddon_ has joined #tripleo | 12:20 | |
| Ng | whatever we do, we need to discuss it out in pretty significant detail, because we need to get it absolutely right | 12:20 |
| Ng | this was a terrible oversight and it's already going to be complicated enough to fix, without trying to do it a second time ;) | 12:20 |
| *** regebro has joined #tripleo | 12:23 | |
| *** jdob has joined #tripleo | 12:24 | |
| tchaypo | i was suggesting we take a json map | 12:35 |
| tchaypo | which elements should be able to add to easily - jq makes it easy to munge them together | 12:35 |
| StevenK | "easy" | 12:35 |
| tchaypo | but its also easy for operators to provide a map if they have preferences | 12:35 |
| tchaypo | includimg scraping from existing images if they have any... | 12:36 |
| *** yamahata has quit IRC | 12:36 | |
| *** yamahata has joined #tripleo | 12:37 | |
| *** dsneddon_ has quit IRC | 12:39 | |
| *** noslzzp has quit IRC | 12:41 | |
| chmouel | sdake: hey there would be nice to get this merged https://review.openstack.org/#/c/126484/ so we can get the gate setup after that https://review.openstack.org/#/c/126485/ | 12:41 |
| *** dsneddon_ has joined #tripleo | 12:41 | |
| *** lucas-hungry is now known as lucasagomes | 12:44 | |
| sdake | chmouel I +2'ed the patch | 12:45 |
| sdake | jpeeler rhallisey larsks radez any of you guys around to ack the patch? | 12:45 |
| chmouel | sdake: tks! | 12:45 |
| larsks | sdake: commuting.... At office soon. | 12:46 |
| sdake | larsks tx | 12:47 |
| *** openstackgerrit has joined #tripleo | 12:48 | |
| *** julim has joined #tripleo | 12:56 | |
| *** pensu has quit IRC | 12:56 | |
| *** morazi has joined #tripleo | 12:58 | |
| *** noslzzp has joined #tripleo | 12:58 | |
| *** rhallisey has joined #tripleo | 12:58 | |
| *** spzala has joined #tripleo | 12:58 | |
| *** pblaho has joined #tripleo | 12:59 | |
| *** radez_g0n3 is now known as radez | 12:59 | |
| *** _crobertsrh is now known as crobertsrh | 13:01 | |
| *** pradk has quit IRC | 13:04 | |
| openstackgerrit | therese-mchale proposed a change to openstack/tripleo-heat-templates: Secure the MySQL bind-address on the undercloud https://review.openstack.org/126551 | 13:04 |
| *** jistr|english is now known as jistr | 13:07 | |
| *** sdake_ has joined #tripleo | 13:09 | |
| openstackgerrit | A change was merged to openstack/tuskar: Refactor the Tuskar storage tests https://review.openstack.org/125660 | 13:13 |
| *** dprince has quit IRC | 13:15 | |
| *** dprince has joined #tripleo | 13:16 | |
| openstackgerrit | A change was merged to openstack/tuskar-ui: Add a generic function for filtering API objects https://review.openstack.org/126259 | 13:17 |
| *** jprovazn_ has quit IRC | 13:19 | |
| *** morazi has quit IRC | 13:20 | |
| *** jprovazn has joined #tripleo | 13:21 | |
| *** nosnos has quit IRC | 13:24 | |
| openstackgerrit | A change was merged to openstack/tuskar-ui: Changed service configuration table to multiple tabs https://review.openstack.org/126087 | 13:25 |
| *** morazi has joined #tripleo | 13:26 | |
| *** sdake_ has quit IRC | 13:30 | |
| *** pradeep has joined #tripleo | 13:34 | |
| *** dsneddon_ has quit IRC | 13:43 | |
| *** jprovazn has quit IRC | 13:44 | |
| openstackgerrit | A change was merged to stackforge/kolla: Validate JSON https://review.openstack.org/126484 | 13:44 |
| *** dsneddon_ has joined #tripleo | 13:45 | |
| *** jprovazn_ has joined #tripleo | 13:45 | |
| larsks | sdake: here now; what review did you want me to look at? | 13:48 |
| *** killer_prince has joined #tripleo | 13:48 | |
| *** killer_prince is now known as lazy_prince | 13:48 | |
| jpeeler | i think the one i just approved | 13:49 |
| larsks | jpeeler: spiffy. | 13:54 |
| larsks | chmouel: can we use a local tool (like "jq") for reformatting, rather than jsonlint.com? Then we could even build that into local scripts. | 13:55 |
| *** andreaf is now known as andreaf_ | 13:57 | |
| *** marun has joined #tripleo | 13:58 | |
| rbrady | marios: thanks for the recheck...it seems to be failing for an auth issue that has nothing to do with my patch. I've been trying to gather enough info to file a bug. | 13:59 |
| chmouel | larsks: yeah sure, what about adding it to tools/validate-samples.sh (only if jq is in the path) | 13:59 |
| larsks | chmouel: I was thinking of some sort of "reformat-json" script. | 13:59 |
| chmouel | larsks: I see, yeah, good idea! | 14:00 |
| larsks | Actually, just use 'python -mjson.tool', since that is pretty much guaranteed to be available. | 14:00 |
| larsks | Don't bother making it conditonal on other stuff. | 14:00 |
| larsks | And then we can have consistent formatting for json files, rather than weird whitespace conflicts because your editor formats json differently from my editor. | 14:01 |
| *** julim has quit IRC | 14:01 | |
| *** julim has joined #tripleo | 14:04 | |
| *** dsneddon_ has quit IRC | 14:05 | |
| *** dsneddon_ has joined #tripleo | 14:06 | |
| *** radez is now known as radez_g0n3 | 14:08 | |
| openstackgerrit | Ladislav Smola proposed a change to openstack/tripleo-heat-templates: Adding SNMP related parameters to storage templates https://review.openstack.org/126562 | 14:14 |
| bnemec | SpamapS: tchaypo: I'm basically our de facto Oslo liaison since I'm already at all of the meetings. :-) | 14:15 |
| derekh | Hi All, nova is proposing to remove the baremetal driver https://review.openstack.org/#/c/126560/ | 14:16 |
| *** pradk has joined #tripleo | 14:17 | |
| derekh | So I've propsed this https://review.openstack.org/#/c/126563/ | 14:17 |
| derekh | Basically only running CI using ironic, anybody think we should do something else ? | 14:17 |
| *** pblaho has quit IRC | 14:18 | |
| *** pblaho has joined #tripleo | 14:18 | |
| *** radez_g0n3 is now known as radez | 14:19 | |
| *** ramishra has joined #tripleo | 14:26 | |
| *** regebro is now known as regebro|afk | 14:27 | |
| lxsli | +1 | 14:29 |
| GheRivero | long live baremetal! | 14:31 |
| *** dprince has quit IRC | 14:35 | |
| *** pcrews has joined #tripleo | 14:36 | |
| sdake | larsks sorry went back to bed for a couple hours | 14:38 |
| sdake | looks like jpeeler got it tho | 14:38 |
| openstackgerrit | Lars Kellogg-Stedman proposed a change to stackforge/kolla: remove "master" from mariadb k8s configs https://review.openstack.org/126454 | 14:40 |
