gRPC proxying in Nginx - Practice Problems & Coding Challenges

Challenge - 5 Problems

💻 Command Output

intermediate

gRPC Proxying Basic Configuration Output

What will be the output status code when NGINX proxies a gRPC request to a backend that is up and running with this configuration?

Nginx

server {
    listen 50051 http2;
    location / {
        grpc_pass grpc://backend:50052;
    }
}

A HTTP/2 200 OK

B HTTP/1.1 500 Internal Server Error

C HTTP/2 404 Not Found

D HTTP/1.1 502 Bad Gateway

💡 Hint

Check if the grpc_pass directive points to a valid backend and if HTTP/2 is enabled.

When NGINX proxies a gRPC request correctly to a healthy backend, it returns HTTP/2 200 OK status.

❓ Troubleshoot

intermediate

Identifying the Cause of gRPC 502 Bad Gateway

Given this NGINX snippet, what is the most likely cause of receiving a 502 Bad Gateway error when proxying gRPC requests?

Nginx

server {
    listen 443 ssl http2;
    ssl_certificate /etc/ssl/certs/server.crt;
    ssl_certificate_key /etc/ssl/private/server.key;

    location / {
        grpc_pass grpc://127.0.0.1:50051;
    }
}

A NGINX does not support grpc_pass on port 443

B Missing proxy_set_header directives for gRPC metadata

C SSL certificates are invalid causing handshake failure

D The backend gRPC server is not running or unreachable on 127.0.0.1:50051

💡 Hint

502 Bad Gateway usually means NGINX cannot connect to the backend service.

A 502 error indicates NGINX could not reach the backend gRPC server at 127.0.0.1:50051. The SSL config is for client connections, not backend connectivity.

❓ Configuration

advanced

Correct gRPC Timeout Configuration

Which NGINX configuration snippet correctly sets a 10-second timeout for gRPC proxying to avoid hanging requests?

A

location / {
    grpc_pass grpc://backend:50051;
    grpc_read_timeout 10s;
}

B

location / {
    grpc_pass grpc://backend:50051;
    proxy_read_timeout 10s;
}

C

location / {
    grpc_pass grpc://backend:50051;
    grpc_send_timeout 10s;
}

D

location / {
    grpc_pass grpc://backend:50051;
    proxy_send_timeout 10s;
}

💡 Hint

Use the directive specific to gRPC read timeout, not generic proxy timeouts.

grpc_read_timeout sets the timeout for reading a response from the gRPC backend. proxy_read_timeout is for HTTP proxying, not gRPC.

🔀 Workflow

advanced

Order of Steps to Enable gRPC Proxying in NGINX

Arrange the steps in the correct order to enable gRPC proxying in NGINX for a new service.

A 1,3,2,4

B 2,1,3,4

C 1,2,3,4

D 2,3,1,4

💡 Hint

Think about starting backend first, then configuring proxy, then applying config, then testing.

You must have the backend running before configuring NGINX to proxy it. Then reload NGINX and test the connection.

✅ Best Practice

expert

Best Practice for Handling gRPC Errors in NGINX Proxy

Which configuration snippet correctly handles gRPC backend errors by returning a custom error message to clients?

A

error_page 502 = /grpc_fallback;
location = /grpc_fallback {
    internal;
    default_type application/grpc;
    add_header grpc-status 14;
    add_header grpc-message "Unavailable backend";
    return 204;
}

B

error_page 502 = /grpc_fallback;
location = /grpc_fallback {
    internal;
    default_type application/grpc;
    add_header grpc-status 14;
    add_header grpc-message "Unavailable backend";
    return 200;
}

C

error_page 502 = /grpc_fallback;
location = /grpc_fallback {
    internal;
    default_type application/grpc;
    add_header grpc-status 14;
    add_header grpc-message "Unavailable backend";
    return 503;
}

D

error_page 502 = /grpc_fallback;
location = /grpc_fallback {
    internal;
    default_type application/grpc;
    add_header grpc-status 14;
    add_header grpc-message "Unavailable backend";
    return 500;
}

💡 Hint

gRPC error responses use HTTP/2 200 with grpc-status trailer (14 = UNAVAILABLE) for proper client handling.

gRPC responses including errors use HTTP/2 200 with grpc-status trailer set to 14 (UNAVAILABLE). Returning 503 or other 5xx makes it an HTTP transport error, not a proper gRPC status.