Development and testing with MicroShift¶
This document provides instructions on how to deploy a MicroShift instance on a CentOS 9 Stream host, from your development computer.
We use MicroShift as the target OpenShift instance for SF-Operator when deploying, developing locally, or testing in our CI.
The deployment will be performed via the sf-operator CLI.
Requirements¶
Host¶
Here are the minimal and recommended specs for your MicroShift host:
| Property | Minimum | Recommended |
|---|---|---|
| CPUs/vCPUS | 2 | 4 |
| RAM | 8GB | 16GB |
| HDD | 40GB | 100GB |
| OS | RHEL 9.4 | RHEL 9.4 |
You must also be able to reach the following ports on the MicroShift host:
- TCP/6443
- TCP/22 (SSH)
Access your machine via SSH, and then ensure that your user has sudo privileges:
Ensure that your system is registered and up to date:
sudo rhc connect --activation-key <my-key> --organization <my-org-id>
# pin to 9.4
sudo subscription-manager release --set 9.4
sudo dnf update -y
sudo reboot
Note that you can register and connect to https://console.redhat.com/insights/connector/activation-keys to get an activation key for your RHEL machine.
Pull Secret¶
MicroShift requires a pull secret to access its container registry.
You can generate a pull secret here and copy it to your clipboard.
export OS_PULL_SECRET="<paste-the-pull-secret-content-here>"
cat << EOF > ~/openshift-pull-secret.yaml
openshift_pull_secret: ${OS_PULL_SECRET}
EOF
Install MicroShift¶
From your MicroShift machine:
cd sf-operator
hack/microshit/setup-microshift.sh localhost
ℹ️ This command logs into /home/fboucher/.cache/setup-microshift.log
▶️ == Running preparation steps ==
⏳ Running ensure_pull_secret ... ✅
⏳ Running ensure_basic_tools ... ✅
⏳ Running ensure_ansible_galaxy_collections ... ✅
⏳ Running ensure_ansible_inventory ... ✅
⏳ Running ensure_microshift_ansible_role ... ✅
▶️ == Deploying MicroShift on cloud-user@microshift.dev (~5 minutes) ==
⏳ Running ensure_microshift ... ✅
⏳ Running ensure_local_kubeconfig ... ✅
🚀 Deploying Microshift done 🚀
To access the deployment, run: KUBECONFIG=~/.kube/microshift-config kubectl -n sf get pods
Once the deployment has ended successfully, you are now ready to deploy and hack the SF-Operator. Congratulations!
Note that the script can also be run from a remote machine with (where