fix: resolve okd 3.10 installation issues (#77)

* fix: resolve okd 3.10 installation issues

This bugfix resolves two issues. The first is that `openshift_hostname`
is no longer a valid configuration value. The second is that `systemctl
restart docker` seems to hang on the openshift nodes, it has been
replaced with a shutdown/restart.

Fixes #75, fixes #72, fixes #71.

* chore: cleaner comments in scripts

* chore: improve makefile output

inventory.template.cfg

@@ -52,14 +52,14 @@ openshift_node_groups=[{'name': 'node-config-master', 'labels': ['node-role.kube
 # is a bug with the installer when using the AWS cloud provider.
 # Note that we use the master node as an infra node as well, which is not recommended for production use.
 [masters]
-${master_hostname} openshift_hostname=${master_hostname}
+${master_hostname}
 
 # host group for etcd
 [etcd]
-${master_hostname} openshift_hostname=${master_hostname}
+${master_hostname}
 
 # all nodes - along with their openshift_node_groups.
 [nodes]
-${master_hostname} openshift_hostname=${master_hostname} openshift_node_group_name='node-config-master-infra' openshift_schedulable=true
-${node1_hostname} openshift_hostname=${node1_hostname} openshift_node_group_name='node-config-compute'
-${node2_hostname} openshift_hostname=${node2_hostname} openshift_node_group_name='node-config-compute'
+${master_hostname} openshift_node_group_name='node-config-master-infra' openshift_schedulable=true
+${node1_hostname} openshift_node_group_name='node-config-compute'
+${node2_hostname} openshift_node_group_name='node-config-compute'

makefile

@@ -1,6 +1,6 @@
 infrastructure:
 	# Get the modules, create the infrastructure.
-	terraform init && terraform get && terraform apply
+	terraform init && terraform get && terraform apply -auto-approve
 
 # Installs OpenShift on the cluster.
 openshift:
@@ -17,9 +17,12 @@ openshift:
 	cat install-from-bastion.sh | ssh -o StrictHostKeyChecking=no -A ec2-user@$$(terraform output bastion-public_ip)
 
 	# Now the installer is done, run the postinstall steps on each host.
-	cat ./scripts/postinstall-master.sh | ssh -A ec2-user@$$(terraform output bastion-public_ip) ssh master.openshift.local
-	cat ./scripts/postinstall-node.sh | ssh -A ec2-user@$$(terraform output bastion-public_ip) ssh node1.openshift.local
-	cat ./scripts/postinstall-node.sh | ssh -A ec2-user@$$(terraform output bastion-public_ip) ssh node2.openshift.local
+	# Note: these scripts cause a restart, so we use a hyphen to ignore the ssh
+	# connection termination.
+	- cat ./scripts/postinstall-master.sh | ssh -A ec2-user@$$(terraform output bastion-public_ip) ssh master.openshift.local
+	- cat ./scripts/postinstall-node.sh | ssh -A ec2-user@$$(terraform output bastion-public_ip) ssh node1.openshift.local
+	- cat ./scripts/postinstall-node.sh | ssh -A ec2-user@$$(terraform output bastion-public_ip) ssh node2.openshift.local
+	echo "Complete! Wait a minute for hosts to restart, then run 'make browse-openshift' to login."
 
 # Destroy the infrastructure.
 destroy:

scripts/postinstall-master.sh

@@ -6,9 +6,14 @@ sudo su
 
 # Create an htpasswd file, we'll use htpasswd auth for OpenShift.
 htpasswd -cb /etc/origin/master/htpasswd admin 123
+echo "Password for 'admin' set to '123'"
 
 # Update the docker config to allow OpenShift's local insecure registry. Also
 # use json-file for logging, so our Splunk forwarder can eat the container logs.
 # json-file for logging
 sed -i '/OPTIONS=.*/c\OPTIONS="--selinux-enabled --insecure-registry 172.30.0.0/16 --log-driver=json-file --log-opt max-size=1M --log-opt max-file=3"' /etc/sysconfig/docker
-systemctl restart docker
+echo "Docker configuration updated..."
+
+# It seems that with OKD 3.10, systemctl restart docker will hang. So just reboot.
+echo "Restarting host..."
+shutdown -r now "restarting post docker configuration"

scripts/postinstall-node.sh

@@ -8,4 +8,8 @@ sudo su
 # use json-file for logging, so our Splunk forwarder can eat the container logs.
 # json-file for logging
 sed -i '/OPTIONS=.*/c\OPTIONS="--selinux-enabled --insecure-registry 172.30.0.0/16 --log-driver=json-file --log-opt max-size=1M --log-opt max-file=3"' /etc/sysconfig/docker
-systemctl restart docker
+echo "Docker configuration updated..."
+
+# It seems that with OKD 3.10, systemctl restart docker will hang. So just reboot.
+echo "Restarting host..."
+shutdown -r now "restarting post docker configuration"