Skip to main content
CHEAT SHEET #03Quick Reference Architecture•5 min read

Cloud Database Decision Matrix & Trade-Offs Cheat Sheet

Systematic evaluation framework for choosing between Relational, Key-Value, Document, Columnar, Time-Series, and Vector databases.

FreeCheat SheetAWSArchitecture

Systematic evaluation framework for choosing between Relational, Key-Value, Document, Columnar, Time-Series, and Vector databases.


1. The Database Family Matrix

CategoryTypical EnginesCore Data StructureScaling ModelOptimal Use Cases
Relational (RDBMS), MySQL, , CockroachDBB+ Tree on Disk / Shared StorageScale-up (or distributed NewSQL)Financial transactions, inventory, strict ACID schemas
Key-Value, , RocksDB, AerospikeHash Table / Inverted Memory IndexHorizontal consistent hash partitionSession store, user profiles, rate limiting counters
Document Store, Amazon DocumentDB, CouchbaseB-Tree / JSON BSON treesSharded document collectionsContent management, product catalogs, dynamic schemas
Wide-Column (Columnar), , Bigtable (Log-Structured Merge)Masterless peer-to-peer ringMassive write ingestion, clickstream analytics
Time-SeriesInfluxDB, TimescaleDB, Amazon TimestreamAppend-only delta-compressed chunksTime-based partitioning & retentionIoT metrics, server health telemetry, stock tickers
Vector DBPinecone, Milvus, pgvector, ()Sharded vector index partitionsLLM embeddings, similarity search, recommendation engines

2. Storage Engine Mechanics: B+ Tree vs. LSM-Tree

Interactive Architecture Diagram
Synthesizing vector architecture diagram...
DimensionB+ Tree Architecture Architecture
Write CostHigh: Random in-place disk page overwrites; write amplification.Low: Sequential append-only writes to and .
Read CostLow: Predictable O(log N) lookup; single leaf page read.Moderate to High: Checks , , and multiple .
Compaction OverheadNone: Pages updated directly on disk.High: Background merging of can cause periodic I/O latency spikes.
Space AmplificationMedium: Page fragmentation (typically 50-70% fill factor).Low: are immutable and sequentially compressed.

3. Database Selection Decision Tree

Interactive Architecture Diagram
Synthesizing vector architecture diagram...

4. Replication & Consistency Cheat Sheet

CAP Theorem Realities

  • CP (Consistency + Partition Tolerance): System refuses writes or halts if cannot guarantee latest data (e.g., , , single-primary).
  • AP (Availability + Partition Tolerance): System accepts writes anywhere; replicas sync asynchronously, resulting in eventual consistency (e.g., , ).

Isolation Levels Ranked from Weakest to Strongest

  1. Read Uncommitted: Suffers from dirty reads, non-repeatable reads, and phantom reads.
  2. Read Committed: Default in ; queries see only committed snapshots.
  3. Repeatable Read: Guarantees snapshot consistency across a single transaction.
  4. Serializable: Strict two-phase locking (2PL) or SSI; eliminates write skew and phantom rows.

5. Architectural Cross-References