TLDR Data · DEEP DIVES

BENCHMARKING ROCKSDB'S THREE AMPLIFICATION FACTORS

hackernoon.com8 min read

Summary from the TLDR Data newsletter

RocksDB amplification depends heavily on workload shape, compaction capacity, and key order. Random inserts produced 4.56x write amplification while sequential keys stayed the same, and delete-heavy workloads temporarily pushed space amplification to 1.58x. Bloom filters kept L0 read latency nearly flat as file counts grew.