This is a viewer only at the moment see the article on how this works.
To update the preview hit Ctrl-Alt-R (or ⌘-Alt-R on Mac) or Enter to refresh. The Save icon lets you save the markdown file to disk
This is a preview from the server running through my markdig pipeline
Sunday, 09 November 2025
We've explored the hosting layer and Kestrel server in Part 2. Now we arrive at the heart of ASP.NET Core: the middleware pipeline. This is where your application's logic lives, where requests are processed, and where responses are built. Understanding middleware is fundamental to mastering ASP.NET Core.
Middleware is elegantly simple in concept—each component is a function that processes an HTTP request—yet powerful enough to handle everything from authentication to error handling to serving files. In this part, we'll explore how middleware works, examine the built-in components, and learn to create our own.
NOTE: This is part of a series of experimental articles (hence being dated in the past). It's a combination of AI and my own editing and tweaking.
Middleware is software that's assembled into an application pipeline to handle requests and responses. Each component:
Think of it as a series of nested function calls, where each middleware wraps the next one:
graph LR
A[Request] --> B[Middleware 1]
B --> C[Middleware 2]
C --> D[Middleware 3]
D --> E[Endpoint]
E --> D
D --> C
C --> B
B --> F[Response]
Each middleware can:
At its core, middleware is just a function with this signature:
public delegate Task RequestDelegate(HttpContext context);
Each middleware receives:
HttpContext context - Contains all request and response informationRequestDelegate next - The next middleware in the pipelineLet's visualize how a request flows through middleware:
sequenceDiagram
participant Client
participant MW1 as Middleware 1
participant MW2 as Middleware 2
participant MW3 as Middleware 3
participant Endpoint
Client->>MW1: HTTP Request
Note over MW1: Before logic
MW1->>MW2: next()
Note over MW2: Before logic
MW2->>MW3: next()
Note over MW3: Before logic
MW3->>Endpoint: next()
Note over Endpoint: Execute endpoint
Endpoint-->>MW3: Return
Note over MW3: After logic
MW3-->>MW2: Return
Note over MW2: After logic
MW2-->>MW1: Return
Note over MW1: After logic
MW1-->>Client: HTTP Response
Key Points:
next()next()Use()The simplest approach uses lambda expressions:
var builder = WebApplication.CreateBuilder(args);
var app = builder.Build();
app.Use(async (context, next) =>
{
// Before the next middleware
Console.WriteLine($"Request: {context.Request.Path}");
var startTime = DateTime.UtcNow;
// Call the next middleware
await next(context);
// After the next middleware
var elapsed = DateTime.UtcNow - startTime;
Console.WriteLine($"Response: {context.Response.StatusCode} ({elapsed.TotalMilliseconds}ms)");
});
app.MapGet("/", () => "Hello World!");
app.Run();
Output when accessing /:
Request: /
Response: 200 (15.3ms)
Run()Run() creates terminal middleware that ends the pipeline:
var builder = WebApplication.CreateBuilder(args);
var app = builder.Build();
app.Use(async (context, next) =>
{
Console.WriteLine("This executes");
await next(context);
});
app.Run(async context =>
{
Console.WriteLine("This is terminal - no next() to call");
await context.Response.WriteAsync("End of pipeline");
});
// This never executes because Run() terminates the pipeline
app.Use(async (context, next) =>
{
Console.WriteLine("This never executes");
await next(context);
});
app.Run();
For complex middleware, use a class:
// Middleware class
public class RequestTimingMiddleware
{
private readonly RequestDelegate _next;
private readonly ILogger<RequestTimingMiddleware> _logger;
public RequestTimingMiddleware(RequestDelegate next, ILogger<RequestTimingMiddleware> logger)
{
_next = next;
_logger = logger;
}
public async Task InvokeAsync(HttpContext context)
{
var sw = Stopwatch.StartNew();
// Store start time in HttpContext.Items for other middleware to access
context.Items["RequestStartTime"] = DateTime.UtcNow;
try
{
await _next(context);
}
finally
{
sw.Stop();
_logger.LogInformation(
"Request {Method} {Path} completed in {ElapsedMs}ms with status {StatusCode}",
context.Request.Method,
context.Request.Path,
sw.ElapsedMilliseconds,
context.Response.StatusCode
);
}
}
}
// Extension method for convenience
public static class RequestTimingMiddlewareExtensions
{
public static IApplicationBuilder UseRequestTiming(this IApplicationBuilder builder)
{
return builder.UseMiddleware<RequestTimingMiddleware>();
}
}
// Usage
var builder = WebApplication.CreateBuilder(args);
var app = builder.Build();
app.UseRequestTiming();
app.MapGet("/", () => "Hello World!");
app.Run();
Output:
info: RequestTimingMiddleware[0]
Request GET / completed in 12ms with status 200
The order in which you add middleware to the pipeline is critical. Here's the recommended order:
graph TD
A[Exception Handler] --> B[HSTS]
B --> C[HTTPS Redirection]
C --> D[Static Files]
D --> E[Routing]
E --> F[CORS]
F --> G[Authentication]
G --> H[Authorization]
H --> I[Custom Middleware]
I --> J[Session]
J --> K[Response Caching]
K --> L[Response Compression]
L --> M[Endpoints]
Why this order?
