Aller au contenu

Cross-Cutting Concerns

The data architecture must include robust security measures, such as encryption, access controls, and audit logging.

Stripe must ensure that its data storage and processing meet regulatory requirements across different jurisdictions.

The system should support automated compliance reporting and real-time monitoring for security breaches.

Security & Compliance

Requirement Solution Tools
Encryption AES-256 (at rest), TLS 1.3 (in transit) Snowflake, PostgreSQL
Access Control RBAC + ABAC HashiCorp Vault
Audit Logging Immutable logs Snowflake Audit Logs, pgAudit
PCI-DSS Compliance Tokenization of PII Protegrity
GDPR Compliance Right to erasure scripts Custom Automation

Monitoring

  • Prometheus collects metrics from PostgreSQL, Kafka, Snowflake, and MongoDB.

  • Grafana visualizes these metrics for observability across the pipeline.

Metric Tool Threshold Applies To
Query Latency Prometheus + Grafana > 500ms OLAP
Query Failure Rate Snowflake Query History > 1% OLAP
Data Freshness (Replication Lag) Kafka Lag Script > 5 minutes Pipeline (OLTP→OLAP)
Storage Usage Snowflake Storage Metrics > 80% capacity OLAP
Transaction Throughput pg_stat_activity / Prometheus OLTP
Replication/WAL Lag pg_stat_replication OLTP
Connection Pool Saturation PgBouncer stats OLTP

Tech Stack

Component Technology Why?
OLTP Database Normalized 3NF, PostgreSQL ACID compliance, transactional integrity, minimize redundancy, support high throughput.
OLAP Database Star schema, PostgreSQL Handle unstructured data, support ML features and flexible queries.
NoSQL Database MongoDB, Document Flexible schema, JSON support, storage for unstructured data and ML features, horizontal scaling.
Data Pipeline Kafka + Debezium (CDC), Spark (batch) Real-time and batch data flow, error handling, transformation.
Batch Orchestration Apache Airflow Schedule and monitor ETL jobs.
Storage Snowflake (Parquet) Optimized for analytical queries.
Security & Compliance AES-256, RBAC, Audit Logging, Protegrity Ensure GDPR, PCI-DSS, CCPA compliance and data protection.
Machine Learning TensorFlow Serving, MLflow, Kubernetes Real-time fraud detection, customer personalization, model monitoring.
Caching Redis Low-latency caching.
Monitoring Prometheus + Grafana Track performance and costs.