| *** mrunge has quit IRC | 14:40 | |
| larsks | sdake: yup, think that is all set. can you take a look at a couple of trivial changes? https://review.openstack.org/#/c/126454/ and https://review.openstack.org/#/c/126453/? | 14:40 |
| openstackgerrit | stephen-ma proposed a change to openstack/tripleo-image-elements: Add Nova Admin access to neutron.conf https://review.openstack.org/126570 | 14:42 |
| sdake | larsks I dont think that rename to public keystone services will work - isn't there a reference in the ENV to one of those keystone services? | 14:43 |
| larsks | sdake: You might be right. Let me go fix that. | 14:44 |
| openstackgerrit | A change was merged to stackforge/kolla: Remove the trailing space in README.md https://review.openstack.org/126491 | 14:50 |
| chmouel | is using supervise within the containers would be a no-no? I understand that in some cases it's one service for one container but in the case of swift it's not something doable to do | 14:51 |
| larsks | chmouel: what services need to run in the same container? | 14:51 |
| * larsks <-- is not familiar with swift. | 14:52 | |
| chmouel | larsks: there is bunch of them, but at least the service type i.e: account or container or object daemon and the replication service on the same node | 14:52 |
| *** chuckC_ has quit IRC | 14:52 | |
| chmouel | larsks: there is the integrity one as well and some others probably | 14:52 |
| larsks | chmouel: what actually needs to be shared, though? remember that all the containers in the pod are sharing the same network namespace. Is that sufficient? Or do they need to share storage? And can that be handled with volumes-from? | 14:53 |
| chmouel | larsks: well that's another thing to figure out cause if we don't use host volumes directly mapped to a disk within containers we would need to use loopback which has some very bad performances | 14:54 |
| openstackgerrit | A change was merged to stackforge/kolla: remove "master" from mariadb k8s configs https://review.openstack.org/126454 | 14:54 |
| larsks | chmouel: I didn't follow all that...specfically, what do you mean by "use loopback"? | 14:56 |
| larsks | I was envisioning that one container the pod would handle the storage, and all the other containers would use volumes-from that container. So now you have shared state and shared network namespace, so maybe you can have one service/container...? | 14:56 |
| *** jprovazn_ has quit IRC | 14:57 | |
| larsks | I am ignoring the lack of supportin k8s for persisnent host storage at the moment (much as we are for the glance and mariadb containers right now). | 14:57 |
| openstackgerrit | Radomir Dopieralski proposed a change to openstack/tuskar-ui: Merge register nodes and autodiscover nodes forms https://review.openstack.org/126576 | 14:58 |
| sdake | larsks I agree there is nothing we can do on that point yet | 14:59 |
| sdake | I think where we should be headed is make it work non-persistent | 14:59 |
| radez | larsks: hey are your glance changes posted anywhere? I'm not sure I've seen them come through yet | 14:59 |
| sdake | by that time, upstream k8s will have persistent storager added to their code base | 14:59 |
| sdake | jpeeler was rabbitmq busted in our repo or did you sort that out? | 15:00 |
| *** untriaged-bot has joined #tripleo | 15:00 | |
| untriaged-bot | No untriaged bugs so far! \o/ | 15:00 |
| *** untriaged-bot has quit IRC | 15:00 | |
| larsks | radez: well, they are still posted where they were yesterday! :) But they are not in the review chain yet. | 15:00 |
| larsks | I will try to get to that today, meeting permitting. | 15:00 |
| radez | sure, where were they posted yesterday? | 15:00 |
| openstackgerrit | Lars Kellogg-Stedman proposed a change to stackforge/kolla: use renamed mariadb services https://review.openstack.org/126578 | 15:00 |
| *** chuckC_ has joined #tripleo | 15:02 | |
| *** pensu has joined #tripleo | 15:02 | |
| jpeeler | sdake: i know jlabocki looked at it some yesterday, but i haven't checked yet | 15:03 |
| openstackgerrit | Ryan Hallisey proposed a change to stackforge/kolla: Initial setup for nova-controller https://review.openstack.org/126305 | 15:03 |
| *** pensu has quit IRC | 15:03 | |
| *** rushiagr is now known as rushiagr_away | 15:04 | |
| larsks | radez: email to openstack-dev | 15:04 |
| radez | larsks: ah, yes I remmeber seeing that now... thx | 15:04 |
| larsks | radez: let me get the link... | 15:04 |
| larsks | radez: https://github.com/larsks/kolla/tree/larsks/hautoproxy | 15:05 |
| sdake | jpeeler I dont think jlabocki will have much time this week to look at it, so if yo ucould tackle it that would rock :) | 15:07 |
| *** pradeep1 has joined #tripleo | 15:07 | |
| *** dsneddon_ has quit IRC | 15:08 | |
| *** pradeep has quit IRC | 15:09 | |
| *** dsneddon_ has joined #tripleo | 15:10 | |
| openstackgerrit | Lars Kellogg-Stedman proposed a change to stackforge/kolla: renamed keystone services https://review.openstack.org/126453 | 15:11 |
| *** dshulyak_ has quit IRC | 15:14 | |
| *** dshulyak_ has joined #tripleo | 15:15 | |
| *** akrivoka has quit IRC | 15:16 | |
| openstackgerrit | Steven Dake proposed a change to stackforge/kolla: Remove -master from README.md https://review.openstack.org/126590 | 15:17 |
| openstackgerrit | Alexis Lee proposed a change to openstack/tripleo-image-elements: Migrate hard-coded values from deploy-ci-overcloud https://review.openstack.org/115862 | 15:22 |
| chmouel | larsks: about container and data, we could potentially do that but that would make the cluster a only one replica cluster | 15:22 |
| larsks | chmouel: Things that can be replicated presumably don't need the shared network/storage, so would not need to run in the same pod. | 15:22 |
| larsks | So we should be able to split those services into separate pods. | 15:23 |
| larsks | ....right? | 15:23 |
| clarkb | greghaynes: https://review.openstack.org/#/c/125475/ see comment | 15:27 |
| chmouel | larsks: yeah i think for replication and things but that's going to make a lot of containers tho | 15:30 |
| openstackgerrit | Jon-Paul Sullivan (jp_at_hp) proposed a change to openstack/tripleo-incubator: Revert "Wait for memory and vcpus from hypervisor stats" https://review.openstack.org/126596 | 15:30 |
