Helpers for the NSX-T BETA integration with Openshift 4.3
1 Destroy machines using terraform terraform destroy python nsx_policy_cleanup.py --mgr-ip=10.114.209.48 -u admin -p VMwareVMware1! -c ocp4 --no-warning -r 2 run the first script, that will generate the ignition files, and copy over NCP manifests and create manifests. ocp4_script.sh 4 You need to copy the contents of master.ign and worker.ign to terraform.tfvars cat vsphere/master.ign >> installer/upi/vsphere/terraform.tfvars cat vsphere/worker.ign >> installer/upi/vsphere/terraform.tfvars 5 Run terraform apply -auto-approve 6 Ping all machines and make sure they are pingable 7 run the copy and load script ocp4_script2.sh python tag_policy_ports.py While the files are being copied - you need to update tags for the logical ports. #On second terminalL
do export KUBECONFIG=/root/openshift43/vsphere/auth/kubeconfig
9 Set resolving for api to the bootstrap
10 run openshift-install --dir=/root/openshift43/vsphere wait-for bootstrap-complete --log-level=DEBUG
./check_network_status.sh
11 Set api to control-nodes
12 run terraform apply -auto-approve -var 'bootstrap_complete=true'
oc get nodes
oc get pods -n nsx-system
oc get pods --all-namespaces
oc patch configs.imageregistry.operator.openshift.io cluster --type merge --patch '{"spec":{"storage":{"emptyDir":{}}}}'
watch -n5 oc get co
## when all completes
openshift-install --dir=/root/openshift43/vsphere wait-for install-complete --log-level=DEBUG
If API connection refused:
ssh to bootstrap
sudo su -
mv /etc/kubernetes/manifests/* /tmp/
systemctl restart bootkube