Object.Supervisor (object v0.1.2)
Enhanced supervision tree for the AAOS Object system with comprehensive fault tolerance.
Implements advanced supervision strategies including circuit breaker integration, progressive restart policies, health monitoring integration, and intelligent failure isolation to ensure maximum system resilience.
Features
- Layered Supervision: Multi-tier supervision with different restart strategies
- Intelligent Restart Policies: Context-aware restart decisions
- Circuit Breaker Integration: Automatic service isolation during failures
- Health Monitoring: Continuous supervision health assessment
- Progressive Backoff: Increasing restart delays for persistent failures
- Failure Isolation: Prevent cascading failures across components
- Resource Protection: Monitor and limit resource consumption
Supervision Hierarchy
Object.Supervisor (rest_for_one)
├── Infrastructure Supervisor (one_for_one)
│ ├── Registry
│ ├── Error Handling
│ ├── Health Monitor
│ └── Resource Monitor
├── Core Services Supervisor (one_for_one)
│ ├── Schema Registry
│ ├── Dead Letter Queue
│ ├── Graceful Degradation
│ └── Coordination Service
├── Object Management Supervisor (rest_for_one)
│ ├── Dynamic Supervisor
│ ├── Performance Monitor
│ └── Schema Evolution Manager
└── Integration Services Supervisor (one_for_one)
├── Message Router
└── AI Reasoning (if available)
Restart Strategies
- Infrastructure: Critical components that must be restarted immediately
- Core Services: Essential services with progressive restart delays
- Object Management: Components that may require dependent restarts
- Integration: Optional services that can fail without affecting core functionality
Summary
Functions
Returns a specification to start this module under a supervisor.
Triggers emergency shutdown with graceful degradation.
Gets the current supervision tree health status.
Gracefully restarts a specific child process with backoff.
Starts the enhanced supervision tree with comprehensive error handling.
Functions
Returns a specification to start this module under a supervisor.
See Supervisor
.
Triggers emergency shutdown with graceful degradation.
Gets the current supervision tree health status.
Gracefully restarts a specific child process with backoff.
Starts the enhanced supervision tree with comprehensive error handling.
Options
:restart_strategy
- Overall restart strategy (default: :rest_for_one):max_restarts
- Maximum restarts in time period (default: 5):max_seconds
- Time period for restart counting (default: 60):enable_circuit_breakers
- Enable circuit breaker integration (default: true):enable_health_monitoring
- Enable health monitoring (default: true):enable_progressive_backoff
- Enable progressive restart delays (default: true)