| openstackgerrit | Alan Pevec proposed openstack-infra/tripleo-ci: Force distro branch reset in DLRN builds https://review.openstack.org/345106 | 00:04 |
|---|---|---|
| *** tzumainn has quit IRC | 00:21 | |
| *** limao has joined #tripleo | 00:24 | |
| *** lblanchard has joined #tripleo | 00:30 | |
| *** tzumainn has joined #tripleo | 00:33 | |
| *** jerrygb has joined #tripleo | 00:36 | |
| *** saneax is now known as saneax_AFK | 00:37 | |
| *** mbound has joined #tripleo | 00:38 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder: [WIP] Move elements into diskimage_builder package https://review.openstack.org/344017 | 00:40 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder: [WIP] Move diskimage-image-create to an entry point https://review.openstack.org/344532 | 00:40 |
| *** jerrygb has quit IRC | 00:41 | |
| *** shivrao has quit IRC | 00:41 | |
| *** mbound has quit IRC | 00:43 | |
| *** beagles_l8r is now known as beagles | 00:49 | |
| *** rhallisey has quit IRC | 00:54 | |
| *** wfoster has quit IRC | 01:03 | |
| *** Guest49585 has quit IRC | 01:03 | |
| openstackgerrit | wes hayutin proposed openstack/tripleo-quickstart: Stop using deprecated network range https://review.openstack.org/343443 | 01:04 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder: [WIP] Move diskimage-image-create to an entry point https://review.openstack.org/344532 | 01:08 |
| *** bfournie has joined #tripleo | 01:09 | |
| *** wfoster has joined #tripleo | 01:10 | |
| *** rook_ has joined #tripleo | 01:10 | |
| *** shivrao has joined #tripleo | 01:37 | |
| *** jerrygb has joined #tripleo | 01:37 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder: [WIP] Move diskimage-image-create to an entry point https://review.openstack.org/344532 | 01:41 |
| *** jerrygb has quit IRC | 01:42 | |
| *** egafford has quit IRC | 02:03 | |
| *** rajinir has quit IRC | 02:15 | |
| *** jerrygb has joined #tripleo | 02:21 | |
| *** shivrao has quit IRC | 02:22 | |
| *** lblanchard has quit IRC | 02:22 | |
| openstackgerrit | Emilien Macchi proposed openstack/tripleo-heat-templates: Deploy Horizon with composable apache service https://review.openstack.org/345031 | 02:26 |
| *** weshay has quit IRC | 02:40 | |
| openstackgerrit | Emilien Macchi proposed openstack/instack-undercloud: Deploy MySQL using composable role https://review.openstack.org/345002 | 02:41 |
| *** yamahata has quit IRC | 02:44 | |
| *** lblanchard has joined #tripleo | 02:49 | |
| *** lblanchard has quit IRC | 02:54 | |
| *** saneax_AFK is now known as saneax | 02:59 | |
| *** julim has quit IRC | 03:31 | |
| *** julim has joined #tripleo | 03:44 | |
| *** masco has joined #tripleo | 03:58 | |
| *** saneax is now known as saneax_AFK | 04:00 | |
| *** julim has quit IRC | 04:05 | |
| *** radek__ has joined #tripleo | 04:25 | |
| openstackgerrit | Marios Andreou proposed openstack/tripleo-heat-templates: Repurpose KeystoneWorkers add keystone::wsgi::apache::threads https://review.openstack.org/336520 | 04:31 |
| *** d0ugal has quit IRC | 04:56 | |
| *** yamahata has joined #tripleo | 04:57 | |
| *** d0ugal has joined #tripleo | 04:58 | |
| *** jaosorior has joined #tripleo | 05:01 | |
| *** liverpooler has quit IRC | 05:04 | |
| *** shivrao has joined #tripleo | 05:09 | |
| *** jerrygb has quit IRC | 05:09 | |
| *** shivrao_ has joined #tripleo | 05:11 | |
| *** shivrao has quit IRC | 05:14 | |
| *** shivrao_ is now known as shivrao | 05:14 | |
| *** tzumainn has quit IRC | 05:14 | |
| *** pgadiya has joined #tripleo | 05:16 | |
| *** jprovazn has joined #tripleo | 05:22 | |
| *** tzumainn has joined #tripleo | 05:26 | |
| *** afazekas|dentist is now known as afazekas | 05:32 | |
| *** akuznetsov has joined #tripleo | 05:33 | |
| *** dsariel has joined #tripleo | 05:42 | |
| jaosorior | marios: Are you around yet? | 05:46 |
| marios | o/ | 05:47 |
| openstackgerrit | Babu Shanmugam proposed openstack/puppet-tripleo: Add ML2/OVN support for neutron profile https://review.openstack.org/314875 | 05:50 |
| *** dsariel has quit IRC | 05:50 | |
| *** oshvartz has joined #tripleo | 05:51 | |
| jaosorior | marios hey dude, how's it going? | 05:51 |
| jaosorior | marios hey dude do you have an upstream deployment around? | 05:52 |
| marios | just fine jaosorior and how are you today | 05:52 |
| jaosorior | marios: tired, but caffeine will make it aaaalll better | 05:53 |
| marios | jaosorior: not at the moment man, i have an osp8..9 setup on my devbox at the moment for upgrades work | 05:53 |
| jaosorior | marios; Alright, no biggie, I'll wait for ccamacho to come online | 05:54 |
| *** liverpooler has joined #tripleo | 06:13 | |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack/puppet-tripleo: Add ironic inspector as a terminated HAProxy endpoint https://review.openstack.org/345183 | 06:14 |
| openstackgerrit | Merged openstack-infra/tripleo-ci: Do not explicitly set inspector iPXE port https://review.openstack.org/344796 | 06:15 |
| *** dtantsur|afk is now known as dtantsur | 06:15 | |
| *** ccamacho has joined #tripleo | 06:15 | |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack-infra/tripleo-ci: TEST IRONIC SSL - DO NOT MERGE https://review.openstack.org/342669 | 06:16 |
| jaosorior | ccamacho hey dude, how's it going>? | 06:16 |
| ccamacho | jaosorior! morning :) | 06:16 |
| jaosorior | ccamacho hey, so in the end, did you get the introspection working with that patch? | 06:16 |
| ccamacho | nice, the mistral submission I think it's almost finished | 06:17 |
| ccamacho | jaosorior, yeahp | 06:17 |
| ccamacho | It solved the timeout issue | 06:17 |
| jaosorior | ccamacho hey dude, so now that it's working, can you test out this patch? https://review.openstack.org/#/c/332641/ it's working for me, but apparently bnemec had problems with it | 06:17 |
| *** saneax_AFK is now known as saneax | 06:18 | |
| *** florianf has joined #tripleo | 06:18 | |
| ccamacho | rebuilt and reinstall instack-undercloud in the undercloud should apply the submission correctly, right? | 06:20 |
| ccamacho | s/rebuilt/rebuild | 06:20 |
| openstackgerrit | Babu Shanmugam proposed openstack/puppet-tripleo: Add ML2/OVN support for neutron profile https://review.openstack.org/314875 | 06:20 |
| *** tobias_fiberdata has joined #tripleo | 06:20 | |
| ccamacho | jaosorior Ill do it from scratch, now provisioning | 06:21 |
| jaosorior | awesome | 06:22 |
| jaosorior | thanks dude | 06:23 |
| jaosorior | ccamacho, yep | 06:23 |
| ccamacho | jaosorior, np | 06:23 |
| jaosorior | ccamacho: Just gotta make sure that hiera ironic::inspector::http_port gives a correct value (with the fix) | 06:24 |
| *** akuznetsov has quit IRC | 06:24 | |
| *** mcornea has joined #tripleo | 06:30 | |
| jaosorior | marios: Hey dude, can you check this commit out? https://review.openstack.org/#/c/344851/ I've been using this code to test the certmonger-managed TLS in the undercloud. It was initially in-line the main puppet manifest for the udnercloud but now I started breaking it down into smaller pieces | 06:30 |
| *** devvesa has joined #tripleo | 06:30 | |
| marios | ack jaosorior | 06:31 |
| jaosorior | marios: Here's how it's used https://review.openstack.org/#/c/332774/26..27/elements/puppet-stack-config/puppet-stack-config.pp | 06:31 |
| *** tesseract- has joined #tripleo | 06:39 | |
| *** ccamacho is now known as ccamacho|afk | 06:44 | |
| jaosorior | marios, ccamacho have you guys seem an error like this in nova? | 06:50 |
| jaosorior | InstanceDeployFailure: Failed to provision instance 86ed0b24-35d6-422d-91c4-f03ebbe12ac0: Failed to prepar | 06:50 |
| jaosorior | e to deploy. Error: 400 Bad Request: Unknown scheme 'swift+config' found in URI (HTTP 400) | 06:50 |
| marios | not seen that jaosorior | 06:54 |
| *** shivrao has quit IRC | 06:54 | |
| openstackgerrit | Flavio Percoco proposed openstack/tripleo-common: WIP: Create new docker command hook. https://review.openstack.org/312723 | 06:57 |
| *** jubapa has joined #tripleo | 06:57 | |
| *** karthiks has quit IRC | 06:57 | |
| openstackgerrit | Flavio Percoco proposed openstack/tripleo-common: Create new docker command hook. https://review.openstack.org/312723 | 06:57 |
| *** ohamada has joined #tripleo | 06:57 | |
| openstackgerrit | Flavio Percoco proposed openstack/tripleo-common: Create new docker command hook. https://review.openstack.org/312723 | 06:58 |
| *** bootsha has joined #tripleo | 07:00 | |
| *** pcaruana has joined #tripleo | 07:03 | |
| *** athomas has joined #tripleo | 07:05 | |
| *** leanderthal|afk is now known as leanderthal | 07:05 | |
| *** rcernin has joined #tripleo | 07:07 | |
| *** rcernin has quit IRC | 07:10 | |
| *** karthiks has joined #tripleo | 07:10 | |
| *** rcernin has joined #tripleo | 07:10 | |
| *** fzdarsky has joined #tripleo | 07:12 | |
| tobias_fiberdata | The postdeployment step in the deployment, what does that mean? i assume it's all the configurations? | 07:13 |
| jaosorior | tobias_fibertdata: Yes, the puppet configuration runs there | 07:14 |
| jaosorior | dtantsur are you around? | 07:15 |
| dtantsur | jaosorior, more or less | 07:15 |
| jaosorior | dtantsur: Hey dude, any idea what would this mean in the ironic-conductor log? ERROR ironic.conductor.manager [req-38e32b2c-141b-4a12-be30-7d226eb22f08 - - - - -] Error while preparing to deploy to node e93aa7f9-5209-4154-94c5-ab84031d5f39: 400 Bad Request: Unknown scheme 'swift+config' found in URI (HTTP 400) | 07:16 |
| dtantsur | jaosorior, I remember something like that, I think it was malformed config drive in the node properties | 07:16 |
| *** ifarkas_afk is now known as ifarkas | 07:17 | |
| jaosorior | dtantsur: Is it a configuration issue? I'm pretty puzzled about that error | 07:17 |
| dtantsur | jaosorior, I don't quite remember, I haven't had it personally, but I think it is. Maybe worth checking the node-show for the node | 07:18 |
| tobias_fiberdata | how do i turn off the LbaaS service in neutron while deploying? | 07:19 |
| tobias_fiberdata | is this possible? | 07:19 |
| tobias_fiberdata | i mean | 07:19 |
| tobias_fiberdata | before deploying | 07:19 |
| tobias_fiberdata | i get this error | 07:19 |
| tobias_fiberdata | [Overcloudtesting-ControllerNodesPostDeployment-6fe4cokn6mfh-ControllerLoadBalancerDeployment_Step1-bkaibyibcsre]: CREATE_FAILED Resource CREATE failed: Error: resources[0]: Deployment to server failed: deploy_status_code : Deployment exited with non-zero status code: 1 | 07:19 |
| jaosorior | tobias_fiberdata: That isn't actually Neutron's LBaaS but it's actually the overcloud's loadbalancer, which is HAProxy | 07:20 |
| tobias_fiberdata | ah okey | 07:20 |
| tobias_fiberdata | that one is gonna be hard to actually turn off :D | 07:20 |
| jaosorior | tobias_fiberdata doing this: heat resource-list -n 5 overcloud | grep FAIL | 07:20 |
| jaosorior | what do you get? | 07:21 |
| tobias_fiberdata | sec | 07:21 |
| tobias_fiberdata | | ComputeNodesPostDeployment | 7143efe6-6f77-4160-9085-7833d51ed429 | OS::TripleO::ComputePostDeployment | CREATE_FAILED | 2016-07-21T06:39:57 | Overcloudtesting | | 07:21 |
| tobias_fiberdata | | ControllerNodesPostDeployment | 7a0ada8c-6c16-48e3-9a53-2ff1227e2a2d | OS::TripleO::ControllerPostDeployment | CREATE_FAILED | 2016-07-21T06:39:57 | Overcloudtesting | | 07:21 |
| tobias_fiberdata | | ControllerLoadBalancerDeployment_Step1 | 52ba923c-c294-48a5-9c1f-bd038f778395 | OS::Heat::StructuredDeployments | CREATE_FAILED | 2016-07-21T06:54:02 | Overcloudtesting-ControllerNodesPostDeployment-6fe4cokn6mfh | | 07:21 |
| tobias_fiberdata | | 0 | 193ee0b0-f6cf-4050-9b46-13e0ff574827 | OS::Heat::StructuredDeployment | CREATE_FAILED | 2016-07-21T06:54:37 | Overcloudtesting-ControllerNodesPostDeployment-6fe4cokn6mfh-ControllerLoadBalancerDeployment_Step1-bkaibyibcsre | | 07:21 |
| tobias_fiberdata | could of sent you a privmess | 07:22 |
| tobias_fiberdata | perhaps :p | 07:22 |
| *** fzdarsky has quit IRC | 07:22 | |
| jaosorior | dtantsur: this is the node-show for one of the controllers, I'm not exactly sure what to look for :/ http://paste.openstack.org/show/538927/ | 07:23 |
| jaosorior | tobias_fiberdata: ok, now you could do: heat deployment-output-show 193ee0b0-f6cf-4050-9b46-13e0ff574827 deploy_stderr | 07:24 |
| *** tzumainn has quit IRC | 07:24 | |
| tobias_fiberdata | hmm | 07:25 |
| jaosorior | 193ee0b0-f6cf-4050-9b46-13e0ff574827 is the ID of the failed resource that's of type OS::Heat::StructuredDeployment | 07:25 |
| tobias_fiberdata | ah ye | 07:25 |
| dtantsur | jaosorior, I see nothing suspicious, hmm | 07:25 |
| tobias_fiberdata | hmm interesting | 07:25 |
| dtantsur | jaosorior, I'll check where this error comes from a bit later | 07:25 |
| jaosorior | tobias_fiberdata: You could use pastebin to show me the outputs | 07:25 |
| jaosorior | dtantsur: Thanks dude | 07:26 |
| tobias_fiberdata | http://paste.openstack.org/show/538928/ | 07:26 |
| jaosorior | tobias_fiberdata: Right, so your issue is that the overcloud images have out-of-date puppet manifests | 07:26 |
| dtantsur | jaosorior, hmm, the problem is that your node was cleaned up already | 07:26 |
| jaosorior | dtantsur: crap | 07:27 |
| jaosorior | so I need to get the node-info while the error is happening | 07:27 |
| jaosorior | I guess | 07:27 |
| tobias_fiberdata | jaosorior, but i built them yesterday :P lol | 07:27 |
| dtantsur | jaosorior, I still think that maybe nova passes config drive in a format unreadable by ironic | 07:27 |
| jaosorior | tobias_fiberdata: That's weird, that class has been there for a while now | 07:27 |
| tobias_fiberdata | i did as the documentation said basically | 07:27 |
| tobias_fiberdata | when creating them | 07:27 |
| tobias_fiberdata | but i will do it again | 07:28 |
| tobias_fiberdata | and see what happnes | 07:28 |
| jaosorior | tobias_fiberdata: well, it does depend how you did them. You might have been the latest stable images. but you might be using the latest t-h-t | 07:28 |
| *** ebarrera has joined #tripleo | 07:28 | |
| jaosorior | that's a common problem | 07:28 |
| jaosorior | if you're using the latest t-h-t then you need the latest images, which are not considered stable | 07:28 |
| tobias_fiberdata | aah | 07:29 |
| tobias_fiberdata | but it's using the same packages as on the undercloud node no? | 07:29 |
| jaosorior | tobias_fiberdata: So, you want to use the latest code then? | 07:29 |
| tobias_fiberdata | parts of it | 07:29 |
| tobias_fiberdata | jaosorior, na i want to use the stable ones tbh, it's just a lab for now | 07:29 |
| jaosorior | what about the t-h-t that you're using? where is taht coming from? | 07:30 |
| tobias_fiberdata | jaosorior, na i want to use the stable ones tbh, it's just a lab for now, but i'm just figuring out stuff to make sure i do it correct the 5th time deploying the production | 07:30 |
| *** akuznetsov has joined #tripleo | 07:30 | |
| tobias_fiberdata | is it the heat-templates you are refering to? | 07:30 |
| tobias_fiberdata | tripleO ones? | 07:30 |
| jaosorior | yeah | 07:30 |
| jaosorior | tripleo-heat-templates == t-h-t | 07:31 |
| tobias_fiberdata | yea :p | 07:31 |
| jaosorior | sorry for the confusion haha, just got used to using the shorter version | 07:31 |
| tobias_fiberdata | it's the one that is under /usr/share/openstack-tripleo-heat-templates/ i just copied them to my homefolder | 07:31 |
| jaosorior | I see | 07:31 |
| jaosorior | well that's messed up, if you got the templates from the package they should match the image | 07:33 |
| tobias_fiberdata | hmm | 07:33 |
| jaosorior | one option might be that you then built too new images | 07:33 |
| tobias_fiberdata | i did this one to be able to build the images | 07:33 |
| tobias_fiberdata | export USE_DELOREAN_TRUNK=1 | 07:33 |
| tobias_fiberdata | export DELOREAN_TRUNK_REPO="http://buildlogs.centos.org/centos/7/cloud/x86_64/rdo-trunk-master-tripleo/" | 07:33 |
| tobias_fiberdata | export DELOREAN_REPO_FILE="delorean.repo" | 07:33 |
| *** zoli_gone-proxy is now known as zoliXXL | 07:33 | |
| tobias_fiberdata | that's a not stable one i guess | 07:33 |
| tobias_fiberdata | but i also have been using the not stable one when installing the undercloud, i think. | 07:34 |
| tobias_fiberdata | atleast the repo says mitaka-testing | 07:34 |
| tobias_fiberdata | but mitaka shouldnt that be considered stable, so there should be a stable version of mitaka, that one is not in the documentation on tripleO.org yet if you take a look at this site | 07:35 |
| tobias_fiberdata | http://tripleo.org/basic_deployment/basic_deployment_cli.html#get-images | 07:35 |
| tobias_fiberdata | i dont know :p | 07:35 |
| tobias_fiberdata | i'm confused. I will rebuild the images and see how that goes then | 07:35 |
| jaosorior | tobias_fiberdata: If you're trying to use mitaka | 07:35 |
| *** fzdarsky has joined #tripleo | 07:36 | |
| jaosorior | it seems to me that the images you built are for master | 07:36 |
| jaosorior | which is considered newton | 07:36 |
| tobias_fiberdata | ah okey | 07:36 |
| tobias_fiberdata | then it's wrong :p | 07:36 |
| tobias_fiberdata | hehe | 07:36 |
| tobias_fiberdata | that explains it for sure | 07:36 |
| *** tzumainn has joined #tripleo | 07:37 | |
| *** d0ugal has quit IRC | 07:37 | |
| jaosorior | dtantsur: This is the node-show while it's trying to deploy http://paste.openstack.org/show/538931/ | 07:38 |
| openstackgerrit | Flavio Percoco proposed openstack/python-tripleoclient: Add ability to upload images based on config https://review.openstack.org/255535 | 07:41 |
| tobias_fiberdata | jaosorior, this is how the repo looks like | 07:44 |
| tobias_fiberdata | http://paste.openstack.org/show/538932/ | 07:44 |
| tobias_fiberdata | ah well | 07:44 |
| tobias_fiberdata | break now :p | 07:44 |
| tobias_fiberdata | need coffee | 07:44 |
| tobias_fiberdata | ;D | 07:44 |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack/puppet-tripleo: Add class to use certmonger's local CA https://review.openstack.org/344851 | 07:44 |
| openstackgerrit | mathieu bultel proposed openstack/tripleo-quickstart: WIP gate upgrade https://review.openstack.org/342161 | 07:44 |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack/instack-undercloud: Use certmonger for automatic cert generation https://review.openstack.org/332774 | 07:46 |
