Spaces:
Runtime error
Runtime error
File size: 21,549 Bytes
1d75522 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 |
"""
Agentic Orchestrator for Advanced AI System
-----------------------------------------
Manages and coordinates multiple agentic components:
1. Task Planning & Decomposition
2. Resource Management
3. Agent Communication
4. State Management
5. Error Recovery
6. Performance Monitoring
"""
import logging
from typing import Dict, Any, List, Optional, Union, TypeVar, Generic
from dataclasses import dataclass, field
from enum import Enum
import json
import asyncio
from datetime import datetime
import uuid
from concurrent.futures import ThreadPoolExecutor
import networkx as nx
from collections import defaultdict
import numpy as np
from reasoning import UnifiedReasoningEngine as ReasoningEngine, StrategyType as ReasoningMode
from reasoning.meta_learning import MetaLearningStrategy
T = TypeVar('T')
class AgentRole(Enum):
"""Different roles an agent can take."""
PLANNER = "planner"
EXECUTOR = "executor"
MONITOR = "monitor"
COORDINATOR = "coordinator"
LEARNER = "learner"
class AgentState(Enum):
"""Possible states of an agent."""
IDLE = "idle"
BUSY = "busy"
ERROR = "error"
LEARNING = "learning"
TERMINATED = "terminated"
class TaskPriority(Enum):
"""Task priority levels."""
LOW = 0
MEDIUM = 1
HIGH = 2
CRITICAL = 3
@dataclass
class AgentMetadata:
"""Metadata about an agent."""
id: str
role: AgentRole
capabilities: List[str]
state: AgentState
load: float
last_active: datetime
metrics: Dict[str, float]
@dataclass
class Task:
"""Represents a task in the system."""
id: str
description: str
priority: TaskPriority
dependencies: List[str]
assigned_to: Optional[str]
state: str
created_at: datetime
deadline: Optional[datetime]
metadata: Dict[str, Any]
class AgentOrchestrator:
"""Advanced orchestrator for managing agentic system."""
def __init__(self, config: Dict[str, Any] = None):
self.config = config or {}
# Core components
self.agents: Dict[str, AgentMetadata] = {}
self.tasks: Dict[str, Task] = {}
self.task_graph = nx.DiGraph()
# State management
self.state_history: List[Dict[str, Any]] = []
self.global_state: Dict[str, Any] = {}
# Resource management
self.resource_pool: Dict[str, Any] = {}
self.resource_locks: Dict[str, asyncio.Lock] = {}
# Communication
self.message_queue = asyncio.Queue()
self.event_bus = asyncio.Queue()
# Performance monitoring
self.metrics = defaultdict(list)
self.performance_log = []
# Error handling
self.error_handlers: Dict[str, callable] = {}
self.recovery_strategies: Dict[str, callable] = {}
# Async support
self.executor = ThreadPoolExecutor(max_workers=4)
self.lock = asyncio.Lock()
# Logging
self.logger = logging.getLogger(__name__)
# Initialize components
self._init_components()
def _init_components(self):
"""Initialize orchestrator components."""
# Initialize reasoning engine
self.reasoning_engine = ReasoningEngine(
min_confidence=0.7,
parallel_threshold=5,
learning_rate=0.1,
strategy_weights={
"LOCAL_LLM": 2.0,
"CHAIN_OF_THOUGHT": 1.0,
"TREE_OF_THOUGHTS": 1.0,
"META_LEARNING": 1.5
}
)
# Initialize meta-learning
self.meta_learning = MetaLearningStrategy()
# Register basic error handlers
self._register_error_handlers()
async def register_agent(
self,
role: AgentRole,
capabilities: List[str]
) -> str:
"""Register a new agent with the orchestrator."""
agent_id = str(uuid.uuid4())
agent = AgentMetadata(
id=agent_id,
role=role,
capabilities=capabilities,
state=AgentState.IDLE,
load=0.0,
last_active=datetime.now(),
metrics={}
)
async with self.lock:
self.agents[agent_id] = agent
self.logger.info(f"Registered new agent: {agent_id} with role {role}")
return agent_id
async def submit_task(
self,
description: str,
priority: TaskPriority = TaskPriority.MEDIUM,
dependencies: List[str] = None,
deadline: Optional[datetime] = None,
metadata: Dict[str, Any] = None
) -> str:
"""Submit a new task to the orchestrator."""
