[Bug]: unable to connect the NGINX Plus agent (agentv3 image) to the NGINX One Console.

[syousF5]

Describe the problem you have identified

The customer is unable to connect the NGINX Plus agent (agentv3 image) to the NGINX One Console.

The error encountered is:rpc error: code = Unauthenticated desc = missing authorization

As a result, the NGINX instance does not appear in the console, preventing management and monitoring. The issue is persistent and impacts the customer's ability to use the product as intended.

Which product or products does this problem relate to?

NGINX Instance Manager

If necessary, explain any additional steps to reproduce the problem.

Troubleshooting Steps Performed:

Initial analysis identified the error as related to the authorization token (NGINX_AGENT_COMMAND_AUTH_TOKEN) not being recognized, possibly due to extra quotes or formatting issues.

Recommendations provided:
Remove quotes from the token in the environment variable.
Validate the token's correctness and activity.
Check for YAML parsing errors in the OTel collector config.
Ensure correct hostname and network connectivity.
Verify certificate trust or adjust TLS verification settings for testing.
Confirm no port conflicts with container network mode.
Customer followed the advice to remove quotes from the token, which immediately resolved the issue. The instance became visible in the console.

Customer suggested updating documentation to clarify that quotes should not be used around the token.

What is the expected or desired behaviour?

Summary:
The core technical issue has been resolved by correcting the environment variable formatting. The only pending action is to update the documentation to prevent similar issues for other users. The case remains open for follow-up.

POA:
Documentation Update:
The customer requested that the documentation be updated to clarify that quotes should not be used around the token in the environment variable. This is the only outstanding action item.
sudo docker run
--env=NGINX_LICENSE_JWT="YOUR_JWT_HERE"
--env=NGINX_AGENT_COMMAND_SERVER_PORT=443
--env=NGINX_AGENT_COMMAND_SERVER_HOST=agent.connect.nginx.com
--env=NGINX_AGENT_COMMAND_AUTH_TOKEN="k+bDekEICMkhqL7SkbN6jfLol34z1QOBktPfTyRnoD4="
--env=NGINX_AGENT_COMMAND_TLS_SKIP_VERIFY=false
--restart=always
--runtime=runc
-d private-registry.nginx.com/nginx-plus/agentv3:debian

Is this problem specific to a particular platform or software version?

Not sure.

Any additional information

No response

[mjang]

@syousF5 Which token? I see two in the issue:

• NGINX_LICENSE_JWT
• NGINX_AGENT_COMMAND_AUTH_TOKEN

(It could be both. Have you tested the command?)

Best I can tell, the doc ID'd by the customer is https://docs.nginx.com/nginx-one-console/connect-instances/connect-nginx-plus-container-images-to-nginx-one/

Whatever the solution, it likely applies in multiple locations

[syousF5]

@mjang not sure. That is a good question for the original reporter, Nithin.

[mjang]

@mjang not sure. That is a good question for the original reporter, Nithin.

@syousF5 can you share their GH account, so they can track the issue?

[syousF5]

@mjang not sure. That is a good question for the original reporter, Nithin.

@syousF5 can you share their GH account, so they can track the issue?

Could not find it in GH, unfortunately.

[n-krishna-z]

Hi @syousF5 .. There is no GH or GL created it was in SF.. Case - 01013016
The environment variable NGINX_AGENT_COMMAND_AUTH_TOKEN must contain a valid authentication token without extra quotes.