Ref: https://www.distributedsystemscourse.com/

MapReduce: Simplified Data Processing on Large Clusters

The Chord Protocol and Consistent Hashing

The Google File System paper was a lot of fun to read. This is what I got out of it while trying to figure out its main design points, parts summarized from the paper.

This article is on Amazon’s Dynamo, a highly distributed key-value storage system, designed to be always available. Here is the paper if you want to read it all in full! -> Amazon’s Dynamo <-

The first thing that jumps out at you when you read the paper is bigtable’s flexibility in terms of the data size and latency requirements it supports. It handles data from web indexing, Google Finance and Google Earth, from urls to images. (And that Orkut was a Google product, who knew)

Cassandra is a popular decentralized distributed key value store, designed for write heavy workloads. Cassandra was designed at Facebook to meet its needs for reliability and scalalbility in the Inbox Search problem. Cassandra’s design seems to be heavily inspired by Amazon’s Dynamo.

ACID

Atomicity, Consistency, Isolation and Durability. Relational DBs

NOTES FROM CLOUD GURU’s AWS CERTIFIED SOLUTIONS ARCHITECT COURSE and AWS Certified Solutions Guide by Ben Piper and David Clinton

What is Elasticsearch?