task_id = str(uuid.uuid4())
task = Task(
id=task_id,
description=description,
priority=priority,
dependencies=dependencies or [],
assigned_to=None,
state="pending",
created_at=datetime.now(),
deadline=deadline,
metadata=metadata or {}
)
async with self.lock:
self.tasks[task_id] = task
self._update_task_graph(task)
# Trigger task planning
await self._plan_task_execution(task_id)
return task_id
async def _plan_task_execution(self, task_id: str) -> None:
"""Plan the execution of a task."""
task = self.tasks[task_id]
# Check dependencies
if not await self._check_dependencies(task):
self.logger.info(f"Task {task_id} waiting for dependencies")
return
# Find suitable agent
agent_id = await self._find_suitable_agent(task)
if not agent_id:
self.logger.warning(f"No suitable agent found for task {task_id}")
return
# Assign task
await self._assign_task(task_id, agent_id)
async def _check_dependencies(self, task: Task) -> bool:
"""Check if all task dependencies are satisfied."""
for dep_id in task.dependencies:
if dep_id not in self.tasks:
return False
if self.tasks[dep_id].state != "completed":
return False
return True
async def _find_suitable_agent(self, task: Task) -> Optional[str]:
"""Find the most suitable agent for a task."""
best_agent = None
best_score = float('-inf')
for agent_id, agent in self.agents.items():
if agent.state != AgentState.IDLE:
continue
score = await self._calculate_agent_suitability(agent, task)
if score > best_score:
best_score = score
best_agent = agent_id
return best_agent
async def _calculate_agent_suitability(
self,
agent: AgentMetadata,
task: Task
) -> float:
"""Calculate how suitable an agent is for a task."""
# Base score on capabilities match
capability_score = sum(
1 for cap in task.metadata.get("required_capabilities", [])
if cap in agent.capabilities
)
# Consider agent load
load_score = 1 - agent.load
# Consider agent's recent performance
performance_score = sum(agent.metrics.values()) / len(agent.metrics) if agent.metrics else 0.5
# Weighted combination
weights = self.config.get("agent_selection_weights", {
"capabilities": 0.5,
"load": 0.3,
"performance": 0.2
})
return (
weights["capabilities"] * capability_score +
weights["load"] * load_score +
weights["performance"] * performance_score
)
async def _assign_task(self, task_id: str, agent_id: str) -> None:
"""Assign a task to an agent."""
async with self.lock:
task = self.tasks[task_id]
agent = self.agents[agent_id]
task.assigned_to = agent_id
task.state = "assigned"
agent.state = AgentState.BUSY
agent.load += 1
agent.last_active = datetime.now()
self.logger.info(f"Assigned task {task_id} to agent {agent_id}")
# Notify agent
await self.message_queue.put({
"type": "task_assignment",
"task_id": task_id,
"agent_id": agent_id,
"timestamp": datetime.now()
})
def _update_task_graph(self, task: Task) -> None:
"""Update the task dependency graph."""
self.task_graph.add_node(task.id, task=task)
for dep_id in task.dependencies:
self.task_graph.add_edge(dep_id, task.id)
async def _monitor_system_state(self):
"""Monitor overall system state."""
while True:
try:
# Collect agent states
agent_states = {
agent_id: {
"state": agent.state,
"load": agent.load,
"metrics": agent.metrics
}
for agent_id, agent in self.agents.items()
}
# Collect task states
task_states = {
task_id: {
"state": task.state,
"assigned_to": task.assigned_to,
"deadline": task.deadline
}
for task_id, task in self.tasks.items()
}
# Update global state
self.global_state = {
"timestamp": datetime.now(),
"agents": agent_states,
"tasks": task_states,
"resource_usage": self._get_resource_usage(),
"performance_metrics": self._calculate_performance_metrics()
}
# Archive state
self.state_history.append(self.global_state.copy())
# Trim history if too long
if len(self.state_history) > 1000:
self.state_history = self.state_history[-1000:]
# Check for anomalies
await self._check_anomalies()
await asyncio.sleep(1) # Monitor frequency
except Exception as e:
self.logger.error(f"Error in system monitoring: {e}")
await self._handle_error("monitoring_error", e)
def _get_resource_usage(self) -> Dict[str, float]:
"""Get current resource usage statistics."""