| openstackgerrit | Merged openstack/tripleo-heat-templates: Fix KeystoneWorkers default value https://review.openstack.org/344795 | 07:46 |
| *** jpich has joined #tripleo | 07:47 | |
| dtantsur | jaosorior, still nothing suspicious, hmm | 07:50 |
| *** pino|work has quit IRC | 07:51 | |
| marios | there is something mildly therapeutic in watch -d pcs status while services are being brought up | 07:51 |
| marios | i need holidays | 07:51 |
| dtantsur | jaosorior, take a look at your overcloud-full image details in glance | 07:51 |
| *** dsariel has joined #tripleo | 07:54 | |
| *** radek__ has quit IRC | 07:55 | |
| jaosorior | dtantsur: http://paste.openstack.org/show/538934/ | 07:55 |
| jaosorior | marios; Can you take another look at the CA patch? I updated it with your suggestion | 07:56 |
| *** pino|work has joined #tripleo | 07:56 | |
| marios | jaosorior: ack thanks | 07:57 |
| *** fzdarsky has quit IRC | 07:57 | |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed openstack/tripleo-docs: Remove discover from test-requirements https://review.openstack.org/345224 | 07:57 |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed openstack/tripleo-common: Remove discover from test-requirements https://review.openstack.org/345225 | 07:59 |
| *** shardy has joined #tripleo | 07:59 | |
| openstackgerrit | Flavio Percoco proposed openstack/tripleo-heat-templates: WIP: Containerized Services for Composable Roles https://review.openstack.org/330659 | 08:00 |
| *** akuznetsov has quit IRC | 08:01 | |
| jaosorior | shardy: Hey dude, by any chance have you seen something like this? 2016-07-21 07:40:07.503 2178 ERROR ironic.conductor.manager [req-7dc12170-746a-42c9-a0a7-a960126ff2dc - - - - -] Error while preparing to deploy to node d07cae3e-fb0f-4692-b4d0-74a45bfa3954: 400 Bad Request: Unknown scheme 'swift+config' found in URI (HTTP 400) | 08:02 |
| jaosorior | I'm unable to deploy nodes because nova gets that from ironic | 08:02 |
| shardy | jaosorior: Hi, no not yet seen that | 08:03 |
| *** karthiks has quit IRC | 08:04 | |
| openstackgerrit | Babu Shanmugam proposed openstack/tripleo-heat-templates: OVN heat templates https://review.openstack.org/307734 | 08:05 |
| dtantsur | jaosorior, I have to run, will get back to it in a few hours. please feel free to grab lucasagomes in the meanwhile :) | 08:05 |
| *** dtantsur is now known as dtantsur|bbl | 08:05 | |
| jaosorior | dtantsur|bbl: Thanks | 08:05 |
| jaosorior | lucasgomes: Are you around? | 08:06 |
| *** stendulker has joined #tripleo | 08:06 | |
| *** ccamacho|afk is now known as ccamacho | 08:08 | |
| *** ohamada has quit IRC | 08:08 | |
| tobias_fiberdata | the documentation on tripleo.org is not updated for mitaka | 08:09 |
| tobias_fiberdata | what's the repo for mitaka stable repos? | 08:09 |
| tobias_fiberdata | for tripleO | 08:09 |
| tobias_fiberdata | i guess i can just replace stuff? | 08:11 |
| shardy | tobias_fiberdata: Ah, we should fix that, yeah it should be s/liberty/mitaka | 08:11 |
| *** radek__ has joined #tripleo | 08:11 | |
| shardy | the tripleo.sh tool we use in CI points at https://trunk.rdoproject.org/centos7-$STABLE_RELEASE/current/ | 08:11 |
| tobias_fiberdata | shardy, yea it caused my images to be built on the wrong release :p | 08:12 |
| tobias_fiberdata | caused my deployment to fail as it built the newton images :p | 08:12 |
| shardy | Another option is to consume TripleO via RDO https://www.rdoproject.org/tripleo/ | 08:12 |
| tobias_fiberdata | doing it on physnodes | 08:13 |
| shardy | https://github.com/openstack-infra/tripleo-ci/blob/master/scripts/tripleo.sh | 08:13 |
| tobias_fiberdata | or i guess that one works for physnodes aswell | 08:13 |
| shardy | you can run that script via export STABLE_RELEASE=mitaka | 08:13 |
| shardy | then tripleo.sh --repo-setup | 08:13 |
| tobias_fiberdata | hmm | 08:14 |
| tobias_fiberdata | i wonder if the undercloud node we have installed has the mitaka packages then or if it's based on newton aswell | 08:14 |
| shardy | how did you install it? | 08:15 |
| tobias_fiberdata | follow this docs http://tripleo.org/index.html | 08:15 |
| tobias_fiberdata | using the current ones | 08:16 |
| tobias_fiberdata | repos that is | 08:16 |
| tobias_fiberdata | centos7/current/delorean.repo | 08:16 |
| tobias_fiberdata | from trunk | 08:16 |
| shardy | Yeah, everything is trunk/newton by default | 08:16 |
| tobias_fiberdata | ah | 08:16 |
| tobias_fiberdata | crap :p | 08:16 |
| jaosorior | ccamacho: Did you get a chance to test the patch? | 08:17 |
| *** karthiks has joined #tripleo | 08:17 | |
| *** derekh has joined #tripleo | 08:18 | |
| tobias_fiberdata | shardy, ah well, i'll solve it | 08:18 |
| tobias_fiberdata | how much different is it in the new templates from newton compared to mitaka? I guess it's just the custom thing that you are working on shardy? | 08:21 |
| *** rdo has quit IRC | 08:21 | |
| openstackgerrit | Babu Shanmugam proposed openstack/puppet-tripleo: Add ML2/OVN support for neutron profile https://review.openstack.org/314875 | 08:21 |
| *** ohamada has joined #tripleo | 08:22 | |
| hewbrocca | dtantsur|bbl: morning -- I guess we're still failing introspection in the RDO promote job? | 08:23 |
| *** zoliXXL is now known as zoli|wfh | 08:23 | |
| *** rdo has joined #tripleo | 08:24 | |
| *** sshnaidm|afk is now known as sshnaidm | 08:24 | |
| shardy | hewbrocca: a fix landed on trunk but there's still CI errors that may not be related to the original problem | 08:24 |
| shardy | https://bugs.launchpad.net/tripleo/+bug/1604770 | 08:24 |
| openstack | Launchpad bug 1604770 in tripleo "introspection hangs due to broken ipxe config" [Critical,Fix released] - Assigned to Dmitry Tantsur (divius) | 08:24 |
| shardy | about to pull everything and re-test locally | 08:25 |
| hewbrocca | shardy: *oof* | 08:25 |
| hewbrocca | OK then | 08:25 |
| sshnaidm | shardy, the fix was merged yesterday I think | 08:25 |
| shardy | Yeah https://review.openstack.org/#/c/344792/ was the fix for the issue I saw | 08:26 |
| hewbrocca | the missing close-bracket one | 08:26 |
| hewbrocca | So works for Wes (at least) locally, but not passing in real CI | 08:26 |
| hewbrocca | shardy: well, let me know how it goes | 08:27 |
| shardy | https://ci.centos.org/artifacts/rdo/jenkins-tripleo-quickstart-promote-master-delorean-minimal-379/console.txt.gz | 08:27 |
| shardy | doesn't exactly make it very clear why the job failed | 08:27 |
| *** rdo has quit IRC | 08:28 | |
| *** rdo has joined #tripleo | 08:30 | |
| shardy | https://ci.centos.org/artifacts/rdo/jenkins-tripleo-quickstart-promote-master-delorean-minimal-379/undercloud/home/stack/undercloud_post_install.log.gz | 08:30 |
| shardy | shows it is still an introspection timeout | 08:30 |
| hewbrocca | Right | 08:31 |
| jaosorior | shardy: I actually don't see any inspector logs there ^^ | 08:32 |
| jaosorior | ah | 08:32 |
| jaosorior | nevermind | 08:32 |
| jaosorior | it was the inspector config | 08:32 |
| jaosorior | does it share config file with ironic? | 08:32 |
| *** gfidente|afk has joined #tripleo | 08:34 | |
| *** rdo has quit IRC | 08:34 | |
| *** rdo has joined #tripleo | 08:35 | |
| mandre | shardy: what's the extent of my new +2 super powers? is it limited to tripleo-validations or will you be mad at me if I +2/+A some container related stuff in tripleo-common? :) | 08:39 |
| tobias_fiberdata | jaosorior, i'll be updating to newton and see if the deployment just works for my hardware i have | 08:40 |
| tobias_fiberdata | i will report back | 08:40 |
| tobias_fiberdata | i'm using Huawei RH1288V3 servers | 08:40 |
| tobias_fiberdata | i guess you are interested in seeing how it goes. | 08:41 |
| jaosorior | haha well, would be cool to know | 08:41 |
| tobias_fiberdata | the production cluster is gonna be based on Huawei RH2288H V3 servers | 08:41 |
| tobias_fiberdata | i hope that will just works when we have figured out all the yaml's | 08:41 |
| jaosorior | hopefully! | 08:41 |
| tobias_fiberdata | me and my programming skills not with me and reading 300 files to get some sort of understanding is not my best side i would say | 08:42 |
| tobias_fiberdata | it's not 300files | 08:42 |
| tobias_fiberdata | but | 08:42 |
| tobias_fiberdata | yea | 08:42 |
| tobias_fiberdata | loads of files | 08:42 |
| jaosorior | tobias_fiberdata: It's understandable, it's a pretty rough learning curve | 08:42 |
| tobias_fiberdata | jaosorior, yea it is for sure! It's easier for you guys that work with them almost every day :P | 08:43 |
| * flaper87 +2's mandre's +2 powers on container stuff | 08:44 | |
| flaper87 | :P | 08:44 |
| shardy | mandre: Hey, we agreed just tripleo-validations but provided the -common stuff is passing CI it's probably fine | 08:45 |
| shardy | what are the patches? | 08:45 |
| mandre | shardy: https://review.openstack.org/#/c/318300/ and https://review.openstack.org/#/q/topic:dockerfile-optimizations | 08:46 |
| *** jubapa has quit IRC | 08:46 | |
| *** dsariel has quit IRC | 08:48 | |
| *** paramite has joined #tripleo | 08:49 | |
| *** aufi has joined #tripleo | 08:50 | |
| flaper87 | shardy: mandre thanks for the reviews. Eventually we'll get all these tested in the CI, so far I've run manual tests | 08:50 |
| shardy | https://review.openstack.org/#/c/344789/1/heat_docker_agent/configure_container.sh | 08:51 |
| *** fzdarsky has joined #tripleo | 08:51 | |
| shardy | mandre, flaper87: why do we do three separate yum installs there? | 08:51 |
| mandre | shardy: I know we agreed on only tripleo-validations, hence me asking. But since I'm working the container stuff right now and testing the above patches I was thinking maybe I can help clean up the review queue a little bit :) | 08:51 |
| shardy | mandre: Yeah, that's fine - feel free to approve container related patches in -common too | 08:52 |
| flaper87 | shardy: it's mostly to group the packages with need in different stages. At the end of that set of patches, I remove 2 of those groups | 08:52 |
| shardy | if you have any doubt just ping me and I'll help with the reviews | 08:52 |
| flaper87 | s/with/we/ | 08:52 |
| mandre | shardy: the separate yum install are not a big issue since they're in a script, it will still be a single docker layer | 08:52 |
| flaper87 | shardy: that group of patches cut the image size by 500 MB | 08:52 |
| flaper87 | which was the final goal of this optimization | 08:52 |
| openstackgerrit | Giulio Fidente proposed openstack/python-tripleoclient: Update overcloud passwords on update command https://review.openstack.org/338213 | 08:52 |
| flaper87 | we were installing way more packages than we needed | 08:52 |
| flaper87 | we are* (until those patches land :P) | 08:53 |
| mandre | shardy: thanks! let's go crazy on the +2 | 08:53 |
| tobias_fiberdata | i get this when trying to build the images. | 08:53 |
| tobias_fiberdata | No package python-hardware-detect available. | 08:53 |
| openstackgerrit | Merged openstack/tripleo-common: Remove unnecessary package dhcp-client https://review.openstack.org/318300 | 08:53 |
| tobias_fiberdata | Failed to build image "ironic-python-agent" | 08:54 |
| openstackgerrit | Merged openstack/tripleo-common: Don't clean after `yum update` in script https://review.openstack.org/344603 | 08:54 |
| openstackgerrit | Merged openstack/tripleo-common: Remove configure_container.sh script https://review.openstack.org/344788 | 08:54 |
| openstackgerrit | Jonas Sticha proposed openstack/diskimage-builder: Correct order of parameters in call to tune2fs https://review.openstack.org/345261 | 08:56 |
| shardy | mandre, flaper87: all except https://review.openstack.org/#/c/312723 look good to me | 08:57 |
| shardy | I'd prefer the new hook didn't go in a tripleo specific repo | 08:58 |
| openstackgerrit | Jonas Sticha proposed openstack/diskimage-builder: Correct order of parameters in call to qemu-img convert https://review.openstack.org/345262 | 08:58 |
| shardy | is there any reason we can't add it to the heat-templates repo where all the other hooks are maintained? | 08:58 |
| flaper87 | shardy: no reason that I can't think of. I'll ping Ian when he's back just to be sure | 08:58 |
| flaper87 | I just took that patch over today | 08:59 |
| flaper87 | :P | 08:59 |
| *** jaosorior has quit IRC | 08:59 | |
| shardy | flaper87: ack, sounds good - mandre if the rest of that series works for you locally then feel free to approve | 08:59 |
| *** jaosorior has joined #tripleo | 08:59 | |
| mandre | shardy: thanks a lot for the reviews | 09:01 |
| *** paramite is now known as paramite|afk | 09:01 | |
| mandre | shardy: no idea why the hook is lives in the docker image, agreed it should be moved to the heat repo | 09:02 |
| shardy | tobias_fiberdata: Are you trying to build images for trunk or mitaka? | 09:03 |
| gfidente|afk | shardy, did you start any work to move the steps in overcloud.yaml ? | 09:03 |
| *** gfidente|afk is now known as gfidente | 09:03 | |
| shardy | gfidente|afk: Not yet, we can't do it until everything is completely switch to use step_config | 09:03 |
| *** bootsha has quit IRC | 09:03 | |
| tobias_fiberdata | i guess it doesnt matter now, the tripleO-node is updated to current trunk | 09:03 |
| gfidente | I am asking because hyperconverged stuff still hits this in ci and I think we'll have to implement it for that to work reliably | 09:03 |
| shardy | and there are still a few remaining patches needed for that | 09:03 |
| tobias_fiberdata | but yea | 09:04 |
| tobias_fiberdata | trunk | 09:04 |
| shardy | tobias_fiberdata: that package comes from the delorean trunk repo - are you doing openstack overcloud image build directly, or via tripleo.sh? | 09:05 |
| tobias_fiberdata | openstack overcloud image build --all | 09:05 |
| tobias_fiberdata | not tripleo.sh | 09:05 |
| *** isq has quit IRC | 09:07 | |
| *** isq has joined #tripleo | 09:08 | |
| shardy | http://docs.openstack.org/developer/tripleo-docs/basic_deployment/basic_deployment_cli.html#get-images | 09:08 |
| shardy | tobias_fiberdata: did you export USE_DELOREAN_TRUNK, DELOREAN_TRUNK_REPO and DELOREAN_REPO_FILE ? | 09:08 |
| shardy | that repo contains the missing package | 09:08 |
| tobias_fiberdata | i'll take a look | 09:09 |
| shardy | alternatively you can do tripleo.sh --repo-setup, then tripleo.sh --overcloud-images | 09:09 |
| tobias_fiberdata | aye | 09:09 |
| shardy | we need to get these docs auto-generated from exactly what we CI test... | 09:09 |
| tobias_fiberdata | will try that aswell | 09:09 |
| shardy | there's a discussion ongoing about doing that | 09:09 |
| shardy | (on the openstack-dev ML) | 09:09 |
| tobias_fiberdata | ah | 09:09 |
| tobias_fiberdata | [admin01@wizard yum.repos.d(openstack_admin)]$ echo $USE_DELOREAN_TRUNK | 09:10 |
| tobias_fiberdata | 1 | 09:10 |
| tobias_fiberdata | [admin01@wizard yum.repos.d(openstack_admin)]$ echo $DELOREAN_TRUNK_REPO | 09:10 |
| tobias_fiberdata | http://buildlogs.centos.org/centos/7/cloud/x86_64/rdo-trunk-master-tripleo/ | 09:10 |
| tobias_fiberdata | [admin01@wizard yum.repos.d(openstack_admin)]$ echo $DELOREAN_REPO_FILE | 09:10 |
| tobias_fiberdata | delorean.repo | 09:10 |
| tobias_fiberdata | that's the output | 09:10 |
| tobias_fiberdata | so yea | 09:10 |
| *** d0ugal has joined #tripleo | 09:10 | |
| tobias_fiberdata | but i'll try your shellscript | 09:10 |
| shardy | I'll take a look at fixing the docs later, that repo does contain the package so I'm not quite sure why it's not getting picked up during the build | 09:11 |
| tobias_fiberdata | yea shardy it does that's a weird thing tbh. | 09:12 |
| *** zoli|wfh is now known as zoli|brb | 09:12 | |
| *** jerrygb has joined #tripleo | 09:13 | |
| *** electrofelix has joined #tripleo | 09:15 | |
| *** paramite|afk is now known as paramite | 09:16 | |
| *** jerrygb has quit IRC | 09:18 | |
| *** d0ugal has quit IRC | 09:22 | |
| *** mgould|afk is now known as mgould | 09:23 | |
| openstackgerrit | Giulio Fidente proposed openstack/tripleo-heat-templates: Remove depends_on Controller nodes https://review.openstack.org/345277 | 09:24 |
| tobias_fiberdata | shardy, looks like it cannot find my OS-family? | 09:24 |
| openstackgerrit | Merged openstack/puppet-tripleo: Remove templates/database https://review.openstack.org/344996 | 09:24 |
| tobias_fiberdata | ################# | 09:24 |
| tobias_fiberdata | ./tripleo.sh: line 218: ./set-os-type: No such file or directory | 09:24 |
| tobias_fiberdata | haha | 09:24 |
| tobias_fiberdata | my undercloud-node is very much broken lol | 09:24 |
| tobias_fiberdata | :D | 09:24 |
| openstackgerrit | Giulio Fidente proposed openstack/tripleo-heat-templates: Cleanup templates from the shared CephCluster config https://review.openstack.org/345273 | 09:25 |
| *** zoli|brb is now known as zoli|wfh | 09:29 | |
| *** d0ugal has joined #tripleo | 09:29 | |
| openstackgerrit | Merged openstack/puppet-tripleo: Implement Apache pacemaker profile https://review.openstack.org/345013 | 09:29 |
| *** fzdarsky has quit IRC | 09:30 | |
| *** jubapa has joined #tripleo | 09:30 | |
| *** jubapa has quit IRC | 09:31 | |
| openstackgerrit | Merged openstack/puppet-tripleo: nova/api: purge archive_deleted_rows via cron https://review.openstack.org/343660 | 09:32 |
| openstackgerrit | Giulio Fidente proposed openstack/puppet-tripleo: Remove unused redis_vip parameter https://review.openstack.org/345285 | 09:37 |
| openstackgerrit | Martin Mágr proposed openstack-infra/tripleo-ci: Add opstools candidate repo https://review.openstack.org/325095 | 09:40 |
| openstackgerrit | Giulio Fidente proposed openstack/tripleo-heat-templates: Remove unused redis_vip parameter https://review.openstack.org/336573 | 09:40 |
| *** akrivoka has joined #tripleo | 09:43 | |
| *** fzdarsky has joined #tripleo | 09:44 | |
| shardy | Ok so we've fixed one introspection issue but now we're getting an "Unrecognised option "--timeout" error trying to boot the image via ipxe | 09:44 |
| shardy | https://github.com/openstack/instack-undercloud/commit/474abe263cb68ab239003354474d9bb6bc22ee3b | 09:44 |
| shardy | presumably that doesn't work | 09:44 |
| *** jubapa has joined #tripleo | 09:45 | |
| * shardy tries rebuilding images | 09:49 | |
| *** apetrich has quit IRC | 09:49 | |
| *** dbaxps has joined #tripleo | 09:51 | |
| ccamacho | dbaxps welcome! | 09:53 |
