Add Learning DynamoDB (learning-software-engineering#54)

Topics/Tech_Stacks/DynamoDB_for_Dummies.md

@@ -0,0 +1,95 @@
+# Learning DynamoDB
+
+## Table of contents
+
+### [Introduction](#introduction)
+
+### [Why would I want to use it?](#why-would-i-want-to-use-it-1)
+
+### [Why would I not want to use it?](#why-would-i-not-want-to-use-it-1)
+
+### [How do I use it effectively?](#how-do-i-use-it-effectively-1)
+
+### [Additional resources](#additional-resources-1)
+
+## Introduction
+
+This is a very brief overview of DynamoDB, what it is, why you might want to use
+it, why you might not want to use it, and how to go about using it effectively.
+
+## What is it?
+
+DynamoDB is a [NoSQL database](https://en.wikipedia.org/wiki/NoSQL) that is
+designed to be fast at any scale. It is developed by Amazon and is available
+through Amazon Web Services.
+
+## Why would I want to use it?
+
+DynamoDB has a couple of main selling points:
+
+1. Scalability: your database will automatically adjust to the scale required of
+   it.
+1. Speed: due to how the database is implemented, queries will always be fast -
+   as long as your data is modelled appropriately for your access patterns!
+1. Fully managed by AWS: it's pretty easy to quickly spin up a database that
+   will just work, with all the great benefits that come with AWS.
+
+## Why would I not want to use it?
+
+There are a few major details of DynamoDB that may make it the wrong choice for
+your application.
+
+1. Inflexibility: DynamoDB is not designed to be flexible. You need to have a
+   pretty good idea of what your core access patterns will be at the time of
+   making your initial data model. If your application is changing frequently
+   DynamoDB might cause you more headaches than it's worth!
+1. Incompatibility with certain access patterns: if your application has certain
+   collections of access patterns, DynamoDB will either take up a lot of extra
+   storage space or be very slow. An example of such an application would be
+   storing analytics data. Generally our access patterns will be a lot of
+   complex queries involving the attributes of our items. This kind of demand is
+   better suited for something relational that better supports lots of ad-hoc
+   queries.
+
+## How do I use it effectively?
+
+If you've decided DynamoDB might be a good fit for your application, then you
+should try designing a single-table data model! There are a couple of key
+concepts to DynamoDB that you'll need to understand to start creating a data
+model (not an exhaustive list):
+
+1. Partition keys & sort keys
+1. Attributes
+1. Global secondary indexes
+
+Together, an **item**'s partition key and sort key must uniquely identify it.
+They should be pretty generic, since lots of different types of items may need
+to be stored. Attributes are the real data you want to store. DynamoDB is
+schemaless so you can store whatever attributes you want on an item, but as a
+result, you must also keep your data model straight in your application code
+when you're reading and writing to the database.
+
+Global secondary indexes can be defined on your table to account for usecases
+where querying on your partition key + sort key will not be efficient. When you
+define a GSI, all of the data will be replicated and stored (effectively in
+another table) with your GSI as the partition key. For example, if an access
+pattern we wanted to account for was to find all users with the first name
+"Spencer", but it wouldn't make sense to define our partition key to be first
+name, we could say that first name is a GSI on our table. This would allow us to
+make an efficient query to find users with the first name "Spencer", but would
+cost us some space to store the replicated data. This is a trade-off; if we had
+very many situations where we needed to query on attributes, then we would
+probably want to re-evaluate whether DynamoDB is the best choice for us.
+
+## Additional resources
+
+- [AWS Documentation](https://docs.aws.amazon.com/amazondynamodb/latest/developerguide/Introduction.html)
+
+A lot of my learning on DynamoDB has been from Alex DeBrie - he has done a bunch
+of practical talks on DynamoDB and has even written "The DynamoDB Book". I'd
+recommend looking into his stuff if you're interested in learning more about
+DynamoDB and how to best use it.
+
+- [The DynamoDB Book](https://www.dynamodbbook.com/)
+- [Data modelling talk at AWS re:Invent 2020](https://www.youtube.com/watch?v=fiP2e-g-r4g)
+- [Alex DeBrie's website](https://www.alexdebrie.com/)