ragaai-catalyst 2.1.5b30__py3-none-any.whl → 2.1.5b33__py3-none-any.whl

ragaai_catalyst/tracers/agentic_tracing/upload/trace_uploader.py

@@ -0,0 +1,675 @@
+"""
+trace_uploader.py - A dedicated process for handling trace uploads
+"""
+
+import os
+import sys
+import json
+import time
+import signal
+import logging
+import argparse
+import tempfile
+from pathlib import Path
+import multiprocessing
+import queue
+from datetime import datetime
+import atexit
+import glob
+
+# Set up logging
+log_dir = os.path.join(tempfile.gettempdir(), "ragaai_logs")
+os.makedirs(log_dir, exist_ok=True)
+
+logging.basicConfig(
+    level=logging.DEBUG,
+    format='%(asctime)s - %(name)s - %(levelname)s - %(message)s',
+    handlers=[
+        logging.StreamHandler(),
+        logging.FileHandler(os.path.join(log_dir, "trace_uploader.log"))
+    ]
+)
+logger = logging.getLogger("trace_uploader")
+
+try:
+    from ragaai_catalyst.tracers.agentic_tracing.upload.upload_agentic_traces import UploadAgenticTraces
+    from ragaai_catalyst.tracers.agentic_tracing.upload.upload_code import upload_code
+    from ragaai_catalyst.tracers.agentic_tracing.upload.upload_trace_metric import upload_trace_metric
+    from ragaai_catalyst.tracers.agentic_tracing.utils.create_dataset_schema import create_dataset_schema_with_trace
+    from ragaai_catalyst import RagaAICatalyst
+    IMPORTS_AVAILABLE = True
+except ImportError:
+    logger.warning("RagaAI Catalyst imports not available - running in test mode")
+    IMPORTS_AVAILABLE = False
+
+# Define task queue directory
+QUEUE_DIR = os.path.join(tempfile.gettempdir(), "ragaai_tasks")
+os.makedirs(QUEUE_DIR, exist_ok=True)
+
+# Clean up any stale processes
+def cleanup_stale_processes():
+    """Check for stale processes but allow active uploads to complete"""
+    pid_file = os.path.join(tempfile.gettempdir(), "trace_uploader.pid")
+    if os.path.exists(pid_file):
+        try:
+            with open(pid_file, "r") as f:
+                old_pid = int(f.read().strip())
+            try:
+                import psutil
+                if psutil.pid_exists(old_pid):
+                    p = psutil.Process(old_pid)
+                    if "trace_uploader.py" in " ".join(p.cmdline()):
+                        # Instead of terminating, just remove the PID file
+                        # This allows the process to finish its current uploads
+                        logger.info(f"Removing PID file for process {old_pid}")
+                        os.remove(pid_file)
+                        return
+            except Exception as e:
+                logger.warning(f"Error checking stale process: {e}")
+            os.remove(pid_file)
+        except Exception as e:
+            logger.warning(f"Error reading PID file: {e}")
+
+cleanup_stale_processes()
+
+# Status codes
+STATUS_PENDING = "pending"
+STATUS_PROCESSING = "processing"
+STATUS_COMPLETED = "completed"
+STATUS_FAILED = "failed"
+
+class UploadTask:
+    """Class representing a single upload task"""
+    
+    def __init__(self, task_id=None, **kwargs):
+        self.task_id = task_id or f"task_{int(time.time())}_{os.getpid()}_{hash(str(time.time()))}"
+        self.status = STATUS_PENDING
+        self.attempts = 0
+        self.max_attempts = 3
+        self.created_at = datetime.now().isoformat()
+        self.updated_at = self.created_at
+        self.error = None
+        
+        # Task details
+        self.filepath = kwargs.get("filepath")
+        self.hash_id = kwargs.get("hash_id")
+        self.zip_path = kwargs.get("zip_path")
+        self.project_name = kwargs.get("project_name")
+        self.project_id = kwargs.get("project_id")
+        self.dataset_name = kwargs.get("dataset_name")
+        self.user_details = kwargs.get("user_details", {})
+        self.base_url = kwargs.get("base_url")
+        
+    def to_dict(self):
+        """Convert task to dictionary for serialization"""
+        return {
+            "task_id": self.task_id,
+            "status": self.status,
+            "attempts": self.attempts,
+            "max_attempts": self.max_attempts,
+            "created_at": self.created_at,
+            "updated_at": self.updated_at,
+            "error": self.error,
+            "filepath": self.filepath,
+            "hash_id": self.hash_id,
+            "zip_path": self.zip_path,