| *** d0ugal has quit IRC | 09:56 | |
| *** dbaxps has quit IRC | 09:57 | |
| *** bootsha has joined #tripleo | 09:58 | |
| *** dbaxps has joined #tripleo | 09:58 | |
| openstackgerrit | Chris Jones proposed openstack/puppet-tripleo: Next generation HA architecture work https://review.openstack.org/342650 | 10:01 |
| shardy | Ok rebuilding the IPA image/kernel fixes the latest introspection hang error | 10:02 |
| shardy | Hopefully RDO CI is using a cached image and it can just be rebuilt | 10:02 |
| *** d0ugal has joined #tripleo | 10:02 | |
| *** d0ugal has quit IRC | 10:03 | |
| *** d0ugal has joined #tripleo | 10:04 | |
| saneax | FYI, we are having the deep dive session today at 14:00 UTC - https://etherpad.openstack.org/p/tripleo-deep-dive-topics | 10:06 |
| *** yamahata has quit IRC | 10:06 | |
| *** tobias_fiberdata has quit IRC | 10:07 | |
| openstackgerrit | Chris Jones proposed openstack/tripleo-heat-templates: Next generation HA architecture work https://review.openstack.org/314208 | 10:07 |
| saneax | trown|outtypewww, with the master, oooq, I am getting error in overcloud-deploy - https://paste.fedoraproject.org/393393/90943751/ | 10:08 |
| openstackgerrit | mathieu bultel proposed openstack/tripleo-quickstart: WIP gate upgrade https://review.openstack.org/342161 | 10:13 |
| *** jerrygb has joined #tripleo | 10:14 | |
| *** radek__ has quit IRC | 10:18 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder: Move elements into diskimage_builder package https://review.openstack.org/344017 | 10:19 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder: Move diskimage-image-create to an entry point https://review.openstack.org/344532 | 10:19 |
| *** jerrygb has quit IRC | 10:20 | |
| *** jubapa has quit IRC | 10:21 | |
| *** tobias_fiberdata has joined #tripleo | 10:24 | |
| *** limao has quit IRC | 10:24 | |
| *** mbound has joined #tripleo | 10:26 | |
| *** apetrich has joined #tripleo | 10:30 | |
| *** d0ugal has quit IRC | 10:31 | |
| bandini | gfidente: shall we just merge https://bugs.launchpad.net/tripleo/+bug/1604039 or do we still want to wait for mitaka CI to work? | 10:31 |
| openstack | Launchpad bug 1604039 in tripleo "CI: delorean build of tripleo-heat-templates fails" [Critical,Triaged] - Assigned to Alan Pevec (apevec) | 10:31 |
| bandini | dciabrin: ^ | 10:31 |
| bandini | gfidente: I mean https://review.openstack.org/#/c/343253/ | 10:31 |
| *** dtantsur|bbl is now known as dtantsur | 10:32 | |
| dtantsur | hewbrocca, WUT? | 10:33 |
| *** jubapa has joined #tripleo | 10:33 | |
| dtantsur | oh that damn introspection.. | 10:33 |
| dtantsur | shardy, anything I can help with right now? | 10:33 |
| shardy | dtantsur: I don't think so, I tested locally and it works OK | 10:34 |
| dtantsur | awesome | 10:34 |
| gfidente | bandini, I would merge it, I have a couple of submission blocked on same problem and I know of at least another from marios | 10:34 |
| shardy | dtantsur: I suspect the RDO CI may be using a cached ipa image/kernel | 10:34 |
| shardy | as initially my local test failed because of the new --timeout option | 10:34 |
| dtantsur | shardy, well, it does for sure, trown|outtypewww might now the promotion rules | 10:34 |
| gfidente | but we should get mitaka/ci going ASAP | 10:35 |
| dtantsur | shardy, hmmm... I have hard time thinking how these are related... are you sure it's not connected with ipxe ROM versions? | 10:35 |
| gfidente | so I think I will focus on that today | 10:35 |
| shardy | dtantsur: No, it failed trying to boot the IPA kernel AFAICT | 10:35 |
| shardy | "unrecognised option --timeout" | 10:35 |
| gfidente | bandini, I'd like the other guys to weight on what is better to do | 10:35 |
| dtantsur | shardy, that's the iPXE error, not IPA one | 10:36 |
| dtantsur | shardy, there was an update to our shipped ipxe ROM to support it | 10:36 |
| dtantsur | shardy, maybe the version in CI is still old | 10:36 |
| *** jpich has quit IRC | 10:36 | |
| bandini | sshnaidm: what is your take about the ETA for https://bugs.launchpad.net/tripleo/+bug/1604039 to be fixed roughly (just interested in a ballbark figure like hours, days, weeks ;)? | 10:36 |
| openstack | Launchpad bug 1604039 in tripleo "CI: delorean build of tripleo-heat-templates fails" [Critical,Triaged] - Assigned to Alan Pevec (apevec) | 10:36 |
| shardy | dtantsur: doh, you're right | 10:36 |
| shardy | I changed my ipxe config | 10:36 |
| shardy | :facepalm: | 10:36 |
| shardy | I need some coffee :) | 10:36 |
| jaosorior | ccamacho hey dude | 10:37 |
| shardy | dtantsur: what package contains that, I updated everything locally and encountered that problem | 10:37 |
| *** tosky has joined #tripleo | 10:37 | |
| ccamacho | jaosorior introspection finished | 10:38 |
| dtantsur | shardy, depends on your setup actually. ipxe-roms-qemu on the virt host machine. | 10:38 |
| sshnaidm | bandini, I have already a workaround for this: https://review.openstack.org/#/c/345070/ apevec also submitted the similar: https://review.openstack.org/#/c/345106/ | 10:38 |
| shardy | ipxe-roms-qemu-20160127-1.git6366fa7a.el7.noarch | 10:38 |
| sshnaidm | bandini, when one of them will be merged it should work, at least I'm sure in mine - I tested it :) | 10:38 |
| dtantsur | shardy, this seems good | 10:38 |
| jaosorior | ccamacho, can you post the output of: openstack endpoint list --long | 10:39 |
| jaosorior | ? | 10:39 |
| bandini | sshnaidm: awesome!! :) gfidente ^ | 10:39 |
| sshnaidm | bandini, but the right and constant *solution* ball is on delorean side | 10:39 |
| *** aufi has quit IRC | 10:40 | |
| ccamacho | jaosorior yeahp, comparing it with yours right now | 10:40 |
| *** pino|work has quit IRC | 10:40 | |
| *** pino|work_ has joined #tripleo | 10:40 | |
| *** aufi has joined #tripleo | 10:40 | |
| *** pino|work_ is now known as pino|work | 10:40 | |
| jaosorior | ccamacho basically ironic inspector's public endpoint should be https | 10:40 |
| bandini | sshnaidm: yeah I sensed that there is a bit more work needed somewhere. I will take workaround for now and offer beers to whoever unblocks mitaka CI first :) | 10:40 |
| sshnaidm | bandini, let's start the race! :) | 10:41 |
| bandini | ;) | 10:41 |
| shardy | http://people.redhat.com/~shardy/screenshots/Screenshot_baremetalbrbm.png | 10:41 |
| ccamacho | yeahp.. but it's still http the package didnt install correctly /me checking | 10:41 |
| shardy | dtantsur: that is what happens until you remove --timeout from the ipxe config | 10:41 |
| shardy | (forever, it's an anti-timeout ;) | 10:42 |
| gfidente | sshnaidm, the only difference I see with apevac change is you added a git checkout, what is needed for? | 10:42 |
| jaosorior | shardy, dtantsur: Wasn't there a bug filed about ipxe failing because of that timeout option? | 10:42 |
| jaosorior | Seems to be also hardcoded https://github.com/openstack/ironic/blob/785c6aca673ff378eb7c69ff76a6d629d32daa78/ironic/drivers/modules/ipxe_config.template | 10:43 |
| *** rodrigods has quit IRC | 10:43 | |
| sshnaidm | gfidente, because changing a text within a source file is so dirty hack, so it could be broken in any case when this file is changed, I just try to save us from such case by pinning the files to same version | 10:43 |
| *** rodrigods has joined #tripleo | 10:44 | |
| gfidente | but the side effect is we'll pin to that version of rdoinfo.py | 10:44 |
| gfidente | "forever" | 10:44 |
| gfidente | not sure what are the implications | 10:45 |
| *** jpich has joined #tripleo | 10:45 | |
| dtantsur | shardy, could you please try to screenshot the iPXE ROM version when it boots? I'm pretty sure the RPM version you've stated is fine | 10:47 |
| shardy | gfidente, sshnaidm: see the ML reply from derekh, seems like some delorean changes may be needed, but a temporary workaround will be needed while that discussion happens | 10:47 |
| sshnaidm | gfidente, yes, or CI will be broken again when it changes, I don't insist of course, the both evils are pretty equal | 10:47 |
| gfidente | sshnaidm, yeah | 10:48 |
| gfidente | so I am leaving +2 but not +A to have more people looking | 10:48 |
| gfidente | maybe apevac himself | 10:48 |
| gfidente | *thanks* :) | 10:49 |
| *** dsariel has joined #tripleo | 10:50 | |
| *** rhallisey has joined #tripleo | 10:50 | |
| sshnaidm | I just hope the real solution will come fast and I could erase such hack from my code history :) | 10:50 |
| hewbrocca | bandini: +1000 on getting mitaka CI going again | 10:50 |
| hewbrocca | We also haven't had an RDO Mitaka promote in... | 10:51 |
| hewbrocca | six days | 10:51 |
| bandini | hewbrocca: first that fixes is gets a round of beers from me ;) | 10:52 |
| tobias_fiberdata | is it possible to get a force of typing yes or no like when you ssh and add a print, when doing this? Openstack stack delete <ID> | 10:52 |
| jstir____ | Tobias_Fiberdata: ITYM 'OpenStack'. | 10:52 |
| tobias_fiberdata | ? | 10:52 |
| tobias_fiberdata | jstir____, sup? | 10:52 |
| shardy | tobias_fiberdata: Yes, pass --yes IIRC | 10:53 |
| hewbrocca | tobias_fiberdata: heh jstir is a bot | 10:53 |
| hewbrocca | that corrects "misspellings" of Openstack | 10:53 |
| tobias_fiberdata | shardy, what i mean is that it asks for that by default when running the command | 10:53 |
| jstir____ | hewbrocca: ITYM 'OpenStack'. | 10:53 |
| tobias_fiberdata | hewbrocca, why did it do that? :D | 10:54 |
| tobias_fiberdata | ah | 10:54 |
| tobias_fiberdata | lol | 10:54 |
| tobias_fiberdata | you know when sshing to a new host | 10:54 |
| tobias_fiberdata | you get a yes or no question | 10:54 |
| hewbrocca | !chanslap jstir____ | 10:54 |
| *** jstir____ is now known as JSTIR_THE_CHANNE | 10:54 | |
| openstack | hewbrocca: Error: "chanslap" is not a valid command. | 10:54 |
| JSTIR_THE_CHANNE | ><}}}*> jstir____ | 10:54 |
| JSTIR_THE_CHANNE | jstir____: you're like a forty degree day! | 10:54 |
| *** JSTIR_THE_CHANNE is now known as jstir____ | 10:54 | |
| tobias_fiberdata | openstackish jstir____ | 10:54 |
| shardy | tobias_fiberdata: Yeah, that's annoying but expected - I'm not sure if there's an environment variable which can set --yes, will have to check the code | 10:54 |
| tobias_fiberdata | shardy, it's a nice lifesaver | 10:54 |
| tobias_fiberdata | of the man writing funny stuff for fun | 10:55 |
| tobias_fiberdata | and then slips | 10:55 |
| tobias_fiberdata | even though it's no by default | 10:55 |
| tobias_fiberdata | my collegue zephcom is doing rm -rf / --no-preserve just to get me a heart attack on important machines for fun. | 10:56 |
| tobias_fiberdata | ^^ | 10:56 |
| tobias_fiberdata | my life is hard here | 10:56 |
| *** bootsha has quit IRC | 10:58 | |
| *** paramite has quit IRC | 10:58 | |
| *** pkovar has joined #tripleo | 11:00 | |
| shardy | http://people.redhat.com/~shardy/screenshots/Screenshot_baremetalbrbm_1.png | 11:01 |
| shardy | dtantsur: ^^ | 11:01 |
| shardy | looks like 00:03.0 | 11:01 |
| dtantsur | shardy, the version is c4bce43, they use commit hashes | 11:02 |
| dtantsur | and this version looks familiar and it is NOT 6366fa7a which you have in the package | 11:03 |
| shardy | aha | 11:03 |
| dtantsur | so somewhere the package is not updated | 11:03 |
| dtantsur | are you sure you have the update package *on the virt host* (not undercloud!) | 11:03 |
| dtantsur | ? | 11:03 |
| shardy | Ah, no it's the undercloud that I've updated | 11:03 |
| *** bootsha has joined #tripleo | 11:04 | |
| hewbrocca | Man | 11:05 |
| hewbrocca | this is tedious to track down, isn't it... | 11:05 |
| dtantsur | oh it is | 11:05 |
| dtantsur | and wonder if we can have tripleo-quickstart check that | 11:05 |
| dtantsur | trown|outtypewww, ^^^ | 11:05 |
| shardy | I don't see that package in yum check-update on the host, but I'll updating everything anyway | 11:06 |
| hewbrocca | dtantsur: we should for sure have some kind of validation to make sure the ipxe rom is the right version, if it is possible for it to get out of sync | 11:06 |
| *** milan has joined #tripleo | 11:06 | |
| shardy | It's not great if we've got this hard coupling of versions with the underlying host | 11:06 |
| shardy | like, it seems likely we've got different versions in upstream and RDO CI causing different results | 11:07 |
| shardy | some sort of validation would be a good idea if that's possible | 11:07 |
| dtantsur | it's not great at all... | 11:12 |
| jaosorior | jistr: Hey dude, do you have an upstream tripleo deployment? | 11:12 |
| dtantsur | here is the problem for history: | 11:12 |
| dtantsur | old iPXE versions are buggy like hell | 11:12 |
| dtantsur | new iPXE versions are also buggy, but we can at least put some workarounds in place | 11:13 |
| dtantsur | so we went for fixing new versions at the expense of limiting support for new versions | 11:13 |
| dtantsur | shardy, ^^^ | 11:13 |
| jistr | jaosorior: something bad happened to the overcloud overnight, but other than that yea | 11:13 |
| dtantsur | (second new is "old") | 11:13 |
| shardy | GAH | 11:13 |
| jistr | jaosorior: i'll probably need to redeploy at least overcloud | 11:13 |
| shardy | updated everything on the virt host, rebooted and still have the old version | 11:13 |
| hewbrocca | blearrgh | 11:14 |
| shardy | dtantsur: do I need to destroy all the VMs and recreate them? | 11:14 |
| jaosorior | jistr: I need help testing one patch for the undercloud. Got time for that? | 11:14 |
| ccamacho | jaosorior, cant make this change to refresh, if you want you can log in my env | 11:14 |
| shardy | I'm not clear where exactly this version is defined tbh | 11:14 |
| dtantsur | shardy, I don't think so | 11:14 |
| jaosorior | ccamacho: Sure, pm me the credentials | 11:14 |
| dtantsur | shardy, what's the RPM version on the virt host? | 11:14 |
| openstackgerrit | Merged openstack/tripleo-heat-templates: Remove the ControllerCount contraint https://review.openstack.org/344739 | 11:14 |
| shardy | Yeah it's wrong | 11:15 |
| shardy | let me juggle repos and try again | 11:15 |
| jistr | jaosorior: hmm not much, trying to sort out 8->9 upgrades... what patch would that be? (trying to get an estimate how much time is needed to test it) | 11:15 |
| dtantsur | shardy, I think it's in some .*common.* repo | 11:15 |
| hewbrocca | oh god, I forgot we're also using the horrible yum repo priorities hack | 11:15 |
| hewbrocca | what could possibly go wrong | 11:15 |
| dtantsur | nothing \o/ let's ship it and go to the beach | 11:16 |
| jaosorior | jsitr: TLS for inspector https://review.openstack.org/#/c/332641/ | 11:16 |
| *** jerrygb has joined #tripleo | 11:16 | |
| *** stendulker has quit IRC | 11:17 | |
| hewbrocca | LOL | 11:17 |
| hewbrocca | jaosorior: jsitr is a spelling we haven't coverd | 11:17 |
| hewbrocca | we have jstir____ jistr and jrist | 11:17 |
| dtantsur | file a bug! | 11:17 |
| jaosorior | jistr | 11:17 |
| jaosorior | jistr: TLS for inspector https://review.openstack.org/#/c/332641/ | 11:18 |
| *** tzumainn has quit IRC | 11:18 | |
| *** tzumainn has joined #tripleo | 11:18 | |
| *** tzumainn has quit IRC | 11:19 | |
| *** tzumainn has joined #tripleo | 11:19 | |
| tobias_fiberdata | how do i clear a node from it's instance_uuid? | 11:20 |
| dtantsur | tobias_fiberdata, if you really have to do it (usually you don't!) $ ironic node-update UUID remove instance_uuid | 11:20 |
| dtantsur | but beware: it should not be required in a normal flow | 11:20 |
| tobias_fiberdata | i have two nodes that is stuck | 11:20 |
| *** jerrygb has quit IRC | 11:21 | |
| tobias_fiberdata | i've deployed a stack | 11:21 |
| tobias_fiberdata | but when i've removed the stack the machines hasnt got cleared | 11:21 |
| tobias_fiberdata | http://paste.openstack.org/show/538970/ | 11:21 |
| tobias_fiberdata | list this | 11:22 |
| jistr | jaosorior: oof i'd prefer not to reinstall the undercloud ATM if possible... you don't have a machine ATM? | 11:22 |
| jaosorior | jistr: I have one but I can't reproduce the issue bnemec has | 11:22 |
| jaosorior | it's working for me | 11:22 |
| shardy | tobias_fiberdata: what does heat stack-list show? | 11:22 |
| shardy | is the stack completely gone? | 11:23 |
| shardy | also check nova list | 11:23 |
| tobias_fiberdata | empty | 11:23 |
| tobias_fiberdata | empty | 11:23 |
| tobias_fiberdata | both nova list and heat stack-list | 11:23 |
| shardy | Ok, that is strange, the instance UUID should be cleared when the nova server is deleted | 11:23 |
| tobias_fiberdata | and i got error that there wasnt any resources when trying to deploy again | 11:23 |
| tobias_fiberdata | yea it is | 11:24 |
| tobias_fiberdata | there's only 1 that has been cleared | 11:24 |
| shardy | but they are all power on state | 11:24 |
| tobias_fiberdata | hmm | 11:24 |
| jistr | jaosorior: from bnemec paste -- InputException: Invalid input [name=baremetal_introspection.wait_for_finish, class=NoneType, unexpected=[u\'uuids\'] | 11:25 |
| tobias_fiberdata | that's true | 11:25 |
| openstackgerrit | Karthik S proposed openstack/tripleo-heat-templates: Add max_reserved_vfs in THT https://review.openstack.org/345352 | 11:25 |
| tobias_fiberdata | ah ye | 11:25 |
| tobias_fiberdata | i started a introspection on them | 11:25 |
| jistr | jaosorior: at a first glance that doesn't look very related to TLS | 11:25 |
| tobias_fiberdata | to see if that solved the problem | 11:25 |
| jaosorior | jsitr: it's not related at all to TLS. Seems to be a mistral issue | 11:25 |
| jaosorior | *jistr | 11:25 |
| jaosorior | but anyway, trying to reproduce it, ccamacho is helping out too | 11:25 |
| tobias_fiberdata | now they are all power off | 11:25 |
| tobias_fiberdata | introspection is finished | 11:26 |
| tobias_fiberdata | no errors | 11:26 |
| tobias_fiberdata | but still instance UUID there | 11:26 |
| tobias_fiberdata | problem is that even if i remove that instance_uuid i have instance_info and stuff like as properties | 11:26 |
| tobias_fiberdata | is that gonna be a problem in this case? | 11:27 |
| jaosorior | jistr: but who knows, maybe there is some component with a hardcoded endpoint, or something of the sort... gotta see if we can reproduce it first | 11:27 |
| *** dbaxps has left #tripleo | 11:29 | |
| dtantsur | tobias_fiberdata, instance_uuid will be a problem, see command I've posted above for how to remove it | 11:30 |