Let's see this in code:
var builder = WebApplication.CreateBuilder(args);
// Add services
builder.Services.AddAuthentication();
builder.Services.AddAuthorization();
builder.Services.AddResponseCaching();
builder.Services.AddResponseCompression();
builder.Services.AddSession();
builder.Services.AddCors();
var app = builder.Build();
// 1. Exception handling (must be first)
if (app.Environment.IsDevelopment())
{
app.UseDeveloperExceptionPage();
}
else
{
app.UseExceptionHandler("/error");
app.UseHsts(); // HTTP Strict Transport Security
}
// 2. HTTPS redirection
app.UseHttpsRedirection();
// 3. Static files (can short-circuit)
app.UseStaticFiles();
// 4. Routing (matches endpoints)
app.UseRouting();
// 5. CORS (after routing, before auth)
app.UseCors();
// 6. Authentication (who are you?)
app.UseAuthentication();
// 7. Authorization (what can you do?)
app.UseAuthorization();
// 8. Custom middleware
app.UseRequestTiming();
// 9. Session
app.UseSession();
// 10. Response caching
app.UseResponseCaching();
// 11. Response compression
app.UseResponseCompression();
// 12. Endpoints
app.MapGet("/", () => "Hello World!");
app.Run();
Catches exceptions from later middleware and generates error responses:
var builder = WebApplication.CreateBuilder(args);
var app = builder.Build();
// Global exception handler
app.UseExceptionHandler(errorApp =>
{
errorApp.Run(async context =>
{
context.Response.StatusCode = 500;
context.Response.ContentType = "application/json";
var error = context.Features.Get<IExceptionHandlerFeature>();
if (error != null)
{
var ex = error.Error;
await context.Response.WriteAsJsonAsync(new
{
error = new
{
message = ex.Message,
type = ex.GetType().Name,
stackTrace = app.Environment.IsDevelopment() ? ex.StackTrace : null
}
});
}
});
});
// This will be caught by the exception handler
app.MapGet("/error", () =>
{
throw new InvalidOperationException("Something went wrong!");
});
app.MapGet("/", () => "Hello World!");
app.Run();
Testing:
$ curl http://localhost:5000/error
{
"error": {
"message": "Something went wrong!",
"type": "InvalidOperationException",
"stackTrace": "..."
}
}
Serves static files and short-circuits the pipeline:
var builder = WebApplication.CreateBuilder(args);
var app = builder.Build();
// Default: serves files from wwwroot/
app.UseStaticFiles();
// Serve files from additional directory
app.UseStaticFiles(new StaticFileOptions
{
FileProvider = new PhysicalFileProvider(
Path.Combine(builder.Environment.ContentRootPath, "MyStaticFiles")),
RequestPath = "/StaticFiles"
});
// Custom file type (MIME mapping)
var provider = new FileExtensionContentTypeProvider();
provider.Mappings[".myapp"] = "application/x-myapp";
app.UseStaticFiles(new StaticFileOptions
{
ContentTypeProvider = provider
});
// Enable directory browsing (development only!)
