| *** xinliang has joined #openstack-containers | 00:02 | |
| xinliang | Hello, this is xinliang from Linaro Datercenter and Cloud Working Group. | 00:14 |
|---|---|---|
| xinliang | We are working on make magnum running on arm64 now. | 00:14 |
| xinliang | These are things need to do, acording to our ivestigation/validation: | 00:14 |
| xinliang | 1, Build magnum heat agent and k8s images for arm64. | 00:15 |
| xinliang | We found some dockerfiles are hard coded to amd64 arch. | 00:15 |
| xinliang | https://github.com/openstack/magnum/tree/master/dockerfiles | 00:15 |
| xinliang | What else might need to do for arm64, or other things arch specific? | 00:16 |
| *** elenalindq has quit IRC | 00:40 | |
| *** ricolin has joined #openstack-containers | 01:10 | |
| flwang | xinliang: at this moment, let's say fedora coreos will be the driver | 02:05 |
| flwang | for that case, you need to make sure the hyperkube can run on arm64 | 02:06 |
| flwang | and heat-container-agent | 02:06 |
| flwang | the docker images you can see under https://github.com/openstack/magnum/tree/master/dockerfiles, some of them won't be used for fedora coreos | 02:07 |
| flwang | since we will switch to podman | 02:07 |
| flwang | xinliang: ^ | 02:07 |
| xinliang | flwang: Thanks for replying. | 02:07 |
| xinliang | I see from the release notes that federa coreos support is just add in train. And federa atomic official doc says coreos is replaing atomic. | 02:10 |
| xinliang | So shall I start with fedora atomic or fedora coreos? flwang | 02:10 |
| flwang | fedora coreos | 02:11 |
| flwang | fedora atomic has been end of life since 26th Nov 2019 | 02:11 |
| flwang | though it's still supported in Magnum | 02:11 |
| flwang | that's why we are adding the new fedora coreos driver | 02:11 |
| xinliang | oh, I got it. | 02:11 |
| xinliang | so I need to set USE_PODMAN to 'true' right? when I create the cluster template | 02:13 |
| xinliang | flwang: Do you have any example or advice to support multiarch dockerfile? | 02:15 |
| flwang | xinliang: i don't have much experience in this area, but i'm happy to support you. | 02:19 |
| flwang | firstly, i would suggest to have a workable magnum env | 02:19 |
| flwang | it doesn't need to be a arm64 env | 02:19 |
| flwang | just make sure it works | 02:20 |
| flwang | and then you can start to get the arm64 based hyperkube and try to build an arm64 docker image for the heat-container-agent and upload to your personal dockerhub | 02:20 |
| flwang | then start to test more | 02:20 |
| xinliang | ok, thanks flwang. | 02:22 |
| xinliang | I now have a setup of magnum dev env deployed by kolla. Or if I need to use devstack to deploy magnum for developing? | 02:24 |
| xinliang | I will look into hyperkub then. thanks flwang :-) | 02:24 |
| flwang | as long as it works, it doesn't need to be a devstack, kolla is OK | 02:25 |
| xinliang | ok, got it | 02:25 |
| *** kevinz has joined #openstack-containers | 02:50 | |
| *** openstackstatus has joined #openstack-containers | 03:41 | |
| *** ChanServ sets mode: +v openstackstatus | 03:41 | |
| *** udesale has joined #openstack-containers | 04:00 | |
| *** ykarel|away has joined #openstack-containers | 04:01 | |
| *** dave-mccowan has quit IRC | 04:04 | |
| *** xinliang has quit IRC | 04:09 | |
| *** ricolin has quit IRC | 04:16 | |
| *** ricolin has joined #openstack-containers | 04:18 | |
| *** ykarel|away is now known as ykarel | 04:21 | |
| *** rcernin has quit IRC | 06:03 | |
| *** elenalindq has joined #openstack-containers | 06:33 | |
| openstackgerrit | Andrei Nistor proposed openstack/magnum stable/train: Change k8s-keystone-auth docker repo https://review.opendev.org/697624 | 07:08 |
