GCP Clustering Overview #

This article describes how to set up advanced load balancing and high availability using the RELIANOID Load Balancer in Google Cloud Platform (GCP).
We’ll demonstrate the full process of deploying and configuring a RELIANOID Cluster with a virtual IP (VIP) for failover and a farm of backend servers for HTTP load balancing.
The objective is to ensure that if one node (master) fails or reboots, the virtual IP and services automatically fail over to the secondary node, providing uninterrupted application availability.

Cloud Load Balancer Clustering Environment #

We are going to demonstrate how it can be implemented in Google Cloud Platform (GCP) using native GCP services. The following diagram illustrates the architecture we aim to deploy a web load balancing setup using RELIANOID Application Delivery Controller in GCP for high availability and failover.

Architecture Diagram #

Each RELIANOID VM will be configured with one network interface (nic0) connected to a VPC network and assigned both an internal and a static external IP.

• LB1:
• Internal IP: 10.0.1.9
• External IP: 34.122.145.90
• LB2:
• Internal IP: 10.0.1.8
• External IP: 35.233.210.70

An additional secondary internal IP (VIP) will be created on LB1’s network interface:

• VIP (for Virtual Service):10.0.1.11
• This IP will move between nodes automatically (master/slave) during failover.
• It will be published through an external static IP, e.g. 34.172.119.80, for user access to the virtual service.

The backend servers (real applications) are deployed in the same private network (10.0.1.0/24) and are not directly exposed to the Internet.

Implement Cloud Load Balancer Clustering in GCP #

1. Deploy RELIANOID Load Balancer Instances #

You can deploy the RELIANOID Load Balancer image directly from the Google Cloud Marketplace.

1. Go to Google Cloud Console → Compute Engine → VM Instances → Create Instance.
2. Select the RELIANOID Load Balancer image under “Custom images” then “Marketplace”.
3. Configure:
• Machine type: e2-micro (sufficient for testing; use n2-standard for production)
• Region / Zone: same region for both nodes
• Network Interface:
• Network: relianoid-vpc
• Subnetwork: relianoid_subnet
• External IP: Static (assign a new static IP)
• Firewall:Allow TCP 22 (SSH), TCP 444 (Web GUI), and TCP 80 (for HTTP services).
4. Click Advanced options → Security → Access Scopes
• Choose Allow full access to all Cloud APIs or set custom service account (recommended).
• Assign a Service Account with Compute Admin and Compute Network Admin roles.

Deploy both nodes (LB1 and LB2) in the same VPC and subnet.

2. Enable and Configure Service Accounts (Identity Access) #

RELIANOID uses the Google Cloud API for failover management, similar to Azure Managed Identities.

• Go to IAM & Admin → Service Accounts.
• Create a new service account named relianoid-cluster-sa.
• Assign the following roles:
• Compute Instance Admin (v1)
• Compute Network Admin
• Service Account User

Attach this service account to both LB1 and LB2 VMs (edit → change service account).

Then verify access from each node:

sudo systemctl status google-guest-agent
gcloud auth list
gcloud projects get-iam-policy <YOUR_PROJECT_ID>

If the agent is not installed, run:

sudo apt-get install google-cloud-sdk google-compute-engine

3. Configure Network Interfaces and Virtual IP #

In GCP, secondary internal IPs can be added manually via the Console or CLI.

On LB1 (Master Node):

• Go to VPC Network → VM instances → LB1 → Edit → Network Interface → Add Secondary Internal IP
• Assign:
• IP: 10.0.1.11
• Name: vip-lb
• Type: Static

On LB2 (Backup Node):

• Do not assign this secondary IP.
• The RELIANOID cluster service will automatically reassign it during failover.

4. Security Rules (Firewall Configuration) #

In VPC Network → Firewall, create rules to allow:

Protocol	Port	Purpose
TCP	22	SSH access
TCP	444	Web GUI
TCP	80	Backend HTTP services

Apply these rules to instances tagged as relianoid-lb.

5. Activate the RELIANOID License #

Access each node:

• LB1: https://34.122.145.90:444
• LB2: https://35.233.210.70:444

Use:

• Username: root
• Password: Instance ID

Retrieve hostname and certificate key → Visit
https://www.relianoid.com/activate-enterprise-edition-cloud-evaluation/
Upload the downloaded .pem license file to each node.

6. Configure RELIANOID Cluster #

From LB1 Web GUI → System → Cluster, fill out:

Field	Value
Local IP	Select nic0 (10.0.1.9)
Remote IP	10.0.1.8
Remote Node Password	(LB2 instance ID)

Click Generate.

LB1 becomes Master, LB2 becomes Slave automatically.

Create a load balancing service in GCP #

Now, let’s configure a simple web load-balancing farm.
Go to:
LSLB → Farms → Create Farm

Parameters:

• Farm Name: webfarm
• VIP: 10.0.1.11
• Port: 80
• Algorithm: Round Robin
• Persistence: Source IP (60 sec)
• Health Check: check_tcp

Add backend servers:

• Backend 1: 10.0.1.20:80
• Backend 2: 10.0.1.21:80

Apply configuration.

Now access the virtual service at:

http://34.172.119.80/

Traffic will be distributed between your backend servers via the RELIANOID cluster.

To test failover:

• Reboot the master node (LB1).
• Observe that LB2 becomes active, taking control of the VIP (10.0.1.11).
• Connection through public IP remains uninterrupted.

Troubleshooting: This Host Has Not Assigned a Service Account #

If you receive the message:

    “This host has not assigned a Service Account. Please configure it in GCP console”

It indicates the GCP service account is not configured or the Google Guest Agent is missing.

To verify:

sudo systemctl status google-guest-agent
gcloud auth login --brief
gcloud auth print-access-token

If the agent or SDK is missing:

sudo apt-get install google-cloud-sdk google-compute-engine

Ensure that:

• The instance has a valid Service Account attached.
• The Service Account has the Compute Admin and Network Admin roles.

Enjoy Advanced Load Balancing and Clustering in GCP with RELIANOID! #