| jp_at_hp | slagle: there is an issue with https://review.openstack.org/#/c/124105/ - the nodes array outside of tripleo CI can contain the undercloud, meaning that we would be waiting on at least 1 more node than is ever going to register, I think - hence the reversion | 15:30 |
| larsks | chmouel: I think that's probably okay. | 15:31 |
| chmouel | larsks: but i don't think i'll have replication at first working, there is still things to figure out about the rings and placement, infromation i don't know in kubernetes | 15:31 |
| *** Hefeweizen has joined #tripleo | 15:31 | |
| jp_at_hp | slagle: I think it needs a check of -z "${ALL_NODES:-}" to determine whether or not to include .nodes[0] in the calculations. | 15:31 |
| larsks | chmouel: Okay, so, let's ignore replication then, and outline what we need as a starting point. | 15:31 |
| *** k4n0 has quit IRC | 15:33 | |
| openstackgerrit | Ladislav Smola proposed a change to openstack/tripleo-heat-templates: Adding SNMP related parameters to storage templates https://review.openstack.org/126562 | 15:34 |
| openstackgerrit | Lars Kellogg-Stedman proposed a change to stackforge/kolla: fix issues with glance scripts https://review.openstack.org/126599 | 15:35 |
| larsks | radez: there they are ^^^^ | 15:35 |
| radez | woot | 15:35 |
| slagle | jp_at_hp: yea, good point | 15:35 |
| slagle | jp_at_hp: for some reason i was thinking the nodes array was updated to remove the ones used | 15:35 |
| radez | larsks: I'll try and pull the changes and test them after lunch... barring meeting avilablilty :) | 15:37 |
| jp_at_hp | slagle: gotta love the way that it passed ci so happily though, I knew there was a tradeoff there somewhere ;-) | 15:37 |
| larsks | radez: breakage possible -- I am trying to tease apart my "fixing glance" changes from my "solve service discovery" changes, so there has been a lot of rebasing going on. | 15:38 |
| radez | larsks: I remember you saying that and wanted to test to help flesh out that separation. I'll let you know if there's any updates that need to be made | 15:39 |
| *** ramishra has quit IRC | 15:41 | |
| *** daneyon_ has left #tripleo | 15:46 | |
| *** daneyon_ has joined #tripleo | 15:46 | |
| openstackgerrit | Lars Kellogg-Stedman proposed a change to stackforge/kolla: fix issues with glance scripts https://review.openstack.org/126599 | 15:47 |
| daneyon_ | I am trying to setup my kolla dev env based on https://github.com/larsks/heat-kubernetes Does an F20 image with the updated kernel for vxlan exist anywhere so I don't have to build it myself? | 15:49 |
| larsks | daneyon_: you can just use the vanilla F20 cloud image and then reboot your minions once things are finished installing. | 15:49 |
| larsks | I think Fedora may have unofficial updated cloud images... | 15:50 |
| larsks | ...but I have not had luck trying to find those (the build infra has been having issues every time I look) | 15:50 |
| daneyon_ | I'll do that. Thanks larsks. | 15:50 |
| *** jml has quit IRC | 15:50 | |
| *** jml has joined #tripleo | 15:52 | |
| openstackgerrit | Phil Neal proposed a change to openstack/tripleo-image-elements: Add Ceilometer configuration ttl setting for expirer https://review.openstack.org/126606 | 15:55 |
| openstackgerrit | Phil Neal proposed a change to openstack/tripleo-heat-templates: Add Ceilometer configuration ttl setting for expirer https://review.openstack.org/126607 | 16:01 |
| openstackgerrit | Phil Neal proposed a change to openstack/tripleo-image-elements: Add Ceilometer configuration ttl setting for expirer https://review.openstack.org/126606 | 16:03 |
| sdake | daneyon_ use http://paste.fedoraproject.org/139962/41269795 to create images | 16:06 |
| sdake | this way teardown/setup is fast | 16:06 |
| *** akrivoka has joined #tripleo | 16:06 | |
| sdake | and it has an updated kernel + apiserver | 16:06 |
| daneyon_ | sdake_ I have not used virt-customize before. Do I run virt-customize after downloading and running the standard f20 qcow2 image? | 16:09 |
| sdake | downlaod f20 image | 16:09 |
| sdake | name it fedora-cloud.qcow2 | 16:10 |
| sdake | run that vc script | 16:10 |
| daneyon_ | sdake_ OK. I think I got it. | 16:10 |
| sdake | register fedora-cloud.qcow2 with glance | 16:10 |
| sdake | use that as the image name | 16:10 |
| sdake | daneyon_ my local.yaml file -> http://paste.fedoraproject.org/139966/69837914 | 16:13 |
| sdake | larsks say the developer workflow rocks - nice job on that! | 16:13 |
| larsks | sdake: Thanks! Glad it's working out. | 16:14 |
| sdake | so much better then those hacky build scripts prior | 16:14 |
| sdake | daneyon_ btw, jpeeler wrote that vc script - just passing it around | 16:15 |
| openstackgerrit | A change was merged to stackforge/kolla: use renamed mariadb services https://review.openstack.org/126578 | 16:15 |
| jpeeler | sdake: there's a slightly different script in the repo: https://github.com/larsks/heat-kubernetes/blob/master/get-image.sh | 16:16 |
| sdake | jpeeler ok cool | 16:16 |
| sdake | I just had this on my hd from prior | 16:16 |
| jpeeler | well just thought i would mention it since daneyon_ was asking how to use virt-customize | 16:17 |
| *** pradeep has joined #tripleo | 16:18 | |
| *** yamahata has quit IRC | 16:18 | |
| slagle | lsu | 16:21 |
| *** pradeep1 has quit IRC | 16:21 | |
| slagle | err, ww | 16:21 |
| openstackgerrit | A change was merged to stackforge/kolla: renamed keystone services https://review.openstack.org/126453 | 16:21 |
| *** Ng has quit IRC | 16:22 | |
| *** rushiagr_away is now known as rushiagr | 16:23 | |
| *** pradeep1 has joined #tripleo | 16:23 | |
| *** pradeep has quit IRC | 16:24 | |
| openstackgerrit | Derek Higgins proposed a change to openstack-infra/tripleo-ci: Remove cherry pick's https://review.openstack.org/126612 | 16:26 |
| openstackgerrit | Phil Neal proposed a change to openstack/tripleo-image-elements: Passthrough-enable ceilometer.conf https://review.openstack.org/126614 | 16:27 |
| *** ifarkas has quit IRC | 16:28 | |
| *** jistr has quit IRC | 16:29 | |
| *** dshulyak_ has quit IRC | 16:29 | |
| openstackgerrit | Lars Kellogg-Stedman proposed a change to stackforge/kolla: fix issues with glance scripts https://review.openstack.org/126599 | 16:32 |
| *** daneyon_ has quit IRC | 16:34 | |
| *** dprince has joined #tripleo | 16:34 | |
| openstackgerrit | Lars Kellogg-Stedman proposed a change to stackforge/kolla: do not add image prefix when building https://review.openstack.org/126618 | 16:34 |
