Google I/O 2012 - SQL vs NoSQL: Battle of the Backends | My third view with a lot of stops to take notes | 60+ minutes study

Sept. 6, 2021

Here is the link. 

Ken Ashcraft, Alfred Fuller Google App Engine now offers both SQL and NoSQL data storage -- but which is right for your application? Advocates of each try to settle the issue once and for all, and show some of the tricks for getting the most out of each. For all I/O 2012 sessions, go to https://developers.google.com/io/

Take my notes - I need to work on learning more carefully. 

App engine - Google app engine (GAE) 

• Build apps on Google's infrastructure
• Platform as a Service (PaaS)
• Easy to build
• Easy to scale
• Easy to maintain
• Focus on what makes your app great!

App Engine _ Storage 

App engine works with Cloud SQL or datastore, cloud storage 

App engine - data store 

• Google storage infrastructure
• Same technology we use for our own applications
• Distilled into well documented APIs
• Build for scale (size and traffic)
• 2 trillion operations per month
• Fully managed 'NoSQL' solution

My notes: App engine data store is NoSQL solution fully managed, built for scale (size and traffic) - 2 trillion operations per months

Cloud SQL

Map reduce - Materialized view 

I like this example, city and age. It is so helpful for me to understand the ideas quickly. 

Fan-in and apply 

Datastore - Consistency 

Scalability - Datastore on Megastore on Bigtable on ...

• All the best features of each layer
• Datastore
• Megastore
• Bigtable
• GFS v2

BigTable with load balancing 

BigTable splitting - load balancing 

Megastore

• Works at scale
• 2011 talk "More 9s Please: Under the covers of the high replication datastore" 
• 9's are important at scale 
• Not reliant on a single datacenter
• Handles local issues
• Handles catastrophic failures

Management