+            "project_name": self.project_name,
+            "project_id": self.project_id,
+            "dataset_name": self.dataset_name,
+            "user_details": self.user_details,
+            "base_url": self.base_url
+        }
+        
+    @classmethod
+    def from_dict(cls, data):
+        """Create task from dictionary"""
+        task = cls(task_id=data.get("task_id"))
+        task.status = data.get("status", STATUS_PENDING)
+        task.attempts = data.get("attempts", 0)
+        task.max_attempts = data.get("max_attempts", 3)
+        task.created_at = data.get("created_at")
+        task.updated_at = data.get("updated_at")
+        task.error = data.get("error")
+        task.filepath = data.get("filepath")
+        task.hash_id = data.get("hash_id")
+        task.zip_path = data.get("zip_path")
+        task.project_name = data.get("project_name")
+        task.project_id = data.get("project_id")
+        task.dataset_name = data.get("dataset_name")
+        task.user_details = data.get("user_details", {})
+        task.base_url = data.get("base_url")    
+        return task
+        
+    def update_status(self, status, error=None):
+        """Update task status"""
+        self.status = status
+        self.updated_at = datetime.now().isoformat()
+        if error:
+            self.error = str(error)
+        self.save()
+        
+    def increment_attempts(self):
+        """Increment the attempt counter"""
+        self.attempts += 1
+        self.updated_at = datetime.now().isoformat()
+        self.save()
+        
+    def save(self):
+        """Save task to disk"""
+        task_path = os.path.join(QUEUE_DIR, f"{self.task_id}.json")
+        with open(task_path, "w") as f:
+            json.dump(self.to_dict(), f, indent=2)
+            
+    def delete(self):
+        """Delete task file from disk"""
+        task_path = os.path.join(QUEUE_DIR, f"{self.task_id}.json")
+        if os.path.exists(task_path):
+            os.remove(task_path)
+            
+    @staticmethod
+    def list_pending_tasks():
+        """List all pending tasks"""
+        tasks = []
+        logger.info("Listing pending tasks from queue directory: {}".format(QUEUE_DIR))
+        for filename in os.listdir(QUEUE_DIR):
+            if filename.endswith(".json"):
+                try:
+                    with open(os.path.join(QUEUE_DIR, filename), "r") as f:
+                        task_data = json.load(f)
+                        task = UploadTask.from_dict(task_data)
+                        if task.status in [STATUS_PENDING, STATUS_FAILED] and task.attempts < task.max_attempts:
+                            # Verify files still exist
+                            if (not task.filepath or os.path.exists(task.filepath)) and \
+                               (not task.zip_path or os.path.exists(task.zip_path)):
+                                tasks.append(task)
+                            else:
+                                # Files missing, mark as failed
+                                task.update_status(STATUS_FAILED, "Required files missing")
+                except Exception as e:
+                    logger.error(f"Error loading task {filename}: {e}")
+        return tasks
+
+
+class TraceUploader:
+    """
+    Trace uploader process
+    Handles the actual upload work in a separate process
+    """
+    
+    def __init__(self):
+        self.running = True
+        self.processing = False
+        
+    def start(self):
+        """Start the uploader loop"""
+        logger.info("Trace uploader starting")
+        
+        # Register signal handlers
+        signal.signal(signal.SIGTERM, self.handle_signal)
+        signal.signal(signal.SIGINT, self.handle_signal)
+        
+        # Register cleanup
+        atexit.register(self.cleanup)
+        
+        # Main processing loop
+        while self.running:
+            try:
+                # Get pending tasks
+                tasks = UploadTask.list_pending_tasks()
+                if tasks:
+                    logger.info(f"Found {len(tasks)} pending tasks")
+                    for task in tasks:
+                        if not self.running:
+                            break
+                        self.process_task(task)
+                else:
+                    # No tasks, sleep before checking again
+                    time.sleep(5)
