| *** dardelean_ has quit IRC | 00:00 | |
| *** vijaykc4 has quit IRC | 00:14 | |
| *** vijaykc4 has joined #openstack-containers | 00:30 | |
| *** dardelean_ has joined #openstack-containers | 01:05 | |
| *** dardelean_ has quit IRC | 01:10 | |
| *** pengdake has joined #openstack-containers | 01:18 | |
| *** hongbin has joined #openstack-containers | 01:18 | |
| *** pengdake_ has joined #openstack-containers | 01:38 | |
| *** pengdake has quit IRC | 01:41 | |
| *** ramishra has joined #openstack-containers | 02:14 | |
| *** Nel1x has joined #openstack-containers | 03:29 | |
| *** yolanda has joined #openstack-containers | 03:44 | |
| *** yolanda_ has quit IRC | 03:46 | |
| *** hongbin has quit IRC | 03:48 | |
| *** pengdake_ has quit IRC | 04:22 | |
| *** yamamoto has joined #openstack-containers | 04:28 | |
| *** pengdake_ has joined #openstack-containers | 04:50 | |
| *** pengdake_ has quit IRC | 04:55 | |
| *** pengdake has joined #openstack-containers | 04:56 | |
| *** Nel1x has quit IRC | 04:59 | |
| *** ykarel has joined #openstack-containers | 05:05 | |
| *** udesale has joined #openstack-containers | 05:09 | |
| *** flwang1 has quit IRC | 05:12 | |
| *** vijaykc4 has quit IRC | 05:21 | |
| *** udesale_ has joined #openstack-containers | 05:40 | |
| *** udesale has quit IRC | 05:41 | |
| *** vijaykc4 has joined #openstack-containers | 05:43 | |
| *** vijaykc4 has quit IRC | 05:56 | |
| *** pengdake has quit IRC | 05:57 | |
| *** udesale__ has joined #openstack-containers | 06:01 | |
| *** udesale_ has quit IRC | 06:03 | |
| *** AlexeyAbashkin has joined #openstack-containers | 06:04 | |
| *** iranzo has joined #openstack-containers | 06:10 | |
| *** iranzo has joined #openstack-containers | 06:10 | |
| *** sheel has joined #openstack-containers | 06:19 | |
| *** janki has joined #openstack-containers | 06:24 | |
| *** pcaruana has joined #openstack-containers | 06:26 | |
| *** belmoreira has joined #openstack-containers | 06:27 | |
| *** yamamoto has quit IRC | 06:35 | |
| *** yamamoto has joined #openstack-containers | 06:41 | |
| *** armaan has joined #openstack-containers | 07:01 | |
| *** pengdake has joined #openstack-containers | 07:01 | |
| *** b_bezak has joined #openstack-containers | 07:07 | |
| *** rcernin has quit IRC | 07:10 | |
| *** dardelean_ has joined #openstack-containers | 07:14 | |
| *** gsimondon has joined #openstack-containers | 07:19 | |
| *** dardelean_ has quit IRC | 07:28 | |
| *** armaan has quit IRC | 07:34 | |
| *** armaan has joined #openstack-containers | 07:34 | |
| *** belmoreira has quit IRC | 07:36 | |
| *** janki has quit IRC | 07:36 | |
| *** armaan has quit IRC | 07:44 | |
| *** armaan has joined #openstack-containers | 07:45 | |
| *** ykarel is now known as ykarel|afk | 07:46 | |
| *** yamamoto has quit IRC | 07:50 | |
| *** yamamoto has joined #openstack-containers | 07:51 | |
| *** yamamoto has quit IRC | 07:54 | |
| *** yamamoto has joined #openstack-containers | 07:54 | |
| *** yamamoto has quit IRC | 07:54 | |
| *** yamamoto has joined #openstack-containers | 07:56 | |
| *** ktibi has joined #openstack-containers | 07:58 | |
| *** janki has joined #openstack-containers | 07:59 | |
| brtknr | we have some patches locally which make magnum compatible with Docker-CE for both swarm and k8s. would it be helpful to push them upstream? | 08:00 |
|---|---|---|
| strigazi | brtknr: on atomic? | 08:00 |
