Patterns & Best Practices For Moving From RDBMS to Azure Storage

I recently presented at DotNetConf 2013 on how to make the paradigm shift from the relational data world like SQL to the non-relational world of Azure Storage.  I had a lot of fun giving the talk.  This was my first public technical speaking event, so any constructive feedback is very welcome!

In the talk I discuss:

• A quick overview of Azure Storage with a focus on Table Store
• The trade off between scale and consistency in Azure Table
• The importance of Table Partitions in Azure Table
• A Code example of a Simple Stats Storage Layer in SQL and Azure.
• The benefits of Immutable data, and how you can realistically implement this
• Using the Saga Pattern for distributed or long running transactions

The Slides for the Presentation are posted on SpeakerDeck
All of the code I use in my slides is available on GitHub.

Resources & References

If you want to dig in deeper.  Below is a great list of resources.  Also I include Links to the quotes and papers I referenced in my talk.

Relational vs Non-Relational

• NoSQL: Past, Present, Future (video) Eric Brewer

Getting Started with Azure Table

• How to Use Table Storage Service
• Real World: Designing a Scalable Partitioning Strategy for Windows Azure Table Storage

Saga Pattern

• Sagas (paper) by Hector Garcia-Molina (1987)
• Clarifying the Saga Pattern (blog post)by Kelly Sommers
• Sagas (blog post) by Clemens Vasters

Immutable Data:

• Human Fault Tolerance (video) Nathan Marz @ StrataConf2013

Google Spanner

• Google Spanner Paper (paper) by James C Corbett et al.

 

You should follow me on Twitter here