| ftl_mason | Hi haoiii and brunomuniz[m], I looked at your conversation and haoiii's logs and just wanted to offer a couple thoughts. | 03:49 |
|---|---|---|
| ftl_mason | First, line 27 of the logs (CMD: ssh-keygen -f "/home/hli/.ssh/known_hosts" -R [127.0.0.1]:3122) is just trying to delete any existing entries in known_hosts for 127.0.0.1. If known_hosts doesn't exist or doesn't have an entry for 127.0.0.1 it doesn't look like that's going to cause a problem for the installer. | 03:51 |
| ftl_mason | Regarding line 30 of the logs, the rsync command is failing because the program, sshpass, doesn't exist, which likely means the sshpass package wasn't installed. | 03:55 |
| ftl_mason | In other news - I got my stuck VirtualBox VM started again. In the IRC chat history danielcaires[m] said that there is a problem with VirtualBox VMs and the serial configuration that the pybox installer uses. I changed the serial port port mode to TCP and set the port to 2000 (just an arbitrary port above 1024) and it worked! I see that danielcaires[m] has made this change in under issue 887301, so it looks like this won't be a | 04:00 |
| ftl_mason | problem once his changes are merged. | 04:00 |
| ftl_mason | brunomuniz[m]: I also filed a bug report for the failed k8s dashboard installation, when using the pybox method. The URL for the bug report is https://bugs.launchpad.net/starlingx/+bug/2027722 | 04:02 |
| haoiii | Okay, now I am back to work. Got past yesterdaÿ́ | 07:37 |
| haoiii | yesterday's error. Here is a new one: https://paste.opendev.org/show/b67t75oKzMTJ7xOXRFz7/ | 07:37 |
| haoiii | My controller is now unlocker, only problem is that it is degraded instead of avaliable | 07:55 |
| danielcaires[m] | haoiii: I may be wrong on this one but I think that when a controller is degrated it means that there are active alarms for it, you can check through the cli or using the horizon dashboard | 11:00 |
| haoiii | what is the horizon dashbord url? | 12:24 |
| danielcaires[m] | If you didn't set any other port it should be in localport:8080 | 12:25 |
| haoiii | on the VM i guess? possible to access from the outside? | 12:25 |
| danielcaires[m] | the script automatically sets a port forwarding using port 8080, so from your browser you can see the horizon dashboard | 12:27 |
| haoiii | on the NATnetwork? | 12:28 |
| danielcaires[m] | yes, but you can check, if it didn't create you can create manually. For guest IP you use the floating IP and guest port 8080 | 12:30 |
| haoiii | I am sorry for all the bad questions, but where can I find the IP? | 12:32 |
| danielcaires[m] | haha don't worry, the floating IP is in the localhost.yml file | 12:33 |
| haoiii | For some reason my VM only have an ansible.log and lab_setup files. Cannot find the localhost.yml I have seen in previous tries | 12:37 |
| danielcaires[m] | if that is the case, in the stage config-controller tthe script probably wasn't able to send the file via ssh from the host to the VM, in this case the ansible bootstrap used the defaul values that you can find in /usr/share/ansible/stx-ansible/playbooks/host_vars/bootstrap/default.yml | 12:39 |
| danielcaires[m] | s/tthe/the/, s/defaul/default/, s/host_vars/host\_vars/ | 12:40 |
| brunomuniz[m] | ftl_mason: thanks for the updates! Indeed I noticed this Kubernetes Dashboard error yesterday on one test installation I was doing to help haoiii. Thanks for filing the bug, I'm already on it. | 12:46 |
| brunomuniz[m] | There's a breaking change on the latest version of Kubernetes Dashboard and the instructions just download the latest, hence the failure. This is from their release notes at https://github.com/kubernetes/dashboard/releases/tag/v3.0.0-alpha0: | 13:01 |
| brunomuniz[m] | "Kubernetes Dashboard now requires cert-manager and nginx-ingress-controller to work properly. They will be automatically installed with the Helm chart. In case you already have them installed, simply set --set=nginx.enabled=false and --set=cert-manager.enabled=false when installing the chart to disable installation of those dependencies." | 13:01 |
| brunomuniz[m] | Actually, the excerpt above is from https://artifacthub.io/packages/helm/k8s-dashboard/kubernetes-dashboard | 13:21 |
| brunomuniz[m] | For information, for anyone trying to manually setup the Kubernetes Dashboard currently, if you set the helm chart version to 6.0.8, it corresponds to the Kubernetes Dashboard version 2.7.0, which works just fine. | 18:28 |
| brunomuniz[m] | It's just a matter of adding --version 6.0.8 to the "helm install" or "helm upgrade -i" command. | 18:28 |
| * brunomuniz[m] uploaded an image: (52KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/sIHbsgqqeeONHUAZWGgYlWlC/image.png > | 18:30 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