return {
"cpu_usage": sum(agent.load for agent in self.agents.values()) / len(self.agents),
"memory_usage": len(self.state_history) * 1000, # Rough estimate
"queue_size": self.message_queue.qsize()
}
def _calculate_performance_metrics(self) -> Dict[str, float]:
"""Calculate current performance metrics."""
metrics = {}
# Task completion rate
completed_tasks = sum(1 for task in self.tasks.values() if task.state == "completed")
total_tasks = len(self.tasks)
metrics["task_completion_rate"] = completed_tasks / max(1, total_tasks)
# Average task duration
durations = []
for task in self.tasks.values():
if task.state == "completed" and "completion_time" in task.metadata:
duration = (task.metadata["completion_time"] - task.created_at).total_seconds()
durations.append(duration)
metrics["avg_task_duration"] = sum(durations) / len(durations) if durations else 0
# Agent utilization
metrics["agent_utilization"] = sum(agent.load for agent in self.agents.values()) / len(self.agents)
return metrics
async def _check_anomalies(self):
"""Check for system anomalies."""
# Check for overloaded agents
for agent_id, agent in self.agents.items():
if agent.load > 0.9: # 90% load threshold
await self._handle_overload(agent_id)
# Check for stalled tasks
now = datetime.now()
for task_id, task in self.tasks.items():
if task.state == "assigned":
duration = (now - task.created_at).total_seconds()
if duration > 3600: # 1 hour threshold
await self._handle_stalled_task(task_id)
# Check for missed deadlines
for task_id, task in self.tasks.items():
if task.deadline and now > task.deadline and task.state != "completed":
await self._handle_missed_deadline(task_id)
async def _handle_overload(self, agent_id: str):
"""Handle an overloaded agent."""
agent = self.agents[agent_id]
# Try to redistribute tasks
assigned_tasks = [
task_id for task_id, task in self.tasks.items()
if task.assigned_to == agent_id and task.state == "assigned"
]
for task_id in assigned_tasks:
# Find another suitable agent
new_agent_id = await self._find_suitable_agent(self.tasks[task_id])
if new_agent_id:
await self._reassign_task(task_id, new_agent_id)
async def _handle_stalled_task(self, task_id: str):
"""Handle a stalled task."""
task = self.tasks[task_id]
# First, try to ping the assigned agent
if task.assigned_to:
agent = self.agents[task.assigned_to]
if agent.state == AgentState.ERROR:
# Agent is in error state, reassign task
await self._reassign_task(task_id, None)
else:
# Request status update from agent
await self.message_queue.put({
"type": "status_request",
"task_id": task_id,
"agent_id": task.assigned_to,
"timestamp": datetime.now()
})
async def _handle_missed_deadline(self, task_id: str):
"""Handle a missed deadline."""
task = self.tasks[task_id]
# Log the incident
self.logger.warning(f"Task {task_id} missed deadline: {task.deadline}")
# Update task priority to CRITICAL
task.priority = TaskPriority.CRITICAL
# If task is assigned, try to speed it up
if task.assigned_to:
await self.message_queue.put({
"type": "expedite_request",
"task_id": task_id,
"agent_id": task.assigned_to,
"timestamp": datetime.now()
})
else:
# If not assigned, try to assign to fastest available agent
await self._plan_task_execution(task_id)
async def _reassign_task(self, task_id: str, new_agent_id: Optional[str] = None):
"""Reassign a task to a new agent."""
task = self.tasks[task_id]
old_agent_id = task.assigned_to
if old_agent_id:
# Update old agent
old_agent = self.agents[old_agent_id]
old_agent.load -= 1
if old_agent.load <= 0:
old_agent.state = AgentState.IDLE
if new_agent_id is None:
# Find new suitable agent
new_agent_id = await self._find_suitable_agent(task)
if new_agent_id:
# Assign to new agent
await self._assign_task(task_id, new_agent_id)
else:
# No suitable agent found, mark task as pending
task.state = "pending"
task.assigned_to = None
def _register_error_handlers(self):
"""Register basic error handlers."""
