When In-Memory OLTP was introduced in SQL Server 2014, many developers were initially excited. The hope of a new, ultra-high-performance data processing engine, coupled with a leading relational database engine, offered the potential of significant improvement for many SQL Server developers. However, this excitement quickly turned into mild disappointment due to the number of restrictions assigned to In-Memory OLTP. Many of these restrictions prevented a widespread adoption of the technology and forced it into a niche set of very tight implementation scenarios. Some of these restrictions, such as minimal support for large data object types (LOBs) and missing support for ALTER commands, dampened many people's enthusiasm for the technology.
As with previous features inside SQL Server, In-Memory OLTP has followed a similar pattern. SQL Server 2014 saw the introduction of the In-Memory OLTP Engine. With SQL Server 2016 and again in 2017, the feature has experienced an evolution: many of the restrictions that were present in SQL Server 2014 have been removed and existing functionality has been extended. In this chapter, we will take a look at the improvements that should now make In-Memory OLTP attractive to almost every developer out there.
This chapter will demonstrate these improvements and additions (many of which were made available in SQL Server 2016), including altering existing memory-optimized objects, expanded data type support, expanded functionality/reduced limitations, and the integration of In-Memory OLTP into other areas of the database engine.
We will cover the following topics in this chapter:
- Ch-Ch-Changes
- Feature improvements
- Improvements in the In-Memory OLTP Engine