clonebox 1.1.4__py3-none-any.whl → 1.1.5__py3-none-any.whl

clonebox/di.py

@@ -0,0 +1,176 @@
+"""IoC container for dependency injection in CloneBox."""
+
+import inspect
+import threading
+from dataclasses import dataclass
+from typing import Any, Callable, Dict, Optional, Type, TypeVar
+
+T = TypeVar("T")
+
+
+@dataclass
+class ServiceRegistration:
+    """Registration info for a service."""
+
+    factory: Callable[..., Any]
+    singleton: bool = True
+    instance: Optional[Any] = None
+
+
+class DependencyContainer:
+    """
+    IoC container for dependency injection.
+
+    Usage:
+        container = DependencyContainer()
+
+        # Register services
+        container.register(HypervisorBackend, LibvirtBackend, singleton=True)
+        container.register(DiskManager, QemuDiskManager)
+
+        # Resolve dependencies
+        cloner = container.resolve(SelectiveVMCloner)
+    """
+
+    def __init__(self):
+        self._registrations: Dict[Type, ServiceRegistration] = {}
+        self._lock = threading.Lock()
+
+    def register(
+        self,
+        interface: Type[T],
+        implementation: Type[T] = None,
+        factory: Callable[..., T] = None,
+        singleton: bool = True,
+        instance: T = None,
+    ) -> "DependencyContainer":
+        """
+        Register a service.
+
+        Args:
+            interface: The interface/base class
+            implementation: Concrete implementation class
+            factory: Factory function to create instance
+            singleton: If True, reuse same instance
+            instance: Pre-created instance to use
+        """
+        if instance is not None:
+            self._registrations[interface] = ServiceRegistration(
+                factory=lambda: instance,
+                singleton=True,
+                instance=instance,
+            )
+        elif factory is not None:
+            self._registrations[interface] = ServiceRegistration(
+                factory=factory,
+                singleton=singleton,
+            )
+        elif implementation is not None:
+            self._registrations[interface] = ServiceRegistration(
+                factory=implementation,
+                singleton=singleton,
+            )
+        else:
+            raise ValueError("Must provide implementation, factory, or instance")
+
+        return self  # Enable chaining
+
+    def resolve(self, interface: Type[T]) -> T:
+        """Resolve a service instance."""
+        with self._lock:
+            if interface not in self._registrations:
+                # If it's a class and not an interface, try to auto-resolve it
+                if inspect.isclass(interface):
+                    return self._create_instance(interface)
+                raise KeyError(f"No registration for {interface}")
+
+            reg = self._registrations[interface]
+
+            # Return existing instance for singletons
+            if reg.singleton and reg.instance is not None:
+                return reg.instance
+
+            # Create new instance
+            instance = self._create_instance(reg.factory)
+
+            # Store for singleton
+            if reg.singleton:
+                reg.instance = instance
+
+            return instance
+
+    def _create_instance(self, factory: Callable) -> Any:
+        """Create instance, resolving constructor dependencies."""
+        try:
+            sig = inspect.signature(factory)
+        except ValueError:
+            # Handle cases where signature can't be inspected (e.g., some built-ins)
+            return factory()
+
+        kwargs = {}
+
+        for name, param in sig.parameters.items():
+            if param.annotation != inspect.Parameter.empty:
+                # Try to resolve dependency
+                try:
+                    kwargs[name] = self.resolve(param.annotation)
+                except (KeyError, TypeError):
+                    if param.default == inspect.Parameter.empty:
+                        raise
+                    # Use default if available
+            elif param.default != inspect.Parameter.empty:
+                # Use default if no annotation but default exists
+                pass
+            else:
+                # Can't resolve this parameter
+                pass
+
+        return factory(**kwargs)
+
+    def has(self, interface: Type) -> bool:
+        """Check if service is registered."""
+        return interface in self._registrations
+
+    def reset(self) -> None:
+        """Reset all singleton instances."""
+        with self._lock:
+            for reg in self._registrations.values():
+                reg.instance = None
+
+
+# Global container instance
+_container: Optional[DependencyContainer] = None
+
+
+def get_container() -> DependencyContainer:
+    """Get the global container instance."""
+    global _container
+    if _container is None:
+        _container = create_default_container()
+    return _container
+
+
+def set_container(container: DependencyContainer) -> None:
+    """Set the global container (useful for testing)."""
+    global _container
+    _container = container
+
+
+def create_default_container() -> DependencyContainer:
+    """Create container with default registrations."""
+    from .backends.libvirt_backend import LibvirtBackend
+    from .backends.qemu_disk import QemuDiskManager
+    from .backends.subprocess_runner import SubprocessRunner
+    from .interfaces.disk import DiskManager
+    from .interfaces.hypervisor import HypervisorBackend
+    from .interfaces.process import ProcessRunner
+    from .secrets import SecretsManager
+
+    container = DependencyContainer()
+
+    container.register(HypervisorBackend, LibvirtBackend)
+    container.register(DiskManager, QemuDiskManager)
+    container.register(ProcessRunner, SubprocessRunner)
+    container.register(SecretsManager, SecretsManager)
+
+    return container

clonebox/health/__init__.py

@@ -1,6 +1,6 @@
 """Health check system for CloneBox VMs."""
 
