User need to update required vales in var.tfvars file such as username and password of IBM intranet credentials to get access to the PowerVC.
Note It is recommended to copy the var.tfvars into data/ and use the data/var.tfvars with your terraform commands.
auth_url = "https://host.net:5000/v3/"
user_name = "***@ibm.com"
password = "***"
If you are using a selinux enabled bastion, please set chcon -R -t httpd_sys_content_t /var/www/html/ignition/worker.ign
where worker.ign is the file for the latest ignition.
terraform plan -var-file=data/var.tfvars
terraform apply -var-file=data/var.tfvars
Get the MAC Address and IP Address of the VM created on PowerVC. Update /etc/dhcp/dhcpd.conf with the lease/addresses as shown below.
host workerp-0 { hardware ethernet fa:16:3e:0f:95:6b; fixed-address 10.20.177.243; }
-
Restart the dhcpd.
systemctl restart dhcpd
-
Restart the Worker node for ignition to start.
-
Approve the CSR certificate to add the worker to intel cluster.
oc get csr
oc adm certificate approve <certificate-name>
- If ignition failed, Once you have the VM console, run the following commands:
systemctl --failed
journalctl -u ignition-fetch
- Ignition happens in a couple of steps
ignition -platform openstack -stage mount -log-to-stdout
then fetch
ignition --platform openstack -stage fetch -log-to-stdout
sudo journalctl -u kubelet
- Ignition file properties should has
/etc/hostname
,/etc/resolv.conf
,passwd
section with proper values.
- If you see the error with
no such host
error in thesudo journalctl -u kubelet
add the DNS service points to the right host/ip of the ignition IP. - Using the IBM Cloud Internet Service or IBM Cloud DNS Service link, add a new record to resolve the IP which points to domain.
- Reboot the worker node to flush the dns cache to resolve the DNS and CSR will appear for the approval. once CSR are approved worker will get added to cluster.