Fauna vs Aurora, DynamoDB, MongoDB

Link: The Why and How of Distributed Databases

Dimensions:

• Data Model: Relation vs. Document-Oriented Non-relational
• Query Model
• Transactional Model
  • DynamoDB: Only ACID within one region
• Consistency Model
  • Aurora: Impossible to perform write transactions or consistent read-only queries in non-primary regions.
  • DynamoDB: within one region
  • MongoDB: Developers must limit queries
• Indexing
• Fault tolerance
• Security
• Scalability
  • DynamoDB: significant operational work for customers vs. Fauna

CockroachDB vs. Postgres & MongoDB

Link: CockroachDB in Comparison

Interesting Dimensions:

• Distributed Transactions
  • MongoDB: Lightweight transactions only
• Isolation
  • MongoDB: Eventual consistent default, read uncommitted highest. Allows snapshot read.
• Potential data issues
  • MongoDB: Dirty reads, phantom reads, non-repeatable reads, write skew
• Cost-based optimization
  • MongoDB: No
• Geo-partitioning
  • Cockroach: Yes, row-level
  • MongoDB: No
• Multi-cloud

Yugabyte vs. Cockroach

Link 1: Bringing Truth to Competitive Benchmark Claims – YugabyteDB vs CockroachDB, Part 2

Link 2: Compare YugabyteDB to other databases

Yugabyte benchmark

Link 1: Achieving 10x Better Distributed SQL Performance in YugabyteDB 2.1

tpc-c:

• NewOrder: 154 tps
• Payment: 672 tps
• OrderStatus: 3059 tps
• Delivery: 55 tps
• StockLevel: 118 tps
• Overall: 217 tps

Link 2: Very poor performance in comparison to Postgresql