Better Dev Link

This repository contains a step-by-step guide that teaches how to create a simple operating system (OS) kernel from scratch. I call this OS Raspberry Pi OS or just RPi OS. The RPi OS source code is largely based on Linux kernel, but the OS has very limited functionality and supports only Raspberry PI 3. If you are into this, check Notes on “The Design of the UNIX Operating System” as well.

Release after release we keep hearing about the good part of Postgres, but what is the bad parts? Knowing the bad parts help a lot otherwise you only know about it when you already get into that situation. Part 2 is release as well.

This is an 2014 document of FireFox project to advise their engineer to avoid SQLite, they explain the downside of SQLite at that time, which may not right in 2020. However, I found many things still ring true such as the IO pattern. I recently have to debug a small SQLite database with heavy disk IO due to its fsync in default rollback-journal mode

But what is a file, really? And what does reading a file entail, exactly? See how it’s done in Node, C, Rust, Strace, x86 asm, look to disk layouts, ext4

If you ever use chown, chmod, or even doing any file read, or mkdir, you are making system calls. System calls are how a progrm enters the kernel to perform some tasks. This guide teachs us User programs, the kernel, and CPU privilege levels, Interrupts(kind of like event that is generated by hardware or software), then how we can craft system calls using assembly. A very low-level guide but I found it’s very helpful and easy to follow to know more about our computer.

Lots of interesting info on Slack tech stack: from their languages, cloud provider, how they monitor performance, logging and track security. Basically a summarize of their evolutions and many links to each section: search index, monitoring with Icinga, Grafana, Sharding database with Vitess.

A 81 pages research discusses Docker from a security perspective and looks at how a penetration tester should assess the security of systems that use Docker. Docker internal, data persistences, networking, docker socket, protection mechanisms, attack vector.

Leader election is the simple idea of giving one thing (a process, host, thread, object, or human) in a distributed system some special powers. This post is in AWS Builder’s Library on How Amazon elects a leader. What they do when that process failed.

Jepsen analyses distributed systems and report their finding. This one focus on Raft implement of Redis. The conclustion is this: users who want fault-tolerance and not lost updates need something stronger than existing Redis replication systems. They found data loss on failover, split-brand and lost updates, Stale Reads in Healthy Clusters…

front-end testing, or end to end testing or UI testing is the most expensive test, but they are the one that give you the most confident because it’s what your end user see, not testing a particular code and skip certain thing, UI test tend to run through whole stack. In this article, we will provide an overview of UI testing used in TeamCity for their React app.