Module 15: Logging, Monitoring & Performance

1. Logging with Serilog / NLog

Standard Console.WriteLine is useless in production. You need Structured Logging.

• Structured Logging: Instead of saving a string like "User 5 logged in", Serilog saves a JSON object: {"UserId": 5, "Event": "Login"}. This allows you to search and filter logs easily.
• Sinks: Serilog can send logs to multiple destinations (Sinks) simultaneously—like a local File, a Database, or cloud tools like Azure Application Insights or Seq.
• Log Levels:
  • Information: General flow (User started checkout).
  • Warning: Something unusual but not a crash (Disk space low).
  • Error: A crash in a specific request (Payment failed).
  • Critical: The whole system is down.

2. Global Exception Handling

In a professional app, you don’t put try-catch blocks everywhere. It makes the code messy.

• Middleware Approach: You create a single piece of Middleware that “wraps” the entire request pipeline.
• Benefits:
  1. If any error occurs anywhere, the middleware catches it.
  2. It logs the error details for the developer.
  3. It returns a clean, “user-friendly” JSON response to the client (e.g., "Something went wrong, please try again later") instead of a scary technical stack trace.

3. Health Checks

How do you know if your app is “alive”?

• The Problem: The web server might be running, but the Database is down, or the Redis cache is disconnected.
• The Solution: ASP.NET Core Health Checks provide an endpoint (usually /health) that returns the status of all dependencies.
• Usage: Load balancers (like Kubernetes or Azure) ping this endpoint. If it returns “Unhealthy,” they stop sending traffic to that instance and try to restart it.

4. Caching (In-Memory & Redis)

Caching is the #1 way to improve performance. It stores frequently used data in high-speed memory so you don’t have to query the database every time.

• Cache Aside Pattern: Check the cache first. If data is there (Hit), return it. If not (Miss), get it from the DB, save it to the cache, and then return it.

5. Performance Optimization Techniques

To build scalable systems, you must eliminate bottlenecks:

• Async/Await: Ensure your entire pipeline is asynchronous. This prevents “Thread Starvation,” allowing one server to handle thousands of concurrent users.
• Pagination: Never return List<Product> with 10,000 items. Use .Skip(20).Take(10) to return small pages of data.
• Response Compression: Use middleware to “Gzip” your JSON responses, making them much smaller and faster to send over the internet.
• Database Indexing: As covered in Module 12, ensuring your SQL queries use indexes is critical for speed.

6. Monitoring & Metrics

Monitoring is about looking at the “big picture” over time.

• Metrics: Tracking CPU usage, Memory consumption, and “Requests Per Second.”
• Tracing: Following a single user’s request as it travels through multiple microservices (using tools like OpenTelemetry).