| *** xinliang has joined #openstack-containers | 07:15 | |
| openstackgerrit | Andrei Nistor proposed openstack/magnum stable/train: Change k8s-keystone-auth docker repo https://review.opendev.org/697624 | 07:24 |
| *** rcernin has joined #openstack-containers | 07:25 | |
| andrein | Good morning everyone! | 07:25 |
| *** elenalindq has quit IRC | 07:27 | |
| *** ykarel is now known as ykarel|lunch | 07:46 | |
| xinliang | morning andrein | 07:58 |
| *** pcaruana has joined #openstack-containers | 08:11 | |
| *** elenalindq has joined #openstack-containers | 08:15 | |
| *** ykarel|lunch is now known as ykarel | 08:57 | |
| *** xinliang has quit IRC | 09:45 | |
| brtknr | morning all | 09:49 |
| andrein | hi brtknr! | 09:58 |
| andrein | any ideas why https://storyboard.openstack.org/#!/story/2006967 is not updating with the changes in gerrit? | 09:59 |
| brtknr | andrein: because you removed the reference to task | 10:00 |
| andrein | Hmm, I think I never added that referencce in the first place :( | 10:00 |
| brtknr | andrein: actually, looks like the orginnal story never had a reference to task :) | 10:01 |
| andrein | is there any way to fix this? | 10:01 |
| brtknr | you can add it to the cherrypick | 10:01 |
| andrein | ok | 10:01 |
| openstackgerrit | Andrei Nistor proposed openstack/magnum stable/train: Change k8s-keystone-auth docker repo https://review.opendev.org/697624 | 10:04 |
| andrein | that was it... | 10:05 |
| andrein | brtknr: any thoughts on this? https://storyboard.openstack.org/#!/story/2006991 | 10:06 |
| brtknr | andrein: I think there is an CLI command to rotate certificates | 10:07 |
| andrein | brtknr: 'rotate_ca_certificate' is not supported by this driver (HTTP 400) (Request-ID: req-4bd0ca58-0f9b-4b22-96e5-3354b7b43669) | 10:08 |
| brtknr | which driver are you using? | 10:10 |
| brtknr | coreos? | 10:10 |
| andrein | using fedora atomic | 10:11 |
| andrein | my current use case is with terraform, I'd like to be able to request short-lived certificates (< 1h?) to authenticate to the kubernetes cluster created via terraform using the helm/kubernetes provider. | 10:11 |
| *** rcernin has quit IRC | 10:12 | |
| andrein | Ideally, I'd like an API endpoint to list/revoke certificates, but I don't think I'm up for that. Short-lived certificates is a good enough compromise for me. | 10:13 |
| *** rcernin has joined #openstack-containers | 10:16 | |
| *** rcernin has quit IRC | 10:33 | |
| openstackgerrit | Merged openstack/magnum stable/train: bug: Use configured heat-container-agent tag https://review.opendev.org/696887 | 10:43 |
| *** ramishra has quit IRC | 11:15 | |
| brtknr | andrein: do you have use_podman enabled? | 11:22 |
| openstackgerrit | Bharat Kunwar proposed openstack/magnum master: Add compatibility matrix for kube_tag https://review.opendev.org/685675 | 11:24 |
| *** ramishra has joined #openstack-containers | 11:31 | |
| openstackgerrit | Merged openstack/magnum stable/train: Change k8s-keystone-auth docker repo https://review.opendev.org/697624 | 11:43 |
| openstackgerrit | Bharat Kunwar proposed openstack/magnum master: Add compatibility matrix for kube_tag https://review.opendev.org/685675 | 11:45 |
| openstackgerrit | Merged openstack/magnum master: bug: Only query Cinder API if volume size > 0 https://review.opendev.org/695239 | 11:48 |
| openstackgerrit | Merged openstack/magnum master: Add nginx_ingress_controller_chart_tag https://review.opendev.org/694286 | 11:48 |
| openstackgerrit | Merged openstack/magnum master: [k8s] Fix rolling upgrade with podman https://review.opendev.org/696204 | 11:48 |
| openstackgerrit | Bharat Kunwar proposed openstack/magnum stable/train: bug: Only query Cinder API if volume size > 0 https://review.opendev.org/697650 | 11:49 |
