Better node monitoring with new client metrics

Tailscale has introduced a new feature that allows users to monitor client metrics for enhanced network health and connectivity analysis, particularly beneficial for setups involving subnet routers, Docker containers, or Kubernetes with site-to-site network configurations. Available in Tailscale version 1.78 or later, these metrics can be accessed locally or remotely and are compatible with Prometheus or similar monitoring systems. Users can view the metrics using the command `tailscale metrics print`, which outputs data in a human-readable Prometheus text format, useful for debugging and scripting. The metrics can also be written to a file using `tailscale metrics write` for further integration with Prometheus node exporter, enabling seamless integration into existing monitoring workflows. Additionally, metrics can be accessed over the network via Tailscale IP addresses or exposed to the public internet, providing flexibility for various monitoring needs. This feature aims to improve network management by offering a more stable and accessible alternative to previous methods of log parsing or undocumented interfaces, thereby enhancing the quality of life for both individual users and IT teams.