Load Balancing Policy¶

Overview¶

The Yappes Load Balancer allows you to distribute incoming API requests across multiple backend servers so you can achieve optimal resource utilization, maximize throughput, minimize response time, and avoid server overload.

This enterprise-grade feature enables intelligent traffic distribution using sophisticated algorithms like Round Robin, IP Hash, and Least Connection, ensuring your APIs remain highly available and performant even under heavy load conditions. The Load Balancer integrates seamlessly into the Yappes API management workflow, providing two flexible implementation paths:

You can configure it during API onboarding for immediate load distribution
Or create it separately through the Gateway Policies page

By implementing load balancing, you'll improve application reliability, enable horizontal scaling of your backend infrastructure, and provide better user experiences through reduced latency and increased uptime. This feature fits naturally into typical Yappes workflows by acting as a gateway policy that intercepts API traffic before it reaches your backend servers, intelligently routing each request based on your chosen algorithm and server availability.

Prerequisites¶

Before implementing Load Balancer on the Yappes platform, ensure you have:

Required Access & Permissions¶

Active Yappes Enterprise account with API Manager access
Permission to create and manage gateway policies
Access to onboard new APIs or modify existing ones

Setup Requirements¶

At least 2 backend servers configured and accessible
Server details ready: hostnames, IP addresses, protocols, and ports
Understanding of your API traffic patterns for algorithm selection

Custom error and message

For error scenarios, you can override the default HTTP status code, error code, and message. See Custom Error and Message.

Step-by-Step Implementation¶

Method 1: Creating Load Balancer During API Onboarding¶

Step 1: Navigate to API Management¶

Click Manage APIs in the main dashboard
Select My APIs from the submenu
You should now see your API management interface

Step 2: Initiate API Onboarding¶

Follow onboarding steps.

Step 3: Attach Load Balancer Policy¶

Click the Attach Policy button
In the Policy Name field, enter a descriptive name (e.g., Newloadbalancer)
Add a detailed description in the text area field to document the policy purpose

Step 4: Select Load Balancing Algorithm¶

Click the Algorithm dropdown field
Choose from the available options:
1. Round Robin: Sequential distribution
2. IP Hash: Client IP-based routing
3. Least Connection: Routes to the least busy server
Select the algorithm that best fits your use case

Step 5: Add Backend Servers¶

Click the Add Server button
For each server, enter:
1. Host name: Server hostname or domain
2. Protocol: HTTP or HTTPS
3. IP Address: Server IP address
4. Port: Server port number
Enter Health Check details:
- Protocol: HTTP GET / TCP / PING
a. If HTTP GET is selected:
- Check Interval: Frequency of the health check (seconds)
- Check Timeout: Max wait time for a response (seconds)
- Path: Endpoint path to verify (e.g., /health)
- Expected Status Code: HTTP status code considered healthy (e.g., 200)

b. If TCP is selected:
- Check Interval: Frequency of the health check (seconds)
- Check Timeout: Max wait time before marking host as unhealthy
- Port Number: Port to test connectivity on

c. If PING is selected:
- Check Interval: Frequency of the health check (seconds)
- Check Timeout: Max wait time before declaring failure (seconds)
- Packet Size: Size of each ICMP packet (bytes)
- Packet Count: Number of packets sent per check
Click Add Server to verify the server status
If the server is active, it will be added to the list
If the server is inactive, check that the provided details are correct and confirm that the server is running
Repeat for all backend servers (two servers are mandatory for Load Balancing)
Review your server list

Step 6: Create and Attach Policy¶

Review all policy settings
Click Create Policy
Preview the complete configuration
Expected Result: Load Balancer policy is created and attached to your API

Method 2: Creating Load Balancer from Gateway Policies Page¶

Step 1: Navigate to Gateway Policies¶

Locate Gateway Control → Gateway Policies in the main navigation menu
The policies management interface will display

Step 2: Create Gateway-Level Policy¶

Click Create policy from the gateway page
Set the Control Level to API Level
In the Policy Type dropdown, select Load Balancing Policy

Step 3: Attach Load Balancer Policy¶

In the Policy Name field, enter a descriptive name (e.g., Newloadbalancer)
Add a detailed description in the text area field to document the policy purpose
You'll now configure the load balancing settings

Step 4: Select Load Balancing Algorithm¶

Click the Algorithm dropdown field
Choose from the available options:
1. Round Robin: Sequential distribution
2. IP Hash: Client IP-based routing
3. Least Connection: Routes to the least busy server
Select the algorithm that best fits your use case