| openstackgerrit | Bharat Kunwar proposed openstack/magnum stable/train: [k8s] Fix rolling upgrade with podman https://review.opendev.org/697651 | 11:50 |
| openstackgerrit | Bharat Kunwar proposed openstack/magnum stable/train: Add nginx_ingress_controller_chart_tag https://review.opendev.org/697652 | 11:51 |
| *** udesale has quit IRC | 12:00 | |
| openstackgerrit | Bharat Kunwar proposed openstack/magnum master: Add compatibility matrix for kube_tag https://review.opendev.org/685675 | 12:11 |
| *** mgoddard has quit IRC | 13:44 | |
| andrein | brtknr: no, I don't use podman. Is this related to the CA question earlier? | 13:58 |
| brtknr | andrein: yes thats right | 14:01 |
| andrein | hmmmm | 14:02 |
| brtknr | i havent used the rotate ca option | 14:02 |
| brtknr | perhaps its a bug? | 14:02 |
| andrein | I think it was never implemented: https://opendev.org/openstack/magnum/src/branch/master/magnum/drivers/common/driver.py#L237 | 14:05 |
| andrein | I can't find an implementation on any driver | 14:06 |
| *** ignaziocassano1 has joined #openstack-containers | 14:08 | |
| ignaziocassano1 | Hello, kubernetes containers on magnum stein is not working. Pods in kube-system are in pending state ... An help, PLEASE ? | 14:10 |
| brtknr | andrein: looking at the code, the api exists but no implementation | 14:11 |
| brtknr | andrein: ah you're already on the case :) | 14:12 |
| brtknr | ignaziocassano1: what version kube_tag are you trying to run? | 14:12 |
| brtknr | ignaziocassano1: please provide more context | 14:12 |
| *** mgoddard has joined #openstack-containers | 14:13 | |
| ignaziocassano1 | on mailing list they suggested to use: heat_container_agent_tag=train-stable,kube_tag=v1.14.8 | 14:13 |
| ignaziocassano1 | I do not know why they suggested train-stable | 14:15 |
| brtknr | ignaziocassano1: it is easier to debug | 14:16 |
| brtknr | because it produces output to log | 14:16 |
| brtknr | what is your output of kubectl get pods -A | 14:16 |
| brtknr | kubectl get all -A | 14:16 |
| brtknr | ignaziocassano1: btw that was me who suggest using train-stable :) | 14:17 |
| ignaziocassano1 | on the kube master the cloud-init logs does not reports erros | 14:17 |
| ignaziocassano1 | I am sorry, I did not know who you are by you nick name :-) | 14:19 |
| ignaziocassano1 | kubectl get nodes reports two nodes but without roles :-( | 14:20 |
| ignaziocassano1 | [centos@kubectl ~]$ kubectl get nodes | 14:20 |
| ignaziocassano1 | NAME STATUS ROLES AGE VERSION | 14:20 |
| ignaziocassano1 | kube-qw7jw4ko7ykc-master-0 Ready <none> 59m v1.14.8 | 14:20 |
| ignaziocassano1 | kube-qw7jw4ko7ykc-minion-0 Ready <none> 59m v1.14.8 | 14:20 |
| ignaziocassano1 | [centos@kubectl ~]$ kubectl get pod -n kube-system | 14:21 |
| ignaziocassano1 | NAME READY STATUS RESTARTS AGE | 14:21 |
| ignaziocassano1 | coredns-865bd969f-gdt88 0/1 Pending 0 62m | 14:21 |
| ignaziocassano1 | heapster-7bf5794cc7-gm6bs 0/1 Pending 0 62m | 14:21 |
| ignaziocassano1 | kube-dns-autoscaler-57bd7f54d5-vdn6r 0/1 Pending 0 62m | 14:21 |
| ignaziocassano1 | kube-flannel-ds-amd64-gjdp5 1/1 Running 0 62m | 14:21 |
| ignaziocassano1 | kube-flannel-ds-amd64-mqpdn 1/1 Running 0 62m | 14:21 |
| ignaziocassano1 | kubernetes-dashboard-d48c76949-ng4qb 0/1 Pending 0 62m | 14:21 |
| ignaziocassano1 | [centos@kubectl ~]$ kubectl get pods -A | 14:22 |
| ignaziocassano1 | NAMESPACE NAME READY STATUS RESTARTS AGE | 14:22 |
| ignaziocassano1 | kube-system coredns-865bd969f-gdt88 0/1 Pending 0 63m | 14:22 |
| ignaziocassano1 | kube-system heapster-7bf5794cc7-gm6bs 0/1 Pending 0 63m | 14:22 |
| ignaziocassano1 | kube-system kube-dns-autoscaler-57bd7f54d5-vdn6r 0/1 Pending 0 63m | 14:22 |