if (app.Environment.IsDevelopment())
{
app.UseDirectoryBrowser(new DirectoryBrowserOptions
{
FileProvider = new PhysicalFileProvider(
Path.Combine(builder.Environment.ContentRootPath, "wwwroot")),
RequestPath = "/browse"
});
}
app.Run();
Flow diagram:
graph TD
A[Request: /css/site.css] --> B{Static Files Middleware}
B -->|File exists| C[Serve file]
C --> D[Return 200]
B -->|File not found| E[Call next middleware]
E --> F[Routing/Endpoints]
F --> G[Return 404]
style C stroke:#10b981,stroke-width:3px
style G stroke:#ef4444,stroke-width:3px
Authenticates the user based on the request:
var builder = WebApplication.CreateBuilder(args);
// Add authentication services
builder.Services.AddAuthentication("Cookies")
.AddCookie("Cookies", options =>
{
options.LoginPath = "/login";
options.LogoutPath = "/logout";
options.ExpireTimeSpan = TimeSpan.FromHours(1);
})
.AddJwtBearer("Bearer", options =>
{
options.TokenValidationParameters = new TokenValidationParameters
{
ValidateIssuer = true,
ValidateAudience = true,
ValidateLifetime = true,
ValidateIssuerSigningKey = true,
ValidIssuer = "https://myapp.com",
ValidAudience = "https://myapp.com",
IssuerSigningKey = new SymmetricSecurityKey(
Encoding.UTF8.GetBytes("your-secret-key-here"))
};
});
var app = builder.Build();
app.UseAuthentication(); // Must be before UseAuthorization()
app.MapGet("/public", () => "Anyone can access this");
app.MapGet("/protected", () => "Only authenticated users can access this")
.RequireAuthorization();
app.MapGet("/login", async (HttpContext context) =>
{
var claims = new[]
{
new Claim(ClaimTypes.Name, "testuser"),
new Claim(ClaimTypes.Email, "test@example.com")
};
var identity = new ClaimsIdentity(claims, "Cookies");
var principal = new ClaimsPrincipal(identity);
await context.SignInAsync("Cookies", principal);
return Results.Ok("Logged in");
});
app.Run();
Determines if an authenticated user has permission to access a resource:
var builder = WebApplication.CreateBuilder(args);
builder.Services.AddAuthentication("Cookies")
.AddCookie("Cookies");
// Configure authorization policies
builder.Services.AddAuthorization(options =>
{
// Policy: Requires specific claim
options.AddPolicy("AdminOnly", policy =>
policy.RequireClaim("role", "admin"));
// Policy: Requires age over 18
options.AddPolicy("Adults", policy =>
policy.RequireAssertion(context =>
{
var ageClaim = context.User.FindFirst("age");
if (ageClaim != null && int.TryParse(ageClaim.Value, out var age))
{
return age >= 18;
}
return false;
}));
// Policy: Combines requirements
options.AddPolicy("AdminOrManager", policy =>
policy.RequireAssertion(context =>
context.User.HasClaim("role", "admin") ||
context.User.HasClaim("role", "manager")));
});
var app = builder.Build();
app.UseAuthentication();
app.UseAuthorization();
// No authorization required
app.MapGet("/public", () => "Public endpoint");
// Requires authentication
app.MapGet("/authenticated", () => "Authenticated endpoint")
.RequireAuthorization();
// Requires specific policy
app.MapGet("/admin", () => "Admin only endpoint")
.RequireAuthorization("AdminOnly");
// Multiple policies
app.MapGet("/restricted", () => "Adults and admins only")
.RequireAuthorization("Adults", "AdminOnly");
app.Run();
Handles Cross-Origin Resource Sharing:
var builder = WebApplication.CreateBuilder(args);
// Configure CORS policies
builder.Services.AddCors(options =>
{
// Policy 1: Allow all
options.AddPolicy("AllowAll", policy =>
{
policy.AllowAnyOrigin()
.AllowAnyMethod()
.AllowAnyHeader();
});
// Policy 2: Specific origin
options.AddPolicy("AllowSpecificOrigin", policy =>
{
policy.WithOrigins("https://example.com", "https://app.example.com")
.AllowAnyMethod()
.AllowAnyHeader()
.AllowCredentials(); // Important for cookies/auth
});
// Policy 3: Specific methods and headers
options.AddPolicy("RestrictedAccess", policy =>
{
policy.WithOrigins("https://partner.com")
.WithMethods("GET", "POST")
.WithHeaders("Content-Type", "Authorization")
.WithExposedHeaders("X-Custom-Header")
.SetPreflightMaxAge(TimeSpan.FromMinutes(10));
});
// Default policy
options.AddDefaultPolicy(policy =>
{
