Full Fledged ORM Challenges

When it comes to full-fledged ORM like Entity Framework, we have the below-mentioned challenges

• Classic (N+1) problem due to lazy loading with Entity Framework for database reads.
• Developer’s responsibility to either disable lazy loading or use it sparingly for specific use cases.
• LINQ query expertise required.
• Entity Framework slows down the data-heavy read operations if used incorrectly. Overkill for the reads.
• Developer consistently must verify the queries that get translated within SQL for every LINQ execution.
• Incorrect LINQ queries can have a drastic performance impact on the application.
• Challenges w.r.t to managing migrations and database schema updates.

Micro ORM to the Rescue

Micro ORM like Dapper can solve these issues. Dapper is an open-source, lightweight ORM developed by the Stack Overflow team. Mentioned below are the advantages of using such an ORM.

As seen below, Dapper Custom C# Library can be integrated within .Net layered architecture or within CQRS design.

Micro ORM & Full-fledged ORM’s (Entity Framework or NHibernate) can co-exist within a .NET application.

Note:  Block marked as “Custom Library” is where the Micro ORM custom implementation will integrate.

• Extremely fast compared to other ORMs primarily because of its lightweight.
• Supports easy integration with MSSQL, Oracle, MySQL, SQLite, PostgreSQL (ADO.Net provider databases).
• Built with performance and ease of use in mind.
• Developers get the much faster speed of execution with all read or write operations compared to Entity Framework or NHibernate & performance is almost the same as ADO.Net as it’s a wrapper on top of ADO.Net.
• Fastest ORM among the other open-source micro ORM’s that are available.
• Supports transactions, stored procedures, bulk inserts of data.
• Provides built-in support for .Net Async features (which came with .Net 4.5).
• Leverages C# DLR (dynamic language runtime) which is a huge advantage.
• Works well with the database’s first approach leveraging stored procedures or RAW SQL to the fullest.  
• The developer can use stored procedures with Dapper API leveraging SQL benefits.
• Dapper API helps to rule out SQL injections attacks with parameterized query parameters.
• Dapper, being a ADO.NET wrapper eliminates the need for the developers to write all the complex ADO.NET logic to get data. Thereby acts as a handy replacement to hand written ADO.Net.
• It enforces using SQL stored procedures to its maximum potential.
• Complex logic gets wrapped within Dapper and the developer focuses only on passing the stored procedure name, parameters etc to Dapper API & Dapper does the rest to execute T-SQL.

Other Use case (within a project using CQRS Pattern)

• Most windows or web applications heavily rely on reading operations which accounts for 80% of all operations.

Dapper can be easily integrated for all those read operations within such applications.

• Dapper can help read data from the Materialized / Indexed View in the fastest possible manner.
• Dapper ORM can be used within the Query part of the CQRS (command query responsibility segregation) implementation. The command part rather sticks to domain-driven design (DDD) & supports ACID transactions to facilitate data integrity using an ORM like Entity Framework or NHibernate.  
• Thereby, both a Micro ORM & a full-fledged ORM can co-exist within a CQRS implementation.

Why Dapper is so fast

• It’s fast because it uses dynamic method generation (MSIL) to assign column values to properties.
• It’s a lightweight wrapper and an abstraction on top of ADO.Net.
• It focuses just on executing user-supplied queries and mapping the results to CLR object properties.
• It caches information for every query it runs thus allowing faster object materialization & parameter processing.

Dapper Micro ORM Vs Entity Framework Core benchmark

Note: Entity Framework 6 & 7 are much slower compared to EF core.

Overall a must-have tool for all database related operations in a web or windows application.