oracle-quickstart / oci-wandisco

Terraform module to deploy WANdisco Fusion on Oracle Cloud Infrastructure (OCI)

oci-wandisco

These are Terraform modules that deploy WANdisco on Oracle Cloud Infrastructure (OCI). They are developed jointly by Oracle and WANdisco.

Prerequisites

First off you'll need to do some pre-deploy setup. That's all detailed here.

Clone the Module

Now, you'll want a local copy of this repo. You can make that with the commands:

git clone https://github.com/oracle-quickstart/oci-wandisco.git
cd oci-wandisco
cd terraform1
ls

That should give you this:

Setup

The goal is the keep OCI Object Storage data in sync contained in two regions. For replication with WANdisco Fusion, we need to provision at least two servers - one in each region.

Buckets and Regions

In this example, we will setup object storage data replication between two OCI regions: us-ashburn-1 and us-phoenix-1.

To configure the software, you will need two storage containers (buckets). These buckets will contain the Fusion metadata, so let's name them both fusion_metadata. So for each region, we must make a bucket with this name. Other storage buckets may contain the actual user data. If these do not exist, you can make them later before you establish the replication rules in Fusion.

If you have not set up your OCI user account for use with an S3 API or cli, make your Customer Secret Keys to create an access key and secret key. This is described here. These extra keys are used as shown below in addition to the user-specific keys you discovered in the pre-req exercise.

Before you perform the Terraform tasks, you will need to update the variables.tf file with values specific to your account and needs. You just need enter the variable information in the file in terraform1 as the second version is a linked file. The region is specified in the region.tf file for each instance. Again, information about the various OCI variables and where to obtain them is described in more detail the previous pre-requisite section. Here is list of the variables you will need to supply at this point:

### OCI Profile
variable "tenant"               {default = "your_tenancy_name"}  
variable "tenancy_ocid"         {default = "ocid1.tenancy.oc1..key"}
variable "compartment_ocid"     {default = "ocid1.tenancy.oc1..key"}
variable "user_ocid"            {default = "ocid1.user.oc1..key"}
variable "fingerprint"          {default = "key"}

variable "ssh_public_key"       { <your_ssh_public_key> }

# Object Storage
variable "bucket"               {default = "fusion_metadata"}
variable "accesskey"            {default = "ocid1.credential.oc1..key"}           
variable "secretkey"            {default = "your_secret_key"}
variable "endpointurl" {
   type = "map" 
   default = { 
     us-phoenix-1 = "https://<your_tenancy_name>.compat.objectstorage.us-phoenix-1.oraclecloud.com"
     us-ashburn-1 = "https://<your_tenancy_name>.compat.objectstorage.us-ashburn-1.oraclecloud.com"
   }
}

Init

So we now need to initialize the each directory with the modules in them. This makes the modules aware of the OCI provider.
You can do this by running these commands:

cd ../terraform1
terraform init

cd ../terraform2
terraform init

This gives the following output:

Plan

Let's make sure the terraform plan looks good:

cd ../terraform1
terraform plan

cd ../terraform2
terraform plan

And this output should look like:

Deploy

If that all looks good as a pre-flight check, we can go ahead and do an apply - which deploys the plan:

cd ../terraform1
terraform apply -auto-approve |tee apply.txt

cd ../terraform2
terraform apply -auto-approve |tee apply.txt 

The terraform apply for each server should each take about 2 minutes to run. Once each process is complete, you'll see something like this:

Make note of the Public IP and URL for each session. But don't worry, this info is also captured in the apply.txt file or by running terraform refresh.

When the terraform apply task has finished, the infrastructure will be deployed and cloud-init scripts run to deploy Fusion on the server. Those scripts will "wrap up" asynchronously to the server provisioning process, meaning the cloud-init process runs on the servers right after they are booted up for the first time. So, it'll be a few more minutes (about 3) before the Fusion application is accessible on each server.

Connect to the UI

When the terraform apply completed, it prints the output of the task - a few lines about each server: (1) the URL to access the Fusion UI and (2) the Public IP and hostname.

Now that you have waited a few minutes, let's try accessing the UI on port 8083 of the public IP for first (Phoenix) server. You should see this:

Now enter the username and password you specified in variables.tf. You should now see the Fusion Dashboard.

At this point, the two servers are not yet configured as a replication pair. We must induct the other server to create a dual-zone membership. After that process has completed (takes a minute or two), we can then create the actual replication policies. These are actual rules or definitions of which buckets to replicate. Fusion will create a proxy server with a virtual object storage bucket that mirrors data into the underlying buckets residing in each region.

Click on the "Nodes Tab" along the top, and then click on the "Induct" Button. If you are on the Phoenix server, enter the Pubic IP of the Ashburn server.

SSH to the Server

These machines are using Oracle Enterprise Linux (OEL). The default login user is opc. You can SSH into the machine with a command like this:

ssh -i ~/.ssh/oci opc@<Public IP Address>

Fusion is installed under /opt/wandisco and has configuration files under /etc/wandisco. You can debug deployments by investigating the cloud-init output file in the home directory and/or look in /var/log/messages. Note: You'll need to be root, so run sudo to be able to read it.

For convenience, on each server you can add the other server name to the hosts file. As generated by the apply task, "tail" the apply.txt files, and cut/paste the IP and hostname.

Add the other host to the server in Phoenix:

 ssh [email protected]
 echo "129.213.53.56 fusion-server.ashburn.fusion.oraclevcn.com" | sudo tee -a /etc/hosts

Add the other host to the server in Ashburn:

 ssh [email protected]
 echo "129.146.162.177 fusion-server.phoenix.fusion.oraclevcn.com" | sudo tee -a /etc/hosts

Add both hosts your own /etc/hosts file (if on say macOS):

 echo "129.146.162.177 fusion-server.phoenix.fusion.oraclevcn.com" | sudo tee -a /etc/hosts
 echo "129.213.53.56   fusion-server.ashburn.fusion.oraclevcn.com" | sudo tee -a /etc/hosts

View the Server in the Console

You can also login to the web console here to view info about the server running in OCI. The click on the instance name "fusion_server" to see its details as shown here:

Destroy the Deployment

When you no longer need the deployment, you can run these commands to destroy the OCI infrastructure you just built.

cd ../terraform1
terraform destroy

cd ../terraform2
terraform destroy

You'll need to enter yes when prompted. Once complete, you'll see something like this:

Thanks for testing out Fusion with Multicloud on OCI!