Cassandra

Apache Cassandra is a distributed, NoSQL database engineered for large-scale, real-time applications with high availability, scalability, and robust write performance.

Key Highlights

1. Core Characteristics

• Architecture: Operates on a peer-to-peer model—no master node—ensuring there's no single point of failure.

• Data Partitioning & Replication: Uses consistent hashing to distribute data, along with tunable consistency levels for flexible read/write settings.

• Data Storage & Processing:

  • Write Path: Data is first written to a commit log, then to an in-memory memtable, and finally flushed to immutable SSTables on disk.

  • Read Path: Uses both the memtable and SSTables, optimized via bloom filters.

    
    

2. Use Cases

Best suited for scenarios involving:

• IoT and sensor data

• Time-series data

• Web activity tracking

• Real-time analytics

  
  

3. Pros & Cons

• Pros:

  • High scalability (easily adds new nodes)

  • Exceptional write performance

  • Strong availability and fault tolerance

• Cons:

  • Complex data modeling required for efficiency

  • Eventual consistency may not suit all applications

  • Operational overhead in setup and maintenance

• Origins & Design Philosophy

  Cassandra was born at Facebook to power Inbox Search (by Avinash Lakshman & Prashant Malik), drawing inspiration from Amazon’s Dynamo and Google’s Bigtable.

• AP (Availability & Partition Tolerance) Focus

  Classified as an AP system, Cassandra prioritizes availability and partition tolerance while providing tunable consistency options—even though it may lean toward eventual consistency over strict consistency.

• Scalability & Fault Tolerance Mechanics

  Features like gossip protocol (for node communication), seed nodes, hinted handoff, and the Φ Accrual Failure Detector contribute to its resilient, fault-tolerant architecture.

Summary Table

Category	Summary
What	Distributed NoSQL database with peer-to-peer architecture
Strengths	Highly scalable, fault-tolerant, excels at heavy write operations
Core Design	Commit log → Memtable → SSTable flow, bloom filters for optimized reads
Use Cases	IoT, time-series, logs, web tracking, real-time analytics
Trade-offs	Eventual consistency, data modeling complexity, maintenance overhead