はてなブックマーク

Disclosure: I work with one of the paper’s authors at Turso SQLite is already fast. But can we make it even faster? Researchers at the University of Helsinki and Cambridge began with this question and published a paper, “Serverless Runtime / Database Co-Design With Asynchronous I/O”. They demonstrate up to a 100x reduction in tail latency. These are my notes on the paper. This is the foundational

I will show how we can implement a durable, distributed, and highly available log using S3. This post is the third part in the series: Disaggregated Storage - a brief introductionZero Disk ArchitectureBuilding a distributed log using S3tl;dr The code is open source, comes with tests and open issues to contribute: s3-log LogI love logs. The log is the heart of data and event streaming systems. A da

Current Best: 100M rows inserts in 33 seconds. (you can check the source code on Github) Recently, I ran into a situation where I needed a test database with lots of rows and needed it fast. So I did what any programmer would do: wrote a Python script to generate the DB. Unfortunately, it was slow. Really slow. So I did what any programmer would do: went down the rabbit hole of learning more about

gRPC is an amazing library, however, the documentation lacks details on error handling. The code examples do not contain error handling part either (though few does seem to have them). Some test suits in main repo do have them, but it might be difficult to figure out for a beginner. This project is my attempt to fix the issue. The repository contains examples in all popular languages and contain s