| jistr | marios: bandini: heya i'm in 8->9 upgrade before triggering first stack-update. What unlanded patches do we cherry pick for testing ATM? i guess https://review.openstack.org/#/c/344823/ and https://review.openstack.org/#/c/338879/ ? do you know about some more? | 11:30 |
| tobias_fiberdata | dtantsur, yea will do that | 11:30 |
| *** akuznetsov has joined #tripleo | 11:31 | |
| tobias_fiberdata | dtantsur worked | 11:33 |
| tobias_fiberdata | dtantsur, now i'll see if i can deploy again without getting error that there's no resources found :) | 11:33 |
| tobias_fiberdata | thanks for the help | 11:33 |
| *** ohamada has quit IRC | 11:35 | |
| *** ohamada has joined #tripleo | 11:35 | |
| tobias_fiberdata | great forgot provision-staet | 11:35 |
| tobias_fiberdata | great forgot provision-state | 11:35 |
| *** ohamada has quit IRC | 11:35 | |
| openstackgerrit | Merged openstack/instack-undercloud: Remove ipxe selinux policy https://review.openstack.org/344954 | 11:36 |
| marios | jistr: yah the httpd race one and i also had https://review.openstack.org/#/c/342725/3 to get through keystone | 11:41 |
| tobias_fiberdata | haha my undercloud is a bit tired, takes 20sec to install iotop :D | 11:42 |
| jistr | marios: ah perfect, thanks, i'd miss that one | 11:42 |
| *** lucasagomes is now known as lucas-hungry | 11:42 | |
| marios | jistr: can't think of any more (occ/rabbit resource agents for controller but they aren't fixes to apply from review as such) | 11:42 |
| jistr | marios: right, thanks | 11:42 |
| *** zoli|wfh is now known as zoli|lunch | 11:44 | |
| *** dprince has joined #tripleo | 11:45 | |
| *** bvandenh has quit IRC | 11:47 | |
| sshnaidm | dprince, hi, sorry, I didn't get about a status page, can you explain please? | 11:49 |
| *** bfournie has quit IRC | 11:51 | |
| jistr | matbu, bandini, marios: is it ok if i rebase https://review.openstack.org/#/c/338879 on top of https://review.openstack.org/#/c/342725 ? we need them both but they conflict with each other (i'm currently unable to just fetch patches and apply them, which complicates producing any copy-pasteable workflow quite a bit) | 11:53 |
| dprince | sshnaidm: hi, so the question is would it be possible to add the periodic jobs into a new tab on this report: http://git.openstack.org/cgit/openstack-infra/tripleo-ci/tree/scripts/tripleo-jobs-gerrit.py | 11:54 |
| dprince | sshnaidm: presumably the periodic jobs you are talking about get executed by upstream Zuul right? | 11:55 |
| *** jubapa_ has joined #tripleo | 11:55 | |
| sshnaidm | dprince, actually I was asking about my status page: http://status-tripleoci.rhcloud.com , it's much more than periodic jobs | 11:56 |
| sshnaidm | dprince, yeah, I mean all TripleO CI jobs - gates, periodic, everything | 11:56 |
| dprince | sshnaidm: right, so I think that is probably a separate thing then | 11:56 |
| sshnaidm | dprince, periodic jobs as a side effect there :) | 11:56 |
| marios | jistr: yeah it makes sense to me | 11:57 |
| hewbrocca | dtantsur: So... are we likely to have this same virt-host problem with tripleo-CI? | 11:57 |
| dprince | sshnaidm: let me put it this way. I think tripleo.org is currently for upstream stuff. If the jobs report you have is upstream periodic jobs then I'd recommend putting it into the tripleo-ci/scripts report | 11:57 |
| hewbrocca | derekh: would probably know I guess | 11:57 |
| dtantsur | hewbrocca, if we get a problem with unknown argument --timeout, then yes | 11:57 |
| jistr | marios: ack thx i'll do it. It's just a few added chars ('interleave=true') for the race condition patch. | 11:57 |
| dprince | sshnaidm: if not the separete report is just fine as is I think | 11:58 |
| * dtantsur has lost track of which problem we discuss where | 11:58 | |
| openstackgerrit | Jiri Tomasek proposed openstack/tripleo-ui: Webpack https://review.openstack.org/343592 | 11:59 |
| openstackgerrit | Jiri Tomasek proposed openstack/tripleo-ui: Remove app.conf to use a JS-based config file https://review.openstack.org/343658 | 11:59 |
| *** jubapa_ has quit IRC | 11:59 | |
| hewbrocca | Does anybody know if we are seeing "introspection timeouts" in TripleO CI? | 11:59 |
| sshnaidm | dprince, yeah, it's not scripts, it's a webserver with data, a lot of python and DB will be soon.. | 11:59 |
| sshnaidm | dprince, it's just a different approach - to analyze the fails and show reasons, not like current status page | 12:00 |
| *** jayg|g0n3 is now known as jayg | 12:00 | |
| *** weshay has joined #tripleo | 12:02 | |
| openstackgerrit | Jiri Stransky proposed openstack/tripleo-heat-templates: Remove two races during the L->M Keystone under httpd migration https://review.openstack.org/338879 | 12:03 |
| *** akuznetsov has quit IRC | 12:04 | |
| sshnaidm | dprince, so, could we put a link somewhere on the site..? | 12:05 |
| sshnaidm | ianw, hi | 12:05 |
| dprince | sshnaidm: So your report is just a different way to display upstream jobs then? | 12:06 |
| dprince | sshnaidm: the domain name you are using confused me.... | 12:06 |
| sshnaidm | dprince, yeah | 12:06 |
| dprince | sshnaidm: could we switch your domain name? | 12:06 |
| sshnaidm | dprince, if you have another hosting, I'd happy to take it :) | 12:07 |
| tobias_fiberdata | where can i see the verbose and debug output for tripleO deployments? | 12:07 |
| tobias_fiberdata | is it heat? | 12:07 |
| tobias_fiberdata | my deployment has stalled | 12:07 |
| tobias_fiberdata | for some reason | 12:07 |
| sshnaidm | dprince, I think it's possible with openshift, but you need a paid account for this.. | 12:07 |
| sshnaidm | weshay, ^^ | 12:07 |
| sshnaidm | dprince, or CNAME maybe | 12:08 |
| sshnaidm | tobias_fiberdata, btw, today in 2 hours is "debugging deployments": https://etherpad.openstack.org/p/tripleo-deep-dive-topics | 12:10 |
| tobias_fiberdata | nice nice | 12:10 |
| tobias_fiberdata | i will be there listening! | 12:10 |
| tobias_fiberdata | for sure | 12:10 |
| *** d0ugal has joined #tripleo | 12:12 | |
| openstackgerrit | Jiri Tomasek proposed openstack/tripleo-ui: Use Webpack for application processing https://review.openstack.org/343592 | 12:12 |
| weshay | sshnaidm, re: the domain name.. we're stuck w/ rhcloud.com | 12:12 |
| bandini | jistr: totally fine by me! | 12:12 |
| dprince | sshnaidm: still, the nice thing about our existing report is anyone can edit it. If we link, to or dsiplay yours I think I'd want the sources visible somewhere too | 12:12 |
| weshay | ya.. should be an "edit me" link | 12:12 |
| sshnaidm | dprince, oh, of course they are visible: https://github.com/sshnaidm/sova | 12:13 |
| matbu | jistr: +1 also | 12:13 |
| dtantsur | sova \o/ | 12:13 |
| sshnaidm | dtantsur, owl :) | 12:14 |
| dtantsur | correct :) | 12:14 |
| *** radek__ has joined #tripleo | 12:14 | |
| weshay | morazi, ^ | 12:15 |
| weshay | sshnaidm, nice work on the page btw | 12:15 |
| sshnaidm | weshay, thanks | 12:15 |
| openstackgerrit | Merged openstack-infra/tripleo-ci: Workaround for building stable branches https://review.openstack.org/345070 | 12:16 |
| sshnaidm | dprince, if you have ideas for hosting for this site, I can move there of course. I'll need only about 100MB-200MB disk space and some mongo, that's all | 12:17 |
| *** jerrygb has joined #tripleo | 12:17 | |
| dprince | sshnaidm: I could create you one and set it up under tripleo.org | 12:18 |
| *** trown|outtypewww is now known as trown | 12:19 | |
| sshnaidm | dprince, great, let's try it then | 12:19 |
| dprince | sshnaidm: let me ponder this a bit... I'd rather be out of the cloud resources race honestly | 12:19 |
| sshnaidm | dprince, np, openshift does its work so far | 12:20 |
| dprince | sshnaidm: but it takes time to do things perhaps more correctly so maybe we just do it | 12:20 |
| dprince | sshnaidm: yep, working (and free) is always nice | 12:20 |
| dprince | sshnaidm: if we set this up somewhere would it make sense to create a puppet modules or something to automatically deploy your site? | 12:21 |
| *** jerrygb has quit IRC | 12:21 | |
| dprince | sshnaidm: because as soon as its running automatically posting new changes/additions is going to be a request | 12:21 |
| dprince | sshnaidm: again, this is the nice thing about the current site/model | 12:22 |
| dprince | sshnaidm: pre-sumably openshift does this nicely for you already. If we host it elsewhere we'd loose this | 12:22 |
| sshnaidm | dprince, of course, I thought about docker | 12:22 |
| dprince | sshnaidm: still, something is going to need to rebuild/deploy the new container | 12:23 |
| sshnaidm | dprince, yeah, a little ansible playbook - is it OK? | 12:24 |
| dprince | sshnaidm: I prefer shell or puppet I suppose | 12:24 |
| sshnaidm | dprince, all right, I don't mind | 12:25 |
| dprince | sshnaidm: do what you think works best thogh | 12:25 |
| dprince | sshnaidm: like the existing site/deploy doesn't require Ansible | 12:25 |
| dprince | sshnaidm: so keeping it as simple as possible would be ideal I think | 12:25 |
| sshnaidm | dprince, so shell I suppose | 12:27 |
| *** athomas has quit IRC | 12:27 | |
| * dtantsur would never pick shell over anything else | 12:28 | |
| dtantsur | though DIB authors won't agree with me for sure | 12:28 |
| * dtantsur hides | 12:29 | |
| dprince | sshnaidm: nothing wrong with some good god fearing shell man | 12:29 |
| slagle | dtantsur: you use ansible everytime you want to run "ls"? /me -> troll :) | 12:30 |
| dtantsur | slagle, well, if deploying something involves running "ls", then I'd probably use bash ;) | 12:30 |
| sshnaidm | dtantsur, I'd agree, but it's really two lines in shell :) Or 2 thousand in puppet /trollmode | 12:30 |
| dtantsur | as long as these two lines don't slowly grow to twenty two and futher... :) | 12:31 |
| sshnaidm | dtantsur, yeah, the wisdom is to know when to stop :) | 12:32 |
| dtantsur | once I bothered to create a playbook for configure my personal development environment | 12:32 |
| dtantsur | proved much longer than I expected: https://github.com/dtantsur/config/blob/master/configure.yml | 12:32 |
| dprince | dtantsur: yeah, enjoy that I guess | 12:33 |
| *** bfournie has joined #tripleo | 12:34 | |
| *** athomas has joined #tripleo | 12:34 | |
| sshnaidm | dtantsur, nice, I should have some like this, I did with fabric before | 12:34 |
| *** zoli|lunch is now known as zoli|wfh | 12:35 | |
| dtantsur | I especially love this bit: https://github.com/dtantsur/config/blob/master/configure.yml#L4-L7 | 12:36 |
| dtantsur | it gives me a nice warning that I should use the "dnf" module for this operation (which makes the "dnf" module work.) | 12:37 |
| dtantsur | ¯\_(ツ)_/¯ | 12:37 |
| tobias_fiberdata | do i have to wipe the nodes after a heatstack failure? when redeploying? | 12:37 |
| tobias_fiberdata | or does solve that stuff itself? | 12:37 |
| dtantsur | tobias_fiberdata, what do you mean by wipe? | 12:37 |
| ccamacho | bandini, quick question based on http://paste.openstack.org/show/538998/ Will $sync_db always be true in a non-ha environment? | 12:37 |
| tobias_fiberdata | dtantsur, the hdds | 12:37 |
| tobias_fiberdata | dtantsur, just root | 12:38 |
| tobias_fiberdata | disk | 12:38 |
| tobias_fiberdata | cause it's stuck again | 12:38 |
| *** egafford has joined #tripleo | 12:38 | |
| dtantsur | tobias_fiberdata, it depends on how many of them you have.. we sometime had problems with several disk devices, if a different root device is used for different deployments | 12:38 |
| tobias_fiberdata | i guess i have to boot them up with a live cd and wipe the disks | 12:39 |
| tobias_fiberdata | now i got erorr in this stack saying it's no valid hosts | 12:39 |
| tobias_fiberdata | anyways | 12:39 |
| tobias_fiberdata | break | 12:39 |
| tobias_fiberdata | coffe | 12:39 |
| tobias_fiberdata | coffee | 12:39 |
| bandini | ccamacho: no, only it will be true only on the first controller of the non-ha environment. clearly if you deploy on a single controller, yes it will always be true, but you cannot always assume that | 12:39 |
| *** liverpooler has quit IRC | 12:40 | |
| bandini | ccamacho: does that answer your question? | 12:40 |
| ccamacho | bandini yes, because I'm testing now the mistral composable submission locally and I wanted to have that condition also validated! Thanks!! | 12:41 |
| bandini | ccamacho: excellent ;) | 12:42 |
| *** jcoufal has joined #tripleo | 12:43 | |
| *** aufi_ has joined #tripleo | 12:43 | |
| *** rlandy has joined #tripleo | 12:43 | |
| shardy | https://review.openstack.org/#/c/338551/ (convert swift ringbuilder to composable services format) is ready for review if anyone has a moment | 12:46 |
| shardy | now passing CI | 12:46 |
| dprince | shardy: on it | 12:47 |
| shardy | I know there is more we need to do re ringbuilder, but that just converts to the new format | 12:47 |
| shardy | dprince: thanks :) | 12:48 |
| dprince | shardy: did you test locally? I don't think our CI touches swift-storage.yaml ATM | 12:48 |
| shardy | dprince: Yeah I have tested locally and it seems to work OK - I'll pull it an re-confirm now though as last test was a few days ago | 12:48 |
| *** lucas-hungry is now known as lucasagomes | 12:49 | |
| gfidente | shardy, how went with the hieradata migration from EmilienM ? | 12:49 |
| gfidente | got any chance to try that? | 12:49 |
| shardy | gfidente: I haven't had a chance to look at it yet as I was working on broken introspection all morning | 12:50 |
| gfidente | shardy, ack I tried again and reproduced the issue I had initially | 12:50 |
| gfidente | which isn't same trown is seeing though | 12:50 |
| shardy | gfidente: ack, OK I'll re-test swift then try that patch | 12:51 |
| *** lblanchard has joined #tripleo | 12:51 | |
| openstackgerrit | Merged openstack/tripleo-heat-templates: Include new apache pacemaker profile https://review.openstack.org/345017 | 12:51 |
| dprince | shardy: your call on testing locally. I'm +2 with a minor inline comment | 12:51 |
| dprince | shardy, ccamacho, jaosorior: FWIW it would be nice to go on and add 'service_name' to new services up for review per https://review.openstack.org/#/c/344440/ | 12:52 |
| dprince | I have an explicit unit test there so my patch shouldn't land without every service having them... but catching them sooner would be nice too | 12:53 |
| jaosorior | dprince: What will that service-name be used for? | 12:54 |
| dprince | jaosorior: serveral things I think. It will help upgrades because knowing which services are deployed is useful | 12:55 |
| dprince | jaosorior: and it helps composable firewall rules here https://review.openstack.org/#/c/344985/ | 12:55 |
| jaosorior | I see | 12:55 |
| dprince | jaosorior: the composable firewall rules stuff is a bit of a compremise until we can wrote a proper heat hiera hook | 12:55 |
| shardy | dprince: Looks good, it's a shame it's not compatible with the keys in ServiceNetMap though | 12:56 |
| jaosorior | dprince: it's not entirely clear to me how the format for that list will be though. How does the deployed file look in the end? | 12:56 |
| shardy | if it was we could use it to generate an endpoint map only containing *-api services that are actually enabled | 12:57 |
| shardy | maybe we can have another mapping to help with that in the future | 12:57 |
| *** masco has quit IRC | 12:57 | |
| jaosorior | shardy has an excellent point there | 12:57 |
| *** jeckersb is now known as jeckersb_gone | 12:57 | |
| dprince | shardy: if we CamelCase it I think it could be? | 12:57 |
| dprince | shardy: there are still some differences though | 12:57 |
| shardy | Yeah, ServiceNetMap isn't consistent, with stuff like NeutronTenantNetwork and CeilometerApiNetwork | 12:58 |
| shardy | maybe we can transition to having the service->network mapping defined inside the service templates too? | 12:58 |
| *** akuznetsov has joined #tripleo | 12:58 | |
| dprince | shardy: so making that net map composable can be taken as an item some time | 12:58 |
| shardy | I think there is a chicken/egg problem there with Endpoint Map atm tho | 12:58 |
| dprince | shardy: probably not our most important issue ATM though | 12:58 |
| shardy | dprince: Yeah, agreed, I'm fine with this as-is, just thinking aloud | 12:59 |
| dprince | shardy: plenty of work to do yet :) | 12:59 |
| dprince | shardy: we can leave a lot of things in place for now and still claim composability as a feature I think | 12:59 |
| dprince | shardy: racing to get truely composable roles though is probably the highest priority... so what that means to me is do whatever we need to make your jinja2 template thing possible | 13:00 |
| openstackgerrit | John Trowbridge proposed openstack/tripleo-quickstart: Add workaround for out of date iPXE ROM in CentOS https://review.openstack.org/345410 | 13:00 |
| dprince | shardy: Endpointmap and serviceNetMap refactoring don't impact that effort I think so they can come afterwards | 13:00 |
| *** aufi has quit IRC | 13:01 | |
| *** julim has joined #tripleo | 13:01 | |
| EmilienM | gfidente: it's weird CI is passing on both ha and nonha | 13:01 |
| shardy | dprince: Yeah, agreed | 13:01 |
| *** jubapa_ has joined #tripleo | 13:01 | |
| shardy | I've just been thinking about those maps in the context of moving the bind_ip stuff | 13:02 |
| *** cdearborn has joined #tripleo | 13:02 | |
| derekh | <hewbrocca> derekh: would probably know I guess <--- remember what this was about? | 13:02 |
| gfidente | EmilienM, yeah indeed | 13:02 |
| hewbrocca | derekh: oh, something about the ipxe thing, but I think it's resolved | 13:02 |
| hewbrocca | NM | 13:02 |
| derekh | hewbrocca: ok | 13:03 |
| gfidente | EmilienM, though do you have another environment where we can test those quickly? | 13:03 |
| EmilienM | gfidente: not atm but I'm working on it | 13:03 |
| gfidente | EmilienM, it's an important cleanup to do | 13:04 |
| gfidente | and the sooner we do it the better as if other stuff lands meantime it might miss bits | 13:04 |
| EmilienM | gfidente: I made the huge mistake to do all in one | 13:05 |
| EmilienM | gfidente: usually I split things but I wanted to save CI resources | 13:05 |
| EmilienM | next time I'll split things | 13:05 |
| gfidente | ahahah | 13:06 |
| jaosorior | lol | 13:06 |