| *** daneyon has joined #tripleo | 16:34 | |
| *** dtantsur is now known as dtantsur|afk | 16:35 | |
| SpamapS | bnemec: oh lovely. :-D | 16:37 |
| SpamapS | bnemec: (re oslo liason) | 16:37 |
| *** bvivek has joined #tripleo | 16:38 | |
| openstackgerrit | Lars Kellogg-Stedman proposed a change to stackforge/kolla: fix issues with glance scripts https://review.openstack.org/126599 | 16:39 |
| sdake | radez what runs on port 9292 - rabbitmq? | 16:39 |
| radez | sdake: glance api | 16:40 |
| sdake | how do we expose the container port for rabbitmq without triggering the proxy? | 16:40 |
| *** pradeep2 has joined #tripleo | 16:40 | |
| sdake | http://paste.fedoraproject.org/139976/41270009 | 16:41 |
| sdake | I had thought 9292 was the rabbit port above | 16:42 |
| *** pblaho has quit IRC | 16:42 | |
| *** pradeep3 has joined #tripleo | 16:42 | |
| *** pradeep1 has quit IRC | 16:43 | |
| *** zigo has quit IRC | 16:43 | |
| *** pradeep2 has quit IRC | 16:45 | |
| openstackgerrit | Lars Kellogg-Stedman proposed a change to stackforge/kolla: fix issues with glance scripts https://review.openstack.org/126599 | 16:45 |
| *** zigo has joined #tripleo | 16:46 | |
| *** dshulyak_ has joined #tripleo | 16:49 | |
| larsks | 9292 is glance api. 5672 is rabbitmq or qpid. | 16:49 |
| larsks | btw...Sorry for all the review spam, I am cleaning up those glance changes and keep finding something new. | 16:50 |
| openstackgerrit | Lars Kellogg-Stedman proposed a change to stackforge/kolla: fix issues with glance scripts https://review.openstack.org/126599 | 16:50 |
| *** pensu has joined #tripleo | 16:51 | |
| larsks | sdake: I don't understand your question about rabbitq re: "without triggering the proxy"... | 16:51 |
| *** pradeep3 has quit IRC | 16:53 | |
| *** derekh has quit IRC | 16:53 | |
| *** pradeep1 has joined #tripleo | 16:54 | |
| openstackgerrit | therese-mchale proposed a change to openstack/tripleo-image-elements: Allow configurable garp_master_refresh https://review.openstack.org/126622 | 16:56 |
| *** akrivoka has quit IRC | 16:56 | |
| *** Ng_ has joined #tripleo | 16:59 | |
| *** ChanServ sets mode: +v Ng_ | 16:59 | |
| *** rdopieralski has quit IRC | 17:00 | |
| openstackgerrit | Monty Taylor proposed a change to openstack/diskimage-builder: WIP: Use docker as engine https://review.openstack.org/126194 | 17:07 |
| *** pradeep2 has joined #tripleo | 17:08 | |
| *** rushiagr is now known as rushiagr_away | 17:09 | |
| *** pradeep1 has quit IRC | 17:10 | |
| *** dshulyak_ has quit IRC | 17:14 | |
| *** andreaf has joined #tripleo | 17:17 | |
| *** spzala has quit IRC | 17:19 | |
| sdake | larsks the containers need to communicate to rabbitmq | 17:26 |
| larsks | sdake: Right. | 17:26 |
| sdake | if we put a portcontainer flag, wont it proxy to the wrong place? | 17:26 |
| larsks | the containerPort is for services offered by your container. | 17:26 |
| larsks | To access rabbitmq, you would just use $RABBITMQ_PORT_5672_TCP_ADDR | 17:26 |
| sdake | ok thanks | 17:26 |
| sdake | that should be fine then | 17:27 |
| larsks | (assuming that corresponding service is named "rbabitmq" and is active and the pod is running) | 17:27 |
| *** pradeep1 has joined #tripleo | 17:32 | |
| *** pradeep2 has quit IRC | 17:34 | |
| *** lucasagomes is now known as lucas-dinner | 17:35 | |
| jpeeler | what log file should i be watching to note progress on a pod that is still waiting? | 17:35 |
| *** lucas-dinner has quit IRC | 17:36 | |
| *** lucasagomes has joined #tripleo | 17:36 | |
| *** lucasagomes is now known as lucas-dinner | 17:36 | |
| sdake | jpeeler if a pod is waiting it is likely downloading | 17:37 |
| sdake | you can login to the machine | 17:37 |
| sdake | and run sudo docker images | 17:37 |
| sdake | this will confirm if the image is downloded or not | 17:37 |
| sdake | afaik there are no logs :( | 17:37 |
| sdake | once the image is downloaded | 17:37 |
| sdake | you can do | 17:37 |
| sdake | sudo docker ps -a | 17:37 |
| sdake | this will show all processes that have ever been started by docker | 17:37 |
| sdake | sudo docker logs xyz is anotehr useful command | 17:38 |
| sdake | the container should log to stderr | 17:38 |
| *** jp_at_hp has quit IRC | 17:38 | |
| jpeeler | cool, just thought there might be something for before a container actually is running | 17:38 |
| sdake | if you find it let me know pls :) | 17:38 |
| *** marun has quit IRC | 17:38 | |
| sdake | libvirt installs 140 packages - true story :) | 17:39 |
| *** pradeep2 has joined #tripleo | 17:39 | |
| jpeeler | if my deployment is actually downloading the image, why doesn't docker show the complete picture in docker images? | 17:39 |
| sdake | it will show via docker images | 17:39 |
| sdake | in the left side it won't be labeled | 17:39 |
| sdake | in the right side, the created will count down | 17:40 |
| sdake | to the last created one | 17:40 |
| sdake | logging is clunky - not sure if there is anything better | 17:40 |
| jpeeler | so a <none> is probably the image being downloaded? | 17:40 |
| sdake | yup | 17:40 |
| *** bvivek has quit IRC | 17:40 | |
| *** pradeep1 has quit IRC | 17:40 | |
| sdake | and if you repeat docker images | 17:40 |
| sdake | for each layer of the image the date will update | 17:40 |
| *** marun has joined #tripleo | 17:40 | |
| *** jcoufal has quit IRC | 17:42 | |
| *** ramishra has joined #tripleo | 17:42 | |
| *** marun has quit IRC | 17:42 | |
| jpeeler | sdake: i guess i can't use kollaglue for rabbitmq because nothing has been pushed yet? | 17:43 |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/tuskar: Updated from global requirements https://review.openstack.org/122674 | 17:46 |
| *** marun has joined #tripleo | 17:46 | |
| *** rushiagr_away is now known as rushiagr | 17:46 | |
| *** lucas-dinner has quit IRC | 17:47 | |
| larsks | jpeeler: were you going to work on the rabbitmq container? | 17:47 |
| *** ramishra has quit IRC | 17:47 | |
| jpeeler | larsks: trying to, but i'm still trying to learn the basics here | 17:47 |
| larsks | jpeeler: okay. anything I can help you with? | 17:47 |
| *** pradeep1 has joined #tripleo | 17:48 | |
| jpeeler | well you can confirm that the image is missing from kollaglue because it hasn't been finished, and that i need to build locally and test with that image right? | 17:48 |