-from .models import HealthCheckResult, HealthStatus, ProbeConfig
+from .models import HealthCheckResult, HealthStatus, ProbeConfig, ProbeType
 from .probes import HTTPProbe, TCPProbe, CommandProbe, ScriptProbe
 from .manager import HealthCheckManager
 
@@ -8,6 +8,7 @@ __all__ = [
     "HealthCheckResult",
     "HealthStatus",
     "ProbeConfig",
+    "ProbeType",
     "HTTPProbe",
     "TCPProbe",
     "CommandProbe",

clonebox/health/manager.py

@@ -0,0 +1,328 @@
+#!/usr/bin/env python3
+"""Health check manager for CloneBox VMs."""
+
+import json
+import time
+from datetime import datetime
+from pathlib import Path
+from typing import Any, Dict, List, Optional
+
+from .models import HealthCheckResult, HealthStatus, ProbeConfig, ProbeType, VMHealthState
+from .probes import get_probe
+
+
+class HealthCheckManager:
+    """Manage health checks for VMs."""
+
+    def __init__(self, config_dir: Optional[Path] = None):
+        self._config_dir = config_dir or Path.home() / ".local/share/clonebox/health"
+        self._config_dir.mkdir(parents=True, exist_ok=True)
+        self._vm_states: Dict[str, VMHealthState] = {}
+
+    def check(
+        self,
+        vm_name: str,
+        probes: List[ProbeConfig],
+    ) -> VMHealthState:
+        """Run health checks for a VM.
+
+        Args:
+            vm_name: Name of VM to check
+            probes: List of probe configurations
+
+        Returns:
+            VMHealthState with aggregated results
+        """
+        results = []
+
+        for config in probes:
+            if not config.enabled:
+                continue
+
+            probe = get_probe(config.probe_type)
+            result = self._run_probe_with_retry(probe, config)
+            results.append(result)
+
+        # Calculate overall status
+        overall = self._calculate_overall_status(results)
+
+        # Update state
+        state = self._update_vm_state(vm_name, overall, results)
+
+        return state
+
+    def check_single(self, config: ProbeConfig) -> HealthCheckResult:
+        """Run a single health check.
+
+        Args:
+            config: Probe configuration
+
+        Returns:
+            HealthCheckResult
+        """
+        probe = get_probe(config.probe_type)
+        return self._run_probe_with_retry(probe, config)
+
+    def check_from_config(
+        self,
+        vm_name: str,
+        config_path: Path,
+    ) -> VMHealthState:
+        """Run health checks from YAML config file.
+
+        Args:
+            vm_name: Name of VM
+            config_path: Path to .clonebox.yaml
+
+        Returns:
+            VMHealthState with results
+        """
+        import yaml
+
+        if not config_path.exists():
+            return VMHealthState(
+                vm_name=vm_name,
+                overall_status=HealthStatus.UNKNOWN,
+                last_check=datetime.now(),
+            )
+
+        config = yaml.safe_load(config_path.read_text())
+        health_checks = config.get("health_checks", [])
+
+        probes = []
+        for hc in health_checks:
+            probe_config = ProbeConfig.from_dict(hc)
+            probes.append(probe_config)
+
+        return self.check(vm_name, probes)
+
+    def get_state(self, vm_name: str) -> Optional[VMHealthState]:
+        """Get current health state for a VM."""
+        return self._vm_states.get(vm_name)
+
+    def wait_healthy(
+        self,
+        vm_name: str,
+        probes: List[ProbeConfig],
+        timeout: int = 300,
+        check_interval: float = 5.0,
+    ) -> bool:
+        """Wait until VM becomes healthy.
+
+        Args:
+            vm_name: Name of VM
+            probes: Probe configurations
+            timeout: Maximum wait time in seconds
+            check_interval: Time between checks
+
+        Returns:
+            True if healthy within timeout, False otherwise
+        """
+        start = time.time()
+
+        while time.time() - start < timeout:
+            state = self.check(vm_name, probes)
+
+            if state.overall_status == HealthStatus.HEALTHY:
+                return True
+
+            time.sleep(check_interval)
+
+        return False
+
+    def create_default_probes(self, services: List[str]) -> List[ProbeConfig]:
+        """Create default health probes for common services.
+
+        Args:
+            services: List of service names (e.g., ["nginx", "postgres"])
+
+        Returns:
+            List of ProbeConfig
+        """
+        defaults = {
+            "nginx": ProbeConfig(
+                name="nginx",
+                probe_type=ProbeType.HTTP,
+                url="http://localhost:80/",
+                expected_status=200,
+                timeout_seconds=5.0,
+            ),
+            "apache2": ProbeConfig(
+                name="apache2",
+                probe_type=ProbeType.HTTP,
+                url="http://localhost:80/",
+                expected_status=200,
+                timeout_seconds=5.0,
+            ),
+            "postgres": ProbeConfig(
+                name="postgres",
+                probe_type=ProbeType.TCP,
+                host="localhost",
+                port=5432,
+                timeout_seconds=3.0,
+            ),
+            "postgresql": ProbeConfig(
+                name="postgresql",
+                probe_type=ProbeType.TCP,
+                host="localhost",
+                port=5432,