| jaosorior | yeah, that is a rough commit | 13:06 |
| gfidente | EmilienM, it's never going to work well for everybody | 13:06 |
| jaosorior | but it makes no sense, it should work. And in CI the hieradata is in the correct place and everything | 13:06 |
| gfidente | either one does it in one change and it was better to split | 13:07 |
| gfidente | or split in more changes and it was better a single one | 13:07 |
| openstackgerrit | Florian Fuchs proposed openstack/tripleo-ui: Move plan upload errors from notifications to formerrors https://review.openstack.org/332181 | 13:07 |
| EmilienM | lesson learnt | 13:07 |
| *** dbecker has quit IRC | 13:08 | |
| *** dbecker has joined #tripleo | 13:08 | |
| gfidente | so I wouldn't bother, I'm fine merging if it passes even once locally for me | 13:08 |
| openstackgerrit | Steven Hardy proposed openstack/tripleo-heat-templates: Convert Swift ringbuilder to composable services format https://review.openstack.org/338551 | 13:10 |
| openstackgerrit | Carlos Camacho proposed openstack/puppet-tripleo: Add pacemaker mistral profile https://review.openstack.org/343346 | 13:10 |
| shardy | dprince: ^^ added service_name | 13:10 |
| EmilienM | gfidente: trown said it would break RDO but I wonder how | 13:11 |
| *** jpich has quit IRC | 13:12 | |
| trown | ya it is unclear to me *why* hiera is missing with that patch and quickstart, but it is reproducible with all published images including tripleo-ci ones | 13:14 |
| openstackgerrit | Emilien Macchi proposed openstack-infra/tripleo-ci: Add multi-node job in cistatus.html https://review.openstack.org/345418 | 13:15 |
| EmilienM | slagle: ^ | 13:15 |
| EmilienM | slagle: gate-tripleo-ci-centos-7-nonha-multinode-nv has been created and is now in our gate | 13:17 |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack/puppet-tripleo: Add ironic inspector as a terminated HAProxy endpoint https://review.openstack.org/345183 | 13:18 |
| *** jerrygb has joined #tripleo | 13:18 | |
| slagle | EmilienM: i saw, thanks! | 13:20 |
| slagle | wasn't sure if they'd updated zuul yet or not though | 13:20 |
| *** akshai has joined #tripleo | 13:20 | |
| slagle | EmilienM: oh yea, i see the job running on all patches now | 13:21 |
| slagle | cool ! | 13:21 |
| EmilienM | awesome you mean? :) | 13:21 |
| slagle | yes, very | 13:22 |
| *** jpich has joined #tripleo | 13:22 | |
| *** jerrygb has quit IRC | 13:22 | |
| *** akshai_ has joined #tripleo | 13:22 | |
| *** cdearborn has quit IRC | 13:23 | |
| *** akshai has quit IRC | 13:26 | |
| *** milan has quit IRC | 13:27 | |
| *** milan has joined #tripleo | 13:27 | |
| openstackgerrit | wes hayutin proposed openstack/tripleo-quickstart: Add workaround for out of date iPXE ROM in CentOS https://review.openstack.org/345410 | 13:28 |
| openstackgerrit | John Trowbridge proposed openstack/tripleo-quickstart: Add workaround for out of date iPXE ROM in CentOS https://review.openstack.org/345410 | 13:28 |
| *** jeckersb_gone is now known as jeckersb | 13:28 | |
| openstackgerrit | Merged openstack/tripleo-common: Remove unnecessary packages https://review.openstack.org/344789 | 13:30 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/tripleo-common: Updated from global requirements https://review.openstack.org/339948 | 13:33 |
| *** milan has quit IRC | 13:34 | |
| *** jerrygb has joined #tripleo | 13:34 | |
| *** milan has joined #tripleo | 13:35 | |
| trown | slagle: it doesnt affect tripleo-ci because we dont test introspection in tripleo-ci | 13:36 |
| *** jubapa has left #tripleo | 13:36 | |
| *** jubapa has joined #tripleo | 13:36 | |
| *** trown is now known as trown|mtg | 13:37 | |
| EmilienM | slagle: some multinode jobs are failing though: ssh: connect to host 10.208.101.86 port 22: No route to host | 13:37 |
| sshnaidm | trown, wasn't it fixed recently? | 13:37 |
| EmilienM | have you seen that in your testing before? | 13:37 |
| trown|mtg | sshnaidm: we have one job in tripleo-ci, and that job does not test introspection | 13:38 |
| slagle | trown|mtg: when we turn that back on, it will start affecting tripleo-ci? | 13:38 |
| hewbrocca | the ipxe thing? | 13:38 |
| *** pgadiya has quit IRC | 13:39 | |
| trown|mtg | slagle: well it is something on the virthost, and there is not virthost in tripleo-ci | 13:39 |
| hewbrocca | no, I think it won't, because of the way TripleO CI builds the virt host | 13:39 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-tripleo: Deploy Keystone and Gnocchi API with new Apache profile https://review.openstack.org/345025 | 13:39 |
| hewbrocca | or that | 13:39 |
| hewbrocca | shardy> tripleo.sh --repo-setup is run on the virthost for tripleO CI which | 13:39 |
| hewbrocca | is why we don't see this I think | 13:39 |
| hewbrocca | < | 13:39 |
| hewbrocca | was what shardy said about it | 13:39 |
| trown|mtg | fwiw, there is a RHEL BZ for it and RHEL updated iPXE | 13:39 |
| *** akuznetsov has quit IRC | 13:39 | |
| slagle | trown|mtg: ok. that's the context i was lacking | 13:40 |
| trown|mtg | slagle: I will add the RHEL BZ to the commit message and comment in the patch | 13:40 |
| trown|mtg | iPXE in CentOS is from 2013 | 13:40 |
| slagle | thanks | 13:40 |
| openstackgerrit | Emilien Macchi proposed openstack/tripleo-heat-templates: Deploy Horizon with composable apache service https://review.openstack.org/345031 | 13:41 |
| derekh | trown|mtg: we test intrspection in the nonha job | 13:41 |
| *** milan has quit IRC | 13:41 | |
| EmilienM | ccamacho: I would need your eyes on https://review.openstack.org/#/c/345031/ please | 13:41 |
| derekh | trown|mtg: its run on all the puppet modules and and experimental job on the tripleo-ci repository | 13:41 |
| trown|mtg | derekh: where is this nonha job? | 13:41 |
| trown|mtg | right, not on instack-undercloud | 13:42 |
| ccamacho | EmilienM, oki let me check the submission | 13:42 |
| shardy | hewbrocca: yeah we run it here https://github.com/openstack-infra/tripleo-ci/blob/master/toci_instack.sh#L44 | 13:42 |
| *** milan has joined #tripleo | 13:43 | |
| shardy | then install some stuff, I assume the ipxe stuff is a dep of qemu-kvm or something | 13:43 |
| trown|mtg | shardy: is that even relevant for OVB now though? | 13:43 |
| trown|mtg | there is no virthost in OVB | 13:43 |
| shardy | https://github.com/openstack-infra/tripleo-ci/blob/master/toci_instack_ovb.sh#L58 | 13:44 |
| shardy | we still configure the repos, but yeah I don't know where the ipxe stuff comes from in the ovb environment | 13:44 |
| shardy | the compute nodes? | 13:44 |
| trown|mtg | probably from the undercloud | 13:44 |
| shardy | Not from the undercloud according to dtantsur | 13:44 |
| *** jubapa_ has quit IRC | 13:44 | |
| shardy | I updated my undercloud and I still got the old version from the host running the hypervisor | 13:45 |
| dtantsur | in virtual environment it usually comes from a virtual host | 13:45 |
| shardy | which was confusing because I thought everything was chainloaded from /httpboot on the undercloud | 13:45 |
| slagle | the vm's use the roms on whatever host they're running on | 13:46 |
| shardy | So in ovb we'll need to update all the compute nodes and point them at delorean repos :( | 13:46 |
| slagle | yes i think so | 13:46 |
| trown|mtg | it is a single package that needs to be there, dont need to put dlrn repos on them | 13:47 |
| jaosorior | EmilienM: Can you take a look at this https://review.openstack.org/#/c/344851/ ? | 13:47 |
| trown|mtg | it is not dlrn repos that contain the package either | 13:47 |
| EmilienM | bfournie: almost good, just one last small comment and I'll +2, thx for your patience | 13:47 |
| EmilienM | jaosorior: sure thing | 13:47 |
| trown|mtg | it is in the newton dependency repo | 13:47 |
| shardy | trown|mtg: sure, we can blacklist everything but that package | 13:47 |
| EmilienM | jaosorior: +A | 13:47 |
| shardy | but if we hard-code an install of a specific version, this will break again | 13:48 |
| shardy | I suppose that's going to happen anyway | 13:48 |
| trown|mtg | possibly | 13:48 |
| jaosorior | EmilienM: Thanks dude | 13:48 |
| slagle | EmilienM: i'll look after the dd | 13:49 |
| slagle | (re: multinode jobs) | 13:49 |
| bfournie | EmilienM: Thanks a lot! I appreciate you looking at this. | 13:50 |
| *** krotscheck_dcm is now known as krotscheck | 13:50 | |
| slagle | reminder, deep dive in 10 minutes | 13:51 |
| slagle | on TripleO Overcloud deployment debugging | 13:51 |
| tobias_fiberdata | gief link | 13:52 |
| slagle | by definition...nothing can go wrong | 13:52 |
| tobias_fiberdata | then i have a reason not to go down in the datacenter and erase disks | 13:52 |
| tobias_fiberdata | from data | 13:52 |
| *** tosky has quit IRC | 13:52 | |
| tobias_fiberdata | ;) | 13:52 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-tripleo: mysql: use parameter to lookup the step instead of hiera again https://review.openstack.org/345442 | 13:52 |
| openstackgerrit | Emilien Macchi proposed openstack/instack-undercloud: Deploy MySQL using composable role https://review.openstack.org/345002 | 13:52 |
| *** ebalduf has joined #tripleo | 13:52 | |
| saneax | slagle, acked | 13:52 |
| openstackgerrit | Chris Jones proposed openstack/tripleo-heat-templates: Next generation HA architecture work https://review.openstack.org/314208 | 13:55 |
| EmilienM | bfournie: you introduce whitespaces on https://review.openstack.org/#/c/342467/7/releasenotes/notes/newton_deprecate_enable_tunneling-d7f6e2de1091b458.yaml | 13:56 |
| EmilienM | bfournie: I'm pushing on it myself | 13:56 |
| shardy | EmilienM, trown|mtg, gfidente: FYI I deployed a nonha overcloud with https://review.openstack.org/#/c/341616/ applied | 13:56 |
| shardy | it worked OK AFAICS | 13:56 |
| trown|mtg | shardy: ya my issue was with ha | 13:56 |
| shardy | Ok, well that helps narrow the possible causes down at least :) | 13:57 |
| EmilienM | shardy: I checked HA job logs and all looks good | 13:57 |
| *** dtantsur is now known as dtantsur|pto | 13:57 | |
| gfidente | I was using HA as well | 13:57 |
| shardy | trown|mtg: got any output from the failure we can see? | 13:58 |
| shardy | openstack stack failures list overcloud | 13:58 |
| bfournie | EmilienM: Thank you for fixing that, and for the the +2. | 13:58 |
| trown|mtg | shardy: ok, I can get it later, but the stderr from puppet is posted to the patch | 13:58 |
| trown|mtg | shardy: and I verified that hiera is indeed missing on the controller | 13:59 |
| openstackgerrit | Merged openstack/puppet-tripleo: Add class to use certmonger's local CA https://review.openstack.org/344851 | 13:59 |
| openstackgerrit | mathieu bultel proposed openstack/tripleo-quickstart: WIP gate upgrade https://review.openstack.org/342161 | 13:59 |
| trown|mtg | quickstart is pretty easy to use I think, and shows the issue 100% of the time, while removing that patch the issue does not happen | 13:59 |
| openstackgerrit | Merged openstack/puppet-tripleo: Ensure the Heat Domain parameters before starting heat-engine pcmk service https://review.openstack.org/342032 | 14:01 |
| shardy | reminder, tripleo deep dive is starting now | 14:01 |
| shardy | https://etherpad.openstack.org/p/tripleo-deep-dive-topics | 14:01 |
| derekh | dprince: ping, do you still have the rh1 ssl certs? if so can you still them on the rh1 bastion somewhere please | 14:02 |
| openstackgerrit | John Trowbridge proposed openstack/tripleo-quickstart: Add workaround for out of date iPXE ROM in CentOS https://review.openstack.org/345410 | 14:02 |
| dprince | derekh: I will dig them up | 14:03 |
| derekh | dprince: thanks | 14:03 |
| *** cdearborn has joined #tripleo | 14:05 | |
| openstackgerrit | Merged openstack/tripleo-heat-templates: Remove ::nova::cron::archive_deleted_rows https://review.openstack.org/343661 | 14:05 |
| *** lazy_prince has quit IRC | 14:05 | |
| *** lazy_prince has joined #tripleo | 14:09 | |
| *** radek__ has quit IRC | 14:13 | |
| *** Ryjedo has joined #tripleo | 14:15 | |
| *** milan has quit IRC | 14:16 | |
| *** milan has joined #tripleo | 14:16 | |
| hewbrocca | I so love that we have to bring in the iPXE rom by name | 14:17 |
| hewbrocca | if only someone had a system for grouping RPMs together so that you could be sure they all work... | 14:18 |
| *** bootsha has quit IRC | 14:18 | |
| hewbrocca | trown|mtg: (not a criticism of your patch, just whingeing) | 14:18 |
| trown|mtg | :) | 14:19 |
| openstackgerrit | Flavio Percoco proposed openstack/tripleo-common: Extract `etc` files from openstack packages https://review.openstack.org/345051 | 14:20 |
| openstackgerrit | Flavio Percoco proposed openstack/tripleo-common: Create new docker command hook. https://review.openstack.org/312723 | 14:20 |
| openstackgerrit | Flavio Percoco proposed openstack/tripleo-common: Build heat-docker-agents for specific versions https://review.openstack.org/344805 | 14:20 |
| *** numans has joined #tripleo | 14:27 | |
| *** apetrich has quit IRC | 14:30 | |
| trozet | dsneddon: got a question for you | 14:32 |
| openstackgerrit | Flavio Percoco proposed openstack/tripleo-common: Extract `etc` files from openstack packages https://review.openstack.org/345051 | 14:37 |
| openstackgerrit | Flavio Percoco proposed openstack/tripleo-common: Create new docker command hook. https://review.openstack.org/312723 | 14:37 |
| openstackgerrit | Flavio Percoco proposed openstack/tripleo-common: Build heat-docker-agents for specific versions https://review.openstack.org/344805 | 14:37 |
| *** fzdarsky is now known as fzdarsky|afk | 14:38 | |
| *** bvandenh has joined #tripleo | 14:38 | |
| *** michchap has quit IRC | 14:40 | |
| *** fzdarsky|afk has quit IRC | 14:42 | |
| *** lazy_prince has quit IRC | 14:43 | |
| *** michchap has joined #tripleo | 14:45 | |
| openstackgerrit | Merged openstack/tripleo-heat-templates: Ensure the Heat Domain parameters before starting heat-engine pcmk service https://review.openstack.org/331652 | 14:46 |
| *** lazy_prince has joined #tripleo | 14:47 | |
| *** saneax is now known as saneax_AFK | 14:50 | |
| jaosorior | bnemec: Hey dude, do you have a machine where you get the ironic-inspecto TLS issue? it works in my machine, and we deployed in ccamacho's and it works there too. | 14:52 |
| jaosorior | And it does seem to be more of a mistral issue, more than TLS related | 14:52 |
| ansiwen | EmilienM: ping re memcached config | 14:52 |
| EmilienM | ansiwen: what's up? | 14:53 |
| ansiwen | EmilienM: https://review.openstack.org/#/c/342829/ | 14:53 |
| ansiwen | EmilienM: you said I have to patch the profile, do you mean manifests/profile/base/nova.pp ? | 14:54 |
| EmilienM | ansiwen: which nova requires cache parameters? | 14:54 |
| openstackgerrit | Paul Belanger proposed openstack-infra/tripleo-ci: Use growroot for undercloud image https://review.openstack.org/345042 | 14:56 |
| ansiwen | EmilienM: all that set the memcached_servers now. the config names changed. | 14:56 |
| EmilienM | ansiwen: again, which nova services require cache parameters? | 14:56 |
| flaper87 | dprince: jpeeler: when you've 5 mins, could you take a look here? https://review.openstack.org/#/c/345051/ | 14:56 |
| *** ebalduf has quit IRC | 14:57 | |
| *** bswartz is now known as bswartz_ | 14:57 | |
| ansiwen | EmilienM: before it was memcached_servers in [DEFAULT], now you have to set it in the [cache] section... | 14:57 |
| ansiwen | EmilienM: let me think | 14:57 |
| EmilienM | ok let me repeat my question | 14:57 |
| EmilienM | which service in Nova requires memcache | 14:57 |
| EmilienM | nova api? nova conductor? all? | 14:57 |
| *** bvandenh has quit IRC | 14:57 | |
| hewbrocca | EmilienM: novnc, for sure | 14:59 |
| hewbrocca | but I don't know how that relates | 14:59 |
| EmilienM | we need to know where to include ::nova::cache | 15:00 |
| EmilienM | hewbrocca: novnc? I'm not sure. | 15:00 |
| shardy | gfidente: Hey, are there any examples of using lookup_hiera_hash ? | 15:00 |
| ansiwen | EmilienM: at least the controller nodes need that cache config | 15:01 |
| openstackgerrit | Honza Pokorny proposed openstack/tripleo-ui: Use only the prebuilt version of phantomjs https://review.openstack.org/345500 | 15:01 |
| EmilienM | hewbrocca: novnc doesn't require memcache | 15:01 |
| ansiwen | EmilienM: it does | 15:01 |
| shardy | I want to use it for per-service bind_ip's by looking in hiera net_ip_map | 15:01 |
| EmilienM | nova-consoleauth does require but not novnc I think | 15:01 |
| gfidente | shardy, not in tripleo code that I know | 15:01 |
| ansiwen | EmilienM: novnc needs memcache, because it shares auth tokens over memcached | 15:01 |
| hewbrocca | EmilienM: yeah, that's probably it. It just shows up as a problem with novnc | 15:01 |
| gfidente | shardy, when we tried the issue was that service_configs.yaml is merging keys before hiera so we couldn't use it | 15:02 |
| *** d0ugal has quit IRC | 15:02 | |
| EmilienM | also all services using servicegroup | 15:02 |
| EmilienM | so you need to include it in base/nova.pp | 15:02 |
| EmilienM | and configure hiera parameters in nova-base.yaml in THT | 15:02 |
| *** pcaruana has quit IRC | 15:02 | |
| shardy | gfidente: Ok, I think I've got a solution to that, I just need to figure out the hiera lookup | 15:02 |
| shardy | "%{hiera('net_ip_map')}" | 15:03 |
| shardy | how would I wire in lookup of e.g "internal_api" there? | 15:03 |
| hewbrocca | heh how many escapes can we add? | 15:03 |
| *** akuznetsov has joined #tripleo | 15:03 | |
| jaosorior | jmiu: Are you still having that problem? If you have your node available in beaker or something like that I could take a look | 15:03 |
| ansiwen | EmilienM: https://bugzilla.redhat.com/show_bug.cgi?id=1238336 | 15:03 |
| openstack | bugzilla.redhat.com bug 1238336 in openstack-tripleo-heat-templates "2/3 requests for Nova instance VNC console fail in HA overcloud" [High,Closed: errata] - Assigned to gfidente | 15:03 |
| EmilienM | ansiwen, hewbrocca: https://github.com/openstack/nova/search?utf8=%E2%9C%93&q=cache_utils | 15:03 |
| EmilienM | ansiwen, hewbrocca: I see: consoleauth, and all services that use servicegroup (afik, nova-api, nova compute, etc) but not novnc | 15:04 |
| gfidente | ansiwen, yeah i think we added memcached a while ago but the LP wasn't closed | 15:04 |