| sdake | jpeeler yes, make sure to set tag to latest and your namespace to jpeeler | 17:49 |
| sdake | TAG=latest | 17:49 |
| *** athomas has quit IRC | 17:49 | |
| sdake | NAMESPACE=jpeeler | 17:49 |
| sdake | in the file | 17:49 |
| sdake | see in the docs directory | 17:49 |
| sdake | that will push to your personal docker hub account | 17:50 |
| *** pradeep2 has quit IRC | 17:50 | |
| sdake | iirc its called .buildconf | 17:51 |
| *** funzo has quit IRC | 17:54 | |
| *** funzo has joined #tripleo | 17:55 | |
| *** cwolferh has joined #tripleo | 17:56 | |
| *** ramishra has joined #tripleo | 17:58 | |
| larsks | sdake: correct, re: .buildconf, in the repository root. | 17:58 |
| larsks | jpeeler: yes, nobody else has worked on the rabbitmq image yet. | 17:58 |
| larsks | jpeeler: but note that you will need to push the image remotely before you can use it with kubernetes (kube does a "pull" before starting a pod) | 17:59 |
| *** cwolferh_ has quit IRC | 17:59 | |
| *** lucasagomes has joined #tripleo | 18:01 | |
| *** jcoufal has joined #tripleo | 18:02 | |
| *** pradeep2 has joined #tripleo | 18:02 | |
| *** ramishra has quit IRC | 18:02 | |
| *** pradeep1 has quit IRC | 18:03 | |
| *** uberj_ is now known as uberj | 18:04 | |
| *** uberj has quit IRC | 18:04 | |
| *** uberj has joined #tripleo | 18:04 | |
| *** dshulyak_ has joined #tripleo | 18:08 | |
| *** pradeep1 has joined #tripleo | 18:09 | |
| *** pradeep2 has quit IRC | 18:13 | |
| *** spzala has joined #tripleo | 18:15 | |
| *** jprovazn has joined #tripleo | 18:18 | |
| openstackgerrit | Steven Dake proposed a change to stackforge/kolla: Add service files for libvirt and compute https://review.openstack.org/126648 | 18:21 |
| openstackgerrit | Steven Dake proposed a change to stackforge/kolla: Modify nova-compute pod to operate properly https://review.openstack.org/126649 | 18:21 |
| openstackgerrit | Steven Dake proposed a change to stackforge/kolla: Get to a functional nova-libvirt start https://review.openstack.org/126650 | 18:21 |
| sdake | larsks/jpeeler/radez can you folks review that ^^ | 18:21 |
| jpeeler | since we're on stackforge, can we maybe just do one core reviewer until milestone 2? | 18:24 |
| *** lucasagomes has quit IRC | 18:25 | |
| openstackgerrit | A change was merged to stackforge/kolla: fix issues with glance scripts https://review.openstack.org/126599 | 18:26 |
| openstackgerrit | A change was merged to stackforge/kolla: do not add image prefix when building https://review.openstack.org/126618 | 18:27 |
| openstackgerrit | A change was merged to stackforge/kolla: Remove -master from README.md https://review.openstack.org/126590 | 18:28 |
| openstackgerrit | A change was merged to stackforge/kolla: Add service files for libvirt and compute https://review.openstack.org/126648 | 18:28 |
| openstackgerrit | A change was merged to stackforge/kolla: Modify nova-compute pod to operate properly https://review.openstack.org/126649 | 18:28 |
| *** dshulyak_ has quit IRC | 18:29 | |
| openstackgerrit | A change was merged to stackforge/kolla: Initial setup for nova-controller https://review.openstack.org/126305 | 18:29 |
| openstackgerrit | A change was merged to stackforge/kolla: Get to a functional nova-libvirt start https://review.openstack.org/126650 | 18:29 |
| *** lucasagomes has joined #tripleo | 18:39 | |
| openstackgerrit | Dougal Matthews proposed a change to openstack/tuskar: Update Tuskar Install Documentation https://review.openstack.org/125596 | 18:42 |
| openstackgerrit | Ghe Rivero proposed a change to openstack/diskimage-builder: Move install bin from rpm-distro to yum https://review.openstack.org/126654 | 18:44 |
| *** jtomasek has quit IRC | 18:47 | |
| sdake | rhallisey I think nova-compute should be responsible for creating the nova db | 18:48 |
| sdake | and populating keystone therein | 18:48 |
| *** jtomasek has joined #tripleo | 18:48 | |
| rhallisey | sdake, ok that's fine, I wasn't sure what I needed so I used what you had as a template | 18:51 |
| *** rushiagr is now known as rushiagr_away | 18:55 | |
| *** rhallisey has quit IRC | 18:56 | |
| *** rhallisey has joined #tripleo | 18:57 | |
| *** ramishra has joined #tripleo | 18:58 | |
| gfidente | meeting time? :) | 19:00 |
| *** Ng_ is now known as Ng | 19:02 | |
| SpamapS | aye | 19:02 |
| *** lucasagomes has quit IRC | 19:03 | |
| *** ramishra has quit IRC | 19:04 | |
| openstackgerrit | Steven Dake proposed a change to stackforge/kolla: chmod +x start.sh for nova-compute/nova-compute https://review.openstack.org/126669 | 19:06 |
| openstackgerrit | Steven Dake proposed a change to stackforge/kolla: chmod +x nova-compute/nova-libvirt/start.sh https://review.openstack.org/126670 | 19:06 |
| sdake | well that sort of works | 19:06 |
| sdake | libvirtd needs the debugging | 19:07 |
| sdake | by sort of, I mean libvirtd starts and spits out errors :( | 19:07 |
| *** pradeep2 has joined #tripleo | 19:07 | |
| *** pradeep1 has quit IRC | 19:08 | |
| sdake | now we need a functional rabbitmq and we can start up nova :) | 19:11 |
| * jpeeler is debugging "failed to find an IP for pod" for rabbitmq | 19:13 | |
| larsks | jpeeler: have you pushed your changes somewhere so we can look at/help debug? | 19:13 |
| *** lucasagomes has joined #tripleo | 19:15 | |
| larsks | 046771 | 19:16 |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/os-collect-config: Updated from global requirements https://review.openstack.org/120606 | 19:16 |
| jpeeler | larsks: i haven't actually made any changes yet aside from what's already posted here https://review.openstack.org/#/c/126037/1/docker/rabbitmq/Dockerfile,unified | 19:17 |
| larsks | jpeeler: Okee dokee. | 19:17 |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/python-tuskarclient: Updated from global requirements https://review.openstack.org/126684 | 19:17 |
| jpeeler | larsks: but it doesn't appear that the container is even starting | 19:17 |
| larsks | Let me take a look. | 19:18 |
| jpeeler | i may just try running in docker directly - i assume that's an okay place to start | 19:19 |
| *** larstobi has joined #tripleo | 19:21 | |
| openstackgerrit | Lars Kellogg-Stedman proposed a change to stackforge/kolla: Updated base image in Dockerfile and changed CMD to reference correct file https://review.openstack.org/126037 | 19:23 |
