Server Load Balancing with Nginx

This article assumes you have a basic understanding of Linux server administration and network concepts.

Prerequisites

Before you begin configuring Nginx for load balancing, ensure you have the following:

• **Two or more backend web servers:** These servers will host your actual wiki content. They should be configured to serve the same content. For example, you might have two dedicated servers running your MediaWiki installation.
• **A dedicated load balancer server:** This server will run Nginx and distribute traffic to the backend servers. It should have a public IP address. You can procure powerful dedicated servers with full root access from PowerVPS that are ideal for this role.
• **SSH access to all servers:** You'll need to execute commands on each server.
• **Nginx installed on the load balancer server:** If Nginx is not already installed, you can install it using your distribution's package manager:
```bash sudo apt update && sudo apt install nginx -y # Or for RHEL/CentOS/Fedora sudo dnf update && sudo dnf install nginx -y ```
• **Basic firewall configuration:** Ensure that ports 80 (HTTP) and 443 (HTTPS) are open on your load balancer and backend servers as needed.

Understanding Load Balancing

Load balancing is the distribution of network traffic across multiple backend servers. This offers several benefits:

• **Improved Performance:** By distributing requests, no single server becomes overwhelmed, leading to faster response times for users.
• **Increased Availability and Reliability:** If one backend server fails, the load balancer can redirect traffic to the remaining healthy servers, ensuring continuous service.
• **Scalability:** As your traffic grows, you can easily add more backend servers to handle the increased load without significant downtime.

Nginx is a high-performance web server and reverse proxy that excels at load balancing.

Configuring Nginx as a Load Balancer

The core of Nginx load balancing lies in the `upstream` directive, which defines a group of backend servers.

Step 1: Define the Upstream Group

Edit your Nginx configuration file. Typically, this is located at `/etc/nginx/nginx.conf` or within a file in `/etc/nginx/conf.d/`.

Create a new `upstream` block to define your backend servers. Replace `my_wiki_servers` with a descriptive name for your upstream group and `192.168.1.101:80`, `192.168.1.102:80` with the actual IP addresses and ports of your backend wiki servers.

```nginx # /etc/nginx/nginx.conf or a file in /etc/nginx/conf.d/ http { upstream my_wiki_servers { server 192.168.1.101:80; server 192.168.1.102:80; # Add more servers as needed }

server { listen 80; server_name your_domain.com;

location / { proxy_pass http://my_wiki_servers; proxy_set_header Host $host; proxy_set_header X-Real-IP $remote_addr; proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for; proxy_set_header X-Forwarded-Proto $scheme; } } } ```

• **`upstream my_wiki_servers { ... }`**: This block defines a group of servers named `my_wiki_servers`. Nginx will use this name to refer to the group.
• **`server 192.168.1.101:80;`**: Each `server` directive specifies a backend server's IP address and port.
• **`proxy_pass http://my_wiki_servers;`**: This directive within the `server` block tells Nginx to forward incoming requests to the `my_wiki_servers` upstream group.
• **`proxy_set_header ...`**: These directives are crucial for passing important client information to the backend servers, such as the original `Host` header, the client's real IP address (`X-Real-IP`), a list of IPs that have proxied the request (`X-Forwarded-For`), and the protocol used (`X-Forwarded-Proto`). This is vital for logging and application logic on your backend servers.

Step 2: Configure Load Balancing Methods

By default, Nginx uses a round-robin method, distributing requests sequentially to each server in the upstream group. You can change this behavior.

• **Round Robin (Default):**
```nginx upstream my_wiki_servers { server 192.168.1.101:80; server 192.168.1.102:80; } ```
• **Least Connections:** Distributes requests to the server with the fewest active connections. This is useful if your backend servers have varying processing capabilities or if requests have vastly different durations.
```nginx upstream my_wiki_servers { least_conn; server 192.168.1.101:80; server 192.168.1.102:80; } ```
• **IP Hash:** Distributes requests based on a hash of the client's IP address. This ensures that requests from the same client IP are consistently sent to the same backend server. This is important for applications that maintain session state on the server-side (session persistence).
```nginx upstream my_wiki_servers { ip_hash; server 192.168.1.101:80; server 192.168.1.102:80; } ``` **Security Implication:** While `ip_hash` provides session persistence, it can lead to uneven load distribution if you have many clients behind a single NAT (e.g., a corporate network).

Step 3: Implement Health Checks

Nginx can automatically detect unhealthy backend servers and stop sending traffic to them. This is achieved using the `health_check` directive, which is available in Nginx Plus or through third-party modules. For open-source Nginx, you can achieve basic health checking by marking servers as `down` or by using `max_fails` and `fail_timeout`.

• *Using `max_fails` and `fail_timeout` (Open Source Nginx):**

These parameters are added to the `server` directive within the `upstream` block.