+            except Exception as e:
+                logger.error(f"Error in uploader loop: {e}")
+                time.sleep(5)
+                
+        logger.info("Trace uploader stopped")
+        
+    def process_task(self, task):
+        """Process a single upload task"""
+        logger.info(f"Starting to process task {task.task_id}")
+        logger.debug(f"Task details: {task.to_dict()}")
+
+        # Check if file exists
+        if not os.path.exists(task.filepath):
+            error_msg = f"Task filepath does not exist: {task.filepath}"
+            logger.error(error_msg)
+            task.update_status(STATUS_FAILED, error_msg)
+            return
+
+        if not IMPORTS_AVAILABLE:
+            logger.warning(f"Test mode: Simulating processing of task {task.task_id}")
+            time.sleep(2)  # Simulate work
+            task.update_status(STATUS_COMPLETED)
+            return
+            
+        logger.info(f"Processing task {task.task_id} (attempt {task.attempts+1}/{task.max_attempts})")
+        self.processing = True
+        task.update_status(STATUS_PROCESSING)
+        task.increment_attempts()
+        
+        # Log memory state for debugging
+        try:
+            import psutil
+            process = psutil.Process()
+            logger.debug(f"Memory usage before processing: {process.memory_info().rss / 1024 / 1024:.2f} MB")
+        except ImportError:
+            pass
+        
+        try:
+            # Step 1: Create dataset schema
+            logger.info(f"Creating dataset schema for {task.dataset_name} with base_url: {task.base_url}")
+            response = create_dataset_schema_with_trace(
+                dataset_name=task.dataset_name,
+                project_name=task.project_name,
+                base_url=task.base_url
+            )
+            logger.info(f"Dataset schema created: {response}")
+            
+            # Step 2: Upload trace metrics
+            if task.filepath and os.path.exists(task.filepath):
+                logger.info(f"Uploading trace metrics for {task.filepath}")
+                try:
+                    response = upload_trace_metric(
+                        json_file_path=task.filepath,
+                        dataset_name=task.dataset_name,
+                        project_name=task.project_name,
+                        base_url=task.base_url
+                    )
+                    logger.info(f"Trace metrics uploaded: {response}")
+                except Exception as e:
+                    logger.error(f"Error uploading trace metrics: {e}")
+                    # Continue with other uploads
+            else:
+                logger.warning(f"Trace file {task.filepath} not found, skipping metrics upload")
+            
+            # Step 3: Upload agentic traces
+            if task.filepath and os.path.exists(task.filepath):
+                logger.info(f"Uploading agentic traces for {task.filepath}")
+                try:
+                    upload_traces = UploadAgenticTraces(
+                        json_file_path=task.filepath,
+                        project_name=task.project_name,
+                        project_id=task.project_id,
+                        dataset_name=task.dataset_name,
+                        user_detail=task.user_details,
+                        base_url=task.base_url,   
+                    )
+                    upload_traces.upload_agentic_traces()
+                    logger.info("Agentic traces uploaded successfully")
+                except Exception as e:
+                    logger.error(f"Error uploading agentic traces: {e}")
+                    # Continue with code upload
+            else:
+                logger.warning(f"Trace file {task.filepath} not found, skipping traces upload")
+            
+            # Step 4: Upload code hash
+            if task.hash_id and task.zip_path and os.path.exists(task.zip_path):
+                logger.info(f"Uploading code hash {task.hash_id}")
+                try:
+                    response = upload_code(
+                        hash_id=task.hash_id,
+                        zip_path=task.zip_path,
+                        project_name=task.project_name,
+                        dataset_name=task.dataset_name,
+                        base_url=task.base_url
+                    )
