Scale write performance on Amazon DocumentDB elastic clusters

This article discusses how Amazon DocumentDB elastic clusters can scale write performance by increasing the number of shards. It provides a benchmark using a Python script to measure write throughput on instance-based and elastic clusters.

Specifically, the article covers:

• Overview of Amazon DocumentDB elastic cluster architecture and its ability to scale compute and storage independently
• Benchmark configuration details like infrastructure, duration, batch size, concurrency, and data schema
• Steps to set up the client environment and run the benchmark
• Benchmark results showing write throughput scaling from 80K inserts/sec on a single-shard cluster to over 1.2M inserts/sec on a 16-shard cluster
• Conclusion highlighting the ability of elastic clusters to support high throughput, petabyte-scale workloads by adding shards