policy.WithOrigins("https://trustedsite.com")
.AllowAnyMethod()
.AllowAnyHeader();
});
});
var app = builder.Build();
app.UseRouting();
// Apply CORS globally
app.UseCors("AllowSpecificOrigin");
// Or apply per endpoint
app.MapGet("/api/data", () => new { data = "Hello" })
.RequireCors("AllowAll");
app.MapGet("/api/restricted", () => new { data = "Restricted" })
.RequireCors("RestrictedAccess");
app.Run();
CORS preflight request flow:
sequenceDiagram
participant Browser
participant CORS as CORS Middleware
participant Endpoint
Browser->>CORS: OPTIONS /api/data<br/>(Preflight Request)
Note over Browser,CORS: Origin: https://example.com<br/>Access-Control-Request-Method: POST<br/>Access-Control-Request-Headers: Content-Type
Note over CORS: Check CORS policy
alt Policy allows
CORS-->>Browser: 204 No Content<br/>Access-Control-Allow-Origin: https://example.com<br/>Access-Control-Allow-Methods: POST<br/>Access-Control-Allow-Headers: Content-Type
Browser->>CORS: POST /api/data<br/>(Actual Request)
CORS->>Endpoint: Forward request
Endpoint-->>CORS: Response
CORS-->>Browser: Response + CORS headers
else Policy denies
CORS-->>Browser: 403 Forbidden
end
Caches responses to improve performance:
var builder = WebApplication.CreateBuilder(args);
builder.Services.AddResponseCaching();
var app = builder.Build();
app.UseResponseCaching();
// Cache this endpoint
app.MapGet("/cached", (HttpContext context) =>
{
context.Response.Headers.CacheControl = "public,max-age=60"; // Cache for 60 seconds
return $"Generated at {DateTime.UtcNow:HH:mm:ss}";
});
// Don't cache this
app.MapGet("/no-cache", (HttpContext context) =>
{
context.Response.Headers.CacheControl = "no-cache";
return $"Generated at {DateTime.UtcNow:HH:mm:ss}";
});
// Conditional caching
app.MapGet("/data", (HttpContext context, string? cache) =>
{
if (cache == "yes")
{
context.Response.Headers.CacheControl = "public,max-age=30";
}
return new
{
timestamp = DateTime.UtcNow,
data = "Some data"
};
});
app.Run();
Testing:
# First request - generates response and caches it
$ curl -i http://localhost:5000/cached
Date: Mon, 15 Jan 2024 10:30:00 GMT
Cache-Control: public,max-age=60
Generated at 10:30:00
# Second request within 60 seconds - served from cache
$ curl -i http://localhost:5000/cached
Date: Mon, 15 Jan 2024 10:30:00 GMT # Same time!
Cache-Control: public,max-age=60
Age: 15 # Cache age in seconds
Generated at 10:30:00 # Same response!
Compresses responses to reduce bandwidth:
var builder = WebApplication.CreateBuilder(args);
builder.Services.AddResponseCompression(options =>
{
options.EnableForHttps = true; // Enable for HTTPS (be aware of CRIME attack)
// Providers (order matters - tried in sequence)
options.Providers.Add<BrotliCompressionProvider>();
options.Providers.Add<GzipCompressionProvider>();
// MIME types to compress
options.MimeTypes = ResponseCompressionDefaults.MimeTypes.Concat(
new[] { "application/json", "text/plain", "text/css", "application/javascript" });
});
// Configure compression levels
builder.Services.Configure<BrotliCompressionProviderOptions>(options =>
{
options.Level = CompressionLevel.Fastest; // Fastest, Optimal, SmallestSize
});
builder.Services.Configure<GzipCompressionProviderOptions>(options =>
{
options.Level = CompressionLevel.Optimal;
});
var app = builder.Build();
app.UseResponseCompression();
app.MapGet("/large", () =>
{
// Generate large response
var data = string.Join("", Enumerable.Repeat("Hello World! ", 1000));
return Results.Text(data, "text/plain");
});
app.Run();
Testing:
# Without compression
$ curl -H "Accept-Encoding:" http://localhost:5000/large
# Response: ~13KB
# With gzip
$ curl -H "Accept-Encoding: gzip" http://localhost:5000/large
# Response: ~100 bytes (compressed)
# Header: Content-Encoding: gzip
# With brotli (better compression)
$ curl -H "Accept-Encoding: br" http://localhost:5000/large
# Response: ~60 bytes (compressed)
# Header: Content-Encoding: br
Adds a unique identifier to each request for tracing:
public class RequestIdMiddleware
{
private readonly RequestDelegate _next;
private readonly ILogger<RequestIdMiddleware> _logger;
