Monthly Readings #18: 2017 March

09 Apr 2017 by fleuria

Spark GraphX源码分析: 分布式图计算

Impala Concepts and Architecture

Monthly Readings #17: 2017 Jan

31 Jan 2017 by fleuria

Semi-Synchronous Replication at Facebook:

The Architecture of Schemaless, Uber Engineering’s Trip Datastore Using MySQL

Designing Schemaless, Uber Engineering’s Scalable Datastore Using MySQL

Why Uber Engineering Switched from Postgres to MySQL

Sharding Pinterest: How we scaled our MySQL fleet

MySQL Backup in Facebook

If You Must Deploy Multi-Master Replication, Read This First

Best Practices for Amazon RDS

gh-ost: Triggerless design

gh-ost: Sub-second replication lag throttling

Monthly Readings #16: December

Monthly Readings #15: November

04 Dec 2016 by fleuria

Druid: A Real-time Analytical Data Store

Error Handling in Node.js

Putting Apache Kafka To Use: A Practical Guide to Building a Stream Data Platform

Making "Push on Green" a Reality

Big Data in Real-Time at Twitter

Monthly Readings #14: October

29 Oct 2016 by fleuria

FollowFeed: LinkedIn's Feed Made Faster and Smarter

Design Decisions For Scaling Your High Traffic Feeds

Etsy Activity Feeds Architecture

What I Wish I Had Known Before Scaling Uber to 1000 Services

Wasting Time TDDing The Wrong Things

There is not Fork: an Abstraction for Efficient, Concurrent, and Concise Data Access


Comparing Redux and Relay

Monthly Readings #13: September

Monthly Readings #12

20 Aug 2016 by fleuria

Building Timeline: Scaling up to hold your life story

Facebook News Feed: Social data at scale

ZooKeeper Resilience at Pinterest

Top 5 Docker Logging Methods to Fit Your Container Deployment Strategy

Web Service Efficiency at Instagram with Python

Kubernates: Pods

Monthly Readings #11

Monthly Readings #10

29 Feb 2016 by fleuria

How we ended up with microservices

Logging v. instrumentation

Advanced Caching: Part 2 - Using Caching Strategies

Increasing Application Performance with HTTP Cache Headers

What design decisions make the GC for Go so much faster?

Monthly Readings #9

28 Jan 2016 by fleuria

过去一年尝试每月整理一次 evernote 中新建的笔记到 blog 中来,没有结构,也没有主线,更没有深入,只能算是个人去阅读一些资料的一种方式。比如,在微博上、在 hackernews 上看到一篇可能会有意思的文章,扫一眼只能留个印象,足够见识观点,但不一定清楚作者为什么抱着这样的观点;留到 pocket 里面,往往只是留下一个 "有时间再看" 的心理暗示就没有后续了。看懂一篇文章终究是一件很累的事。记录笔记,就成了一个 "我已经看到这里" 了的标记,下次再遇到类似的文章可以不用从头读起。

转眼已 2016 年,"Readings in XXX" 这样的标题就要不好使了,先换一个更容易扩展的标题。

TAO: Facebook's Distributed Data Store for the Social Graph

Building Microservices: Testing

Feature Toggle

Readings in November and December

Readings in October

01 Nov 2015 by fleuria

A Critique of the CAP Theorem

Nonvolatile Memory is a Broken Time Machine

Memory Barriers: a Hardware View for Software Hackers

Why Cassandra doesn’t need vector clocks

The SO_REUSEPORT socket option