| ignaziocassano1 | kube-system kube-flannel-ds-amd64-gjdp5 1/1 Running 0 63m | 14:22 |
| ignaziocassano1 | kube-system kube-flannel-ds-amd64-mqpdn 1/1 Running 0 63m | 14:22 |
| ignaziocassano1 | kube-system kubernetes-dashboard-d48c76949-ng4qb 0/1 Pending 0 | 14:22 |
| ignaziocassano1 | NAMESPACE NAME READY STATUS RESTARTS AGE | 14:24 |
| ignaziocassano1 | kube-system pod/coredns-865bd969f-gdt88 0/1 Pending 0 65m | 14:24 |
| ignaziocassano1 | kube-system pod/heapster-7bf5794cc7-gm6bs 0/1 Pending 0 65m | 14:24 |
| ignaziocassano1 | kube-system pod/kube-dns-autoscaler-57bd7f54d5-vdn6r 0/1 Pending 0 65m | 14:24 |
| ignaziocassano1 | kube-system pod/kube-flannel-ds-amd64-gjdp5 1/1 Running 0 65m | 14:24 |
| ignaziocassano1 | kube-system pod/kube-flannel-ds-amd64-mqpdn 1/1 Running 0 65m | 14:24 |
| ignaziocassano1 | kube-system pod/kubernetes-dashboard-d48c76949-ng4qb 0/1 Pending 0 65m | 14:24 |
| ignaziocassano1 | NAMESPACE NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE | 14:24 |
| ignaziocassano1 | default service/kubernetes ClusterIP 10.254.0.1 <none> 443/TCP 65m | 14:24 |
| ignaziocassano1 | kube-system service/heapster ClusterIP 10.254.103.116 <none> 80/TCP 65m | 14:24 |
| ignaziocassano1 | kube-system service/kube-dns ClusterIP 10.254.0.10 <none> 53/UDP,53/TCP,9153/TCP 65m | 14:24 |
| ignaziocassano1 | kube-system service/kubernetes-dashboard ClusterIP 10.254.164.121 <none> 443/TCP 65m | 14:24 |
| ignaziocassano1 | NAMESPACE NAME DESIRED CURRENT READY UP-TO-DATE AVAILABLE NODE SELECTOR AGE | 14:24 |
| ignaziocassano1 | kube-system daemonset.apps/k8s-keystone-auth 0 0 0 0 0 node-role.kubernetes.io/master= 65m | 14:24 |
| ignaziocassano1 | kube-system daemonset.apps/kube-flannel-ds-amd64 2 2 2 2 2 beta.kubernetes.io/arch=amd64 65m | 14:24 |
| ignaziocassano1 | kube-system daemonset.apps/npd 0 0 0 0 0 <none> 65m | 14:24 |
| ignaziocassano1 | kube-system daemonset.apps/openstack-cloud-controller-manager 0 0 0 0 0 node-role.kubernetes.io/master= 65m | 14:24 |
| ignaziocassano1 | NAMESPACE NAME READY UP-TO-DATE AVAILABLE AGE | 14:24 |
| ignaziocassano1 | kube-system deployment.apps/coredns 0/1 1 0 65m | 14:24 |
| ignaziocassano1 | kube-system deployment.apps/heapster 0/1 1 0 65m | 14:24 |
| ignaziocassano1 | kube-system deployment.apps/kube-dns-autoscaler 0/1 1 0 65m | 14:24 |
| ignaziocassano1 | kube-system deployment.apps/kubernetes-dashboard 0/1 1 0 65m | 14:24 |
| ignaziocassano1 | NAMESPACE NAME DESIRED CURRENT READY AGE | 14:24 |
| ignaziocassano1 | kube-system replicaset.apps/coredns-865bd969f 1 1 0 65m | 14:24 |
| ignaziocassano1 | kube-system replicaset.apps/heapster-7bf5794cc7 1 1 0 65m | 14:25 |
| ignaziocassano1 | kube-system replicaset.apps/kube-dns-autoscaler-57bd7f54d5 1 1 0 65m | 14:25 |
| ignaziocassano1 | kube-system replicaset.apps/kubernetes-dashboard-d48c76949 1 1 0 65m | 14:25 |
| ignaziocassano1 | [centos@kubectl ~]$ | 14:25 |
| ignaziocassano1 | Do you need I share my cluster template ? | 14:28 |
| ignaziocassano1 | help /nick | 14:31 |
| brtknr | I know what your problem is | 14:34 |
| brtknr | ignaziocassano1: you need to use stein-8.2.0 release | 14:34 |
| brtknr | we have the fix for your issue there | 14:34 |
| ignaziocassano1 | I installed stein via yum tool .....Is there any rpm package update ? | 14:36 |
| ignaziocassano1 | I am using centos | 14:37 |
| brtknr | ignaziocassano1: yep https://trunk.rdoproject.org/centos7-stein/1d/ce/1dcedf17314ff111ce5458896a447c2dd2e7f50d_9591888a/rpmbuild.log | 14:37 |
| brtknr | it was released this afternoon | 14:37 |
| brtknr | 2 hours ago | 14:37 |
| ignaziocassano1 | OK | 14:37 |
| ignaziocassano1 | I will test it. Many thanks | 14:38 |
| ignaziocassano1 | mmmm ... if I run yum updtate on my controllers it reports no updates are available :-( | 14:40 |
