Driver-Specific Health Checks

Each repository driver provides optimized health checks tailored to its underlying data store.

Entity Framework Core

Package: Kista.HealthChecks.EntityFramework

What It Checks

1. Database Connectivity: Calls DbContext.Database.CanConnectAsync() to verify the database is reachable
2. Optional Query Test: If enabled, runs context.Set<TEntity>().AnyAsync() to verify query execution

Configuration

builder.Services
    .AddRepositoryContext()
    .UseEntityFramework<MyDbContext>(ef => ef
        .WithHealthChecks(options => {
            options.Timeout = TimeSpan.FromSeconds(10);
            options.TestQuery = true;  // Run a test query
            options.Tags = ["data", "sql"];
        }));

Diagnostic Data

When healthy:

{
  "EntityType": "ProductEntity",
  "KeyType": "Int32",
  "DbContextType": "MyApp.Data.MyDbContext",
  "EntityExists": true,  // Only if TestQuery = true
  "ResponseType": "Healthy"
}

When unhealthy:

{
  "EntityType": "ProductEntity",
  "KeyType": "Int32",
  "DbContextType": "MyApp.Data.MyDbContext",
  "ConnectionState": "Disconnected",  // or ExceptionType
  "ResponseType": "Unhealthy"
}

Best Practices

✅ Enable TestQuery for critical entities - Ensures queries can execute, not just connect
✅ Use longer timeouts for cold databases - Some databases take time to wake up
✅ Monitor DbContext type - Helps identify which context is failing in multi-context scenarios

⚠️ Avoid heavy queries - TestQuery runs AnyAsync(), which is lightweight
⚠️ Consider connection pooling - Health checks can exhaust pools if timeout is too short

---

MongoDB

Package: Kista.HealthChecks.MongoFramework

What It Checks

1. Cluster Connectivity: Runs a MongoDB ping command to verify the cluster is reachable
2. Database Accessibility: Confirms the configured database can be accessed

Configuration

builder.Services
    .AddRepositoryContext()
    .UseMongoDB<MyMongoContext>(mongo => mongo
        .WithHealthChecks(options => {
            options.Timeout = TimeSpan.FromSeconds(5);
            options.PingTimeout = TimeSpan.FromSeconds(3);
            options.Tags = ["data", "mongo"];
        }));

Diagnostic Data

When healthy:

{
  "EntityType": "CustomerEntity",
  "KeyType": "String",
  "ResponseType": "Healthy"
}

When unhealthy:

{
  "EntityType": "CustomerEntity",
  "KeyType": "String",
  "ExceptionType": "MongoDB.Driver.MongoConnectionException",
  "ResponseType": "Unhealthy"
}

Best Practices

✅ Set PingTimeout shorter than Timeout - Fail fast on MongoDB issues
✅ Monitor replica set status - Health checks work with replica sets and sharded clusters
✅ Use tags for environment separation - Differentiate prod vs. dev MongoDB instances

⚠️ Network latency - MongoDB health checks are network-dependent
⚠️ Authentication failures - Will show as connection failures

---

In-Memory

Package: Kista.HealthChecks.InMemory

What It Checks

The In-Memory health check always returns Healthy if the repository is registered. Since there are no external dependencies, the check simply confirms the repository is available in the DI container.

Configuration

builder.Services
    .AddRepositoryContext()
    .UseInMemory(inMem => inMem
        .WithHealthChecks(options => {
            options.Tags = ["cache", "live"];
        }));

Diagnostic Data

Always healthy:

{
  "EntityType": "CacheEntity",
  "KeyType": "Int32",
  "ResponseType": "Healthy"
}

Use Cases

✅ Testing - Perfect for integration tests that need health check infrastructure
✅ Development - Run the full application stack without external dependencies
✅ Caching layers - Monitor in-memory caches alongside persistent stores

⚠️ Not for production monitoring - Doesn't verify actual data store health
⚠️ Always healthy - Won't detect application-level issues

---

Driver Comparison

Feature	Entity Framework	MongoDB	In-Memory
Connectivity Test	✅ CanConnectAsync()	✅ Ping command	❌ N/A
Query Test	✅ Optional (TestQuery)	❌ N/A	❌ N/A
External Dependency	✅ Database server	✅ MongoDB cluster	❌ None
Typical Timeout	5-10 seconds	3-5 seconds	1 second
Failure Modes	Connection, auth, timeout	Connection, auth, timeout	None
Best For	Relational data	Document storage	Testing, caching

---

Multi-Driver Scenarios

When using multiple drivers in the same application:

builder.Services
    .AddRepositoryContext()
    .UseEntityFramework<MyDbContext>(ef => ef
        .WithHealthChecks(options => {
            options.Tags = ["data", "sql", "ready"];
        }))
    .UseMongoDB<MyMongoContext>(mongo => mongo
        .WithHealthChecks(options => {
            options.Tags = ["data", "mongo", "ready"];
        }))
    .UseInMemory(inMem => inMem
        .WithHealthChecks(options => {
            options.Tags = ["cache", "live"];
        }));

builder.Services
    .AddHealthChecks()
    .AddKistaRepositories();

Health check names will be:

• "Kista:EntityFramework:ProductEntity"
• "Kista:MongoDB:CustomerEntity"
• "Kista:InMemory:CacheEntity"

---

Next Steps

• Advanced Scenarios - Readiness/liveness probes, custom naming
• Configuration Guide - All configuration options