SQL Server 2017 Developer's Guide

Looking behind the curtain of concurrency

So far, we have seen that we can easily create a table with indexes, query, and insert data into that table with our traditional T-SQL knowledge. It is now also clear that we are able to change an existing implementation of disk-based tables into memory-optimized tables with some simple changes. If we want to push the performance to the max, we can also consider natively compiled stored procedures to further improve upon some quite large performance gains. We have also seen how the two types of index; hash index and non-clustered index, are implemented in the storage engine. However, to better understand how the magic happens, we need to have a closer look at what is happening to our sample table ...