| ansiwen | EmilienM: https://bugzilla.redhat.com/show_bug.cgi?id=1340475 | 15:04 |
| openstack | bugzilla.redhat.com bug 1340475 in rhel-osp-director "Nova novnc console doesn't load 2/3 times: Failed to connect to server (code: 1006)" [High,On_dev] - Assigned to svanders | 15:04 |
| EmilienM | gfidente: we did | 15:05 |
| gfidente | shardy, net_ip_map is a dict where keys are the network name and values the local ip addresses | 15:05 |
| EmilienM | gfidente: ansiwen wants to move it to composable services I guess | 15:05 |
| ansiwen | gfidente: well, now the controllers can't share the tokens anymore, because memcached is not used anymore, because the way of the configuration changed | 15:05 |
| EmilienM | we need to include ::nova::cache in base profile | 15:05 |
| EmilienM | but you need IPv6 params | 15:05 |
| *** d0ugal has joined #tripleo | 15:05 | |
| ansiwen | EmilienM: no, I just wanna fix this bug: https://bugzilla.redhat.com/show_bug.cgi?id=1340475 | 15:05 |
| openstack | bugzilla.redhat.com bug 1340475 in rhel-osp-director "Nova novnc console doesn't load 2/3 times: Failed to connect to server (code: 1006)" [High,On_dev] - Assigned to svanders | 15:05 |
| EmilienM | ansiwen: https://github.com/openstack/puppet-tripleo/blob/master/manifests/profile/base/nova.pp#L45 | 15:06 |
| *** ebarrera has quit IRC | 15:06 | |
| EmilienM | ansiwen: I know and your fix in THT is correct to use ::nova::cache, it's just the wrong place | 15:07 |
| shardy | gfidente: Yeah, I know that, I'm just not sure of the hiera syntax to lookup a key | 15:07 |
| EmilienM | you need to 1) move it to puppet-tripleo 2) make it work for IPv6 | 15:07 |
| gfidente | shardy, ah the actual function | 15:07 |
| gfidente | shardy, lookup_hiera_hash('hash_name', 'key_name') | 15:08 |
| shardy | gfidente: Great, thanks | 15:08 |
| ansiwen | EmilienM: ok, so what I was doing now was basically the non-composable way? | 15:08 |
| gfidente | shardy, the problem is we can call it from .pp but not from hieradata .yaml | 15:09 |
| shardy | https://github.com/openstack/tripleo-heat-templates/blob/master/puppet/services/keystone.yaml#L133 | 15:09 |
| gfidente | shardy, from hieradata .yaml only hiera() and a couple more functions are allowed to be called | 15:09 |
| shardy | gfidente: can we call it e.g like we define those hiera lookups in tht? | 15:09 |
| * shardy tries | 15:09 | |
| *** ebalduf has joined #tripleo | 15:09 | |
| shardy | ah | 15:09 |
| gfidente | shardy, yeah we can't, that was my initial goal | 15:09 |
| shardy | :( | 15:09 |
| shardy | Ok | 15:10 |
| shardy | no worries, I have a heat based approach with map_replace I can use instead | 15:10 |
| gfidente | shardy, it restricts calls to scope() and hiera() | 15:10 |
| gfidente | https://docs.puppet.com/hiera/1/variables.html#using-lookup-functions | 15:10 |
| shardy | gfidente: Ok, thanks, this is what I was unsure on | 15:10 |
| EmilienM | ansiwen: right | 15:11 |
| shardy | we could have a bind_host_key in hiera I guess, then do the lookup in the service profiles | 15:11 |
| shardy | tripleo::profile::base::$service::bind_host_network or something | 15:12 |
| gfidente | shardy, well at this time the ServiceNetMap is not known to the service | 15:12 |
| gfidente | so I was thinking to pass it to services | 15:13 |
| shardy | gfidente: we can pass the ServiceNetMap into the puppet/services/* templates | 15:13 |
| gfidente | and the dump net_ip_map_internal_api | 15:13 |
| shardy | yeah | 15:13 |
| gfidente | net_ip_map_external | 15:13 |
| gfidente | and so on | 15:13 |
| gfidente | shardy, I was doing some of this work in https://review.openstack.org/#/c/318344/ | 15:14 |
| gfidente | sounds like time to reprise it? | 15:15 |
| flaper87 | shardy: mind restoring your +2 here ? https://review.openstack.org/#/c/344805/ fixed some nits, no major changes to the patch | 15:15 |
| *** trown|mtg is now known as trown | 15:15 | |
| *** panda|zZ is now known as panda | 15:15 | |
| panda | keep calm and deep dive | 15:16 |
| * flaper87 deep dives | 15:16 | |
| flaper87 | panda: what are we deep diving into? | 15:16 |
| * flaper87 likes scuba | 15:16 | |
| shardy | gfidente: Ah, yeah, that's pretty similar to what I was thinking, thanks | 15:16 |
| * tobias_fiberdata deep dives | 15:16 | |
| jmiu | jaosorior: i don't quite know what beaker is | 15:16 |
| shardy | gfidente: one thing I did wonder is should we just pass the ServiceNetMap via the EndpointMap mapping | 15:17 |
| gfidente | shardy, unless we come up with a sort of a mixin for the service templates which allow us to pass an instance of NetIpMap coming from the base host | 15:17 |
| shardy | that might make the sedding of the services templates easier | 15:17 |
| *** apetrich has joined #tripleo | 15:17 | |
| panda | flaper87: tripleo of course | 15:17 |
| gfidente | and then we can do it entirely in heat | 15:17 |
| shardy | gfidente: well the NetIpMap is already in hiera on each host | 15:17 |
| jmiu | jaosorior: and i know it's kind of avoiding the problem, but i've just been iterating on that freeipa template. | 15:17 |
| shardy | so we just need to pass the ServiceNetMap value I think | 15:17 |
| shardy | then use your function to do the lookup inside the profiles? | 15:18 |
| gfidente | shardy, yeah I meant to do it in heat instead of hiera interpolation | 15:18 |
| EmilienM | trown: oooq on my laptop takes 100%CPU to run on remote node | 15:18 |
| shardy | gfidente: Ok, cool - are you planning to rebase that patch? | 15:18 |
| gfidente | shardy, yes | 15:18 |
| gfidente | looks like it's time to | 15:18 |
| shardy | we need this for custom roles so I'm keen to help get it done :) | 15:18 |
| gfidente | let's agree on the approach though | 15:18 |
| jaosorior | jmiu: I see | 15:19 |
| gfidente | pass it to all services or via endpoint map? | 15:19 |
| jaosorior | jmiu: Is the node you're working on a machine that you have at home, or is it in the red hat network? | 15:19 |
| gfidente | and also, looks like dumping net_ip_map_$network in flat keys is the only option, agreed? | 15:19 |
| jmiu | jaosorior: on the red hat network. it's on warp | 15:19 |
| shardy | dprince: what are your thoughts re https://review.openstack.org/#/c/318344/ ? | 15:19 |
| jaosorior | jmiu: warp is in beaker ;) | 15:19 |
| jaosorior | anyway | 15:19 |
| jmiu | i don't have enough ram to handle all of that...goodness | 15:19 |
| jaosorior | jmiu: if you give me access I can take a look | 15:19 |
| tobias_fiberdata | slagle, i think you will have to prepare templates for the 4th session ;) | 15:20 |
| jmiu | jaosorior: what is beaker? and how can i do that | 15:20 |
| jmiu | that was a great session ;_; | 15:20 |
| EmilienM | matbu: hey, did you already have high cpu usage when running oooq on your laptop to deploy a remote host? | 15:20 |
| tobias_fiberdata | jmiu, i agree, alot of good tips to debug problems | 15:20 |
| *** skramaja has quit IRC | 15:20 | |
| shardy | gfidente: can't we just leave the maps in hieradata/networks.yaml and lookup e.g "internal_api" etc? | 15:20 |
| shardy | gfidente: the other option I considered is we can do a value replacement on config_settings | 15:21 |
| shardy | so e.g "internal_api" gets replaced with the value from NetIpMap | 15:21 |
| gfidente | shardy, we can lookup hash keys from within the .pp files | 15:22 |
| gfidente | so we'd have to pass the binding ips to the classes from puppet-tripleo | 15:22 |
| gfidente | and pass the network to lookup to puppet-tripleo | 15:22 |
| jaosorior | EmilienM: well, the hierdata patch has been quite hard to merge, I think I'll rebase my patches to not depend on that one | 15:23 |
| jaosorior | so in the meantime I'll also fix the merge conflict in the patch | 15:23 |
| jaosorior | makes sense? | 15:23 |
| EmilienM | jaosorior: well, the patch is passing CI nonha and ha, I see no reason to not merge it. | 15:24 |
| openstackgerrit | Steven Hardy proposed openstack/tripleo-heat-templates: WIP show map_replace syntax https://review.openstack.org/345510 | 15:24 |
| shardy | gfidente: ^^ that is an alternative | 15:24 |
| jaosorior | EmilienM: well, right now cause it's in merge conflict | 15:24 |
| EmilienM | we were waiting to shardy to test it, and AFIK he tested it and it worked on non-ha | 15:24 |
| shardy | we could use the new map_replace function I wrote for heat | 15:24 |
| EmilienM | and gfidente tested it on ha | 15:24 |
| jaosorior | EmilienM: didn't trown say it would break RDO? | 15:24 |
| shardy | to replace e.g internal_api with the IP from NetIpMap | 15:24 |
| shardy | gfidente: we'd still need to pass ServiceNetMap to all the services tho | 15:25 |
| shardy | so we could e.g do foo::bind_ip: internal_api by a lookup from ServiceNetMap | 15:25 |
| trown | jaosorior: ya but since when does tripleo care | 15:25 |
| EmilienM | jaosorior: I'm still curious to know why and how | 15:25 |
| EmilienM | trown: ... we do care | 15:25 |
| shardy | trown: we do care, but nobody can reproduce your issue yet | 15:26 |
| *** liverpooler has joined #tripleo | 15:26 | |
| EmilienM | trown: do you have an example where we don't care? | 15:26 |
| dprince | shardy: nice one on the map_replace +2 | 15:26 |
| EmilienM | if we wouldn't care, my patch would be merged already | 15:26 |
| trown | shardy: has someone tried to reproduce it with quickstart other than me? | 15:27 |
| openstackgerrit | Merged openstack/tripleo-heat-templates: Create openstack-core resource regardless of enable_load_balancer https://review.openstack.org/339100 | 15:27 |
| openstackgerrit | Emilien Macchi proposed openstack/tripleo-heat-templates: Migrate Puppet Hieradata to composable services https://review.openstack.org/341616 | 15:27 |
| shardy | dprince: Cool, if gfidente likes it we can go with that, what are your thoughts to passing the ServiceNetMap into each service e.g https://review.openstack.org/#/c/318344/ | 15:27 |
| EmilienM | jaosorior: rebased ^ | 15:28 |
| matbu | EmilienM: no i don't | 15:28 |
| jaosorior | EmilienM: I was doing that, lol. too slow | 15:28 |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack/tripleo-heat-templates: Enable keystone to use the SSL middleware https://review.openstack.org/341354 | 15:28 |
| matbu | EmilienM: (sorry missed your ping) | 15:28 |
| openstackgerrit | Juan Antonio Osorio Robles proposed openstack/tripleo-heat-templates: Enable glance to use the SSL middleware https://review.openstack.org/341367 | 15:28 |
| matbu | EmilienM: you need to change your dell xps ;) | 15:28 |
| EmilienM | matbu: how do you know I have XPS :P | 15:29 |
| matbu | EmilienM: i'm watching you :D | 15:29 |
| EmilienM | matbu: :-O | 15:30 |
| gfidente | shardy, cool | 15:30 |
| *** bootsha has joined #tripleo | 15:30 | |
| *** akuznetsov has quit IRC | 15:30 | |
| matbu | EmilienM: i got the same in the past, that's why i remember it :) | 15:30 |
| openstackgerrit | Steven Hardy proposed openstack/tripleo-heat-templates: WIP show map_replace syntax https://review.openstack.org/345510 | 15:31 |
| gfidente | shardy, do I get it right we need in the array an empty dict to *not* replace foo::bind_ip ? | 15:31 |
| tobias_fiberdata | i need that video uploaded asap ;D | 15:32 |
| shardy | gfidente: Sorry, the syntax was wrong, we pass a "values" map, which will lookup e.g internal_api, and replace it with the value | 15:32 |
| gfidente | shardy, and values needs to be the key for the net_ip_map dict | 15:32 |
| jmiu | i took NOTES :p | 15:32 |
| shardy | e.g exactly what net_ip_map gives us | 15:32 |
| gfidente | shardy, right ok so that should be something like | 15:32 |
| shardy | http://docs.openstack.org/developer/heat/template_guide/hot_spec.html#map-replace | 15:33 |
| gfidente | yeah fine | 15:33 |
| hewbrocca | So, like trown says, could somebody over here try a oooq deployment with the patch in question and see if it fails for you? | 15:34 |
| openstackgerrit | Chris Jones proposed openstack/puppet-tripleo: Next generation HA architecture work https://review.openstack.org/342650 | 15:34 |
| gfidente | shardy, but we can only override it with a single map | 15:35 |
| *** bootsha has quit IRC | 15:35 | |
| gfidente | shardy, for example we can't lookup network subnets with this mechanism | 15:35 |
| gfidente | because the outputs in net_ip_subnet_map have same name | 15:35 |
| tobias_fiberdata | a quick question, if i deploy only three controllers to start with is it possible to add computenodes afterwards or do i have to deploy everything at the same time? I mean initially? like 3controller and 1 compute? | 15:36 |
| openstackgerrit | Chris Jones proposed openstack/tripleo-heat-templates: Next generation HA architecture work https://review.openstack.org/314208 | 15:36 |
| shardy | gfidente: Yeah, if we need to do that, we'll have to flatten the map output by net_ip_map | 15:36 |
| shardy | sorry NetIpMap | 15:37 |
| shardy | into one map | 15:37 |
| openstackgerrit | Merged openstack/tripleo-quickstart: Add workaround for out of date iPXE ROM in CentOS https://review.openstack.org/345410 | 15:37 |
| shardy | internal_api_subnet | 15:37 |
| shardy | etc | 15:37 |
| ccamacho | EmilienM Im finishing some local tests, the next submission will be yours | 15:37 |
| gfidente | and net_ip_uri_map I suppose is the same map with addresses in brackets for ipv6 | 15:37 |
| shardy | Yeah, so that becomes internal_api_uri | 15:38 |
| EmilienM | ccamacho: cool | 15:38 |
| openstackgerrit | Jiri Stransky proposed openstack/tripleo-heat-templates: Remove two races during the L->M Keystone under httpd migration https://review.openstack.org/338879 | 15:40 |
| openstackgerrit | Jiri Stransky proposed openstack/tripleo-heat-templates: openstack-core resource does not have interleave=true https://review.openstack.org/342725 | 15:40 |
| *** leanderthal is now known as leanderthal|afk | 15:40 | |
| ansiwen | EmilienM: so, do I have to write "include ::nova::cache", or do I instantiate the class with "class { '::nova::cache' :..." , in nova.pp? | 15:40 |
| ansiwen | EmilienM: I still don't understand a basic concept here, I guess | 15:41 |
| *** sshnaidm has quit IRC | 15:42 | |
| EmilienM | ansiwen: in a call, brb | 15:43 |
| gfidente | shardy, so I like the map_replace thing | 15:43 |
| gfidente | but we need to restructure the maps a bit | 15:43 |
| gfidente | cause I have seen a few places where we need _subnet and _uri as well, so merging those looks like a requirement | 15:43 |
| gfidente | shardy, would you create a resource which outputs them all or merge them into a single resource? | 15:44 |
| *** adarazs has quit IRC | 15:45 | |
| *** cschwede has quit IRC | 15:45 | |
| *** slagle has quit IRC | 15:45 | |
| *** marios has quit IRC | 15:45 | |
| openstackgerrit | Steven Hardy proposed openstack/tripleo-heat-templates: Allow map_replace substitution of network names https://review.openstack.org/345510 | 15:45 |
| shardy | gfidente: we could just do that ^^ | 15:46 |
| *** zoli|wfh has quit IRC | 15:46 | |
| shardy | add them all into net_ip_map, and eventually remove the other two outputs? | 15:46 |
| *** jschlueter has quit IRC | 15:46 | |
| tobias_fiberdata | interesting | 15:46 |
| tobias_fiberdata | i ran a deploy without any templates | 15:46 |
| tobias_fiberdata | and it's stuck on networkdeployment | 15:46 |
| *** psanchez has quit IRC | 15:46 | |
| *** psanchez has joined #tripleo | 15:47 | |
| tobias_fiberdata | how do i deep dive even further? | 15:47 |
| *** slagle has joined #tripleo | 15:47 | |
| tobias_fiberdata | i know what resource type it is | 15:47 |
| *** jschlueter has joined #tripleo | 15:47 | |
| *** adarazs has joined #tripleo | 15:48 | |
| gfidente | shardy, ya :) | 15:48 |
| *** marios has joined #tripleo | 15:48 | |
| tobias_fiberdata | i check in the heat temps? | 15:48 |
| openstackgerrit | Steven Hardy proposed openstack/tripleo-heat-templates: Allow map_replace substitution of network names https://review.openstack.org/345510 | 15:48 |
| bandini | marios: can we +A https://review.openstack.org/343253 now? | 15:48 |
| *** zoliXXL has joined #tripleo | 15:48 | |
| *** cschwede has joined #tripleo | 15:49 | |
| bandini | and then remove the -1 on https://review.openstack.org/#/c/337304/ after | 15:49 |
| shardy | tobias_fiberdata: log onto one of the nodes stuck on NetworkDeployment, and check the /etc/os-net-config/config.json | 15:49 |
| tobias_fiberdata | thanks steven | 15:49 |
| *** serverascode has joined #tripleo | 15:49 | |
| shardy | tobias_fiberdata: you can debug the config locally by running os-net-config --debug -c /etc/os-net-config/config.json | 15:49 |
| marios | bandini: in call gimme few | 15:49 |
| tobias_fiberdata | awesum! | 15:49 |
| bandini | marios: np. thanks ;) | 15:49 |
| shardy | that step unfortunately doesn't signal when it fails, so the deployment gets stuck (I have a patch up which aims to fix that) | 15:49 |
| *** penick has joined #tripleo | 15:50 | |
| tobias_fiberdata | ah yea, you mentioned that in your deep dive? | 15:50 |
| shardy | Yeah I think I did | 15:50 |
| tobias_fiberdata | hehe, something actually got stuck after checking it out :P | 15:50 |
| openstackgerrit | Jiri Tomasek proposed openstack/tripleo-ui: Use Webpack for application processing https://review.openstack.org/343592 | 15:51 |
| shardy | tobias_fiberdata: probably it's eiter run /usr/libexec/os-refresh-config/configure.d/20-os-net-config and failed, or it's configured the network in a way that means it can't signal success | 15:51 |
| tobias_fiberdata | ah could be that ye | 15:51 |
| tobias_fiberdata | the latest part | 15:51 |
| tobias_fiberdata | believe i have to configure some network isolation :P | 15:51 |
| *** jubapa has quit IRC | 15:52 | |
| tobias_fiberdata | interesting, there's no output at all in the config.json file you are refering to | 15:53 |
| tobias_fiberdata | atleast not on the controller | 15:53 |
| gfidente | anyone willing to merge this https://review.openstack.org/#/c/339444/ ? it's a backport that landed in master already | 15:53 |
| *** yamahata has joined #tripleo | 15:53 | |
| *** numans has quit IRC | 15:54 | |
| tobias_fiberdata | actually shardy, it's also stuck on the UpdateDeployment aswell | 15:55 |
| tobias_fiberdata | so i'll be looking into that tomorrow | 15:55 |
| tobias_fiberdata | will go home from work now :P | 15:55 |
