Repository navigation
Fall: Progress Report 2
- Bootstrap a Python server using HTTP to talk to and from clients
- Make at least a barebones database structure for storing endpoint & task info (**NOTE: ** DB schema (syntax) is on Draft Methodology, will eventually become it's own page
- Bootstraped a Go client program to ping the server, and print task data
| Step | Direction | Description | Purpose | Implementation Details |
|---|---|---|---|---|
| 1 | Backend Server → CapyDaemon | Backend server sends task action via API call | Initiate tasks on endpoint agent (e.g., run scripts, execute commands, collect data) | • Authentication: Mutual TLS (mTLS) for bidirectional certificate validation • CapyDaemon receives task payload with action type and parameters • Task types include: powershell/cmd/bash scripts or predefined tasks • Built-in retry logic in management program for failed dispatch attempts to backend API |
| 2 | CapyDaemon → Backend Server | CapyDaemon updates results and status of command to backend server | Report task execution results and maintain agent heartbeat with backend | • CapyDaemon calls backend API with task completion status (success/failure) • Includes output data, logs, or error messages • Periodic status updates: CapyDaemon regularly reports general health and checks for new tasks (heartbeat mechanism) • Built-in retry logic in CapyDaemon for failed API calls to backend • mTLS authentication for secure communication |
| 3 | CapyDaemon → Backend Server | CapyDaemon pulls additional resources based on task requirements | Pull-on-demand model for configuration files needed by tasks or services | • Pull-based model: Task payload in Step 1 may include references to required files • CapyDaemon initiates separate pull request(s) to backend API to retrieve needed resources • Resources may include: service configs (HTTP server, etc.), scripts, data files • State safety: Pull model prevents unknown/inconsistent states if CapyDaemon is offline • Config files delivered to monitored host's filesystem • No concern for missed updates—CapyDaemon only pulls what it needs when active |
| Debug Function | Local (Monitored Host) | Debug function within CapyDaemon for ad-hoc command execution | Development/troubleshooting capability to execute commands without full task workflow | • Debug function built into CapyDaemon • Bypasses normal task workflow • Direct powershell/cmd/bash command execution • Security: Should be disabled/restricted in production environments |