| brtknr | strigazi: yes | 08:00 |
| strigazi | then yes! | 08:00 |
| brtknr | okay! great | 08:01 |
| strigazi | brtknr: we have a stoty (previously bug) since some time | 08:01 |
| strigazi | brtknr: we have a story (previously bug) since some time | 08:01 |
| brtknr | on storyboard? | 08:02 |
| strigazi | brtknr: https://storyboard.openstack.org/#!/story/1727700 | 08:02 |
| brtknr | strigazi: also nice to see the move to storyboard :) | 08:05 |
| strigazi | brtknr: I'm passing all the blueprints now | 08:05 |
| strigazi | brtknr: do you have on github or somewhere the patches? | 08:05 |
| brtknr | strigazi: yes, https://github.com/stackhpc/magnum/commits/stackhpc/queens | 08:08 |
| brtknr | strigazi: it is specifically commit https://github.com/stackhpc/magnum/commit/2272ac84f29ff6cc5c6a19577b75c33377bf0193 that relates to supporting k8s and swarm | 08:09 |
| brtknr | with docker-ce | 08:09 |
| *** jmlowe has quit IRC | 08:10 | |
| strigazi | this assuems that docker is installed in the OS, no? | 08:10 |
| brtknr | but unlike the blueprint, we install docker via rpm-ostree rather than inside a container | 08:10 |
| brtknr | yes it does | 08:10 |
| strigazi | what do you think about atomic install docker? | 08:12 |
| brtknr | how would that work? | 08:13 |
| brtknr | i haven't tried that approach but sounds like that would be a better approach | 08:13 |
| strigazi | https://gitlab.cern.ch/cloud-infrastructure/magnum/commit/6aecfdcd237363a5dac2d0945bbb4da0c96d67f5 | 08:13 |
| strigazi | we can combine the two patches | 08:14 |
| strigazi | our patch with yours that configures daemon.json | 08:14 |
| *** ispp has joined #openstack-containers | 08:17 | |
| brtknr | how does running docker inside a container affect performance? | 08:18 |
| strigazi | it doesn't, why should it? | 08:18 |
| brtknr | i thought atomic install = run docker inside runc | 08:19 |
| strigazi | it does | 08:19 |
| strigazi | brtknr: what is the overhead of running a container with runc? | 08:23 |
| brtknr | im not entirely sure, thats why i was asking you | 08:24 |
| brtknr | strigazi: i imagine its minimal | 08:24 |
| brtknr | strigazi: i imagine its minimal | 08:32 |
| openstackgerrit | Piotr Mrowczynski proposed openstack/magnum stable/queens: Strip signed certificate https://review.openstack.org/574167 | 08:33 |
| strigazi | brtknr: i think there won't be any, not minimal | 08:33 |
| *** belmoreira has joined #openstack-containers | 08:34 | |
| *** janki has quit IRC | 08:38 | |
| *** janki has joined #openstack-containers | 08:40 | |
| *** janki has quit IRC | 08:52 | |
| *** salmankhan has joined #openstack-containers | 09:28 | |
| *** ykarel|afk is now known as ykarel | 09:30 | |
| brtknr | strigazi: what is a valid DOCKER_CE_VERSION? | 09:31 |
| brtknr | I've tried 18.03 17.03, 17.03.1, none are working | 09:31 |
| strigazi | You try to pull from where? | 09:32 |
| brtknr | From docker hub | 09:38 |
| brtknr | atomic install --system --system-package no --name docker hub.docker.com/docker/docker:stable | 09:38 |
| brtknr | is this not correct? | 09:38 |
| strigazi | it is not | 09:39 |
| strigazi | see https://storyboard.openstack.org/#!/story/1727700 | 09:39 |
| strigazi | we need a special container | 09:39 |
| *** udesale_ has joined #openstack-containers | 09:45 | |
| strigazi | tryL | 09:46 |
| strigazi | try: | 09:46 |