| *** goldyfruit___ has joined #openstack-containers | 14:41 | |
| ignaziocassano1 | If I understood well I must enable delorean stein ? | 14:42 |
| *** ykarel is now known as ykarel|afk | 14:50 | |
| brtknr | ignaziocassano1: i am not sure how long releases take to get published | 14:51 |
| andrein | brtknr: is there a reason most tags can't be updated on a running cluster? | 14:51 |
| ignaziocassano1 | Many thanks for your help. I'll check when updates will be released | 14:54 |
| andrein | brtknr: I've read in the 9.1.0 release notes that kube_tag can be changed during a cluster upgrade, why can't we do the same for tags like min_node_count, max_node_count, autoscaling_enabled, monitoring_enabled etc | 14:56 |
| openstackgerrit | Bharat Kunwar proposed openstack/magnum master: bug: cluster creation without docker_volume_size https://review.opendev.org/697671 | 14:58 |
| openstackgerrit | Bharat Kunwar proposed openstack/magnum master: bug: cluster creation without docker_volume_size https://review.opendev.org/697671 | 15:00 |
| brtknr | andrein: those pieces of code get deployed during cluster bootstrapping | 15:01 |
| brtknr | I didnt design this :) | 15:01 |
| brtknr | The way things work under the hood, its imperative, not declaritive like kubernetes | 15:01 |
| brtknr | there is no "controller-manager" that tries to converge desired state with actual state | 15:02 |
| brtknr | it would be possible to do but would require time and commitment | 15:02 |
| brtknr | hence some things can be changed like node_count etc. | 15:03 |
| andrein | where can I read more about how magnum works internally? | 15:03 |
| brtknr | but things like max_node_count, are cluster variables etc. cannot be changed by magnum | 15:03 |
| brtknr | you can still edit the autoscaling daemonset | 15:03 |
| brtknr | and change the min/max node count | 15:03 |
| brtknr | or deployment... | 15:04 |
| brtknr | not sure what it is | 15:04 |
| brtknr | andrein: probably the docs? | 15:04 |
| *** dave-mccowan has joined #openstack-containers | 15:05 | |
| *** goldyfruit___ is now known as goldyfruit | 15:07 | |
| *** dave-mccowan has quit IRC | 15:10 | |
| andrein | docs are pretty dry on the internals, unfortunately | 15:13 |
| brtknr | strigazi: need a quick review for a semi-urgent bugfix: https://review.opendev.org/697671 | 15:13 |
| brtknr | andrein: can you show me an example of what a good internals doc looks like for another openstack project? | 15:17 |
| andrein | brtknr: that's a good question, I've found https://docs.openstack.org/octavia/latest/reference/introduction.html enlightening when installing octavia for the first time :) | 15:19 |
| brtknr | andrein: the wiki might be useful, there is a diagram of the architecture: https://wiki.openstack.org/wiki/Magnum | 15:20 |
| andrein | thanks! | 15:22 |
| *** ignaziocassano1 has quit IRC | 15:24 | |
| *** KeithMnemonic has joined #openstack-containers | 15:29 | |
| *** ykarel|afk is now known as ykarel|away | 15:38 | |
| *** ykarel|away has quit IRC | 15:49 | |
| *** elenalindq has quit IRC | 16:26 | |
| *** lpetrut has quit IRC | 16:35 | |
| *** ykarel|away has joined #openstack-containers | 17:16 | |
| *** KeithMnemonic has quit IRC | 17:22 | |
| *** pcaruana has quit IRC | 17:23 | |
| *** ricolin has quit IRC | 17:55 | |
| *** lpetrut has joined #openstack-containers | 20:11 | |
| *** lpetrut has quit IRC | 20:56 | |
| *** irclogbot_3 has quit IRC | 21:45 | |
| *** elenalindq has joined #openstack-containers | 22:04 | |
| *** ykarel|away has quit IRC | 22:11 | |
| *** goldyfruit has quit IRC | 22:12 | |
| *** ramishra has quit IRC | 23:48 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!