| jpeeler | larsks: if the container immediately exits and docker logs returns nothing, how do you determine why a container exited? | 19:23 |
| larsks | jpeeler: ^^^ that is just a rebase on master. | 19:23 |
| larsks | jpeeler: start a shell in the container, then run the start script by hand. | 19:24 |
| sdake | jpeeler is your start.sh +x? | 19:35 |
| jpeeler | sdake: it is | 19:35 |
| sdake | run manually | 19:35 |
| sdake | it will show you the error that way | 19:35 |
| sdake | (dont run with kube) | 19:35 |
| jpeeler | i am, an the broker seems to be going | 19:35 |
| *** lucasagomes has quit IRC | 19:36 | |
| jpeeler | don't see anything listening on port 5672 though | 19:36 |
| larsks | jpeeler: check logs? /var/log/rabbitmq I think. | 19:38 |
| larsks | jpeeler: I just built that container with no changes; if I run /start.sh inside the container, I end up with a service listening on port 5672... | 19:40 |
| openstackgerrit | Steven Dake proposed a change to stackforge/kolla: chmod +x nova-compute/nova-libvirt/start.sh https://review.openstack.org/126670 | 19:40 |
| openstackgerrit | Steven Dake proposed a change to stackforge/kolla: chmod +x start.sh for nova-compute/nova-compute https://review.openstack.org/126669 | 19:40 |
| openstackgerrit | Steven Dake proposed a change to stackforge/kolla: Modify start.sh for nova-compute/nova-compute https://review.openstack.org/126700 | 19:40 |
| openstackgerrit | Steven Dake proposed a change to stackforge/kolla: Change sdake/image to kollaglue/image https://review.openstack.org/126701 | 19:40 |
| larsks | 2# ss -tlnp | grep 5672 | 19:40 |
| larsks | LISTEN 0 128 :::5672 :::* users:(("beam.smp",pid=8,fd=15)) | 19:40 |
| jpeeler | larsks: yeah, i forgot -l to ss >_< | 19:41 |
| larsks | :) | 19:41 |
| sdake | jlabocki had indicated it does not work however | 19:41 |
| sdake | time for lunch for me, bbiaf :) | 19:41 |
| jpeeler | right, at the container level i'm not seeing anything wrong | 19:41 |
| sdake | do you have containerPort in your json? | 19:42 |
| jpeeler | that is present | 19:42 |
| sdake | and expose? | 19:42 |
| larsks | jpeeler: the .json file is using the wrong image. | 19:42 |
| larsks | "image": "docker.usersys.redhat.com/larsks/rabbitmq", | 19:42 |
| jpeeler | right, i changed it to mine | 19:43 |
| sdake | oh ya that will do it :) | 19:43 |
| larsks | Okay. | 19:43 |
| sdake | ok lunch bbiaf | 19:43 |
| sdake | good luck :) | 19:43 |
| jpeeler | not seeing an expose anywhere yet | 19:43 |
| *** pradeep has joined #tripleo | 19:44 | |
| *** pradeep2 has quit IRC | 19:44 | |
| larsks | jpeeler: you wouldn't necessarily need an expose as long as the .json file has the correct containerPort. And in any case, it should still run. | 19:47 |
| jpeeler | ok, if the image is fine as is, time to look into the kubernetes side then | 19:48 |
| *** marun has quit IRC | 19:52 | |
| jpeeler | larsks: this is what i'm getting on the kube master: http://fpaste.org/140047/11551141/ | 19:53 |
| *** alexisli has joined #tripleo | 19:53 | |
| alexisli | Can we update the topic since 1375641 is fixed please? | 19:57 |
| alexisli | Also I'd like feedback on https://review.openstack.org/#/c/125025/ please | 19:57 |
| *** ramishra has joined #tripleo | 19:59 | |
| openstackgerrit | A change was merged to openstack/tripleo-incubator: Revert "Wait for memory and vcpus from hypervisor stats" https://review.openstack.org/126596 | 20:00 |
| *** SpamapS changes topic to "CI: Presumed O-K | Using OpenStack to deploy OpenStack; meetings Tuesday 1900/0800 UTC in #openstack-meeting-alt" | 20:01 | |
| SpamapS | alexisli: ty | 20:01 |
| *** ramishra has quit IRC | 20:04 | |
| *** jprovazn has quit IRC | 20:05 | |
| *** weshay has quit IRC | 20:06 | |
| tchaypo | well that was boring | 20:07 |
| tchaypo | new ptl, just like the old ptl *yawn* | 20:08 |
| tchaypo | ;) | 20:08 |
| SpamapS | tchaypo: what? I sang the whole thing | 20:13 |
| larsks | jpeeler: something is odd. | 20:13 |
| larsks | jpeeler: Oh, never mind. | 20:14 |
| larsks | jpeeler: "docker logs" in my kube environment is showing some errors. | 20:15 |
| jpeeler | well, i don't know why your containers start and mine don't | 20:17 |
| larsks | Well, they promptly fail. It's just that they do show some errors. | 20:22 |
| *** marun has joined #tripleo | 20:22 | |
| larsks | I am doing a little investigation right now. | 20:22 |
| jpeeler | ok this time it did run... i can debug this now that i see the errors too | 20:22 |
| larsks | This is what I see: http://chunk.io/f/097fd5db4447477f81677c5992c58df4 | 20:22 |
| jpeeler | yeah i see that too now. i don't know why i wasn't earlier | 20:23 |
| larsks | jpeeler: I figured out the problem. | 20:26 |
| jpeeler | syntax error? | 20:26 |
| larsks | The rabbitmq container was originally designed to accept the rabbitmq port number via a RABBITMQ_PORT environment variable. This unfortunately conflicts with a kubernetes-provided variable, which is not just a port number. | 20:26 |
| larsks | So the init-rabbitmq.sh script needs updating. | 20:27 |
| *** pensu has quit IRC | 20:27 | |
| larsks | Maybe just replace $RABBITMQ_PORT with something like $RABBITMQ_LISTEN_PORT. | 20:27 |
| jpeeler | how were you able to determine which env variable it was? | 20:27 |
| larsks | well, looking at /etc/rabbitmq/rabbitmq-env.conf and seeing where the bad config was... | 20:28 |
| larsks | Also, looking at the error message and finding which env var had that content. | 20:28 |
| larsks | Ooo, call it RABBITMQ_NODE_PORT, because that's the name in rabbitmq-env.conf. | 20:29 |
| *** julim has quit IRC | 20:32 | |
| *** pelix has quit IRC | 20:32 | |
| larsks | jpeeler: with those changes it seems to boot and run in my kubernetes environment. | 20:34 |
| mordred | SpamapS: did you guys ever look at shocco for docs like devstack uses? | 20:40 |
| mordred | SpamapS: we were just integrating devstack docs with infra and it made me think that, given the fact that dib and devstack _both_ have tons of shell, it might be an alignment point or something | 20:41 |
| *** weshay has joined #tripleo | 20:43 | |
| jpeeler | larsks: after updating and pushing the image, the only thing required would be to delete the pod and redeploy? | 20:43 |
| larsks | jpeeler: should be, yes. | 20:45 |