| strigazi | atomic install --system --system-package no --name docker gitlab-registry.cern.ch/cloud/docker-ce-centos:17.09 | 09:47 |
| strigazi | brtknr: ^^ | 09:47 |
| *** udesale__ has quit IRC | 09:47 | |
| *** udesale_ has quit IRC | 09:50 | |
| brtknr | strigazi: why do we need custom image? i dont fully understand | 09:54 |
| brtknr | also, where is the Dockerfile for docker-ce-centos:17.09? | 09:55 |
| brtknr | may i take a look? | 09:55 |
| strigazi | gitlab.cern.ch/cloud/docker-ce-centos:17.09 | 09:55 |
| strigazi | gitlab.cern.ch/cloud/docker-ce-centos here | 09:55 |
| *** ykarel has quit IRC | 09:58 | |
| *** ykarel has joined #openstack-containers | 10:01 | |
| brtknr | so the client inside the container remains 1.13 but the server is updated to 17.09 | 10:02 |
| brtknr | btw how does cri-o compare to running docker | 10:04 |
| *** pengdake has quit IRC | 10:44 | |
| brtknr | strigazi: when i build docker-centos locally, it appears to install docker-1.13.1 | 10:53 |
| brtknr | not 17.09 | 10:53 |
| strigazi | brtknr: try this one https://gitlab.cern.ch/cloud/atomic-system-containers/tree/cern-qa/docker-centos | 10:57 |
| *** ykarel_ has joined #openstack-containers | 11:00 | |
| *** ykarel has quit IRC | 11:02 | |
| *** udesale has joined #openstack-containers | 11:03 | |
| brtknr | strigazi: this one works! thanks | 11:07 |
| *** dardelean_ has joined #openstack-containers | 11:13 | |
| *** udesale has quit IRC | 11:14 | |
| *** udesale has joined #openstack-containers | 11:29 | |
| *** ykarel_ is now known as ykarel | 11:31 | |
| *** vijaykc4 has joined #openstack-containers | 11:33 | |
| *** rochapor1o has quit IRC | 11:34 | |
| *** udesale has quit IRC | 11:35 | |
| *** armaan has quit IRC | 11:57 | |
| *** armaan has joined #openstack-containers | 11:58 | |
| *** armaan has quit IRC | 12:02 | |
| *** ramishra has quit IRC | 12:09 | |
| *** markguz_ has joined #openstack-containers | 12:10 | |
| *** armaan has joined #openstack-containers | 12:18 | |
| *** armaan has quit IRC | 12:20 | |
| *** armaan has joined #openstack-containers | 12:20 | |
| *** markguz_ has quit IRC | 12:23 | |
| *** markguz_ has joined #openstack-containers | 12:26 | |
| *** ramishra has joined #openstack-containers | 12:30 | |
| *** serlex has joined #openstack-containers | 12:37 | |
| *** udesale has joined #openstack-containers | 12:38 | |
| *** pcaruana has quit IRC | 12:39 | |
| *** AlexeyAbashkin has quit IRC | 12:42 | |
| *** Nel1x has joined #openstack-containers | 12:44 | |
| *** Nel1x has quit IRC | 12:53 | |
| *** mago_ has joined #openstack-containers | 12:58 | |
| *** armaan has quit IRC | 13:01 | |
| *** armaan has joined #openstack-containers | 13:01 | |
| *** markguz_ has quit IRC | 13:03 | |
| *** pengdake has joined #openstack-containers | 13:05 | |
| *** udesale has quit IRC | 13:07 | |
| *** pengdake has quit IRC | 13:12 | |
| brtknr | how do you override daemon.json create by this repo? | 13:12 |
| *** udesale has joined #openstack-containers | 13:15 | |
| *** zul has quit IRC | 13:15 | |
| brtknr | actually, it looks like docker running inside runc respect docker-storage | 13:19 |
| *** zul has joined #openstack-containers | 13:19 | |
| *** pengdake has joined #openstack-containers | 13:21 | |
| *** vabada has joined #openstack-containers | 13:23 | |
| *** pengdake has quit IRC | 13:26 | |
| *** ispp has quit IRC | 13:29 | |
| *** markguz_ has joined #openstack-containers | 13:30 | |
