deploy.sh is a Bash script for deploying OpenShift Origin in Azure, using the template provided by Microsoft, with a little more automation and repeatability.
The ARM templates for deploying OpenShift Origin and OpenShift Container Platform on Azure are available on http://aka.ms/OpenShift.
You will need a Linux system (tested on Ubuntu 16.04 LTS) or Bash on Windows Subsystem for Linux, or a Mac.
Then you need an SSH Key, but by default the script creates one for you. Remember that the private key will be uploaded some of the VMs, so don't use your own key, the best is to let the script handle that for you.
And if not already done, install Azure CLI 2.0, login and select the subscription:
az login
az account set --subscription <subscription name>
You don't need to change anything in the script! Only change something if you need to.
You may change:
- the variables at the top of the script
- the parameters in the JSON portion of the script
Variables:
DEMO_NAME: name used by default for the log file, the SSH key file, the resource group, the Key Vault and the parameters file. Since the Key Vault name must be unique, the use of$RANDOMin this name is convenient.LOCATION: Azure region (full list:az account list-locations).SSH_KEY_FILE: SSH private key file name. The script will create a new one if it doesn't exist.USER_NAME: a user name for the VMs and for OpenShiftUSER_PASSWORD: the user password for OpenShift
Parameters for the JSON parameters file: the most likely to be changed are VM sizes and the numbers of VM.
masterVmSize,nodeVmSize,infraVmSize: for instance,Standard_DS2_v2. These parameters are case sensitive.masterInstanceCount: 1, 3 or 5infraInstanceCount: 1, 2 or 3nodeInstanceCount: 1 to 30
Just run ./deploy.sh to run the deployment. The script will automatically:
- create the new SSH key,
- create the resource group,
- create a Key Vault in the resource group, and store the SSH key as a secret in the vault,
- create a service principal and assign the Contributor role on the resource group,
- generate a parameters file for the ARM template,
- initiate the deployment.
All will be logged in a .log file with the demo name as its base name.
Once the script is finished, the deployment should be in progress. You can track its progress with this command:
az group deployment list -g <resource group name>
Where the resource group name is the same as the demo name.
Example:
$ az group deployment list -g demo-openshift-24843
Name Timestamp State
------------------- -------------------------------- ---------
azuredeploy 2018-10-11T19:20:09.324631+00:00 Running
infraVmDeployment2 2018-10-11T19:22:05.764170+00:00 Succeeded
nodeVmDeployment2 2018-10-11T19:22:07.437872+00:00 Succeeded
nodeVmDeployment1 2018-10-11T19:22:08.142167+00:00 Succeeded
infraVmDeployment0 2018-10-11T19:22:11.415120+00:00 Succeeded
nodeVmDeployment0 2018-10-11T19:22:20.528948+00:00 Succeeded
infraVmDeployment1 2018-10-11T19:22:21.930035+00:00 Succeeded
masterVmDeployment1 2018-10-11T19:22:35.780993+00:00 Succeeded
masterVmDeployment0 2018-10-11T19:22:35.881848+00:00 Succeeded
masterVmDeployment2 2018-10-11T19:22:35.933034+00:00 Succeeded
OpenShiftDeployment 2018-10-11T19:30:03.517852+00:00 Running
When the deployment is completely finished (it may take 50 minutes to complete), you can get the deployment outputs with this command:
az group deployment show -n azuredeploy -g <resource group name> --query [properties.outputs] -o json
Example:
$ az group deployment show -n azuredeploy -g demo-openshift-24843 --query [properties.outputs] -o json
[
{
"openShift Console Url": {
"type": "String",
"value": "https://masterdnsbz7f4nm3ld4bg.francecentral.cloudapp.azure.com/console"
},
"openShift Infra Load Balancer FQDN": {
"type": "String",
"value": "infradnsv3nonlsqjrrb4.francecentral.cloudapp.azure.com"
},
"openShift Master SSH": {
"type": "String",
"value": "ssh -p 2200 openshift@masterdnsbz7f4nm3ld4bg.francecentral.cloudapp.azure.com"
}
}
]
Two interesting outputs are the OpenShift Console URL, and the OpenShift Master SSH command. When connecting with SSH, don't forget to specify the SSH key from before, with the -i parameter:
ssh -i <SSH key file> -p 2200 <username>@<master FQDN>
For instance:
$ ssh -i ~/.ssh/demo-openshift-24843_rsa -p 2200 openshift@masterdnsbz7f4nm3ld4bg.francecentral.cloudapp.azure.com
The authenticity of host '[masterdnsbz7f4nm3ld4bg.francecentral.cloudapp.azure.com]:2200 ([40.89.139.76]:2200)' can't be established.
ECDSA key fingerprint is SHA256:Ye5YNOjVA3R12z5BFQVizxapAiH9vie5ZqNokva56to.
Are you sure you want to continue connecting (yes/no)? yes
Warning: Permanently added '[masterdnsbz7f4nm3ld4bg.francecentral.cloudapp.azure.com]:2200,[40.89.139.76]:2200' (ECDSA) to the list of known hosts.
Last login: Thu Oct 11 20:34:00 2018 from demo-openshift-24843-master-0
[openshift@demo-openshift-24843-master-0 ~]$
[openshift@demo-openshift-24843-master-0 ~]$ oc status
In project default on server https://masterdnsbz7f4nm3ld4bg.francecentral.cloudapp.azure.com
https://docker-registry-default.40.89.143.131.nip.io (passthrough) (svc/docker-registry)
dc/docker-registry deploys docker.io/openshift/origin-docker-registry:v3.9.0
deployment #2 failed about an hour ago: config change
deployment #1 deployed about an hour ago - 1 pod
svc/kubernetes - 172.30.0.1 ports 443, 53->8053, 53->8053
https://registry-console-default.40.89.143.131.nip.io (passthrough) (svc/registry-console)
dc/registry-console deploys docker.io/cockpit/kubernetes:latest
deployment #1 deployed about an hour ago - 1 pod
svc/router - 172.30.28.155 ports 80, 443, 1936
dc/router deploys docker.io/openshift/origin-haproxy-router:v3.9.0
deployment #1 deployed about an hour ago - 3 pods
1 info identified, use 'oc status -v' to see details.
[openshift@demo-openshift-24843-master-0 ~]$
Finally, the log file contains the commands you can use to completely delete the demo, by deleting the resource group and the service principal.