Creating and configuring a Load Balancer
This section describes how to configure a third-party load balancer for an Adeptia Rancher infrastructure deployment. The load balancer distributes incoming traffic across the application nodes, provides high availability, and ensures that requests are routed only to healthy backend servers.
The configuration process generally involves the following high-level steps:
- Creating or configuring a Virtual IP (VIP) on the load balancer.
- Defining frontend listener settings for incoming client traffic.
- Configuring backend server pools consisting of RKE2 nodes.
- Setting up health probes to monitor backend node availability.
- Creating load-balancing rules to route traffic between frontend listeners and backend servers.
Prerequisites
Before configuring the load balancer, ensure that the following requirements are met:
| Item | Description |
|---|---|
| Load Balancer | A supported hardware or software load balancer capable of Layer 4 (TCP) |
| Rancher Nodes | Linux virtual machines or physical servers hosting Adeptia Rancher |
| Virtual IP (VIP) | A dedicated IP address exposed by the load balancer for client access |
| DNS Records | DNS entries that resolve hostname to the load balancer VIP |
| Firewall Rules | Network access between the load balancer and Rancher nodes |
| TLS Certificate | Required if SSL/TLS termination is performed at the load balancer |
Step 1: Configure the Frontend Listener
Create a frontend listener that accepts incoming client requests.
Recommended settings:
| Setting | Value |
|---|---|
| Listener Name | User-defined |
| IP Version | IPv4 |
| Frontend IP | Load Balancer VIP |
| Protocol | TCP or HTTPS |
| Port | 443 (recommended) |
Step 2: Configure the Backend Server Pool
Create a backend server pool containing all the Rancher nodes.
Example:
| Node | IP Address |
|---|---|
| Node 1 | 10.32.4.6 |
| Node 2 | 10.32.4.7 |
| Node 3 | 10.32.4.10 |
 | Use the private IP addresses of the Rancher nodes. Do not use public IP addresses. |
|---|
Step 3: Configure Health Probes
Create health probes to monitor the availability of backend nodes.
Recommended settings:
| Setting | Value |
|---|---|
| Protocol | TCP or HTTPS |
| Port | 443 (HTTPS) |
| Probe Interval | 5–30 seconds |
| Unhealthy Threshold | 2–3 failures |
If your load balancer supports HTTP or HTTPS health checks, configure an application-specific health endpoint as recommended by your deployment architecture.
Step 4: Configure Load-Balancing Rules
Create load-balancing rules that associate the frontend listener with the backend server pool.
Recommended settings:
| Setting | Value |
|---|---|
| Frontend Listener | Previously created frontend listener |
| Backend Pool | Rancher server pool |
| Protocol | TCP or HTTPS |
| Frontend Port | 443, 9345, 6443 |
| Backend Port | 443, 9345, 6443 |
| Health Probe | Configured health probe |
| Session Persistence | None |
| Idle Timeout | 5 minutes |
| Connection Reset | Disabled (recommended) |
Step 5: Configure DNS
Create DNS records that resolve the application and Rancher UI hostnames to the load balancer VIP.
Example:
| Hostname | Type | Value |
|---|---|---|
| automate.mycompany.org | A | Load Balancer VIP |
| rancher.mycompany.org | A | Load Balancer VIP |
Step 6: Validate Connectivity
Verify that the load balancer can communicate with all backend nodes and that health probes report them as healthy.
Confirm that:
- All backend nodes are marked healthy by the load balancer.
- Traffic is distributed across the configured backend pool.
- The application is accessible using the configured hostname.
- SSL/TLS certificates are presented correctly (if SSL termination is enabled).