| *** ispp has joined #openstack-containers | 13:31 | |
| *** pcaruana has joined #openstack-containers | 13:32 | |
| *** belmorei_ has joined #openstack-containers | 13:32 | |
| *** belmoreira has quit IRC | 13:34 | |
| strigazi | you can clear /etc/sysconfig/docker-storage | 13:43 |
| strigazi | brtknr: ^^ | 13:44 |
| *** sheel has quit IRC | 13:49 | |
| brtknr | any reason why you havent pushed your patch upstream as it looks like it already works without any issues... | 14:00 |
| *** ianychoi has quit IRC | 14:02 | |
| *** markguz_ has quit IRC | 14:02 | |
| *** AlexeyAbashkin has joined #openstack-containers | 14:04 | |
| *** armaan has quit IRC | 14:05 | |
| *** armaan has joined #openstack-containers | 14:06 | |
| *** zhubingbing has joined #openstack-containers | 14:06 | |
| *** Alexey_Abashkin has joined #openstack-containers | 14:11 | |
| *** zul is now known as zul_ | 14:11 | |
| *** yamamoto has quit IRC | 14:11 | |
| *** yamamoto has joined #openstack-containers | 14:12 | |
| *** AlexeyAbashkin has quit IRC | 14:13 | |
| *** Alexey_Abashkin is now known as AlexeyAbashkin | 14:13 | |
| *** zul_ is now known as zul | 14:14 | |
| *** yamamoto has quit IRC | 14:19 | |
| *** chhagarw has joined #openstack-containers | 14:20 | |
| *** ricolin__ has joined #openstack-containers | 14:21 | |
| *** jmlowe has joined #openstack-containers | 14:31 | |
| *** yamamoto has joined #openstack-containers | 14:32 | |
| *** ykarel is now known as ykarel|away | 14:43 | |
| *** hongbin has joined #openstack-containers | 14:46 | |
| *** ispp has quit IRC | 14:51 | |
| *** b_bezak has quit IRC | 14:53 | |
| *** b_bezak has joined #openstack-containers | 14:53 | |
| *** ispp has joined #openstack-containers | 14:54 | |
| *** vabada has quit IRC | 14:54 | |
| *** b_bezak has quit IRC | 14:58 | |
| *** gsimondon has quit IRC | 15:03 | |
| *** jmlowe has quit IRC | 15:08 | |
| *** zhubingbing has quit IRC | 15:12 | |
| *** vijaykc4 has quit IRC | 15:30 | |
| *** jmlowe has joined #openstack-containers | 15:32 | |
| *** markguz_ has joined #openstack-containers | 15:36 | |
| *** markguz_ has quit IRC | 15:37 | |
| *** vijaykc4 has joined #openstack-containers | 15:37 | |
| *** markguz_ has joined #openstack-containers | 15:38 | |
| *** mago_ has quit IRC | 15:38 | |
| *** armaan has quit IRC | 15:39 | |
| *** ispp has quit IRC | 15:40 | |
| *** belmorei_ has quit IRC | 15:42 | |
| *** ramishra has quit IRC | 15:44 | |
| *** vijaykc4 has quit IRC | 15:48 | |
| *** ktibi has quit IRC | 15:55 | |
| *** dardelean_ has quit IRC | 15:59 | |
| *** dtruong_ has quit IRC | 16:05 | |
| *** AlexeyAbashkin has quit IRC | 16:11 | |
| *** zhubingbing has joined #openstack-containers | 16:12 | |
| *** AlexeyAbashkin has joined #openstack-containers | 16:13 | |
| *** dardelean_ has joined #openstack-containers | 16:14 | |
| *** ramishra has joined #openstack-containers | 16:16 | |
| *** zhubingbing has quit IRC | 16:17 | |
| *** ramishra has quit IRC | 16:21 | |
| *** ricolin__ has quit IRC | 16:27 | |
| *** Alexey_Abashkin has joined #openstack-containers | 16:35 | |
| *** AlexeyAbashkin has quit IRC | 16:36 | |
| *** Alexey_Abashkin is now known as AlexeyAbashkin | 16:36 | |
| *** Alexey_Abashkin has joined #openstack-containers | 16:42 | |
| *** AlexeyAbashkin has quit IRC | 16:44 | |
| *** Alexey_Abashkin is now known as AlexeyAbashkin | 16:44 | |