+                    logger.info(f"Code hash uploaded: {response}")
+                except Exception as e:
+                    logger.error(f"Error uploading code hash: {e}")
+            else:
+                logger.warning(f"Code zip {task.zip_path} not found, skipping code upload")
+            
+            # Mark task as completed
+            task.update_status(STATUS_COMPLETED)
+            logger.info(f"Task {task.task_id} completed successfully")
+            
+            # Clean up task file
+            task.delete()
+            
+        except Exception as e:
+            logger.error(f"Error processing task {task.task_id}: {e}")
+            if task.attempts >= task.max_attempts:
+                task.update_status(STATUS_FAILED, str(e))
+                logger.error(f"Task {task.task_id} failed after {task.attempts} attempts")
+            else:
+                task.update_status(STATUS_PENDING, str(e))
+                logger.warning(f"Task {task.task_id} will be retried (attempt {task.attempts}/{task.max_attempts})")
+        finally:
+            self.processing = False
+            
+    def handle_signal(self, signum, frame):
+        """Handle termination signals"""
+        logger.info(f"Received signal {signum}, shutting down gracefully")
+        self.running = False
+        
+    def cleanup(self):
+        """Cleanup before exit"""
+        logger.info("Performing cleanup before exit")
+        self.running = False
+        
+
+def submit_upload_task(filepath, hash_id, zip_path, project_name, project_id, dataset_name, user_details, base_url):
+    """
+    Submit a new upload task to the queue.
+    This function can be called from the main application.
+    
+    Returns:
+        str: Task ID
+    """
+    logger.info(f"Submitting new upload task for file: {filepath}")
+    logger.debug(f"Task details - Project: {project_name}, Dataset: {dataset_name}, Hash: {hash_id}, Base_URL: {base_url}")
+    
+    # Verify the trace file exists
+    if not os.path.exists(filepath):
+        logger.error(f"Trace file not found: {filepath}")
+        return None
+
+    # Create task with absolute path to the trace file
+    filepath = os.path.abspath(filepath)
+    logger.debug(f"Using absolute filepath: {filepath}")
+
+    task = UploadTask(
+        filepath=filepath,
+        hash_id=hash_id,
+        zip_path=zip_path,
+        project_name=project_name,
+        project_id=project_id,
+        dataset_name=dataset_name,
+        user_details=user_details,
+        base_url=base_url
+    )
+    
+    # Save the task with proper error handling
+    task_path = os.path.join(QUEUE_DIR, f"{task.task_id}.json")
+    logger.debug(f"Saving task to: {task_path}")
+    
+    try:
+        # Ensure queue directory exists
+        os.makedirs(QUEUE_DIR, exist_ok=True)
+        
+        with open(task_path, "w") as f:
+            json.dump(task.to_dict(), f, indent=2)
+            
+        logger.info(f"Task {task.task_id} created successfully for trace file: {filepath}")
+    except Exception as e:
+        logger.error(f"Error creating task file: {e}", exc_info=True)
+        return None
+    
+    # Ensure uploader process is running
+    logger.info("Starting uploader process...")
+    pid = ensure_uploader_running()
+    if pid:
+        logger.info(f"Uploader process running with PID {pid}")
+    else:
+        logger.warning("Failed to start uploader process, but task was queued")
+    
+    return task.task_id
+
+
+def get_task_status(task_id):
+    """
+    Get the status of a task by ID.
+    This function can be called from the main application.
+    
+    Returns:
+        dict: Task status information
+    """
+    task_path = os.path.join(QUEUE_DIR, f"{task_id}.json")
+    if not os.path.exists(task_path):
+        # Check if it might be in completed directory
+        completed_path = os.path.join(QUEUE_DIR, "completed", f"{task_id}.json")
+        if os.path.exists(completed_path):
+            with open(completed_path, "r") as f:
+                return json.load(f)
+        return {"status": "unknown", "error": "Task not found"}
+    
+    with open(task_path, "r") as f:
+        return json.load(f)
+
+
+def ensure_uploader_running():
+    """
+    Ensure the uploader process is running.
+    Starts it if not already running.
+    """
+    logger.info("Checking if uploader process is running...")
