AWS Neptune Is Killing Relational Databases and you need to know why and how if you are going to stay competitive in the age of AI/ML.

In my earlier post exploringGraph Databases I tried desperately to give you some basic insights as far as how they might scale on AWS but found each DB has its own unique properties.

In this post we will take a quick look at AWS Neptune. Here are some of the key pieces of information I have gathered so far:

Joins: We all have come across those monstrous SQL queries with JOINS, INNER, OUTER, etc. This is what they way about joins in Neptune:

Traversing the joins or relationships is a very fast process, as the relationships between nodes are not calculated at query times but are persisted in the database.

Which makes sense. It’s not a whole lot different from my Pre-Built Indexes. Instead of trying to grab records from a bunch of different tables and filter through them when the user runs the query Neptune will look at the data when it is Created, Update, Deleted and decide what records it is likely to be joined with then does some magic to ensure they system knows that before the query comes in and is ready to serve up those related records.

It sounds like instead of me explicitly coding how those prebuilt indexes will be populated it has AI learn how to populate those indexes (Great… another AI tool coming for my job).

What is Neptune not good at: I found this next bit interesting:

A dedicated graph database provides the most value for highly connected datasets and any analyses that require searching for hidden and apparent relationships. If this doesn’t fit your use case, other database types may be better suited.

This means value key stores like Redis are safe(For now…).

Wrapping It Up:

Neptune is really interesting and I think it deserves a deeper dive. Let me know if you agree. If so I will make a series out of this or a full length video.

Question For You:

Are you considering AWS Neptune or any other Graph DBs for your infrastructure? If so, what is your use case?