| larsks | Heading out now; back online later this evening. | 20:45 |
| *** jdob has quit IRC | 20:50 | |
| openstackgerrit | A change was merged to openstack/tripleo-heat-templates: Add SwiftMinPartHours parameter https://review.openstack.org/121094 | 20:51 |
| *** untriaged-bot has joined #tripleo | 21:00 | |
| untriaged-bot | No untriaged bugs so far! \o/ | 21:00 |
| *** untriaged-bot has quit IRC | 21:00 | |
| *** ramishra has joined #tripleo | 21:00 | |
| *** jcoufal has quit IRC | 21:03 | |
| *** crobertsrh is now known as _crobertsrh | 21:04 | |
| *** ramishra has quit IRC | 21:05 | |
| sdake | nice work jpeeler larsks | 21:06 |
| *** pradeep has quit IRC | 21:06 | |
| *** pradeep has joined #tripleo | 21:06 | |
| jpeeler | sdake: i'm considering this "practice" for my next container | 21:06 |
| sdake | yup | 21:06 |
| sdake | might as well tackle heat next unless someone already has it | 21:07 |
| jpeeler | i'll verify what larsks said and if all is well will commit it | 21:07 |
| sdake | so is the container definately operational then? | 21:07 |
| sdake | and a git review out? | 21:07 |
| *** pradeep1 has joined #tripleo | 21:08 | |
| sdake | jpeeler cool | 21:09 |
| *** pradeep2 has joined #tripleo | 21:10 | |
| *** pradeep has quit IRC | 21:11 | |
| *** pradeep1 has quit IRC | 21:12 | |
| *** pradeep has joined #tripleo | 21:13 | |
| *** pradeep2 has quit IRC | 21:14 | |
| *** pradeep1 has joined #tripleo | 21:14 | |
| *** rhallisey has quit IRC | 21:17 | |
| *** pradeep has quit IRC | 21:17 | |
| *** pradeep has joined #tripleo | 21:17 | |
| *** pradeep1 has quit IRC | 21:19 | |
| *** pradeep1 has joined #tripleo | 21:20 | |
| *** pradeep has quit IRC | 21:21 | |
| *** dprince has quit IRC | 21:22 | |
| *** rlandy_ has quit IRC | 21:22 | |
| *** pradeep has joined #tripleo | 21:23 | |
| *** sdake_ has joined #tripleo | 21:23 | |
| *** sdake_ has quit IRC | 21:23 | |
| *** sdake_ has joined #tripleo | 21:23 | |
| *** pradeep1 has quit IRC | 21:24 | |
| *** pradeep1 has joined #tripleo | 21:25 | |
| *** pradeep2 has joined #tripleo | 21:27 | |
| *** pradeep has quit IRC | 21:27 | |
| sdake_ | rhallisey/jpeeler can you two review https://review.openstack.org/#/q/status:open+project:stackforge/kolla,n,z plz | 21:27 |
| sdake_ | i want to reorg the repo a bit | 21:27 |
| sdake_ | and need the changes merged before i do so | 21:28 |
| *** pradeep has joined #tripleo | 21:29 | |
| *** pradeep1 has quit IRC | 21:29 | |
| sdake_ | chmouel around? | 21:30 |
| chmouel | sdake_: yeah | 21:30 |
| sdake_ | for the json blueprint can you mark it done? | 21:30 |
| chmouel | sdake_: almost before going to sleep :) | 21:30 |
| chmouel | sdake_: yep will do | 21:30 |
| sdake_ | just 1 click needed :) | 21:30 |
| sdake_ | thanks | 21:30 |
| chmouel | heh perfect! | 21:30 |
| sdake_ | two more blueprints implemented yay :) | 21:30 |
| chmouel | party time! | 21:31 |
| *** pradeep2 has quit IRC | 21:31 | |
| *** pradeep1 has joined #tripleo | 21:32 | |
| *** pradeep has quit IRC | 21:33 | |
| *** pradeep has joined #tripleo | 21:36 | |
| sdake_ | larsks when you get a chance can you mark the workflow blueprint as done in the launchpad tracker? | 21:37 |
| *** pradeep1 has quit IRC | 21:38 | |
| *** jang1 has quit IRC | 21:38 | |
| *** pradeep1 has joined #tripleo | 21:39 | |
| *** pradeep has quit IRC | 21:40 | |
| *** ccrouch is now known as ccrouch-brb | 21:45 | |
| *** morazi has quit IRC | 21:50 | |
| *** morazi has joined #tripleo | 21:50 | |
| openstackgerrit | Jeff Peeler proposed a change to stackforge/kolla: Updated base image in Dockerfile and changed CMD to reference correct file https://review.openstack.org/126037 | 21:51 |
| *** pradeep has joined #tripleo | 21:53 | |
| *** rhallisey has joined #tripleo | 21:53 | |
| *** pradeep2 has joined #tripleo | 21:56 | |
| *** pradeep1 has quit IRC | 21:56 | |
| rhallisey | sdake, why do you think that nova-compute should be creating the database? | 21:57 |
| sdake_ | it is the first thing that starts | 21:57 |
| rhallisey | sdake, looking into this more I was thinking the nova-controller should create it | 21:58 |
| sdake_ | the stuff your working on accesses compute | 21:58 |
| *** pradeep has quit IRC | 21:58 | |
| sdake_ | why? | 21:58 |
| rhallisey | well there will be essentially only one controller | 21:58 |
| *** pradeep has joined #tripleo | 21:58 | |
| rhallisey | vs many many computes I guess | 21:59 |
| *** pradeep2 has quit IRC | 22:01 | |
| sdake_ | ya that makes sense | 22:01 |
| rhallisey | also, shouldn't the controller be deployed first | 22:01 |
| sdake_ | it shouldn't matter | 22:02 |
| sdake_ | if either is started first, nova should still operate | 22:02 |
| rhallisey | compute won't really do anything without nova-api ect.. | 22:03 |
| *** pradeep1 has joined #tripleo | 22:03 | |
| sdake_ | cool | 22:04 |
| sdake_ | well make some patches then :) | 22:04 |
| rhallisey | ok! | 22:04 |
| rhallisey | :) | 22:04 |
| *** pradeep has quit IRC | 22:04 | |
| sdake_ | what services should be in the controller vs compute? | 22:04 |
| sdake_ | does copmute currently contain everyting needed? | 22:04 |
| sdake_ | can you and jpeeler work through the review queue, I want to mov ethe k8s files around | 22:05 |
| *** jtomasek has quit IRC | 22:05 | |
| rhallisey | ya sure | 22:05 |
| * jpeeler already did | 22:05 | |
| rhallisey | sdake_, ya I think compute has everything it needs | 22:06 |
| sdake_ | thanks jpeeler | 22:06 |
| sdake_ | rhallisey can you ack the review queue if the patches look good | 22:06 |
| rhallisey | on it | 22:06 |
| sdake_ | i would like to get nova wrapped up by wed, I am out of office thur/fri | 22:06 |
| *** jtomasek has joined #tripleo | 22:07 | |
| *** pradeep has joined #tripleo | 22:07 | |
| openstackgerrit | A change was merged to stackforge/kolla: chmod +x start.sh for nova-compute/nova-compute https://review.openstack.org/126669 | 22:07 |
| *** pradeep1 has quit IRC | 22:08 | |
| openstackgerrit | A change was merged to stackforge/kolla: chmod +x nova-compute/nova-libvirt/start.sh https://review.openstack.org/126670 | 22:10 |
| openstackgerrit | A change was merged to stackforge/kolla: Modify start.sh for nova-compute/nova-compute https://review.openstack.org/126700 | 22:10 |
