fix: Add timeout to plugin stop requests, correctly propagate termination signals during graceful shutdown

[TheCrazyGM]

Description

I run the node as a systemd service and I noticed it always timed out waiting for shutdown or restart. So I looked into it.

The current signal handling logic in app.js is flawed. It catches SIGINT/SIGTERM, performs cleanup, and then re-sends the signal to the process using process.kill(process.pid, signal). However, since the custom signal handler is still active, it catches the re-sent signal, checks the shuttingDown flag (which is now true), and simply returns, leaving the process running indefinitely.

Additionally, I found that during cleanup, some plugins (like P2P or Streamer) might not respond to the IPC message if they are in a bad state, causing unloadPlugin to hang indefinitely while awaiting the response.

Changes

1. Refined Signal Propagation: In stopApp, explicitly remove the SIGINT/SIGTERM listeners before re-sending the signal. This ensures the default signal handler takes over and terminates the process correctly.
2. Added Plugin Timeout: Added a 2-second timeout to the unloadPlugin function. If a plugin does not acknowledge the stop request within 2 seconds, the main process logs an error and force-kills the plugin, preventing the entire shutdown process from hanging.

Verification

• Tested manually by running the node and sending SIGINT.
• Verified that the process now exits immediately after "Saving config" with the correct exit signal, even if plugins are unresponsive.

[TheCrazyGM]

After a few more tests, I believe 2 seconds is a bit harsh, should probably be increased for a bit more timeout.