public RequestIdMiddleware(RequestDelegate next, ILogger<RequestIdMiddleware> logger)
{
_next = next;
_logger = logger;
}
public async Task InvokeAsync(HttpContext context)
{
// Check if request already has an ID (from client)
var requestId = context.Request.Headers["X-Request-ID"].FirstOrDefault()
?? Guid.NewGuid().ToString();
// Add to response headers
context.Response.Headers["X-Request-ID"] = requestId;
// Store in HttpContext for other middleware/endpoints to access
context.Items["RequestId"] = requestId;
// Use scope for structured logging
using (_logger.BeginScope(new Dictionary<string, object>
{
["RequestId"] = requestId
}))
{
_logger.LogInformation("Processing request {RequestId}", requestId);
await _next(context);
_logger.LogInformation("Completed request {RequestId}", requestId);
}
}
}
Custom authentication for API keys:
public class ApiKeyMiddleware
{
private readonly RequestDelegate _next;
private readonly IConfiguration _configuration;
private const string API_KEY_HEADER = "X-API-Key";
public ApiKeyMiddleware(RequestDelegate next, IConfiguration configuration)
{
_next = next;
_configuration = configuration;
}
public async Task InvokeAsync(HttpContext context)
{
// Skip authentication for public endpoints
if (context.Request.Path.StartsWithSegments("/public"))
{
await _next(context);
return;
}
// Check for API key in header
if (!context.Request.Headers.TryGetValue(API_KEY_HEADER, out var providedKey))
{
context.Response.StatusCode = 401;
await context.Response.WriteAsJsonAsync(new { error = "API Key is missing" });
return;
}
// Validate API key (in production, check against database)
var validApiKey = _configuration["ApiKey"];
if (providedKey != validApiKey)
{
context.Response.StatusCode = 403;
await context.Response.WriteAsJsonAsync(new { error = "Invalid API Key" });
return;
}
// Set user identity based on API key
var claims = new[] { new Claim("ApiKey", providedKey!) };
var identity = new ClaimsIdentity(claims, "ApiKey");
context.User = new ClaimsPrincipal(identity);
await _next(context);
}
}
Limits requests per IP address:
public class RateLimitingMiddleware
{
private readonly RequestDelegate _next;
private readonly IMemoryCache _cache;
private readonly int _requestLimit;
private readonly TimeSpan _timeWindow;
public RateLimitingMiddleware(
RequestDelegate next,
IMemoryCache cache,
int requestLimit = 100,
int timeWindowSeconds = 60)
{
_next = next;
_cache = cache;
_requestLimit = requestLimit;
_timeWindow = TimeSpan.FromSeconds(timeWindowSeconds);
}
public async Task InvokeAsync(HttpContext context)
{
var clientIp = context.Connection.RemoteIpAddress?.ToString() ?? "unknown";
var cacheKey = $"RateLimit_{clientIp}";
// Get current request count
var requestCount = _cache.GetOrCreate(cacheKey, entry =>
{
entry.AbsoluteExpirationRelativeToNow = _timeWindow;
return 0;
});
if (requestCount >= _requestLimit)
{
context.Response.StatusCode = 429; // Too Many Requests
context.Response.Headers["Retry-After"] = _timeWindow.TotalSeconds.ToString();
await context.Response.WriteAsJsonAsync(new
{
error = "Rate limit exceeded",
retryAfter = _timeWindow.TotalSeconds
});
return;
}
// Increment request count
_cache.Set(cacheKey, requestCount + 1, _timeWindow);
// Add rate limit headers
context.Response.Headers["X-RateLimit-Limit"] = _requestLimit.ToString();
context.Response.Headers["X-RateLimit-Remaining"] = (_requestLimit - requestCount - 1).ToString();
context.Response.Headers["X-RateLimit-Reset"] = DateTimeOffset.UtcNow.Add(_timeWindow).ToUnixTimeSeconds().ToString();
await _next(context);
}
}
// Usage
var builder = WebApplication.CreateBuilder(args);
builder.Services.AddMemoryCache();
var app = builder.Build();
app.UseMiddleware<RateLimitingMiddleware>(100, 60); // 100 requests per 60 seconds
app.Run();
Testing:
# First request - OK
$ curl -i http://localhost:5000/api/data
HTTP/1.1 200 OK
X-RateLimit-Limit: 100
X-RateLimit-Remaining: 99
X-RateLimit-Reset: 1705318860
# After 100 requests
$ curl -i http://localhost:5000/api/data
HTTP/1.1 429 Too Many Requests
Retry-After: 60
X-RateLimit-Limit: 100
X-RateLimit-Remaining: 0
X-RateLimit-Reset: 1705318860