| shardy | gfidente: done | 15:55 |
| tobias_fiberdata | tomorrow there's going to be hardware to setup in the datacenter for the productionsetup (i'm thrilled!) | 15:56 |
| shardy | tobias_fiberdata: it's probably related, if the network's down yum can't run | 15:56 |
| tobias_fiberdata | aye | 15:56 |
| EmilienM | ansiwen: yes you'll need to declare the parameters for ipv6, with the same logic I showed you | 15:56 |
| gfidente | shardy, :) | 15:56 |
| EmilienM | ansiwen: https://github.com/openstack/puppet-tripleo/blob/master/manifests/profile/base/nova.pp#L45 | 15:56 |
| ansiwen | EmilienM: yes, but in which case i just include and in which case I instantiate the class? | 15:57 |
| tobias_fiberdata | shardy, it does have internet though. so it's a bit weird, that's why i asked earlier if i had to erase the data on the disks if i deployed the machines to many times. I will do that otmorrow though | 15:57 |
| tobias_fiberdata | now i have 140km to go home by car, so see ya ;) | 15:57 |
| shardy | o/ | 15:58 |
| *** tobias_fiberdata is now known as tobias_fb_away | 15:58 | |
| openstackgerrit | Merged openstack/tripleo-heat-templates: Repurpose KeystoneWorkers add keystone::wsgi::apache::threads https://review.openstack.org/336520 | 15:59 |
| bandini | bnemec: I have no idea how to fix https://review.openstack.org/#/c/304858/ I toyed a bit this morning with it, but no luck. Any hints appreciated ;) | 15:59 |
| ansiwen | EmilienM: normally I would think I first have to include ::nova::cache, and after that I can instantiate the class ::nova::cache | 15:59 |
| *** aufi_ has quit IRC | 16:02 | |
| openstackgerrit | Merged openstack/tripleo-heat-templates: Merge pacemaker_master/sync_db conditionals https://review.openstack.org/339444 | 16:02 |
| *** d0ugal has quit IRC | 16:04 | |
| EmilienM | ansiwen: no, you need to declare class { '::nova::cache': ... } | 16:04 |
| jaosorior | bnemec: Around? | 16:05 |
| marios | bandini: done | 16:08 |
| *** fzdarsky|afk has joined #tripleo | 16:09 | |
| EmilienM | gfidente: can you look https://review.openstack.org/#/c/330785/ please? | 16:10 |
| EmilienM | shardy: ^ | 16:10 |
| shardy | gfidente, dprince: radical idea, what if we didn't pass EndpointMap into service templates, and instead did that via a value substitution too? | 16:10 |
| bandini | marios: thanks a bunch. I rechecked the liberty one now | 16:10 |
| shardy | e.g swift::proxy::authtoken::auth_uri: EndpointKeystoneInternalUri | 16:10 |
| marios | bandini: np man just clicked 2 buttons | 16:10 |
| bandini | marios: it takes skill for that too ;) | 16:11 |
| shardy | that would remove the chicken/egg problem and enable us to get rid of endpoint_data | 16:11 |
| shardy | instead putting the per-service endpoint stuff into the service templates | 16:11 |
| shardy | ServiceNetMap would then become an empty override for network mappings that are defaulted in the service templates | 16:12 |
| serverascode | hi, is it possible to disable openstack services from being deployed by tripleo? I have been looking for docs but can't find anything obvious -- for example what if I don't want to deploy swift at all? | 16:14 |
| shardy | serverascode: on recent trunk builds, yes, you can override the ControllerServices list to remove the services you don't want | 16:15 |
| *** tesseract- has quit IRC | 16:15 | |
| shardy | https://github.com/openstack/tripleo-heat-templates/blob/master/overcloud.yaml#L367 | 16:15 |
| serverascode | ok thanks shardy | 16:16 |
| shardy | serverascode: https://github.com/openstack/tripleo-heat-templates/blob/stable/mitaka/overcloud.yaml#L430 | 16:16 |
| shardy | if you're on stable branches, the interface is different, you'd pass ControllerEnableSwiftStorage: false instead | 16:16 |
| *** akuznetsov has joined #tripleo | 16:17 | |
| openstackgerrit | Merged openstack/tripleo-heat-templates: Minor updates might fail with missing MysqlClustercheckPassword property https://review.openstack.org/343253 | 16:17 |
| *** rcernin has quit IRC | 16:19 | |
| *** rajinir has joined #tripleo | 16:22 | |
| *** ayoung has joined #tripleo | 16:24 | |
| *** ifarkas is now known as ifarkas_afk | 16:24 | |
| *** ayoung has quit IRC | 16:25 | |
| *** ayoung has joined #tripleo | 16:25 | |
| *** akuznetsov has quit IRC | 16:26 | |
| *** derekh has quit IRC | 16:27 | |
| *** julim has quit IRC | 16:28 | |
| *** derekjhyang has joined #tripleo | 16:29 | |
| *** jpich has quit IRC | 16:32 | |
| *** tobias_fb_away has quit IRC | 16:33 | |
| *** dsariel has quit IRC | 16:34 | |
| bandini | has anyone deployed newton recently with ipv6 and was successful? | 16:34 |
| *** lucasagomes is now known as lucas-afk | 16:35 | |
| openstackgerrit | Ben Nemec proposed openstack/diskimage-builder: Nothing to see here https://review.openstack.org/345539 | 16:36 |
| *** mcornea has quit IRC | 16:36 | |
| gfidente | bandini, I think there are a couple of services incorrectly setting :host to MysqlInternal instead of MysqlNoBracketsInternal for that to work | 16:40 |
| jaosorior | hey guys | 16:40 |
| jaosorior | if somehow we lost contact with the nodes... is it possible to do again the openstack baremetal import instackenv.json | 16:40 |
| trozet | bandini: there is a bug at least on stable/mitaka where rabbit doesn't work with ipv6 if you are using ipv6 api endpoints | 16:40 |
| jaosorior | and such? | 16:40 |
| trozet | bandini: https://bugzilla.redhat.com/show_bug.cgi?id=1347802#c14 | 16:41 |
| openstack | bugzilla.redhat.com bug 1347802 in rabbitmq-server "RabbitMQ resources fail to start in HA IPv6 deployment" [Urgent,On_qa] - Assigned to jeckersb | 16:41 |
| openstackgerrit | Merged openstack/python-tripleoclient: Generate CephX keys and FSID regardless of CephStorageCount https://review.openstack.org/343794 | 16:43 |
| *** d0ugal has joined #tripleo | 16:45 | |
| *** mcornea has joined #tripleo | 16:45 | |
| openstackgerrit | Giulio Fidente proposed openstack/tripleo-heat-templates: Move mysq::host param from MysqlInternal to MysqlNoBracketsInternal https://review.openstack.org/345542 | 16:45 |
| gfidente | bandini, ^^ | 16:45 |
| openstackgerrit | Giulio Fidente proposed openstack/tripleo-heat-templates: Move mysql::host param from MysqlInternal to MysqlNoBracketsInternal https://review.openstack.org/345542 | 16:45 |
| openstackgerrit | Merged openstack/tripleo-ui: Use only the prebuilt version of phantomjs https://review.openstack.org/345500 | 16:47 |
| *** dsariel has joined #tripleo | 16:48 | |
| openstackgerrit | Merged openstack/tripleo-common: Build heat-docker-agents for specific versions https://review.openstack.org/344805 | 16:49 |
| *** trown is now known as trown|lunch | 16:49 | |
| openstackgerrit | Merged openstack/python-tripleoclient: Don't put --builder in image_group https://review.openstack.org/333013 | 16:51 |
| bandini | trozet: yeah I am working on the THT part of that BZ, but I don't even get to the rabbit step ;) it fails much earlier on | 16:52 |
| bandini | gfidente: HA! thank you sir! | 16:52 |
| *** paramite has joined #tripleo | 16:53 | |
| openstackgerrit | Florian Fuchs proposed openstack/tripleo-ui: Move plan upload errors from notifications to formerrors https://review.openstack.org/332181 | 16:54 |
| EmilienM | jaosorior: can you review https://review.openstack.org/#/c/330785/ please? | 16:55 |
| openstackgerrit | Honza Pokorny proposed openstack/tripleo-ui: Add packaging spec and config files https://review.openstack.org/344932 | 16:55 |
| jaosorior | EmilienM: Did you test it? | 16:56 |
| EmilienM | jaosorior: yep | 16:57 |
| openstackgerrit | Emilien Macchi proposed openstack/instack-undercloud: Deploy MySQL using composable role https://review.openstack.org/345002 | 16:57 |
| *** yamahata has quit IRC | 16:58 | |
| *** yamahata has joined #tripleo | 16:58 | |
| *** jaosorior has quit IRC | 16:59 | |
| *** zoliXXL is now known as zoli|gone | 16:59 | |
| *** jaosorior has joined #tripleo | 16:59 | |
| *** zoli|gone is now known as zoli_gone-proxy | 17:00 | |
| *** dsariel has quit IRC | 17:01 | |
| *** florianf has quit IRC | 17:01 | |
| *** pkovar has quit IRC | 17:02 | |
| openstackgerrit | Merged openstack/puppet-tripleo: deploy composable firewall rules for HAproxy https://review.openstack.org/330785 | 17:06 |
| openstackgerrit | wes hayutin proposed openstack/tripleo-quickstart: add debug if introspection fails https://review.openstack.org/345559 | 17:10 |
| openstackgerrit | Swapnil Kulkarni (coolsvap) proposed openstack/python-tripleoclient: Remove discover from test-requirements https://review.openstack.org/345562 | 17:12 |
| *** alhernan has joined #tripleo | 17:16 | |
| *** athomas has quit IRC | 17:19 | |
| *** numans has joined #tripleo | 17:25 | |
| *** trown|lunch is now known as trown | 17:27 | |
| *** athomas has joined #tripleo | 17:30 | |
| openstackgerrit | Ben Nemec proposed openstack-infra/tripleo-ci: Skip pingtest cleanup https://review.openstack.org/345570 | 17:31 |
| openstackgerrit | Emilien Macchi proposed openstack/tripleo-heat-templates: Add 'service_name' to composable services https://review.openstack.org/344440 | 17:33 |
| *** julim has joined #tripleo | 17:34 | |
| *** oshvartz has quit IRC | 17:34 | |
| openstackgerrit | Emilien Macchi proposed openstack/tripleo-heat-templates: Composable firewall rules https://review.openstack.org/344985 | 17:34 |
| EmilienM | dprince: just rebased ^ | 17:34 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-tripleo: Add firewall::service_rules define https://review.openstack.org/344860 | 17:35 |
| *** electrofelix has quit IRC | 17:35 | |
| openstackgerrit | Emilien Macchi proposed openstack/tripleo-heat-templates: Composable firewall rules https://review.openstack.org/344985 | 17:36 |
| *** jcoufal has quit IRC | 17:36 | |
| dprince | EmilienM: thanks :) | 17:36 |
| EmilienM | my pleasure | 17:37 |
| EmilienM | I want this feature :) | 17:37 |
| dprince | EmilienM: I know, I want you to have it :) | 17:37 |
| bnemec | jaosorior: Sorry, was heads down on some stuff. I pushed a patch to test the SSL inspector in CI: https://review.openstack.org/#/c/345539/ | 17:37 |
| dprince | EmilienM: I'm working on cleaning out compute.yaml now, and then controller.yaml if need be too | 17:37 |
| bnemec | I'm just concerned about merging that when I can't get it to work. | 17:38 |
| jaosorior | bnemec: I did the same | 17:38 |
| jaosorior | and I see your issue there | 17:38 |
| jaosorior | but I have no idea how to reproduce it | 17:38 |
| bnemec | Heh, great minds. :-) | 17:38 |
| jaosorior | and we tried with several machines now | 17:38 |
| bnemec | jaosorior: I install https://repos.fedorapeople.org/repos/openstack-m/jenkins/rpm-build/332641/7/instack-undercloud-999.9.9-99999.noarch.rpm before doing the undercloud install. | 17:38 |
| bnemec | It has reproduced consistently for me. | 17:38 |
| jaosorior | bnemec: https://review.openstack.org/#/c/342669/ | 17:38 |
| bnemec | Maybe it has to do with the fact that I'm in OVB? | 17:38 |
| bnemec | That shouldn't matter, but who knows. | 17:39 |
| jaosorior | it shouldn't | 17:40 |
| jaosorior | bnemec: and I still can't explain why mistral fails | 17:40 |
| jaosorior | that seems extremely unrelated | 17:40 |
| EmilienM | dprince: ack | 17:40 |
| bnemec | rbrady: dprince: If you have any suggestions on why we're seeing http://logs.openstack.org/69/342669/3/experimental-tripleo/gate-tripleo-ci-centos-7-ovb-nonha/76b3781/console.html#_2016-07-21_07_44_39_176647 help would be appreciated. | 17:41 |
| bnemec | This is when testing https://review.openstack.org/#/c/332641/ | 17:41 |
| *** Goneri has quit IRC | 17:41 | |
| bnemec | bandini: Re https://review.openstack.org/#/c/304858/ - my guess is that the quotes are getting stripped in the transition from template to hieradata. | 17:42 |
| bnemec | Possibly wrapping it in single quotes would help? I'll give it a try. | 17:43 |
| *** fzdarsky|afk has quit IRC | 17:44 | |
| *** ebalduf has quit IRC | 17:45 | |
| *** padkrish has joined #tripleo | 17:49 | |
| trown | how is tripleo-ci creating opt/stack dir now? https://github.com/openstack/instack-undercloud/commit/2d76a8de512bcf3aa76a5e3b0e881431c3a6245e removed the creation of it and breaks quickstart | 17:53 |
| dprince | bnemec: which patch? | 17:53 |
| bnemec | dprince: This is the one that actually has changes: https://review.openstack.org/#/c/332641/ | 17:55 |
| bnemec | Since introspection doesn't get run on that repo there's a test patch here too: https://review.openstack.org/#/c/342669/ | 17:55 |
| openstackgerrit | Miles Gould proposed openstack/diskimage-builder: Allow installing proliant tools from an OS package https://review.openstack.org/344214 | 17:56 |
| openstackgerrit | Steven Hardy proposed openstack/tripleo-heat-templates: Allow map_replace substitution of network names https://review.openstack.org/345510 | 17:58 |
| *** mgould is now known as mgould|afk | 17:59 | |
| *** julim_ has joined #tripleo | 18:01 | |
| jaosorior | bnemec, dprince: Note that that patch needs dtantsur's fix to the ironic introspection's hieradata | 18:01 |
| jaosorior | without it it won't work | 18:01 |
| *** mcornea has quit IRC | 18:02 | |
| bnemec | jaosorior: That merged yesterday though, right? | 18:02 |
| openstackgerrit | John Trowbridge proposed openstack/instack-undercloud: Make sure we have an opt/stack dir https://review.openstack.org/345588 | 18:02 |
| jaosorior | that is correct | 18:02 |
| bnemec | trown: We're using the Jenkins slave as the undercloud now, and I believe /opt/stack is already populated when we get it. | 18:02 |
| jaosorior | it's this patch https://review.openstack.org/#/c/344792/ | 18:02 |
| trown | bnemec: k, so the bug I found would be encountered by anyone outside of CI I think | 18:03 |
| trown | not just quickstart | 18:03 |
| trown | bnemec: https://review.openstack.org/345588 fixes it, I can file a launchpad if we must, but really trying to get a promote in RDO and this is the latest in a string of blocking issues | 18:04 |
| *** julim has quit IRC | 18:04 | |
| bnemec | Interesting. I've installed multiple underclouds today with no trouble. | 18:05 |
| bnemec | tripleo.sh uses puppet modules from source though, which I think might also create /opt/stack. | 18:05 |
| trown | oh right | 18:05 |
| slagle | i stand alone | 18:08 |
| slagle | but you're still net +6 :) | 18:08 |
| slagle | trown: check my review, i think we can just get rid of that $OK_FILE | 18:08 |
| trown | slagle: I get wanting a bug... but it has been 15 days since we promoted master in RDO... I am just trying to get things working | 18:08 |
| * trown files a bug | 18:08 | |
| slagle | then don't file one | 18:09 |
| *** padkrish has quit IRC | 18:09 | |
| slagle | that's not why i -1'd | 18:09 |
| openstackgerrit | Jiri Tomasek proposed openstack/tripleo-ui: Use Webpack for application processing https://review.openstack.org/343592 | 18:09 |
| openstackgerrit | Jiri Tomasek proposed openstack/tripleo-ui: Remove app.conf to use a JS-based config file https://review.openstack.org/343658 | 18:09 |
| trown | slagle: ah right, I think a bug is getting filed in any case, but what did we ever use it for? | 18:10 |
| slagle | trown: the script wasn't always rerunabble | 18:10 |
| slagle | i made it so a wile back, and probably forgot to clean this up | 18:10 |
| trown | k | 18:10 |
| slagle | sorry | 18:11 |
| slagle | i left this there, and broke you with my commit yesterday | 18:11 |
| *** yamahata has quit IRC | 18:11 | |
| openstackgerrit | John Trowbridge proposed openstack/instack-undercloud: Remove "OK_FILE" from undercloud-install element https://review.openstack.org/345588 | 18:13 |
| trown | now with 100% more bugs attached | 18:13 |
| *** xinwu has joined #tripleo | 18:14 | |
| *** akuznetsov has joined #tripleo | 18:14 | |
| *** akuznetsov has quit IRC | 18:14 | |
| bnemec | Remove all the cruft! :-) | 18:15 |
| *** padkrish has joined #tripleo | 18:16 | |
| openstackgerrit | Sarath Kumar Sankaran Kutty proposed openstack/os-net-config: Enable os-net-config to support and configure NFVSwitch https://review.openstack.org/345599 | 18:17 |
| openstackgerrit | Merged openstack/tripleo-quickstart: Do not exit prior to the inventory of the deployment. https://review.openstack.org/341506 | 18:19 |
| openstackgerrit | James Slagle proposed openstack-infra/tripleo-ci: Add multi-node job in cistatus.html https://review.openstack.org/345418 | 18:24 |
| *** numans has quit IRC | 18:25 | |
| slagle | EmilienM: ^ check my change | 18:25 |
| slagle | it works for me now locally | 18:25 |
| EmilienM | slagle: I was writting the patch :) | 18:25 |
| slagle | I WIN | 18:25 |
| EmilienM | I was about doing a regex | 18:25 |
| slagle | oh ok well, i won't deprive you of doing a regex if you want | 18:25 |
| EmilienM | your patch looks good to me | 18:26 |
| slagle | you can overwrite what i did you want | 18:26 |
| slagle | *if | 18:26 |
| EmilienM | slagle: do you know the feeling when someone interrupts you when doing a regex? | 18:27 |
| slagle | is that like a SEGFAULT in your brain? | 18:27 |
| EmilienM | slagle: http://goo.gl/WDKPtq | 18:27 |
| *** jaosorior has quit IRC | 18:27 | |
| *** eggmaster has quit IRC | 18:28 | |
| openstackgerrit | Steven Hardy proposed openstack/instack-undercloud: Add missing db sync includes https://review.openstack.org/345604 | 18:28 |
| trown | is tripleo-ci pinning the Ironic API somewhere? I just started getting errors when trying to register nodes to Ironic that they are in enroll statee | 18:29 |
| slagle | we ought to be getting it from current-tripleo, so it's pinned in that regard | 18:30 |
| *** mcornea has joined #tripleo | 18:30 | |
| shardy | trown: we do export OS_BAREMETAL_API_VERSION=1.11 in stackrc? | 18:30 |
| *** jcoufal has joined #tripleo | 18:30 | |
| trown | shardy: interesting, I have that in my stackrc | 18:31 |
| bnemec | I don't know that stackrc is used for Ironic calls at registration anymore though. | 18:31 |
| trown | oh right, it is in mistral somewhere | 18:32 |
| bnemec | Mistral might pass that sort of thing through, but I wouldn't assume it does. | 18:32 |
| *** cwolferh has quit IRC | 18:32 | |
| pabelanger | What is needed to stop killing zuul-console on tripleo-ci jobs? https://review.openstack.org/#/c/340339/ | 18:36 |
