Project CloudBridge – Day 1 & 2 Summary (Architect Consolidation)

Project CloudBridge – Brief Summary (Day 1 & Day 2)

This short recap consolidates the core learning from Day 1 and Day 2, so we can move into Day 3 (hands-on build) with a clear production mindset.

---

✅ Day 1 – Why Enterprises Separate OLTP and Analytics

Day 1 established the foundational enterprise principle: Transactional systems (OLTP) and analytical systems (Analytics/OLAP) are built for different workload patterns.

OLTP	Analytics
Small, frequent transactions	Large scans & aggregations
Low latency (milliseconds)	High throughput reporting
Predictable response time	Long-running queries

Enterprise Rule: Never run heavy analytics workload on the production OLTP database. It creates CPU/I/O contention, lock waits, and SLA impact.

The enterprise solution pattern introduced: separate OLTP and Analytics and connect them using a replication layer.

---

✅ Day 2 – Enterprise Real-Time Architecture Design (Deep Mode)

Day 2 moved from principle to production design thinking. We designed a real-time architecture using: RDS PostgreSQL (OLTP), AWS DMS (Full Load + CDC), and Amazon Redshift (Analytics).

Customer Applications ↓ Amazon RDS PostgreSQL (OLTP) ↓ AWS DMS (Full Load + CDC) ↓ Amazon Redshift (Analytics Warehouse) ↓ BI / Dashboards

Key technical decisions and learnings:

• RDS (OLTP): design for low latency, high availability (Multi-AZ), and predictable storage performance.
• CDC (Change Data Capture): DMS reads PostgreSQL WAL to capture INSERT/UPDATE/DELETE with minimal OLTP impact.
• WAL/Slot Risk: if DMS lags or stops, WAL can accumulate (replication slot retention), creating storage pressure.
• DMS Sizing Matters: under-sized replication instances lead to replication lag and stale analytics.
• Redshift (Analytics): purpose-built warehouse for heavy aggregations and large scans; scales independently from OLTP.

Day 2 Outcome: We shifted from “tool learning” to architect thinking — workload isolation, CDC design, sizing awareness, and failure scenario planning.

---

🔜 Day 3 – What Comes Next (Hands-on Build)

Day 3 is execution with a production mindset:

• Create RDS PostgreSQL with proper sizing and storage
• Configure parameter group for CDC (WAL/logical replication)
• Plan networking (subnets, security groups) and access model
• Prepare DMS replication instance sizing and monitoring approach

Tags: #AWS #CloudBridge #RDS #PostgreSQL #DMS #CDC #Redshift #RealTimeAnalytics #DataEngineering