EF Core Data Seeding: Production-Ready Patterns for .NET

Two battle-tested seeding strategies in one runnable sample — Model-based HasData for static reference data and custom runtime seeders with full idempotency guards for dynamic datasets.

---

🚀 Support the Channel — Join on Patreon

If this sample saved you time, consider joining our Patreon community. You'll get exclusive .NET tutorials, premium code samples, and early access to new content — all for the price of a coffee.

👉 Join CodingDroplets on Patreon

Prefer a one-time tip? Buy us a coffee ☕

---

🎯 What You'll Learn

• How to use model-based HasData seeding for static reference data (roles, categories, lookup tables)
• How to build custom runtime seeders with IDataSeeder for dynamic or large datasets
• How to write idempotency guards so seeders are safe to run on every startup
• How to orchestrate multiple seeders in a controlled order with DatabaseSeeder
• How to switch from InMemory to SQL Server with a two-line change

---

📋 Seeding Strategies Compared

Strategy	Best For	Pros	Cons
Model-based HasData	Static reference data (lookup tables, roles, categories)	Migrations-aware, zero startup overhead, deterministic	Hard-coded IDs required; large datasets inflate migration files
Custom Runtime Seeders	Dynamic/computed data, large datasets, external source data	Full C# expressiveness, no migration bloat, DB-generated IDs	Runs at startup; must guard against duplicates

---

🗺️ Architecture Overview

App Startup (Program.cs)
        │
        ▼
┌───────────────────────────────────────────────────┐
│               DatabaseSeeder                      │
│  Orchestrates all IDataSeeder implementations     │
│  in order (Order property)                        │
│                                                   │
│  ┌──────────────────────────────────────────────┐ │
│  │  Strategy 1: AppDbContext.OnModelCreating    │ │
│  │  modelBuilder.Entity<T>().HasData(...)       │ │
│  │  → Applied by EF migrations / EnsureCreated  │ │
│  └──────────────────────────────────────────────┘ │
│                                                   │
│  ┌──────────────────────────────────────────────┐ │
│  │  Strategy 2: ProductSeeder : IDataSeeder     │ │
│  │  → Guard: AnyAsync() check before inserting  │ │
│  │  → Resolves FK by slug (no hard-coded IDs)   │ │
│  │  → SaveChangesAsync()                        │ │
│  └──────────────────────────────────────────────┘ │
└───────────────────────────────────────────────────┘
        │
        ▼
  Database ready with seeded reference + product data

---

📁 Project Structure

efcore-data-seeding-seeddata/
├── src/
│   └── EFCore.DataSeeding.Api/
│       ├── Controllers/
│       │   └── ProductsController.cs        # REST endpoints to verify seeded data
│       ├── Data/
│       │   ├── AppDbContext.cs               # DbContext with HasData configuration
│       │   └── Seeding/
│       │       ├── IDataSeeder.cs            # Seeder marker interface
│       │       ├── DatabaseSeeder.cs         # Orchestrates all seeders in order
│       │       └── ProductSeeder.cs          # Custom runtime seeder (Strategy 2)
│       ├── Models/
│       │   ├── Category.cs
│       │   └── Product.cs
│       ├── Repositories/
│       │   ├── IProductRepository.cs
│       │   └── ProductRepository.cs         # Repository pattern with AsNoTracking
│       └── Program.cs                       # DI wiring + startup seeding
└── tests/
    └── EFCore.DataSeeding.Tests/
        ├── HasDataSeedingTests.cs            # Tests for Strategy 1
        └── ProductSeederTests.cs            # Tests for Strategy 2 (idempotency verified)

---

🛠️ Prerequisites

• .NET 10 SDK (or .NET 8/9 — update TargetFramework in .csproj)
• Any IDE: Visual Studio 2022+, VS Code, or JetBrains Rider

---

⚡ Quick Start

# Clone the repo
git clone https://github.com/codingdroplets/efcore-data-seeding-seeddata.git
cd efcore-data-seeding-seeddata

# Run (uses InMemory database — no SQL Server setup needed)
dotnet run --project src/EFCore.DataSeeding.Api

# Open Swagger UI → https://localhost:{port}/swagger

---

🔧 How It Works

Strategy 1 — Model-Based HasData (Static Reference Data)

Configured inside OnModelCreating in AppDbContext.cs. Applied automatically by EF Core on EnsureCreated or migration:

modelBuilder.Entity<Category>().HasData(
    new Category { Id = 1, Slug = "electronics", Name = "Electronics" },
    new Category { Id = 2, Slug = "books",        Name = "Books" },
    new Category { Id = 3, Slug = "clothing",     Name = "Clothing" }
);

⚠️ Important: HasData requires hard-coded primary keys. Navigation properties are not supported — use flat scalar values only.

Strategy 2 — Custom Runtime Seeder (Dynamic Data)

ProductSeeder implements IDataSeeder and is called at startup. It uses an idempotency guard so it is safe to run on every startup:

public async Task SeedAsync(CancellationToken cancellationToken = default)
{
    // Guard: skip if already seeded
    bool alreadySeeded = await _db.Products
        .AnyAsync(p => p.Stock > 500, cancellationToken);

    if (alreadySeeded) return;

    // Resolve FK by slug — no hard-coded IDs
    var electronics = await _db.Categories
        .FirstAsync(c => c.Slug == "electronics", cancellationToken);

    await _db.Products.AddRangeAsync(new List<Product>
    {
        new() { Name = "USB-C Hub (7-in-1)", Price = 49.99m, Stock = 750, CategoryId = electronics.Id },
        // ...
    }, cancellationToken);

    await _db.SaveChangesAsync(cancellationToken);
}

Register Seeders in Program.cs

builder.Services.AddScoped<IDataSeeder, ProductSeeder>();
builder.Services.AddScoped<DatabaseSeeder>();

Add more seeders by implementing IDataSeeder and registering them. Control execution order via the Order property.

---

🔄 Switching to SQL Server

The sample uses InMemory for zero-friction local runs. To use SQL Server:

1. Update Program.cs:

   builder.Services.AddDbContext<AppDbContext>(options =>
       options.UseSqlServer(builder.Configuration.GetConnectionString("DefaultConnection")));

2. Replace EnsureCreatedAsync() with MigrateAsync().
3. Add your connection string to appsettings.json.

---

🧪 Running Tests

dotnet test -c Release

Test	Strategy Covered
HasData_Seeds_ThreeCategories	Strategy 1
HasData_Seeds_ElectronicsCategory_WithCorrectSlug	Strategy 1
HasData_Seeds_FourBaseProducts	Strategy 1
HasData_Products_HaveCorrectCategoryAssignments	Strategy 1
ProductSeeder_InsertsHighStockProducts	Strategy 2
ProductSeeder_IsIdempotent_DoesNotDuplicate	Strategy 2
ProductSeeder_AssignsCorrectCategories	Strategy 2

---

📚 References

• EF Core — Data Seeding (Microsoft Docs)
• EF Core — Migrations Overview
• EF Core — InMemory Database Provider

---

📄 License

This project is licensed under the MIT License.

---

🔗 Connect with CodingDroplets

Platform	Link
🌐 Website	https://codingdroplets.com/
📺 YouTube	https://www.youtube.com/@CodingDroplets
🎁 Patreon	https://www.patreon.com/CodingDroplets
☕ Buy Me a Coffee	https://buymeacoffee.com/codingdroplets
💻 GitHub	http://github.com/codingdroplets/

Want more samples like this? Support us on Patreon or buy us a coffee ☕ — every bit helps keep the content coming!