| openstackgerrit | Clenimar Filemon proposed openstack/magnum master: federation-api: add api reference https://review.openstack.org/559732 | 16:48 |
| *** udesale has quit IRC | 16:56 | |
| *** armaan has joined #openstack-containers | 16:57 | |
| *** gsimondon has joined #openstack-containers | 17:07 | |
| *** dardelean_ has quit IRC | 17:08 | |
| *** dardelean_ has joined #openstack-containers | 17:09 | |
| *** dardelean_ has quit IRC | 17:13 | |
| *** armaan has quit IRC | 17:13 | |
| *** armaan has joined #openstack-containers | 17:14 | |
| *** armaan has quit IRC | 17:22 | |
| *** armaan has joined #openstack-containers | 17:23 | |
| *** armaan has quit IRC | 17:27 | |
| *** armaan has joined #openstack-containers | 17:27 | |
| *** AlexeyAbashkin has quit IRC | 17:35 | |
| *** spiette has quit IRC | 17:44 | |
| *** spiette has joined #openstack-containers | 17:46 | |
| *** armaan has quit IRC | 17:52 | |
| *** ykarel|away has quit IRC | 18:00 | |
| *** flwang1 has joined #openstack-containers | 18:05 | |
| *** iranzo has quit IRC | 18:06 | |
| *** salmankhan has quit IRC | 18:12 | |
| *** zhubingbing has joined #openstack-containers | 18:14 | |
| *** zhubingbing has quit IRC | 18:19 | |
| *** imdigitaljim has quit IRC | 18:20 | |
| *** imdigitaljim has joined #openstack-containers | 18:20 | |
| *** armaan has joined #openstack-containers | 18:24 | |
| *** dave-mcc_ has quit IRC | 18:25 | |
| *** armaan has quit IRC | 18:27 | |
| *** gsimondon has quit IRC | 18:29 | |
| *** dave-mccowan has joined #openstack-containers | 18:30 | |
| *** dtruong has joined #openstack-containers | 18:31 | |
| *** oikiki has joined #openstack-containers | 18:34 | |
| *** chhagarw has quit IRC | 18:50 | |
| *** gsimondon has joined #openstack-containers | 18:56 | |
| *** dardelean_ has joined #openstack-containers | 18:59 | |
| *** dardelean_ has quit IRC | 19:03 | |
| *** jmlowe has quit IRC | 19:04 | |
| *** flwang1 has quit IRC | 19:29 | |
| *** gsimondo1 has joined #openstack-containers | 19:35 | |
| *** gsimondon has quit IRC | 19:37 | |
| *** yamamoto has quit IRC | 19:51 | |
| *** armaan has joined #openstack-containers | 19:55 | |
| -openstackstatus- NOTICE: Zuul was restarted for a software upgrade; changes uploaded or approved between 19:30 and 19:50 will need to be rechecked | 19:56 | |
| *** armaan has quit IRC | 20:00 | |
| imdigitaljim | Regarding the continuation of applying primary/secondary masters: managing clusters, adding features, and maintaining code with a primary master resource and a secondary master resource for Kubernetes clusters (that may exist for long periods of time) will become burdensome and more prone to issues in terms of scaling masters without having a +1/-1 consideration, in-place upgrading clusters for a master pool and a snowflake, | 20:02 |
| imdigitaljim | federation for similar concerns, and failure recovery (HA) between different resource groups to name a few. Following the intended Kubernetes paradigm of a homogenous pools for masters and minions (not: 1 special master, other/more masters, and minions), makes everything more manageable and less prone to error because you treat all masters as masters and all minions as minions and heat would reflect this cleanly. However, at cluster | 20:02 |