+    
+    # Check if we can find a running process
+    pid_file = os.path.join(tempfile.gettempdir(), "trace_uploader.pid")
+    logger.debug(f"PID file location: {pid_file}")
+    
+    if os.path.exists(pid_file):
+        try:
+            with open(pid_file, "r") as f:
+                pid_str = f.read().strip()
+                logger.debug(f"Read PID from file: {pid_str}")
+                pid = int(pid_str)
+            
+            # Check if process is actually running
+            # Use platform-specific process check
+            is_running = False
+            try:
+                if os.name == 'posix':  # Unix/Linux/Mac
+                    logger.debug(f"Checking process {pid} on Unix/Mac")
+                    os.kill(pid, 0)  # This raises an exception if process doesn't exist
+                    is_running = True
+                else:  # Windows
+                    logger.debug(f"Checking process {pid} on Windows")
+                    import ctypes
+                    kernel32 = ctypes.windll.kernel32
+                    SYNCHRONIZE = 0x00100000
+                    process = kernel32.OpenProcess(SYNCHRONIZE, False, pid)
+                    if process:
+                        kernel32.CloseHandle(process)
+                        is_running = True
+            except (ImportError, AttributeError) as e:
+                logger.debug(f"Platform-specific check failed: {e}, falling back to cross-platform check")
+                # Fall back to cross-platform check
+                try:
+                    import psutil
+                    is_running = psutil.pid_exists(pid)
+                    logger.debug(f"psutil check result: {is_running}")
+                except ImportError:
+                    logger.debug("psutil not available, using basic process check")
+                    # If psutil is not available, make a best guess
+                    try:
+                        os.kill(pid, 0)
+                        is_running = True
+                    except Exception as e:
+                        logger.debug(f"Basic process check failed: {e}")
+                        is_running = False
+            
+            if is_running:
+                logger.debug(f"Uploader process already running with PID {pid}")
+                return pid
+        except (ProcessLookupError, ValueError, PermissionError):
+            # Process not running or other error, remove stale PID file
+            try:
+                os.remove(pid_file)
+            except:
+                pass
+    
+    # Start new process
+    logger.info("Starting new uploader process")
+    
+    # Get the path to this script
+    script_path = os.path.abspath(__file__)
+    
+    # Start detached process in a platform-specific way
+    try:
+        # First, try the preferred method for each platform
+        if os.name == 'posix':  # Unix/Linux/Mac
+            import subprocess
+            
+            # Use double fork method on Unix systems
+            try:
+                # First fork
+                pid = os.fork()
+                if pid > 0:
+                    # Parent process, return
+                    return pid
+                    
+                # Decouple from parent environment
+                os.chdir('/')
+                os.setsid()
+                os.umask(0)
+                
+                # Second fork
+                pid = os.fork()
+                if pid > 0:
+                    # Exit from second parent
+                    os._exit(0)
+                    
+                # Redirect standard file descriptors
+                sys.stdout.flush()
+                sys.stderr.flush()
+                si = open(os.devnull, 'r')
+                so = open(os.path.join(tempfile.gettempdir(), 'trace_uploader_stdout.log'), 'a+')
+                se = open(os.path.join(tempfile.gettempdir(), 'trace_uploader_stderr.log'), 'a+')
+                os.dup2(si.fileno(), sys.stdin.fileno())
+                os.dup2(so.fileno(), sys.stdout.fileno())
+                os.dup2(se.fileno(), sys.stderr.fileno())
+                
+                # Execute the daemon process
+                os.execl(sys.executable, sys.executable, script_path, '--daemon')
+                
+            except (AttributeError, OSError):
+                # Fork not available, try subprocess
+                process = subprocess.Popen(
+                    [sys.executable, script_path, "--daemon"],
+                    stdout=subprocess.PIPE,
+                    stderr=subprocess.PIPE,
+                    stdin=subprocess.PIPE,
+                    start_new_session=True  # Detach from parent
+                )
