[container] Nginx not receiving signals due to background process handling

[timo-reymann]

In the current Docker entrypoint, Nginx is started in the background, and the main process (statsig_forward_proxy) is executed afterward:

nginx > /dev/null 2>&1 &
exec statsig_forward_proxy "$@"

This setup prevents Nginx from receiving container stop signals (SIGTERM, SIGINT) directly, which can cause:

• Graceful shutdown not happening.
• Orphaned Nginx processes.
• Containers taking longer to stop.

Expected behavior:

• Nginx should receive signals from Docker/Kubernetes directly.
• Both Nginx and the proxy should shut down cleanly when the container stops.

Possible solutions:

• Use a process manager (e.g., supervisord, s6, multirun) to run both processes.
• Or, run Nginx as the main process and have the proxy in a separate container.
• Or, have the entrypoint handle signals and forward them to Nginx.