Portainer Agent connectivity and log retrieval for failed deployments

Container Log Accessibility via Portainer Agent

Portainer provides a UI for accessing container and service logs by proxying requests to the underlying runtime or orchestrator API. When managing remote environments, this process relies heavily on the Portainer Agent's connectivity to the primary instance.

In scenarios where a stack deployment is marked as failed, the availability of logs depends on whether the container persists and if the Agent remains reachable. Because Portainer does not store logs internally, the visibility of diagnostic data is tied to the runtime logging driver and the current state of the endpoint connection.

There is uncertainty regarding how the UI handles log requests when the Agent is disconnected but the container daemon is still operational, particularly during the convergence phase of a failed deployment.

• Does the Portainer UI provide a fallback mechanism for log retrieval if the Agent is unreachable but the Docker socket is accessible?
• How does the log metadata and timestamping differ when switching between standalone Docker endpoints and Kubernetes endpoints for the same container image?