After working on the HP Chromebox G1 – I discovered that a single 32 GB DDR3 SODIMM was going to cost 3 times what the Chromebox itself cost me to begin with. It quickly became evident my openshift experiment was going to be limited using the Chromebox, so I decided to try on another PC I had available, this was a bee-link GTR5. In addition to the internal SSD, I also added a 1 TB NVME drive.
The chromebox G1’s might be possible to use as a microshift cluster but still waiting on the parts to really determine if that’s possible.
The GTR5 was previously used as a desktop machine running the i3 respin of Fedora. First step was to back up everything and then off to the races with openshift.
I started out following this guide.
https://www.redhat.com/sysadmin/low-cost-openshift-cluster
Installation followed pretty closely, I’m only going to note any special steps I did on my side.
I’m running a pretty simple consumer grade router, but it let me configured the DHCP hostname – I set the GTR5 as “hive.geolaw.loc” and used that in the cluster details.
Cluster Name: hive
Base Domain: geolaw.loc
Copied my ssh .pub and then generated the discovery iso
DNS entries : like I said, I’ve got a cheap consumer class router, does not support adding DNS entries.
So on the machines I plan on accessing the web GUI or ‘oc’ – I plan on just using the following /etc/hosts entries :
$ grep hive /etc/hosts
192.166.29.7 api.hive.geolaw.loc *.apps.hive.geolaw.loc api-int.hive.geolaw.loc
Booting the discovery.iso
I had an existing Ventoy USB drive that I first tried just dropping the iso file into the Ventoy partition – this did not boot properly for me and went to an emergency shell. I then just used dd to write the discovery iso to the thumb drive:
$ sudo dd if=discovery_image_hive.iso of=/dev/sdb bs=1024
Once this finished I rebooted the GTR5 and from the UEFI level selected the USB to boot from.
After booting, the agent.service was failing due to it being unable to pull from the redhat.io registry:
Jun 22 14:26:26 hive podman[17680]: Error: initializing source docker://registry.redhat.io/rhai-tech-preview/assisted-installer-agent-rhel8:v1.0.0-264: unable to retrieve auth token: invalid username/password: unauthorized: Please login to the Red Hat Registry using your Customer Portal credentials. Further instructions can be found here: https://access.redhat.com/articles/3399531
Jun 22 14:26:29 hive podman[17749]: Trying to pull registry.redhat.io/rhai-tech-preview/assisted-installer-agent-rhel8:v1.0.0-264…
To fix this I ssh’d into the openshift installer, su’d to root, and then logged into to registry.redhat.io. Once I logged in, I restarted the agent.service and away it went!
$ ssh core@hive
** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** **
This is a host being installed by the OpenShift Assisted Installer.
It will be installed from scratch during the installation.
The primary service is agent.service. To watch its status, run:
sudo journalctl -u agent.service
To view the agent log, run:
sudo journalctl TAG=agent
** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** **
Last login: Thu Jun 22 14:26:22 2023 from 192.168.29.16
[core@hive ~]$ sudo su –
Last login: Thu Jun 22 14:17:22 UTC 2023 on pts/0
[root@hive ~]# podman login registry.redhat.io
Authenticating with existing credentials for registry.redhat.io
Existing credentials are invalid, please enter valid username and password
Username (|uhc-pool-81ec5a21-635b-4c43-8409-63e45c46ad51): glaw@redhat.com
Password:
Login Succeeded!
[root@hive ~]# systemctl restart agent
The discovered host eventually popped up in the assisted installer and I was able to select my network and continue the install.
The host rebooted several times along the way as it was processing the install.
Watching the console I could see where it was pulling down the containers and starting them.
but again getting the registry errors and the containers going into a ImagePullBackOff state
Jun 22 15:42:07 hive kubenswrapper[2978]: E0622 15:42:07.423504 2978 pod_workers.go:965] “Error syncing pod, skipping” err=”failed to \”StartContainer\” for \”registry-server\” with ImagePullBackOff: \”Back-off pulling image \\\”registry.redhat.io/redhat/certified-operator-index:v4.13\\\”\”” pod=”openshift-marketplace/certified-operators-bb2nx” podUID=0f76c0fa-cb11-436f-9e7e-77357117b313
I tried doing the podman register again, as root, as core, as containers .. no bueno 🙁
Oh well, good first test, will have to retry later.