Into the Stack

Async I/O on Linux in databases

I've been working on a complex multi-model database for a few weeks now, and recently I took time to simplify and test out an idea I had on a simple key-value database. I started with the basics: A hash table in memory, a simple append-only log