+                pid = process.pid
+                
+        else:  # Windows
+            import subprocess
+            # Use the DETACHED_PROCESS flag on Windows
+            startupinfo = subprocess.STARTUPINFO()
+            startupinfo.dwFlags |= subprocess.STARTF_USESHOWWINDOW
+            startupinfo.wShowWindow = 0  # SW_HIDE
+            
+            # Windows-specific flags
+            DETACHED_PROCESS = 0x00000008
+            CREATE_NO_WINDOW = 0x08000000
+            
+            process = subprocess.Popen(
+                [sys.executable, script_path, "--daemon"],
+                stdout=subprocess.PIPE,
+                stderr=subprocess.PIPE,
+                stdin=subprocess.PIPE,
+                startupinfo=startupinfo,
+                creationflags=DETACHED_PROCESS | CREATE_NO_WINDOW
+            )
+            pid = process.pid
+        
+        # Write PID to file
+        with open(pid_file, "w") as f:
+            f.write(str(pid))
+            
+        logger.info(f"Started uploader process with PID {pid}")
+        return pid
+        
+    except Exception as e:
+        logger.error(f"Error starting uploader process using primary method: {e}")
+        
+        # Fallback method using multiprocessing (works on most platforms)
+        try:
+            logger.info("Trying fallback method with multiprocessing")
+            import multiprocessing
+            
+            def run_uploader():
+                """Run the uploader in a separate process"""
+                # Redirect output
+                sys.stdout = open(os.path.join(tempfile.gettempdir(), 'trace_uploader_stdout.log'), 'a+')
+                sys.stderr = open(os.path.join(tempfile.gettempdir(), 'trace_uploader_stderr.log'), 'a+')
+                
+                # Run daemon
+                run_daemon()
+                
+            # Start process
+            process = multiprocessing.Process(target=run_uploader)
+            process.daemon = True  # Daemonize it
+            process.start()
+            pid = process.pid
+            
+            # Write PID to file
+            with open(pid_file, "w") as f:
+                f.write(str(pid))
+                
+            logger.info(f"Started uploader process with fallback method, PID {pid}")
+            return pid
+            
+        except Exception as e2:
+            logger.error(f"Error starting uploader process using fallback method: {e2}")
+            
+            # Super fallback - run in the current process if all else fails
+            # This is not ideal but better than failing completely
+            try:
+                logger.warning("Using emergency fallback - running in current process thread")
+                import threading
+                
+                thread = threading.Thread(target=run_daemon, daemon=True)
+                thread.start()
+                
+                # No real PID since it's a thread, but we'll create a marker file
+                with open(pid_file, "w") as f:
+                    f.write(f"thread_{id(thread)}")
+                    
+                return None
+            except Exception as e3:
+                logger.error(f"All methods failed to start uploader: {e3}")
+                return None
+
+
+def run_daemon():
+    """Run the uploader as a daemon process"""
+    # Write PID to file
+    pid_file = os.path.join(tempfile.gettempdir(), "trace_uploader.pid")
+    with open(pid_file, "w") as f:
+        f.write(str(os.getpid()))
+        
+    try:
+        uploader = TraceUploader()
+        uploader.start()
+    finally:
+        # Clean up PID file
+        if os.path.exists(pid_file):
+            os.remove(pid_file)
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(description="Trace uploader process")
+    parser.add_argument("--daemon", action="store_true", help="Run as daemon process")
+    parser.add_argument("--test", action="store_true", help="Submit a test task")
+    args = parser.parse_args()
+    
+    if args.daemon:
+        run_daemon()
+    elif args.test:
+        # Submit a test task
+        test_file = os.path.join(tempfile.gettempdir(), "test_trace.json")
+        with open(test_file, "w") as f:
+            f.write("{}")
+        
+        task_id = submit_upload_task(
+            filepath=test_file,
+            hash_id="test_hash",
+            zip_path=test_file,
+            project_name="test_project",
+            project_id="test_id",
+            dataset_name="test_dataset",
+            user_details={"id": "test_user"}
+        )
+        print(f"Submitted test task with ID: {task_id}")
+    else:
+        print("Use --daemon to run as daemon or --test to submit a test task")