This is a simple TCP and UDP echo server and client, designed for testing container lifetime and reachability.

In server mode, it listens on both TCP and UDP on the port specified by the PORT environment variable. It reads data sent by clients, and echos back the data along with some additional context that's useful for debugging.

In client mode, you specify the protocol (TCP or UDP), the target host and an interval, and it will create a new connection for every echo request.

• Client and server in one binary (controlled by env vars)
• TCP and UDP mode for both client and server
• JSON logs with lots of detail
• SIGINT/SIGTERM handler with extra logging
• The AWS ECS task ARN is included in the logs, if available

All configuration happens through environment variables:

• APP_MODE (default: client) – client or server

• PORT (default: 2222) – the TCP/UDP port to use

• CLIENT_PROTO (default tcp) – tcp or udp

• CLIENT_TARGET_HOST (default: 127.0.0.1) – hostname or IP to target in client mode.

• CLIENT_CONN_INTERVAL_MS (default: 1000) – connection interval in milliseconds

• HANDLE_SIGNALS (default: unset) – if set to any value, SIGINT and SIGTERM are caught and timing info is logged every 200 milliseconds for as long as possible

Some things are not yet configurable:

• Client TCP/UDP dial timeout is 800 milliseconds

• Client/server TCP/UDP read/write timeout is 800 milliseconds

• Client/server read buffer is 4 kB

• Go issues spurious SIGURG signals that can be a bit noisy.