| imdigitaljim | creation, I understand identifying a single bootstrapping master to construct the control plane and provide some services. flwang1 has put forth an excellent solution for this problem and can be adapted to fit almost any scenario for bootstrapping and I look forward to using it soon. Regarding the discoveryurl concerns, I think using the current default pattern as the primary choice in production still serves the best cluster delivery | 20:02 |
| imdigitaljim | method. However, I don’t like using the (external) public discoveryurl for availability and security reasons and propose adding as part of the Magnum setup the ability/documentation for configuring a project level etcd discovery service (we would be glad to put forth what we’ve done for this service), further eliminating the necessity/desire to use the two resource groups. Additional thoughts: we could consider a separate e | 20:02 |
| imdigitaljim | tcd cluster creation resource group to make scaling etcd and masters separated concerns. This has pros/cons we could discuss but I’m less concerned about this in the short term. | 20:02 |
| *** gsimondon has joined #openstack-containers | 20:04 | |
| *** gsimondo1 has quit IRC | 20:05 | |
| *** oikiki has quit IRC | 20:06 | |
| *** oikiki has joined #openstack-containers | 20:07 | |
| *** ispp has joined #openstack-containers | 20:12 | |
| *** gsimondon has quit IRC | 20:15 | |
| *** zhubingbing has joined #openstack-containers | 20:16 | |
| *** serlex has quit IRC | 20:17 | |
| *** zhubingbing has quit IRC | 20:21 | |
| *** AlexeyAbashkin has joined #openstack-containers | 20:32 | |
| *** Alexey_Abashkin has joined #openstack-containers | 20:46 | |
| *** AlexeyAbashkin has quit IRC | 20:47 | |
| *** Alexey_Abashkin is now known as AlexeyAbashkin | 20:47 | |
| *** Alexey_Abashkin has joined #openstack-containers | 20:51 | |
| *** yamamoto has joined #openstack-containers | 20:51 | |
| *** AlexeyAbashkin has quit IRC | 20:51 | |
| *** Alexey_Abashkin is now known as AlexeyAbashkin | 20:51 | |
| *** AlexeyAbashkin has quit IRC | 20:56 | |
| *** yamamoto has quit IRC | 20:58 | |
| *** flwang1 has joined #openstack-containers | 21:05 | |
| *** AlexeyAbashkin has joined #openstack-containers | 21:08 | |
| *** oikiki has quit IRC | 21:13 | |
| *** ispp has quit IRC | 21:15 | |
| *** oikiki has joined #openstack-containers | 21:17 | |
| *** oikiki has quit IRC | 21:20 | |
| flwang | imdigitaljim: still around? I'm keen to know the idea of separated etcd resource group or any idea removing the dependency of discovery.etcd.io | 21:21 |
| imdigitaljim | yeah im around | 21:23 |
| imdigitaljim | we've been using a discovery etcd service that we have control over instead of the public one, which I'd recommend using as the primary approach for production level cluster bootstrapping | 21:25 |
| imdigitaljim | https://github.com/coreos/discovery.etcd.io | 21:25 |
| imdigitaljim | I can get some documentation on the setup pushed upstream | 21:26 |
| imdigitaljim | by me or others on my team | 21:26 |
| imdigitaljim | The idea of separated etcd would involve another resource group that is built prior to the masters that would separately provide etcd for the cluster. This would reflect in heat with just resource groups for etcd, masters, and minions. I havent POC'd this idea yet personally but I've researched similar configurations | 21:30 |
| flwang | imdigitaljim: will those 3 etcd server be built in the same time? | 21:41 |
| flwang | then how to resolve the discovery issue? | 21:42 |
