NGINX Agent V3 Open Telemetry Configuration is not set to use NGINX_AGENT_COMMAND_TLS_SERVER_NAME

[thomas-germano]

Bug Overview

I've setup NGINX agent to use a NGINX proxy to contact NGINX One Console like in the documentation:

Prepare - Set up NGINX as a proxy

Unfortunately, the documentation is not up to date with NGINX Agent v3 so I've opened an issue:

Documentation Issue 822

Unfortunately, the generated Open Telemetry configuration isn't setup to use this configuration of nginx agent:

...
  tls:
    server_name: agent.connect.nginx.com
...

And we have a the following TLS issue:

2025-07-16T11:23:39.416+0200    warn    [grpc@v1.71.1/clientconn.go:1406](mailto:grpc@v1.71.1/clientconn.go:1406) [core] [Channel #2 SubChannel #3]grpc: addrConn.createTransport failed to connect to {Addr: "PROXY_IP:5000", ServerName: "PROXY_IP:5000", }. Err: connection error: desc = "transport: authentication handshake failed: remote error: tls: unrecognized name"    {"grpc_log": true}

Expected Behavior

The generated Open Telemetry configuration should use the server name specified in the nginx agent configuration for the TLS connection.

Steps to Reproduce the Bug

setup a nginx proxy:

stream {
    upstream stream_backend {
        zone tcp_servers 64k;
        server agent.connect.nginx.com:443;
    }

    log_format basic '$remote_addr [$time_local] '
                    '$protocol $status $bytes_sent $bytes_received '
                    '$session_time "$upstream_addr" '
                    '"$upstream_bytes_sent" "$upstream_bytes_received" "$upstream_connect_time"';

    access_log /var/log/nginx/nginx1_access.log basic;
    error_log /var/log/nginx/nginx1_error.log;

    server {
        listen 5000;
        status_zone tcp_server;
        proxy_pass stream_backend;
        proxy_next_upstream on;
    }
}

Configure your agent to use this proxy:

command:
  server:
    host: YOUR_PROXY_IP_ADDRESS_HERE
    port: 5000
  auth:
    token: "YOUR_DATA_PLANE_KEY_HERE"
  tls:
    server_name: agent.connect.nginx.com
    skip_verify: false

Environment Details

• Target deployment platform: Local cluster
• Target OS: RHEL 8
• Version of this project or specific commit: 3.1.0
• Version of any relevant project languages: Open Telemetry Configuration

Additional Context

No response

[oCHRISo]

Hi @thomas-germano

Thanks for opening this issue and bringing it to our attention. We'll look into getting it resolved.

To ensure your issue is handled as efficiently as possible, we kindly ask that you also open a new support request through the NGINX One Console UI. You can do this by clicking the Support button in the top-right corner and selecting Contact Support.

Thank you for your understanding and for helping us improve

[thomas-germano]

Hi @oCHRISo

Thank you for your response.

I've opened the case on the NGINX One Console UI.

Best regards

[JTorreG]

Duplicate of
nginx/documentation#822

[thomas-germano]

Hi @JTorreG,

This is not a duplicate of nginx/documentation#822.

The mentioned issue is about the documentation not being up-to-date, this one is about Open Telemetry not using the server name for the TLS Handshake and, because of that, the TLS handshake failed. Can you re-open it, please?

Best regards

[thomas-germano]

related Pull Request: #1203

[JTorreG]

@oCHRISo can you clarify what is still required here?

[dhurley]

@JTorreG we have released agent v3.2.1 with a fix for this today and docs are updated by this PR nginx/documentation#930

[thomas-germano]

Hi @JTorreG, everything looks good with the release of agent v3.2.1. I just referenced the PR to keep a trace of the issue.

I will close the issue, thanks to the NGINX Team for the work!