Data Pipeline Observability

Overview

Monitoring and alerting for ML data pipeline health. This guide covers practical implementation for production ML systems.

Why This Matters in MLOps

Modern ML systems require rigorous operations practices:

Reliability: Models must perform consistently in production

Reproducibility: Experiments must be tracked and reproducible

Scalability: Systems must handle growing data and traffic

Observability: You need visibility into what models are doing

Setup

bash
Install required tools
pip install opentelemetry mlflow pandas numpy scikit-learn
Or with Docker
docker pull python:3.11-slim

Core Implementation

python
import os
import json
import logging
from datetime import datetime
from pathlib import Path
logger = logging.getLogger(__name__)
class DataPipelineObservability:
    """
    Data Pipeline Observability implementation.
    
    Handles: observability
    Tool: opentelemetry
    """
    
    def __init__(self, config: dict = None):
        self.config = config or self._default_config()
        self._setup()
    
    def _default_config(self) -> dict:
        return {
            "tool": "opentelemetry",
            "environment": os.getenv("ENVIRONMENT", "development"),
            "log_level": "INFO",
        }
    
    def _setup(self):
        """Initialize opentelemetry connection and resources."""
        logging.basicConfig(level=self.config.get("log_level", "INFO"))
        logger.info(f"Initialized Data Pipeline Observability with config: {self.config}")
    
    def run(self, **kwargs) -> dict:
        """Execute observability."""
        start = datetime.utcnow()
        
        try:
            result = self._execute(**kwargs)
            
            elapsed = (datetime.utcnow() - start).total_seconds()
            logger.info(f"Data Pipeline Observability completed in {elapsed:.2f}s")
            
            return {
                "status": "success",
                "result": result,
                "elapsed_seconds": elapsed
            }
        
        except Exception as e:
            logger.error(f"Data Pipeline Observability failed: {e}")
            return {
                "status": "failed", 
                "error": str(e)
            }
    
    def _execute(self, **kwargs) -> dict:
        """Core observability logic. Override to customize."""
        return {"completed": True, "tool": "opentelemetry"}
Configuration
config = {
    "tool": "opentelemetry",
    "tracking_uri": os.getenv("MLFLOW_TRACKING_URI", "http://localhost:5000"),
    "artifact_root": "./artifacts",
}
Initialize
processor = DataPipelineObservability(config)
result = processor.run()
print(json.dumps(result, indent=2))

OPENTELEMETRY Integration

python
Specific opentelemetry integration for observability
import subprocess
def setup_opentelemetry():
    """Configure opentelemetry for observability."""
    
    # Initialize project
    print(f"Setting up opentelemetry for observability...")
    
    # Example configuration
    config = {
        "project": "my-ml-project",
        "tool": "opentelemetry",
        "specialty": "observability",
        "version": "1.0.0"
    }
    
    # Save configuration
    Path(".opentelemetry").mkdir(exist_ok=True)
    with open(f".opentelemetry/config.json", "w") as f:
        json.dump(config, f, indent=2)
    
    print(f"opentelemetry configured for observability")
    return config
config = setup_opentelemetry()

Monitoring and Alerting

python
from dataclasses import dataclass
import time
@dataclass
class MetricSnapshot:
    timestamp: float
    metric_name: str
    value: float
    labels: dict
class MLOpsMonitor:
    """Monitor observability metrics."""
    
    def __init__(self):
        self.metrics: list[MetricSnapshot] = []
        self.thresholds = {
            "error_rate": 0.05,
            "latency_p99_ms": 1000,
            "data_drift_score": 0.3
        }
    
    def record(self, metric: str, value: float, labels: dict = None):
        snapshot = MetricSnapshot(
            timestamp=time.time(),
            metric_name=metric,
            value=value,
            labels=labels or {}
        )
        self.metrics.append(snapshot)
        self._check_threshold(metric, value)
    
    def _check_threshold(self, metric: str, value: float):
        threshold = self.thresholds.get(metric)
        if threshold and value > threshold:
            logger.warning(f"ALERT: {metric}={value:.3f} exceeds threshold {threshold}")
monitor = MLOpsMonitor()

CI/CD Integration

yaml
.github/workflows/ml-pipeline.yml
name: ML Pipeline
on:
  push:
    paths: ['src/', 'data/']
jobs:
  train-and-evaluate:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      
      - name: Setup Python
        uses: actions/setup-python@v5
        with:
          python-version: '3.11'
      
      - name: Install dependencies
        run: pip install -r requirements.txt
      
      - name: Run observability
        run: python -m src.data_pipeline_observability
        env:
          MLFLOW_TRACKING_URI: ${{ secrets.MLFLOW_URI }}
      
      - name: Check model quality
        run: python -m src.validate_model

Best Practices

Version everything — models, data, configs, and code

Automate testing — catch regressions before production

Monitor continuously — don't wait for users to report issues

Document experiments — future you will thank present you

Use feature flags — control rollouts without code changes

Resources

MLflow documentation: https://mlflow.org/docs

DVC documentation: https://dvc.org/doc

Kubeflow documentation: https://www.kubeflow.org/docs

Made With ML MLOps course: https://madewithml.com