| imdigitaljim | flwang: there still is a discovery service but its not external | 21:51 |
| imdigitaljim | all clusters in a project use this discovery service on bootstrap | 21:52 |
| imdigitaljim | but its solely controlled by the OpenStack/Magnum admins | 21:53 |
| imdigitaljim | rather than discovery.etcd.io | 21:53 |
| imdigitaljim | http:// | 21:53 |
| *** AlexeyAbashkin has quit IRC | 21:54 | |
| flwang | ok, so you're talking about create a dedicated discovery server in openstack level, right? | 21:57 |
| flwang | which may be not dedicated for magnum, is it? | 21:57 |
| flwang | imdigitaljim: ^ | 21:57 |
| *** armaan has joined #openstack-containers | 21:59 | |
| imdigitaljim | yes the discovery service would be dedicated on an openstack level | 22:02 |
| imdigitaljim | but specifically for all magnum clusters to consume on bootstrap | 22:02 |
| imdigitaljim | so in other words, the configuration of magnum would be to setup api, conductor, and global etcd discovery | 22:03 |
| imdigitaljim | part of the configuration* | 22:03 |
| *** armaan has quit IRC | 22:13 | |
| *** armaan has joined #openstack-containers | 22:14 | |
| openstackgerrit | Jim Bach proposed openstack/magnum master: Allow multimaster lb with no floating ip option https://review.openstack.org/574476 | 22:14 |
| *** rcernin has joined #openstack-containers | 22:16 | |
| flwang | imdigitaljim: cool | 22:16 |
| flwang | that's probably we're going to do as well | 22:16 |
| *** zhubingbing has joined #openstack-containers | 22:18 | |
| flwang1 | imdigitaljim: thanks for the patch https://review.openstack.org/574476 that's what i'm going to propose | 22:21 |
| *** zhubingbing has quit IRC | 22:22 | |
| flwang1 | imdigitaljim: but what happened if user want to enable octavia and without FIP? | 22:23 |
| flwang1 | imdigitaljim: because technically, if user has got LB for multi master, then it's not necessary to assign FIP for master nodes | 22:24 |
| imdigitaljim | flwang1: "enable octavia and without FIP?" I completely agree here and havent made an octavia version yet but I absolutely will, we have a team member getting octavia working soon and I'll be adding that as well | 22:27 |
| imdigitaljim | we definitely will need it too | 22:27 |
| flwang1 | imdigitaljim: i would suggest improving the logic here https://review.openstack.org/#/c/574476/1/magnum/drivers/heat/template_def.py@332 | 22:28 |
| flwang1 | and we can help test it | 22:28 |
| flwang1 | BTW, i'm working on a patch to make the etcd lb optional | 22:28 |
| flwang1 | because it doesn't make much sense to waste 1 FIP and 2 instances for a internal use etcd server | 22:29 |
| imdigitaljim | I will have a look at it :) | 22:30 |
| flwang | cool | 22:31 |
| *** armaan has quit IRC | 22:33 | |
| *** armaan has joined #openstack-containers | 22:33 | |
| imdigitaljim | flwang: would you like me to make an expected change for octavia or can I make a later patch for octavia's no fip version? | 22:34 |
| flwang | imdigitaljim: your call, mate | 22:36 |
| flwang | we can merge this one and make the octavia_no_VIP as a following patch | 22:36 |
| imdigitaljim | i could verify it works later :) | 22:37 |
| imdigitaljim | (the octavia version) on a later patch | 22:37 |
| flwang | cool | 22:39 |
| flwang | thanks | 22:39 |
| imdigitaljim | what changes would you like for now for a +1? :) | 22:46 |