| pabelanger | ^ seems like that wrong approach to CI | 18:36 |
| pabelanger | specifically, I cannot seem to find what is using 8088 | 18:37 |
| slagle | pabelanger: it's httpd for ipxe | 18:37 |
| slagle | we'd need to backport the instack-undercloud patch | 18:38 |
| slagle | and i'm not sure we could | 18:38 |
| slagle | this one: https://review.openstack.org/#/c/339777/ | 18:38 |
| slagle | i dont think that would backport to liberty or mitaka | 18:38 |
| openstackgerrit | Merged openstack/tripleo-puppet-elements: Installing openstack-sahara-ui https://review.openstack.org/344976 | 18:39 |
| pabelanger | slagle: Hmm, I am trying to remember which project also had the problem with 8088. Let me see if I can find that | 18:39 |
| pabelanger | I also think we proposed changing from port 8088 to something else | 18:39 |
| *** mcornea has quit IRC | 18:40 | |
| *** devvesa has quit IRC | 18:40 | |
| trown | dprince: do you know how to keep our pin to Ironic API v1.11 using the new mistral node registration? | 18:41 |
| trown | dprince: in my env nodes are starting in enroll state, which we dont seem to handle well | 18:42 |
| dprince | trown: yeah I think so. let me get a link | 18:42 |
| trown | filed a bug for it https://bugs.launchpad.net/tripleo/+bug/1605359 | 18:42 |
| openstack | Launchpad bug 1605359 in tripleo "[trunk] With master of Ironic, node registration fails" [Critical,Triaged] | 18:42 |
| *** jcoufal_ has joined #tripleo | 18:44 | |
| dprince | trown: http://git.openstack.org/cgit/openstack/mistral/tree/mistral/actions/openstack/actions.py#n31 | 18:45 |
| dprince | trown: http://git.openstack.org/cgit/openstack/mistral/tree/mistral/actions/openstack/actions.py#n372 | 18:46 |
| pabelanger | slagle: 330829 will fix zuul-console, were changing the port to 19885 | 18:46 |
| *** jcoufal has quit IRC | 18:46 | |
| dprince | trown: perhaps we need some options there? | 18:46 |
| dsneddon | trozet, You were looking for me earlier? | 18:46 |
| trozet | dsneddon: yeah I think I found a bug related to neutron THT in stable/mitaka, wanted to run it by you | 18:47 |
| dsneddon | trozet, Uh oh, yeah, what did you find? | 18:47 |
| trown | dprince: or maybe we just hardcode a pin as we had previously | 18:47 |
| dprince | trown: maybe | 18:48 |
| trozet | dsneddon: in stable/mitaka THT there is a heat parameter NeutronL3HA, which is set to type string and defaulted to 'false' | 18:48 |
| trown | oh... except that is in mistral code | 18:48 |
| trozet | dsneddon: but if you look in puppet-neutron stable/mitaka services.pp it is supposed to be a boolean | 18:48 |
| dprince | trown: right, we'd need an option to hard code it | 18:48 |
| trozet | dsneddon: so I think this always returns true | 18:48 |
| trown | dprince: nevermind... dont think that would fly there | 18:48 |
| trown | dprince: so ya we need some option to pin the api... or alternatively fix the workflow to allow for enroll state | 18:49 |
| dsneddon | trozet, Nice catch. Heat does funny things with types sometimes. | 18:49 |
| dprince | trown: right, but if you need an option for this exposed I can work on it | 18:49 |
| *** cwolferh has joined #tripleo | 18:49 | |
| trozet | dsneddon: well I don't think it's heat, i think its just the wrong type | 18:49 |
| trown | dprince: ya, we are broken with master of Ironic | 18:49 |
| dsneddon | trozet, If you point out the template, I can fix it. One of us should open a launchpad bug, too. | 18:49 |
| dprince | trown: perhaps we need both. I'd imagine we are eventually going to want better control over things | 18:49 |
| trozet | dsneddon: I can file the bug and fix it upstream | 18:49 |
| dprince | trown: but if master Ironic is breaking things then we may as well fix them sooner too | 18:49 |
| trozet | dsneddon: but I'm at odds whether to fix it in master or not... let me show you why | 18:50 |
| trozet | dsneddon: in stable/mitaka the value is checked as a conditional: https://github.com/openstack/puppet-neutron/blob/stable/mitaka/manifests/server.pp#L429 | 18:50 |
| trozet | dsneddon: meaning it will always be true with 'false' | 18:50 |
| dprince | trown: I imagine fixing the workflow is probalby easier in this short term case | 18:50 |
| *** derekjhyang has quit IRC | 18:52 | |
| dprince | trown: pinning to a stable Ironic API is probably fine too. But the new actions would need to have a new version then | 18:52 |
| dprince | trown: like when I added the introspection actions I set api_version=1 in that file in Mistral | 18:52 |
| dprince | trown: if you can make a case that Ironic has gone and made things unstable pinning to the old version in a patch there would be acceptable I think | 18:53 |
| trown | dprince: ya, enroll has been the starting state in Ironic API for a long time I think... | 18:53 |
| dprince | trown: file a bug, and push a patch with that 1 line to Mistral | 18:53 |
| trown | dprince: we were just pinning so we didnt see it | 18:53 |
| dprince | trown: I'd +1 that | 18:53 |
| serverascode | shardy: I set ControllerEnableSwiftStorage to false but swift was still deployed, just not the storage part on the controller, which I guess makes sense, but I'm not clear on if it's possible to not deploy swift for glance? Also, what if I don't want to deploy ceilometer? I don't see any options around that. Sorry for newbie questions. Point me to any docs. | 18:53 |
| serverascode | FYI I'm on 0.8.14 of tripleo. | 18:53 |
| dprince | trown: I bet it'll land too | 18:53 |
| *** xinwu has quit IRC | 18:54 | |
| slagle | pabelanger: i am occassionally seeing a POST_FAILURE in the multinode job, like so: http://logs.openstack.org/02/345002/3/check/gate-tripleo-ci-centos-7-nonha-multinode-nv/355c785/ | 18:58 |
| slagle | pabelanger: i guess it can't upload logs for some reason | 18:58 |
| trown | dprince: k, so file a mistral bug for it? or just add mistral to the bug I filed? | 18:58 |
| slagle | pabelanger: do you think that's a problem on the tripleo-ci side? | 18:58 |
| dprince | trown: add it to the bug you filed | 18:58 |
| dprince | trown: just so there is context around why you are pushing it | 18:59 |
| dprince | trown: add me as a reviewer and I'll +1 | 18:59 |
| trown | dprince: k, thanks | 18:59 |
| pabelanger | slagle: hard to say, if I knew which launch it ran on I could check the logs | 19:01 |
| pabelanger | slagle: looking | 19:01 |
| slagle | pabelanger: i guess the error is in ansible_log.txt..."symlink has no referent" during the last rsync command | 19:01 |
| pabelanger | slagle: Yup, looking at that now. We don't support symlinks for log publishers | 19:02 |
| *** yamahata has joined #tripleo | 19:02 | |
| pabelanger | I'd have to see what the job is to see where that is failing | 19:02 |
| slagle | ok so it's probably b/c the packages we've installed must set up some dangling symlinks | 19:03 |
| openstackgerrit | wes hayutin proposed openstack/tripleo-quickstart: allow for a custom number of overcloud service workers https://review.openstack.org/345627 | 19:03 |
| pabelanger | ya, likely | 19:03 |
| openstackgerrit | Ben Nemec proposed openstack/tripleo-heat-templates: Explicitly set nova and neutron host on controllers https://review.openstack.org/304858 | 19:08 |
| *** r-mibu has quit IRC | 19:13 | |
| *** r-mibu has joined #tripleo | 19:13 | |
| trown | dprince: which mistral service actually runs the actions? executor? | 19:22 |
| *** akrivoka has quit IRC | 19:22 | |
| *** akuznetsov has joined #tripleo | 19:28 | |
| EmilienM | dprince: can you also look https://review.openstack.org/#/c/345025/ please? it was the dependency | 19:29 |
| *** beagles is now known as beagles_biab | 19:30 | |
| trown | dprince: hmm or actions are loaded during undercloud install? /me is confused how to test the Ironic API version change | 19:32 |
| openstackgerrit | wes hayutin proposed openstack/tripleo-heat-templates: [WIP] Do not merge, doc update https://review.openstack.org/345636 | 19:32 |
| *** jerrygb has quit IRC | 19:32 | |
| *** ramishra has quit IRC | 19:36 | |
| *** ramishra has joined #tripleo | 19:37 | |
| *** radek_ has joined #tripleo | 19:39 | |
| shardy | trown: actions and workflows are loaded during the undercloud install | 19:40 |
| *** jerrygb has joined #tripleo | 19:40 | |
| shardy | trown: puppet runs mistral-db-populate which loads the actions, and 98-undercloud-setup loads the workflows | 19:42 |
| trown | shardy: ah thanks, do you know how to reload the actions by chance? | 19:43 |
| shardy | trown: mistral-db-populate I think | 19:43 |
| *** shardy has quit IRC | 19:45 | |
| trown | maybe mistral-db-sync | 19:46 |
| trown | dprince: whenever you are around I could use some pointers on how you test mistral changes... hacking on /usr/lib/python2.7/site-packages is not working for me | 19:48 |
| *** radek_ has quit IRC | 19:50 | |
| dprince | trown: sorry, got interrupted | 19:57 |
| *** akuznetsov has quit IRC | 19:57 | |
| dprince | trown: try this http://paste.openstack.org/show/539109/ | 19:59 |
| dprince | trown: or some derivative | 20:00 |
| openstackgerrit | Merged openstack/puppet-tripleo: Include gnocchi config on non-bootstrap nodes https://review.openstack.org/345000 | 20:02 |
| trown | dprince: setup.py install is for tripleo_common ? | 20:02 |
| *** jprovazn has quit IRC | 20:04 | |
| dprince | trown: sure | 20:07 |
| *** lblanchard has quit IRC | 20:07 | |
| *** gfidente is now known as gfidente|afk | 20:11 | |
| *** shardy has joined #tripleo | 20:12 | |
| openstackgerrit | Tim Rozet proposed openstack/tripleo-heat-templates: Fixes type and description for NeutronL3HA https://review.openstack.org/345648 | 20:13 |
| dprince | EmilienM: +2 on your dependency patch | 20:15 |
| EmilienM | ack | 20:16 |
| EmilienM | trozet: nice catch | 20:16 |
| openstackgerrit | Steven Hardy proposed openstack/tripleo-heat-templates: Remove the ControllerCount constraint https://review.openstack.org/345649 | 20:16 |
| trozet | EmilienM: thanks, it is also set in tripleoclient, coming up with a patch for that | 20:17 |
| trozet | EmilienM: one thing maybe you know the answer to | 20:17 |
| openstackgerrit | Steven Hardy proposed openstack/tripleo-heat-templates: Remove the ControllerCount constraint https://review.openstack.org/345650 | 20:17 |
| EmilienM | shardy: if you're still up for reviews : https://review.openstack.org/#/c/345025/ and https://review.openstack.org/#/c/345031/ - thanks! | 20:17 |
| EmilienM | trozet: shoot | 20:18 |
| trozet | EmilienM: in the current python-tripleoclient logic, if number of controllers is greater than 1, they set NeutronL3HA to True. Do we want to preserve this logic? If so, how do I inherit in pacemaker/neutron-server.yaml and override the NeutronL3HA from base? | 20:18 |
| *** jeckersb is now known as jeckersb_gone | 20:19 | |
| EmilienM | trozet: nothing changes AFIK | 20:21 |
| EmilienM | I think we want to preserve this logic, at least to be backward compatible with our architecture | 20:21 |
| shardy | EmilienM: ack will do | 20:21 |
| EmilienM | if you think it's a mistake, you can file a bug | 20:21 |
| EmilienM | shardy: thanks | 20:21 |
| shardy | EmilienM: thanks for your feedback re the db_sync problems, maybe there is some other reason why it's not working for me, will investigate | 20:21 |
| shardy | I've definitely had to manually sync the last few times I've upgraded my undercloud | 20:22 |
| EmilienM | shardy: it was probably not working for you because there is an orchestrationn issue somewhere in the puppe tmodules | 20:22 |
| EmilienM | I need to know what service for example | 20:22 |
| EmilienM | so I can dig in code | 20:22 |
| EmilienM | give me one example of one service | 20:22 |
| trozet | EmilienM: yeah so what I'm wondering is here https://github.com/openstack/tripleo-heat-templates/blob/master/puppet/services/pacemaker/neutron-server.yaml#L26 | 20:22 |
| dprince | trozet: you can override it by creating a new variable, and ordering the hiera such that the new Heat parameter value sets a hiera key to take priority | 20:23 |
| shardy | EmilienM: I added a note to the review - I've had to manually sync Heat, Nova, Neutron and Ironic | 20:23 |
| trozet | dprince: ah! | 20:23 |
| dprince | trozet: that is... if we decide to do it that way | 20:23 |
| dprince | trozet: it can be done though. The last hiera setting will win due to how map_merge works | 20:23 |
| trozet | dprince: so in https://github.com/openstack/tripleo-heat-templates/blob/master/puppet/services/pacemaker/neutron-server.yaml#L26 is there a way to do it in the map_merge? | 20:23 |
| shardy | EmilienM: If you've not upgraded your undercloud today you can probably reproduce, it happened today when I pulled the packages from the most recent current-tripleo promote (updated nova) | 20:24 |
| trozet | dprince: i would think we default it to false in base neutron-server.yaml , but we want to override that to true in pacemaker/neutron-server.yaml | 20:24 |
| *** akshai_ has quit IRC | 20:24 | |
| EmilienM | shardy: ok I found why I think | 20:26 |
| EmilienM | https://github.com/openstack/puppet-nova/blob/master/manifests/db/sync.pp#L25 | 20:26 |
| EmilienM | we subscribe to config only, but let me confirm that | 20:26 |
| trozet | dprince: I guess to make the question more simple. What does map_merge do if you include multiple keys of the same name? | 20:26 |
| shardy | trozet: last one wins | 20:27 |
| EmilienM | shardy: if packages are updated in Puppet, db_sync is applied | 20:27 |
| dprince | trozet: takes the last one :) | 20:27 |
| trozet | shardy, dprince: great, then I can just make the pacemaker version set it to true | 20:27 |
| trozet | thanks | 20:27 |
| EmilienM | shardy: but if not, if you manually runs "yum update" -> puppet won't know package is updated and will not run dbsync | 20:27 |
| shardy | EmilienM: I think we should make db sync happen every time openstack undercloud install runs | 20:28 |
| EmilienM | shardy: wait, I might have a fix | 20:28 |
| EmilienM | shardy: https://github.com/openstack/instack-undercloud/blob/f7b21b7571e4db2ec1099d58c9d5632abb13df0d/elements/puppet-stack-config/puppet-stack-config.pp#L27-L31 | 20:29 |
| EmilienM | this is not goof | 20:29 |
| EmilienM | good* | 20:29 |
| EmilienM | shardy: let me propose something else | 20:29 |
| *** egafford has quit IRC | 20:30 | |
| shardy | EmilienM: Ok, sounds good, thanks :) | 20:30 |
| shardy | I'll abandon my patch | 20:30 |
| EmilienM | shardy: no | 20:30 |
| EmilienM | keep it please | 20:30 |
| EmilienM | I'll use it for my test | 20:31 |
| *** egafford has joined #tripleo | 20:31 | |
| openstackgerrit | Tim Rozet proposed openstack/tripleo-heat-templates: Fixes type and description for NeutronL3HA https://review.openstack.org/345648 | 20:32 |
| openstackgerrit | Emilien Macchi proposed openstack/instack-undercloud: Use Puppet package provider to update OpenStack packages https://review.openstack.org/345658 | 20:34 |
| EmilienM | shardy: ^ it should do the job | 20:34 |
| *** akshai has joined #tripleo | 20:36 | |
| openstackgerrit | Steven Hardy proposed openstack-infra/tripleo-ci: Add SwiftRingBuilder to multinode job https://review.openstack.org/345661 | 20:39 |
| slagle | EmilienM: see my review. this is probably my fault | 20:39 |
| slagle | EmilienM: when i added this to be done in puppet, i wanted to make sure it was the same as it was done in the elements, so that's why the puppet steps are separate | 20:40 |
| slagle | i figured we could improve it later as we went | 20:40 |
| openstackgerrit | Steven Hardy proposed openstack/tripleo-heat-templates: Convert Swift ringbuilder to composable services format https://review.openstack.org/338551 | 20:40 |
| shardy | EmilienM: thanks! | 20:41 |
| openstackgerrit | Tim Rozet proposed openstack/python-tripleoclient: Removes setting NeutronL3HA and NeutronAllowL3AgentFailover https://review.openstack.org/345663 | 20:42 |
| *** dprince has quit IRC | 20:44 | |
| openstackgerrit | Merged openstack/puppet-tripleo: Deploy Keystone and Gnocchi API with new Apache profile https://review.openstack.org/345025 | 20:44 |
| *** jeckersb_gone is now known as jeckersb | 20:51 | |
| EmilienM | slagle: ok I'll look | 20:58 |
| *** bfournie1 has joined #tripleo | 20:59 | |
| openstackgerrit | Merged openstack/tripleo-heat-templates: Deploy Horizon with composable apache service https://review.openstack.org/345031 | 20:59 |
| *** bfournie has quit IRC | 21:01 | |
| *** ccamacho has quit IRC | 21:02 | |
| *** julim_ has quit IRC | 21:14 | |
| *** alhernan has quit IRC | 21:15 | |
| *** bvandenh has joined #tripleo | 21:16 | |
| *** trown is now known as trown|outtypewww | 21:19 | |
| *** bandini has quit IRC | 21:22 | |
| *** penick has quit IRC | 21:26 | |
| *** bvandenh has quit IRC | 21:32 | |
| pabelanger | I asked this yesterday, but no response | 21:33 |
| pabelanger | what are you using the local mirror for in tripleo-ci? To me it looks mirrors of puppet modules | 21:34 |
| *** shardy has quit IRC | 21:39 | |
| *** liverpooler has quit IRC | 21:41 | |
| *** bfournie1 has quit IRC | 21:43 | |
| *** akshai has quit IRC | 21:51 | |
| *** dtrainor_ has joined #tripleo | 21:53 | |
| *** dtrainor has quit IRC | 21:53 | |
| *** cdearborn has quit IRC | 22:11 | |
| *** dtrainor_ has quit IRC | 22:11 | |
| *** jayg is now known as jayg|g0n3 | 22:15 | |
| *** paramite has quit IRC | 22:16 | |
| *** jeckersb is now known as jeckersb_gone | 22:19 | |
| *** jeckersb_gone is now known as jeckersb | 22:27 | |
| *** dtrainor_ has joined #tripleo | 22:29 | |
| bnemec | pabelanger: The puppet modules, and I believe that's also where we cache images. | 22:31 |
| bnemec | I'm mostly going off of what is in http://git.openstack.org/cgit/openstack-infra/tripleo-ci/tree/scripts/mirror-server | 22:31 |
| *** gfidente|afk has quit IRC | 22:38 | |
| *** morazi has quit IRC | 22:51 | |
| *** jerrygb has quit IRC | 22:55 | |
| *** akshai has joined #tripleo | 22:57 | |
| bnemec | Everybody's probably gone at this point, but https://review.openstack.org/#/c/345570/ passed CI and will knock a couple of minutes of all our runs. | 23:05 |
| *** akshai has quit IRC | 23:26 | |
| *** saneax_AFK is now known as saneax | 23:33 | |
| *** jeckersb is now known as jeckersb_gone | 23:53 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
