thoolihan / json_sql_server_explore

SQL Server JSON Performance Playground

Background: The question arose to the impact of storing data as JSON in SQL server. This is a simple to attempt to checkout some common scenarios.

Setup:

1. Create a SQL Server Database Named Test
2. Give permissions to a user and pass you have created
3. Run the script create_table.sql in your preferred SQL Server IDE, or from sqlcmd
4. Make sure you have pyodbc package installed (in virtualenv if preferred) using pip install pyodbc
5. Set environment variables for username and password to connect to db (see comments near the top of load.py). Optional: set number of records in the variable RECORD_COUNT
6. run: python load.py

Now you can use the following scripts:

• json_sample_queries.sql to explore how to query with json
• average.sql to compare a column average with json column average. The timing info will be in the message tab if using SSMS.
• index_test.sql to compare filtering a record by a string with a column, vs a json value, including indexing the json value. The timing info will be in the message tab if using SSMS.

You can email me with any questions or feedback at tim at hoolihan.net, or open a pull request with suggested changes.