{"error":"Rate limit exceeded","retryAfter":60}
You can branch the pipeline based on conditions:
var builder = WebApplication.CreateBuilder(args);
var app = builder.Build();
// Map: Creates a branch based on path prefix
app.Map("/api", apiApp =>
{
apiApp.UseMiddleware<ApiKeyMiddleware>();
apiApp.MapGet("/data", () => new { data = "API data" });
});
// MapWhen: Creates a branch based on custom condition
app.MapWhen(
context => context.Request.Headers.ContainsKey("X-Custom-Header"),
customApp =>
{
customApp.Use(async (context, next) =>
{
context.Response.Headers["X-Custom-Response"] = "Matched!";
await next(context);
});
customApp.MapGet("/special", () => "Special endpoint");
});
// UseWhen: Rejoins the main pipeline after the branch
app.UseWhen(
context => context.Request.Path.StartsWithSegments("/admin"),
adminApp =>
{
adminApp.Use(async (context, next) =>
{
// Log admin access
Console.WriteLine("Admin area accessed");
await next(context);
});
});
app.MapGet("/", () => "Main pipeline");
app.Run();
Branching visualization:
graph TD
A[Request] --> B{Path?}
B -->|/api/*| C[API Branch]
B -->|/admin/*| D[Admin Branch UseWhen]
B -->|Other| E[Main Pipeline]
C --> F[API Key Middleware]
F --> G[API Endpoints]
G --> Z1[Response]
D --> H[Admin Logging]
H --> I[Rejoin Main Pipeline]
I --> J[Main Pipeline Continues]
J --> Z2[Response]
E --> K[Main Endpoints]
K --> Z3[Response]
Each middleware should have a single responsibility:
// ❌ Bad: Doing too much
public class BadMiddleware
{
public async Task InvokeAsync(HttpContext context)
{
// Authentication
// Authorization
// Logging
// Rate limiting
// Response modification
// ... too much!
}
}
// ✅ Good: Single responsibility
public class AuthenticationMiddleware { /* Only authentication */ }
public class LoggingMiddleware { /* Only logging */ }
public class RateLimitingMiddleware { /* Only rate limiting */ }
public class SafeMiddleware
{
private readonly RequestDelegate _next;
private readonly ILogger<SafeMiddleware> _logger;
public SafeMiddleware(RequestDelegate next, ILogger<SafeMiddleware> logger)
{
_next = next;
_logger = logger;
}
public async Task InvokeAsync(HttpContext context)
{
try
{
await _next(context);
}
catch (Exception ex)
{
_logger.LogError(ex, "Error in middleware");
// Don't swallow exceptions - let exception handler middleware handle them
throw;
}
}
}
Don't modify the response after it has started:
public class ResponseSafeMiddleware
{
private readonly RequestDelegate _next;
public ResponseSafeMiddleware(RequestDelegate next)
{
_next = next;
}
public async Task InvokeAsync(HttpContext context)
{
// Good: Check before modifying
if (!context.Response.HasStarted)
{
context.Response.Headers["X-Custom-Header"] = "Value";
}
await _next(context);
// Bad: Response might have already started
// context.Response.Headers["X-After-Header"] = "Value"; // This might throw!
// ✅ Good: Check first
if (!context.Response.HasStarted)
{
context.Response.Headers["X-After-Header"] = "Value";
}
}
}
public class DIAwareMiddleware
{
private readonly RequestDelegate _next;
// Singleton services injected in constructor
private readonly ILogger<DIAwareMiddleware> _logger;
public DIAwareMiddleware(RequestDelegate next, ILogger<DIAwareMiddleware> logger)
{
_next = next;
_logger = logger;
}
// Scoped/Transient services injected in InvokeAsync
public async Task InvokeAsync(HttpContext context, IMyService myService)
{
// myService is scoped to this request
var data = await myService.GetDataAsync();
await _next(context);
}
}
next()app.Use() for inline middleware, classes for complex middlewareMap() and MapWhen()Response.HasStarted before modifying headersInvokeAsync()Understanding middleware gives you complete control over how requests are processed. Every cross-cutting concern—from logging to authentication to custom business logic—can be elegantly handled as middleware.
Continue to Part 4: Routing and Endpoints to learn how requests are matched to endpoints and how the routing system works.
© 2025 Scott Galloway — Unlicense — All content and source code on this site is free to use, copy, modify, and sell.