| imdigitaljim | flwang:^ | 22:46 |
| *** dardelean_ has joined #openstack-containers | 22:49 | |
| flwang | imdigitaljim: at least a comments from you say the octavia with FIP case will be done later, or something like that | 22:49 |
| *** armaan has quit IRC | 22:50 | |
| flwang | so that the other reviewers could be aware of our talk | 22:50 |
| *** armaan has joined #openstack-containers | 22:50 | |
| imdigitaljim | okay sounds good | 22:51 |
| imdigitaljim | will do | 22:51 |
| flwang | imdigitaljim: awesome, thanks | 22:51 |
| flwang | imdigitaljim: another question | 22:57 |
| flwang | can't we just reuse the existing floating_ip_enabled https://review.openstack.org/#/c/574476/1/magnum/drivers/heat/template_def.py@357 ? | 22:57 |
| *** armaan has quit IRC | 23:06 | |
| imdigitaljim | its a different set of conditions | 23:06 |
| imdigitaljim | if you mean env_files.append(COMMON_ENV_PATH + 'disable_floating_ip.yaml') this path | 23:06 |
| imdigitaljim | here: def add_fip_env_file(env_files, cluster_template): | 23:07 |
| flwang | i mean can't we do --master-lb-enabled=True --floating-ip-enabled=False ? | 23:07 |
| flwang | can't it meet your requirement? | 23:07 |
| imdigitaljim | those flags are what we use yes | 23:07 |
| imdigitaljim | but the desired outcome is not | 23:08 |
| flwang | oh, ok | 23:08 |
| imdigitaljim | so what happens is when load balancers are made | 23:08 |
| imdigitaljim | floating: | 23:08 |
| imdigitaljim | type: Magnum::Optional::Neutron::LBaaS::FloatingIP | 23:08 |
| imdigitaljim | in lb.yaml | 23:09 |
| imdigitaljim | is still left with a floating ip | 23:09 |
| imdigitaljim | the no_fip patch present only does floating ip's for master/minions with | 23:09 |
| imdigitaljim | "Magnum::Optional::KubeMaster::Neutron::FloatingIP": "OS::Heat::None" | 23:09 |
| imdigitaljim | "Magnum::Optional::KubeMinion::Neutron::FloatingIP": "OS::Heat::None" | 23:10 |
| imdigitaljim | flwang: ^ | 23:10 |
| flwang | imdigitaljim: ok, i see. | 23:11 |
| flwang | then the fix for octavia scenario should be also easy | 23:11 |
| flwang | btw, why https://review.openstack.org/#/c/574476/1/magnum/drivers/common/templates/environments/with_master_lb_no_floating_ip.yaml@10 ? | 23:11 |
| imdigitaljim | great questions | 23:12 |
| flwang | it shouldn't be changed based on my understanding | 23:12 |
| imdigitaljim | that may be needing to be another patch separately actually with some other changes. | 23:13 |
| imdigitaljim | using a hardware load balancer you may not have health monitors | 23:13 |
| flwang | hmm... for that case, it shouldn't be part of the this change | 23:13 |
| imdigitaljim | agreed | 23:13 |
| imdigitaljim | good find =) | 23:14 |
| flwang | that's my job ;) | 23:14 |
| flwang | thanks for contributing | 23:15 |
| imdigitaljim | thanks for reviewing | 23:15 |
| flwang | I'm happy to +1 after you update you patch | 23:18 |
| flwang | but I may replace it with a -1 after my local testing ;) hah | 23:18 |
| *** hongbin has quit IRC | 23:26 | |
| *** dardelean_ has quit IRC | 23:31 | |
| imdigitaljim | pointing out that change made me see a potentially cleaner more effective way to do this | 23:33 |
| imdigitaljim | ill test it out and push it up | 23:33 |
| flwang | awesome | 23:46 |
| *** Nel1x has joined #openstack-containers | 23:52 | |
| *** markguz_ has quit IRC | 23:54 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!