BeCuriouzLearn something new every day
S3 stores over 350 trillion objects. Here's what most devs get wrong about it.
💡 S3 durability is legendary, but request costs can surprise you — batch small files and use Intelligent-Tiering.
Amazon S3 isn't just a file system in the cloud — it's an object store designed for 99.999999999% (eleven 9s) durability. That means if you store 10 million objects, you'd expect to lose one every 10,000 years.
But here's what catches people: S3 charges you for requests, not just storage. A bucket full of millions of tiny files can cost more in GET requests than a bucket with fewer, larger files holding the same data.
The sweet spot? Batch small files together when possible, use S3 Intelligent-Tiering if access patterns are unpredictable, and always enable versioning on buckets you care about — accidental deletes happen more often than you'd think.