Skip to content

Health check related logs that occur immediately after deploying AgentCore Runtime

1

I am deploying the Amazon Bedrock AgentCore Runtime using the CDK (@aws-cdk/aws-bedrock-agentcore-alpha). The following log appears in CloudWatch Logs immediately after deployment (before invoke).

2026-03-08T02:39:24.228Z Connection failed out to container health check URI

Subsequent invokes complete successfully, with no functional issues.

I have confirmed that this log appears not only with code I created from scratch, but also with code generated with the bedrock-agentcore-starter-toolkit.

[Question]

  1. "Connection failed out to container health check URI" appears immediately after deployment (before invoke). From the documentation, I understand that a session (microVM) is created during the first invoke, but are containers also started and health checked during deployment?

  2. Can I safely ignore this log? Or is there some setting I need to fix this?

asked 2 months ago302 views
1 Answer
0

Based on the information available about Amazon Bedrock AgentCore Runtime, I can provide some context about health checks and container behavior, though the specific timing of health checks during deployment versus invocation isn't explicitly detailed in the documentation.

Regarding your questions:

  1. Health checks and container startup timing: While the documentation confirms that sessions are created during invocations, the specific behavior of health checks immediately after deployment (before the first invoke) isn't explicitly documented. However, it's common practice in container orchestration systems to perform health checks as part of the deployment process to verify that containers are ready to receive traffic.

  2. Whether you can ignore this log: Since you've confirmed that subsequent invocations complete successfully with no functional issues, and this behavior appears consistently even with code generated from the bedrock-agentcore-starter-toolkit, this initial health check failure appears to be a transient issue during the deployment phase. The fact that your agent functions correctly afterward suggests this is likely related to the container initialization timing rather than a configuration problem.

For troubleshooting container-related issues with AgentCore Runtime, the documentation recommends ensuring your Docker container meets the requirements: exposing port 8080, having the /invocations path available, and being ARM64 compatible. If these requirements are met and your agent is functioning correctly after deployment, the initial health check log is likely just capturing the container's startup phase and can be safely ignored as long as your invocations continue to work properly.

If you want to investigate further, you could enable more verbose logging in your agent code to understand the container startup sequence better, or run the agent container locally to observe its initialization behavior.
Sources
Troubleshoot AgentCore Runtime - Amazon Bedrock AgentCore
Failure management - Container Build Lens

answered 2 months ago

You are not logged in. Log in to post an answer.

A good answer clearly answers the question and provides constructive feedback and encourages professional growth in the question asker.