+                timeout_seconds=3.0,
+            ),
+            "mysql": ProbeConfig(
+                name="mysql",
+                probe_type=ProbeType.TCP,
+                host="localhost",
+                port=3306,
+                timeout_seconds=3.0,
+            ),
+            "redis": ProbeConfig(
+                name="redis",
+                probe_type=ProbeType.COMMAND,
+                command="redis-cli ping",
+                expected_output="PONG",
+                timeout_seconds=3.0,
+            ),
+            "mongodb": ProbeConfig(
+                name="mongodb",
+                probe_type=ProbeType.TCP,
+                host="localhost",
+                port=27017,
+                timeout_seconds=3.0,
+            ),
+            "docker": ProbeConfig(
+                name="docker",
+                probe_type=ProbeType.COMMAND,
+                command="docker info",
+                expected_exit_code=0,
+                timeout_seconds=5.0,
+            ),
+            "ssh": ProbeConfig(
+                name="ssh",
+                probe_type=ProbeType.TCP,
+                host="localhost",
+                port=22,
+                timeout_seconds=3.0,
+            ),
+        }
+
+        probes = []
+        for service in services:
+            service_lower = service.lower()
+            if service_lower in defaults:
+                probes.append(defaults[service_lower])
+            else:
+                # Create generic process check
+                probes.append(
+                    ProbeConfig(
+                        name=service,
+                        probe_type=ProbeType.COMMAND,
+                        command=f"pgrep -x {service} || pgrep -f {service}",
+                        expected_exit_code=0,
+                        timeout_seconds=3.0,
+                    )
+                )
+
+        return probes
+
+    def _run_probe_with_retry(
+        self,
+        probe,
+        config: ProbeConfig,
+    ) -> HealthCheckResult:
+        """Run probe with retry logic."""
+        last_result = None
+
+        for attempt in range(config.retries):
+            result = probe.check(config)
+
+            if result.is_healthy:
+                return result
+
+            last_result = result
+
+            if attempt < config.retries - 1:
+                time.sleep(config.retry_delay_seconds)
+
+        return last_result or HealthCheckResult(
+            probe_name=config.name,
+            status=HealthStatus.UNKNOWN,
+            checked_at=datetime.now(),
+            duration_ms=0,
+            error="No result after retries",
+        )
+
+    def _calculate_overall_status(
+        self,
+        results: List[HealthCheckResult],
+    ) -> HealthStatus:
+        """Calculate overall health status from results."""
+        if not results:
+            return HealthStatus.UNKNOWN
+
+        statuses = [r.status for r in results]
+
+        # All healthy = healthy
+        if all(s == HealthStatus.HEALTHY for s in statuses):
+            return HealthStatus.HEALTHY
+
+        # Any unhealthy = unhealthy
+        if any(s == HealthStatus.UNHEALTHY for s in statuses):
+            return HealthStatus.UNHEALTHY
+
+        # Any timeout = degraded
+        if any(s == HealthStatus.TIMEOUT for s in statuses):
+            return HealthStatus.DEGRADED
+
+        # Mix of healthy and unknown = degraded
+        return HealthStatus.DEGRADED
+
+    def _update_vm_state(
+        self,
+        vm_name: str,
+        overall: HealthStatus,
+        results: List[HealthCheckResult],
+    ) -> VMHealthState:
+        """Update VM health state with new results."""
+        if vm_name not in self._vm_states:
+            self._vm_states[vm_name] = VMHealthState(
+                vm_name=vm_name,
+                overall_status=overall,
+                last_check=datetime.now(),
+            )
+
+        state = self._vm_states[vm_name]
+        state.overall_status = overall
+        state.last_check = datetime.now()
+        state.check_results = results
+        state.total_checks += 1
+
+        if overall == HealthStatus.HEALTHY:
+            state.consecutive_successes += 1
+            state.consecutive_failures = 0
+        else:
+            state.consecutive_failures += 1
+            state.consecutive_successes = 0
+            state.total_failures += 1
+
+        return state
+
+    def export_metrics(self, vm_name: str) -> Dict[str, Any]:
+        """Export health metrics in Prometheus format."""
+        state = self._vm_states.get(vm_name)
+        if not state:
+            return {}
+
+        metrics = {
+            "clonebox_health_status": 1 if state.overall_status == HealthStatus.HEALTHY else 0,
+            "clonebox_health_consecutive_failures": state.consecutive_failures,
+            "clonebox_health_consecutive_successes": state.consecutive_successes,
+            "clonebox_health_total_checks": state.total_checks,
+            "clonebox_health_failure_rate": state.failure_rate,
+        }
+
+        for result in state.check_results:
+            probe_name = result.probe_name.replace("-", "_")
+            metrics[f"clonebox_probe_{probe_name}_healthy"] = 1 if result.is_healthy else 0
+            metrics[f"clonebox_probe_{probe_name}_duration_ms"] = result.duration_ms
+
+        return metrics