Health checks in .NET are a way to monitor the health of your application and its dependencies. They help you periodically check the state of critical components, such as databases or external services, to ensure they are working properly.
Health checks are typically implemented as endpoints or middleware in your application that respond to HTTP requests and return a status indicating the health of the application or its dependencies. The status can be one of the following:
By implementing health checks, you can improve the reliability and availability of your application.
This github repository show cases how health checks can be implemented, it covers some of the following:
The github repo was created to show how health checks work when they succeed or fail by simply updating some environment variables. A .http file has also been added for ease of endpoint exportation and manual testing.
Firstly I'll cover the basics of setting up health checks.
This example would be the setup in a Program.cs
file for a minimal API project. This is all that is need to get started.
var builder = WebApplication.CreateBuilder(args);
builder.Services.AddHealthChecks() //Adds health checks
var app = builder.Build();
app.MapHealthChecks("/healthz"); //Registers endpoint for health checks
app.Run();
There are already many health checks available via Nuget. To get a idea of what is available, have a look at this Nuget search result.
In this example I will make use of a heath check that checks a URL for a successful response.
Add this nuget reference
<PackageReference Include="AspNetCore.HealthChecks.Uris" Version="6.0.3" />
Then you can add the health check
services.AddHealthChecks()
.AddUrlGroup(new Uri("https://matthewregis.dev"), name: "UriCheck");
Although there are many health check libraries available, they might not fit your needs, in which case creating your own health check is probably your best option.
Create a health check class that implements the IHealthCheck
interface. In the example below you can see you can make use of dependency injection. This specific example is a health check that checks that a config value has been populated, this is a more realistic example that might be used in a real world application.
using Microsoft.Extensions.Diagnostics.HealthChecks;
namespace Healthz.Api
{
public class MyCustomStartUpHealthCheck : IHealthCheck
{
private readonly IConfiguration configuration;
public MyCustomStartUpHealthCheck(IConfiguration configuration)
{
this.configuration = configuration;
}
public Task<HealthCheckResult> CheckHealthAsync(
HealthCheckContext context,
CancellationToken cancellationToken = default)
{
if (string.IsNullOrWhiteSpace(this.configuration[$"Db:ConnectionString"]))
{
return Task.FromResult(HealthCheckResult.Unhealthy("'Db:ConnectionString' empty!"));
}
return Task.FromResult(HealthCheckResult.Healthy());
}
}
}
Then you can register the health check like the following:
services.AddHealthChecks()
.AddCheck<MyCustomStartUpHealthCheck>(nameof(MyCustomStartUpHealthCheck));
NOTE: The MyCustomStartUpHealthCheck.cs example in the github repo returns a Healthy or Unhealthy result based of what a configuration value is set too. This is not a health check you would use in the real world, it was created for demonstrating purposes of what a Healthy or Unhealthy result looks like by easily updating environment variables and seeing the results.
You can register health checks with tags and filter on those tags so you only run a subset of health checks matching those tags. One use case in having start up checks and liveness checks separate, so you only run start up checks when a application starts and then run liveness checks at a certain timed interval.
An example of this is in the HealthCheckExtensions.cs in the github repo mentioned previously.
First register with tags.
public static void RegisterHealthChecks(this IServiceCollection services, string? healthCheckUri)
{
var url = string.IsNullOrEmpty(healthCheckUri) ? "http://localhost" : healthCheckUri;
services.AddHealthChecks()
.AddUrlGroup(new Uri(url), name: "UriCheck", tags: new[] { "live", "all" })
.AddCheck<MyCustomStartUpHealthCheck>(nameof(MyCustomStartUpHealthCheck), tags: new[] { "ready", "all" });
}
Secondly filter on those tags. Here three separate endpoints were created and only runs health checks with those tags associated with it.
public static void MapHealthChecks(this WebApplication app)
{
app.MapHealthChecks("/healthz/live", new HealthCheckOptions
{
ResponseWriter = WriteResponse,
Predicate = healthCheck => healthCheck.Tags.Contains("live")
});
app.MapHealthChecks("/healthz/ready", new HealthCheckOptions
{
ResponseWriter = WriteResponse,
Predicate = healthCheck => healthCheck.Tags.Contains("ready")
});
app.MapHealthChecks("/healthz/all", new HealthCheckOptions
{
ResponseWriter = WriteResponse,
Predicate = healthCheck => healthCheck.Tags.Contains("all")
});
}
If you are familiar with with Kubernetes, you can configure your deployment with readiness and liveness probes which is a prefect use case for separating your health checks.
This was taken from the Health checks : Customize output documentation as an example.
This highlights how you can customise the output how you see fit, you don't necessarily have to implement it like the following. Also worth noting that there are existing health check frameworks that would expect a response to be in a certain structure, this might help drive your decision on what structure to use.
app.MapHealthChecks("/healthz", new HealthCheckOptions
{
ResponseWriter = WriteResponse
});
private static Task WriteResponse(HttpContext context, HealthReport healthReport)
{
context.Response.ContentType = "application/json; charset=utf-8";
var options = new JsonWriterOptions { Indented = true };
using var memoryStream = new MemoryStream();
using (var jsonWriter = new Utf8JsonWriter(memoryStream, options))
{
jsonWriter.WriteStartObject();
jsonWriter.WriteString("status", healthReport.Status.ToString());
jsonWriter.WriteStartObject("results");
foreach (var healthReportEntry in healthReport.Entries)
{
jsonWriter.WriteStartObject(healthReportEntry.Key);
jsonWriter.WriteString("status",
healthReportEntry.Value.Status.ToString());
jsonWriter.WriteString("description",
healthReportEntry.Value.Description);
jsonWriter.WriteStartObject("data");
foreach (var item in healthReportEntry.Value.Data)
{
jsonWriter.WritePropertyName(item.Key);
JsonSerializer.Serialize(jsonWriter, item.Value,
item.Value?.GetType() ?? typeof(object));
}
jsonWriter.WriteEndObject();
jsonWriter.WriteEndObject();
}
jsonWriter.WriteEndObject();
jsonWriter.WriteEndObject();
}
return context.Response.WriteAsync(
Encoding.UTF8.GetString(memoryStream.ToArray()));
}