Link copied to clipboard!

Group 7: NoSQL & In-Memory

Non-relational and caching databases: Firestore (document database), Bigtable (wide-column), Memorystore (Redis/Memcached), Firebase Realtime Database (mobile sync). Each serves different data models: documents, wide-column, key-value caching, and real-time synchronization.

Data Model Selection: Firestore for documents with rich queries; Bigtable for time-series and wide-column data; Memorystore for caching and session storage; Firebase Realtime Database for mobile real-time sync.

Services & Core Identity

Firestore

Serverless document database with real-time synchronization, offline support, and ACID transactions.

Bigtable

High-performance wide-column database for time-series, IoT, and analytics workloads requiring low latency.

Memorystore

Fully managed Redis and Memcached for caching, session storage, and high-speed data access.

Firebase Realtime Database

NoSQL cloud database optimized for mobile applications with real-time synchronization and offline capabilities.

Key Differences

Dimension	Firestore	Bigtable	Memorystore	Firebase Realtime DB
Data Model	Document	Wide-column	Key-value	JSON tree
Query Capabilities	Rich queries	Row key/scan	Key operations	Limited queries
Consistency	Strong	Eventual	Strong	Eventual
Scaling	Automatic	Manual clusters	Manual/automatic	Automatic
Use Cases	Web/mobile apps	Time-series, IoT	Caching, sessions	Mobile sync
Latency	Single-digit ms	Sub-10ms	Sub-millisecond	Real-time

Mathematical Selection Model

Criteria [0..10]. Higher scores indicate better database fit.

Document Data Time-Series Data Caching Need Mobile App Real-time Sync High Throughput Complex Queries Offline Support

Score_Firestore = 0.30*C_documentData + 0.25*C_complexQueries + 0.20*C_offlineSupport + 0.15*C_transactions + 0.10*C_serverless Score_Bigtable = 0.35*C_timeSeriesData + 0.25*C_highThroughput + 0.20*C_lowLatency + 0.15*C_iotWorkloads + 0.05*(10 - C_complexQueries) Score_Memorystore = 0.40*C_cachingNeed + 0.25*C_sessionStorage + 0.20*C_ultraLowLatency + 0.10*C_dataStructures + 0.05*(10 - C_offlineSupport) Score_FirebaseRealtime = 0.35*C_mobileApp + 0.30*C_realTimeSync + 0.20*C_offlineSupport + 0.10*C_simpleData + 0.05*(10 - C_complexQueries)

{{s.name}}: {{s.val | number:2}}

Interpretation Rules

Firestore: Content management, user profiles, catalog data, collaborative applications
Bigtable: Time-series analytics, IoT data ingestion, financial data, AdTech
Memorystore: Session management, gaming leaderboards, real-time analytics, cache layer
Firebase Realtime DB: Chat applications, gaming, collaborative editing, mobile-first apps

When NOT to Use NoSQL

Complex relational data with ACID requirements (consider Cloud SQL)
Analytical workloads requiring SQL (consider BigQuery)
Strong consistency across global regions (consider Cloud Spanner)
Complex reporting and business intelligence (consider data warehouse solutions)

Summary

GCP NoSQL services address different data patterns and performance requirements. Firestore excels for document-oriented applications, Bigtable handles high-throughput time-series data, Memorystore provides ultra-fast caching, and Firebase Realtime Database enables mobile real-time synchronization. Choose based on data model, consistency requirements, and access patterns.

Scan with Phone