Step 5: Add Backend Servers¶

Click the Add Server button
For each server, enter:
1. Host name: Server hostname or domain
2. Protocol: HTTP or HTTPS
3. IP Address: Server IP address
4. Port: Server port number
Enter Health Check details:
- Protocol: HTTP GET / TCP / PING
a. If HTTP GET is selected:
- Check Interval: Frequency of the health check (seconds)
- Check Timeout: Max wait time for a response (seconds)
- Path: Endpoint path to verify (e.g., /health)
- Expected Status Code: HTTP status code considered healthy (e.g., 200)

b. If TCP is selected:
- Check Interval: Frequency of the health check (seconds)
- Check Timeout: Max wait time before marking host as unhealthy
- Port Number: Port to test connectivity on

c. If PING is selected:
- Check Interval: Frequency of the health check (seconds)
- Check Timeout: Max wait time before declaring failure (seconds)
- Packet Size: Size of each ICMP packet (bytes)
- Packet Count: Number of packets sent per check
Click Add Server to verify the server status
If the server is active, it will be added to the list
If the server is inactive, check that the provided details are correct and confirm that the server is running
Repeat for all backend servers (two servers are mandatory for Load Balancing)
Review your server list

Step 6: Create and Attach Policy¶

Review all policy settings
Click Create Policy
Preview the complete configuration
Expected Result: Load Balancer policy is created

Understanding the Interface¶

Main Policy Configuration Panel¶

The Load Balancer interface consists of several key sections:

Section	Description	Purpose
Policy Name	Text field at the top	Identifies your load balancer policy uniquely
Algorithm Selector	Dropdown menu	Determines traffic distribution method
Server List	Table/grid view	Displays all configured backend servers
Add Server Button	Action button	Opens server configuration dialog
Test Connection	Validation tool	Verifies backend server availability

Configuration and Settings¶

Load Balancing Algorithms: Best Practices¶

Algorithm	Best Practices	Use Cases	Considerations
Round Robin	Use when all backend servers have similar capacity and performance. Avoid for stateful apps.	Stateless applications, uniform workloads, microservices with similar resource profiles	Not suitable if servers differ in capacity or if session persistence is required
IP Hash	Use when session persistence ("sticky sessions") is needed. Ensure a wide range of client IPs.	Applications requiring user session stickiness, e.g., shopping carts, user dashboards	May cause uneven load if client IP diversity is low; not ideal for highly dynamic IP scenarios
Least Connection	Use for workloads with variable or unpredictable request processing times.	Real-time applications, chat servers, APIs with long-lived or variable connections	Requires connection tracking, which adds overhead; best when connection duration varies

Verification and Testing¶

Confirming Successful Implementation¶

Check Policy Status¶

Navigate to your API's policy list
Confirm all servers display "Healthy" status

Test Traffic Distribution¶

Send multiple test requests to your API endpoint
Monitor server logs to verify requests reach different backends
Check response headers for server identification

Validate Algorithm Behavior¶

Round Robin: Requests should cycle through servers sequentially
IP Hash: Same client IP should always reach the same server
Least Connection: New connections should favor less-loaded servers

Success Indicators¶

All backend servers receive traffic
Response times are consistent
No connection errors or timeouts
API endpoints respond correctly

Troubleshooting¶

Common Issues and Solutions¶

Issue	Symptoms	Solution
Connection Test Fails	Error message when testing API connection	• Verify API hostname is correct • Check network connectivity • Ensure backend servers are running
Policy Not Applied	Requests not being distributed	• Confirm policy is attached to API • Check policy status is "Active" • Submit/deploy policy if pending
Uneven Distribution	One server gets most traffic	• Review algorithm selection • Check server health status • Verify all servers are properly configured
Backend Connection Errors	502/503 errors returned	• Validate server IP addresses and ports • Check protocol matches (HTTP/HTTPS) • Verify firewall rules
Session Loss	Users losing session state	• Switch to IP Hash algorithm for session persistence • Configure sticky sessions

Error Messages Guide¶

"Failed to connect to backend server": Check server availability and network path
"Policy deployment failed": Review policy configuration for missing required fields

Pro Tip

Start with Round Robin for simple deployments, then optimize based on your specific traffic patterns and requirements. The Yappes platform makes it easy to adjust algorithms without downtime.