Enhance SQL Query Performance and Tracking Behavior

[vmussak]

This PR introduces two significant behavioral settings for SQL queries in our application, aimed at optimizing performance and managing entity tracking more effectively. These changes include:

QuerySplittingBehavior:

Setting: QuerySplittingBehavior.SplitQuery
Impact: This configuration alters how Entity Framework Core handles queries involving collections. By using SplitQuery, it directs the framework to generate separate SQL queries for each collection in the query. This approach can significantly reduce the complexity of generated SQL and improve performance, especially for queries that include multiple includes or large collections. However, it might lead to multiple round-trips to the database, so it's more suitable for scenarios where network latency is not a primary concern.

NoTrackingWithIdentityResolution:

Setting: QueryTrackingBehavior.NoTrackingWithIdentityResolution
Impact: This setting optimizes the tracking behavior of Entity Framework Core. When entities are fetched from the database, NoTrackingWithIdentityResolution ensures that they are not tracked by the DbContext, which reduces memory usage and increases performance for read-only scenarios. Unlike standard NoTracking, this behavior still performs identity resolution within the scope of a single query. This means if the same entity is encountered more than once in the query, all occurrences will point to the same instance, ensuring consistency while still reaping the performance benefits of no tracking.

These enhancements are expected to improve the overall efficiency and speed of our database operations, particularly in scenarios with complex queries and read-intensive operations.