How Switching S3 Compression to Zstd Cut Storage Costs by 30%
AWS switched S3 log compression from gzip to Zstandard, saving roughly 30% of storage space—equivalent to exabytes—by changing the internal storage method, a move confirmed by employee comments and a coinciding 31% price drop in late 2021.
Former AWS Vice President Adrian Cockcroft revealed on Twitter that Amazon switched the compression algorithm used for S3 logs from gzip to Zstandard (zstd), saving roughly 30% of storage space, equivalent to exabytes.
Amazon employees clarified that the change affects the way S3 stores its own service data (primarily logs), not customer data, and the switch to zstd reduced storage costs by about 30%.
Although AWS did not officially announce the change, a price reduction of 31% at the end of November 2021 aligns with the reported savings.
About Zstd
Zstandard is an open‑source, fast, lossless compression algorithm developed by Facebook, targeting real‑time compression scenarios with better ratios than zlib. It can trade compression speed for higher ratios, and it offers a special “dictionary compression” mode for small data, where a trained dictionary improves compression efficiency.
Zstd aims to combine high compression performance and ratio, distinguishing it from algorithms such as zlib, lz4, and xz. It also supports dictionary compression for small data blocks.
Below is a typical performance comparison of common compression algorithms:
Reference: https://twitter.com/adrianco/status/1560854827810361345
Signed-in readers can open the original source through BestHub's protected redirect.
This article has been distilled and summarized from source material, then republished for learning and reference. If you believe it infringes your rights, please contactand we will review it promptly.
Programmer DD
A tinkering programmer and author of "Spring Cloud Microservices in Action"
How this landed with the community
Was this worth your time?
0 Comments
Thoughtful readers leave field notes, pushback, and hard-won operational detail here.
