Nginx no live upstreams while connecting to upstream

This error occurs when Nginx, acting as a reverse proxy or load balancer, cannot forward a request because all servers defined in an `upstream` block are considered unavailable. It typically results in a 502 Bad Gateway response to the client.

Common Causes

All backend servers (upstreams) are down, unreachable, or not listening on the configured port.
Health checks (if configured) are failing, marking healthy servers as 'down'.
Network issues (firewall rules, routing problems) are blocking communication between Nginx and the upstream servers.
The upstream servers are crashing or rejecting connections due to resource exhaustion (e.g., out of memory, max connections).

How to Fix

1 Check Upstream Server Status

Manually verify the status and connectivity of each backend server defined in the upstream block.

BASH

 $ # Check if the service is running on the backend server
ssh backend-server 'systemctl status your-app-service'

# Test network connectivity from the Nginx host to the backend
curl -v http://backend-server:app-port/
nc -zv backend-server app-port

2 Review Nginx Configuration & Logs

Examine the upstream configuration and the Nginx error log for specific connection failure messages.

BASH

 $ # Check your upstream configuration
sudo nginx -T | grep -A 20 'upstream'

# Check the Nginx error log for details (common path)
sudo tail -f /var/log/nginx/error.log

3 Adjust Upstream Health Check Parameters

If using passive or active health checks, tune the `max_fails`, `fail_timeout`, and `slow_start` parameters to prevent servers from being marked 'down' incorrectly.

BASH

 $ upstream backend {
    server backend1.example.com max_fails=3 fail_timeout=30s;
    server backend2.example.com max_fails=3 fail_timeout=30s;
    # Consider adding a 'backup' server for failover
    server backup.example.com backup;
}

4 Implement a Backup Server or Custom Error Page

Add a `backup` server directive to handle traffic when primary servers are down, or define a custom error page for a better user experience.

BASH

 $ upstream app_backend {
    server primary1.example.com;
    server primary2.example.com;
    server backup-server.example.com backup;
}

server {
    ...
    error_page 502 /maintenance.html;
    location = /maintenance.html {
        root /usr/share/nginx/html;
    }
}