| openstackgerrit | A change was merged to stackforge/kolla: Change sdake/image to kollaglue/image https://review.openstack.org/126701 | 22:10 |
| *** pradeep1 has joined #tripleo | 22:11 | |
| *** radez is now known as radez_g0n3 | 22:12 | |
| sdake_ | later radez | 22:13 |
| *** pradeep has quit IRC | 22:13 | |
| sdake_ | one more rhallisey | 22:19 |
| sdake_ | and then we are rockin :) | 22:19 |
| openstackgerrit | A change was merged to stackforge/kolla: Updated base image in Dockerfile and changed CMD to reference correct file https://review.openstack.org/126037 | 22:20 |
| greghaynes | aye, debugging CI fails for HA without all the logs going to one box is.... not so fun | 22:20 |
| *** pradeep has joined #tripleo | 22:20 | |
| sdake_ | https://review.openstack.org/#/c/126037/ | 22:20 |
| rhallisey | sdake, oh we're rockin | 22:20 |
| *** mikal has quit IRC | 22:21 | |
| *** mikal has joined #tripleo | 22:23 | |
| *** pradeep1 has quit IRC | 22:23 | |
| *** pradeep1 has joined #tripleo | 22:23 | |
| *** sdake_ has quit IRC | 22:25 | |
| *** pradeep has quit IRC | 22:26 | |
| *** pradeep has joined #tripleo | 22:28 | |
| *** pradeep1 has quit IRC | 22:30 | |
| *** pradeep1 has joined #tripleo | 22:30 | |
| *** jtomasek has quit IRC | 22:30 | |
| *** pradeep has quit IRC | 22:33 | |
| openstackgerrit | Steven Dake proposed a change to stackforge/kolla: Reorganize k8s structure https://review.openstack.org/126744 | 22:33 |
| sdake | rhallisey jpeeler ok that should do it if you could have a look at that change | 22:33 |
| *** pradeep has joined #tripleo | 22:35 | |
| *** ccrouch-brb is now known as ccrouch | 22:36 | |
| rhallisey | sdake, ok | 22:36 |
| *** pradeep1 has quit IRC | 22:36 | |
| *** pradeep1 has joined #tripleo | 22:38 | |
| *** andreaf has quit IRC | 22:40 | |
| *** pradeep has quit IRC | 22:40 | |
| openstackgerrit | A change was merged to stackforge/kolla: Reorganize k8s structure https://review.openstack.org/126744 | 22:40 |
| *** andreaf has joined #tripleo | 22:40 | |
| sdake | rhallisey take care to make sure two folks have reviewed a patch :) | 22:41 |
| greghaynes | wow, systemd has its own remote logging / forwarding system | 22:41 |
| *** pradk has quit IRC | 22:41 | |
| rhallisey | accidents happen o.o | 22:42 |
| * greghaynes weeps | 22:42 | |
| *** pradeep has joined #tripleo | 22:42 | |
| nibalizer | greghaynes: and its cryptographically signed, and there are QR codes so you can verify the fingerprint on your phone... | 22:44 |
| *** pradeep1 has quit IRC | 22:44 | |
| greghaynes | so now the question becomes - do we use rsyslog to forward logs in the name of consistency with other distros, or do we try out the systemd-journal-remote | 22:44 |
| greghaynes | nibalizer: oh, well that makes the decision easy then | 22:44 |
| nibalizer | https://plus.google.com/+LennartPoetteringTheOneAndOnly/posts/g1E6AxVKtyc | 22:45 |
| greghaynes | oh my god is that really his g+ name | 22:45 |
| *** pradeep1 has joined #tripleo | 22:45 | |
| *** chuckC_ has quit IRC | 22:46 | |
| *** chuckC_ has joined #tripleo | 22:47 | |
| *** pradeep has quit IRC | 22:47 | |
| *** pradeep has joined #tripleo | 22:49 | |
| *** pradeep1 has quit IRC | 22:50 | |
| sdake | two poterrings | 22:50 |
| sdake | does the world really need two | 22:50 |
| sdake | did I jus tsay that out loud ;-) | 22:50 |
| *** chuckC_ has quit IRC | 22:51 | |
| sdake | jpeeler so mariadb is set then? | 22:52 |
| sdake | rather rabbitmq | 22:52 |
| *** pradeep1 has joined #tripleo | 22:52 | |
| *** pradeep has quit IRC | 22:54 | |
| *** morazi has quit IRC | 22:55 | |
| *** pradeep has joined #tripleo | 22:56 | |
| *** pradeep1 has quit IRC | 22:57 | |
| *** noslzzp has quit IRC | 22:57 | |
| *** pradeep1 has joined #tripleo | 23:00 | |
| *** ChuckC has quit IRC | 23:01 | |
| *** pradeep has quit IRC | 23:01 | |
| *** ramishra has joined #tripleo | 23:02 | |
| *** pradeep has joined #tripleo | 23:02 | |
| *** pradeep2 has joined #tripleo | 23:04 | |
| *** pradeep1 has quit IRC | 23:05 | |
| *** ramishra has quit IRC | 23:06 | |
| jpeeler | sdake: i believe so. i had a sort of crazy idea of making a clients container that tested each deployed service | 23:06 |
| *** pradeep has quit IRC | 23:07 | |
| *** pradeep has joined #tripleo | 23:08 | |
| *** pradeep1 has joined #tripleo | 23:11 | |
| *** pradeep2 has quit IRC | 23:11 | |
| *** pradeep has quit IRC | 23:12 | |
| tchaypo | such poettering | 23:13 |
| tchaypo | much sealing | 23:13 |
| tchaypo | wow | 23:13 |
| *** pradeep has joined #tripleo | 23:14 | |
| *** pradeep1 has quit IRC | 23:15 | |
| *** pradeep1 has joined #tripleo | 23:17 | |
| *** pradeep has quit IRC | 23:18 | |
| openstackgerrit | James Polley proposed a change to openstack/tripleo-incubator: Rotate an SSH key for tchaypo https://review.openstack.org/126750 | 23:19 |
| sdake | jpeeler YES | 23:21 |
| sdake | jpeeler I intend to do that as well | 23:21 |
| *** pradeep has joined #tripleo | 23:21 | |
| sdake | and make it a gate | 23:21 |
| *** pradeep1 has quit IRC | 23:21 | |
| sdake | such a gate would be hugely helpful | 23:21 |
| sdake | if you want to tackle it it would have huge impact | 23:22 |
| sdake | I was also thinking of a tempest style gate along the same lines | 23:23 |
| *** ChuckC has joined #tripleo | 23:23 | |
| sdake | it doesn't even have to be a container, could just be a gate in general | 23:24 |
| *** pradeep1 has joined #tripleo | 23:24 | |
| *** pradeep has quit IRC | 23:25 | |
| *** dsneddon_ has quit IRC | 23:28 | |
| *** pradeep1 has quit IRC | 23:28 | |
| *** dsneddon_ has joined #tripleo | 23:29 | |
| *** pradeep has joined #tripleo | 23:32 | |
| *** pradeep1 has joined #tripleo | 23:35 | |
| sdake | zugh 350 bashate error(s) found | 23:35 |
| *** pradeep has quit IRC | 23:37 | |
| *** pradeep has joined #tripleo | 23:38 | |
| *** alexisli has quit IRC | 23:39 | |
| *** pradeep1 has quit IRC | 23:39 | |
| *** pradeep1 has joined #tripleo | 23:40 | |
| *** pradeep has quit IRC | 23:43 | |
| *** pradeep has joined #tripleo | 23:44 | |
| *** pradeep1 has quit IRC | 23:45 | |
| *** pradeep1 has joined #tripleo | 23:47 | |
| *** pradeep has quit IRC | 23:48 | |
| *** pradeep has joined #tripleo | 23:52 | |
| *** pradeep1 has quit IRC | 23:55 | |
| *** pradeep1 has joined #tripleo | 23:57 | |
| *** pradeep has quit IRC | 23:58 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!