Module Engine

ModuleEngine (Lib/moduleengine.py) is responsible for automatically discovering, loading, and continuously running alert processing modules.

Discovery Mechanism

At startup, it scans all .py files under the MODULES/ directory (excluding those starting with _). The filename serves as the module name and also as the Redis Stream name.

Run Model

1. For each module, read THREAD_NUM (default 1) to determine the number of concurrent threads
2. Each thread creates an independent Module instance and runs as a daemon thread
3. Threads loop calling instance.run(), continue after exception capture, with a 100ms interval

ModuleEngine.start()
  └── _load_initial_modules()
        └── MODULES/*.py → importlib dynamic loading → Module class
              └── range(THREAD_NUM):
                    instance = Module()
                    Thread(target=run_loop, daemon=True).start()

Message Consumption

BaseModule.read_stream_message() provides two modes:

Mode	Condition	Redis Command	Behavior
Normal Mode	debug_message_id is empty	XREADGROUP (block 5s, count=1, noack=True)	Blocks waiting for new messages
Debug Mode	debug_message_id has a value	XRANGE (exact ID)	Non-blocking, reads the specified message

• Consumer Group: AI_SOC_FRAMEWORK_GROUP, auto-created
• Consumer Name: Thread name (e.g. Cloud-01-..._thread_0), independent per thread within the same module
• noack=True: Fire-and-forget messages, no retries

Concurrency Configuration

Modules can set THREAD_NUM in the class to control concurrency:

class Module(BaseModule):
    THREAD_NUM = 3  # Start 3 consumer threads