self.error_handlers.update({
"monitoring_error": self._handle_monitoring_error,
"agent_error": self._handle_agent_error,
"task_error": self._handle_task_error,
"resource_error": self._handle_resource_error
})
self.recovery_strategies.update({
"agent_recovery": self._recover_agent,
"task_recovery": self._recover_task,
"resource_recovery": self._recover_resource
})
async def _handle_error(self, error_type: str, error: Exception):
"""Handle an error using registered handlers."""
handler = self.error_handlers.get(error_type)
if handler:
try:
await handler(error)
except Exception as e:
self.logger.error(f"Error in error handler: {e}")
else:
self.logger.error(f"No handler for error type: {error_type}")
self.logger.error(f"Error: {error}")
async def _handle_monitoring_error(self, error: Exception):
"""Handle monitoring system errors."""
self.logger.error(f"Monitoring error: {error}")
# Implement recovery logic
pass
async def _handle_agent_error(self, error: Exception):
"""Handle agent-related errors."""
self.logger.error(f"Agent error: {error}")
# Implement recovery logic
pass
async def _handle_task_error(self, error: Exception):
"""Handle task-related errors."""
self.logger.error(f"Task error: {error}")
# Implement recovery logic
pass
async def _handle_resource_error(self, error: Exception):
"""Handle resource-related errors."""
self.logger.error(f"Resource error: {error}")
# Implement recovery logic
pass
async def _recover_agent(self, agent_id: str):
"""Recover a failed agent."""
try:
agent = self.agents[agent_id]
# Log recovery attempt
self.logger.info(f"Attempting to recover agent {agent_id}")
# Reset agent state
agent.state = AgentState.IDLE
agent.load = 0
agent.last_active = datetime.now()
# Reassign any tasks that were assigned to this agent
for task_id, task in self.tasks.items():
if task.assigned_to == agent_id:
await self._reassign_task(task_id)
# Update metrics
agent.metrics["recovery_attempts"] = agent.metrics.get("recovery_attempts", 0) + 1
self.logger.info(f"Successfully recovered agent {agent_id}")
return True
except Exception as e:
self.logger.error(f"Failed to recover agent {agent_id}: {e}")
return False
async def _recover_task(self, task_id: str):
"""Recover a failed task."""
try:
task = self.tasks[task_id]
# Log recovery attempt
self.logger.info(f"Attempting to recover task {task_id}")
# Reset task state
task.state = "pending"
task.assigned_to = None
# Try to reassign the task
await self._reassign_task(task_id)
self.logger.info(f"Successfully recovered task {task_id}")
return True
except Exception as e:
self.logger.error(f"Failed to recover task {task_id}: {e}")
return False
async def _recover_resource(self, resource_id: str):
"""Recover a failed resource."""
try:
# Log recovery attempt
self.logger.info(f"Attempting to recover resource {resource_id}")
# Release any locks on the resource
if resource_id in self.resource_locks:
lock = self.resource_locks[resource_id]
if lock.locked():
lock.release()
# Reset resource state
if resource_id in self.resource_pool:
self.resource_pool[resource_id] = {
"state": "available",
"last_error": None,
"last_recovery": datetime.now()
}
self.logger.info(f"Successfully recovered resource {resource_id}")
return True
except Exception as e:
self.logger.error(f"Failed to recover resource {resource_id}: {e}")
return False
async def create_agent(self, role: AgentRole, capabilities: List[str]) -> str:
"""Create a new agent with specified role and capabilities."""
agent_id = str(uuid.uuid4())
agent_metadata = AgentMetadata(
id=agent_id,
role=role,
capabilities=capabilities,
state=AgentState.IDLE,
load=0.0,
last_active=datetime.now(),
metrics={
"tasks_completed": 0,
"success_rate": 1.0,
"avg_response_time": 0.0,
"resource_usage": 0.0
}
)
self.agents[agent_id] = agent_metadata
self.logger.info(f"Created new agent {agent_id} with role {role}")
return agent_id
|