```nginx upstream my_wiki_servers { server 192.168.1.101:80 max_fails=3 fail_timeout=30s; server 192.168.1.102:80 max_fails=3 fail_timeout=30s; } ```

• **`max_fails=3`**: If a server fails to respond 3 times within the `fail_timeout` period, it will be considered unhealthy.
• **`fail_timeout=30s`**: The duration during which `max_fails` attempts are counted. After this period, Nginx will attempt to send a request to the server again to see if it has recovered.
• *Nginx Plus Health Checks (More Advanced):**

Nginx Plus offers more sophisticated health checks. You can define a specific URI to check and the expected response code.

```nginx upstream my_wiki_servers { server 192.168.1.101:80; server 192.168.1.102:80;

health_check uri=/healthcheck.html interval=5s fails=3 passes=2; } ```

• **`uri=/healthcheck.html`**: Nginx will send requests to this path on each backend server. You should create a simple static file (e.g., `healthcheck.html`) at this location on your backend servers that always returns a 200 OK status.
• **`interval=5s`**: How often to perform the health check.
• **`fails=3`**: Number of unsuccessful health checks before marking a server as unhealthy.
• **`passes=2`**: Number of successful health checks before marking a server as healthy again.

Step 4: Enable Session Persistence (If Required)

If your wiki application relies on server-side sessions (e.g., user login states are stored directly on the web server), you need to ensure that a user's requests are always directed to the same backend server. As mentioned earlier, the `ip_hash` method is the simplest way to achieve this with open-source Nginx.

```nginx upstream my_wiki_servers { ip_hash; server 192.168.1.101:80; server 192.168.1.102:80; } ```

• *Alternative: Sticky Sessions with Cookies (Nginx Plus):**

Nginx Plus offers a more robust sticky session mechanism using cookies, which is less susceptible to NAT issues than `ip_hash`.

```nginx upstream my_wiki_servers { sticky cookie srv_id expires=1h domain=.your_domain.com path=/; server 192.168.1.101:80; server 192.168.1.102:80; } ```

• **`sticky cookie srv_id`**: This directive tells Nginx to set a cookie named `srv_id` on the client's browser.
• **`expires=1h`**: The cookie will expire after 1 hour.
• **`domain=.your_domain.com`**: The cookie will be sent to all subdomains of `your_domain.com`.
• **`path=/`**: The cookie will be sent for all paths on the domain.

Step 5: Test and Reload Nginx

After making changes to your Nginx configuration, it's crucial to test the syntax and then reload Nginx for the changes to take effect.

1. **Test configuration syntax:** ```bash sudo nginx -t ``` **Expected Output:** ``` nginx: the configuration file /etc/nginx/nginx.conf syntax is ok nginx: configuration file /etc/nginx/nginx.conf test is successful ``` If you see errors, carefully review your configuration file for typos or syntax mistakes.

2. **Reload Nginx:** ```bash sudo systemctl reload nginx # Or for older systems sudo service nginx reload ```

Step 6: Monitor Load Balancing

Monitor your Nginx access logs and backend server logs to ensure traffic is being distributed as expected. You can also use tools like `htop` or `nload` on your backend servers to observe CPU and network usage.

For more advanced monitoring, consider using Nginx Plus's built-in metrics or integrating with external monitoring solutions like Prometheus and Grafana.

Performance Benchmarks

Nginx is known for its high performance and low resource footprint. When acting as a load balancer, its performance will depend on the complexity of your configuration, the number of upstream servers, and the traffic volume.

• **Throughput:** A well-configured Nginx load balancer can handle tens of thousands of requests per second on modest hardware. For very high-traffic sites, consider using multiple Nginx instances or leveraging Nginx Plus with its advanced features.
• **Latency:** The added latency introduced by Nginx as a load balancer is typically very low, often in the sub-millisecond range, especially when Nginx and backend servers are on the same high-speed network.
• **Resource Usage:** Nginx is event-driven and uses a non-blocking I/O model, making it very efficient with CPU and memory. You can often run a robust load balancer on a small VPS or a dedicated server. For example, a dedicated server from PowerVPS can easily handle significant load balancing tasks.
• *Benchmarking Tip:** Use tools like `ab` (ApacheBench) or `wrk` to simulate traffic and measure your load balancer's performance under load.

```bash # Example using wrk wrk -t4 -c100 -d30s http://your_domain.com/ ```

This command will run 4 threads, maintain 100 concurrent connections for 30 seconds, targeting your Nginx load balancer. Analyze the output for requests per second, latency, and error rates.

Troubleshooting

• **"502 Bad Gateway" Error:** This is a common error indicating that Nginx could not connect to any of the backend servers or that the backend servers returned an invalid response.
* **Check Backend Server Status:** Ensure your backend servers are running and accessible from the load balancer server. * **Verify IP Addresses and Ports:** Double-check that the IP addresses and ports in your `upstream` block are correct. * **Firewall Issues:** Ensure firewalls on both the load balancer and backend servers are not blocking traffic between them on the specified ports. * **Backend Application Errors:** The backend application itself might be crashing or returning errors. Check the logs on your backend servers.
• **"504 Gateway Timeout" Error:** This means Nginx connected to a backend server, but the server did not send a response within the configured timeout period.
* **Increase `proxy_read_timeout`:** You might need to increase the timeout values in your Nginx configuration: ```nginx location / { proxy_pass http://my_wiki_servers; proxy_read_timeout 120s; # Increase timeout # ... other proxy_set_header directives } ``` * **Optimize Backend Performance:** The backend application might be too slow to respond. Investigate performance bottlenecks on your wiki servers.
• **"500 Internal Server Error":** This usually originates from the backend application itself. Check the application logs on your backend servers for detailed error messages.
• **Uneven Load Distribution:** If you're not using `ip_hash` or sticky sessions and notice one server is consistently overloaded while others are idle, review your load balancing method. If using `ip_hash`, check if multiple clients are behind the same NAT.

Related Articles

• Nginx Configuration
• Reverse Proxy Setup
• MediaWiki Installation Guide

Category:Optimization Category:Networking Category:Web Server