Hi!
Great project we started implementing it recently and we love it so far!
Question: is support for iot/event hub in the pipes?
Thanks!

I can't seem to find the HealthChecks.UI nuget package. I checked the last AppVeyor build and it looks like the package is getting built, but not uploaded to MyGet.

Hi
i have a problem with healthchecksdb file !
how can i config the location of this file to make right permissions ?

Application startup exception: System.AggregateException: One or more errors occurred. (Access to the path 'C:\apps\healthchecksdb' is denied.) ---> System.UnauthorizedAccessException: Access to the path 'C:\apps\healthchecksdb' is denied. at System.IO.FileSystem.DeleteFile(String fullPath) at System.IO.File.Delete(String path) at Microsoft.EntityFrameworkCore.Storage.RelationalDatabaseCreator.DeleteAsync(CancellationToken cancellationToken) at Microsoft.EntityFrameworkCore.Storage.RelationalDatabaseCreator.EnsureDeletedAsync(CancellationToken cancellationToken) at Microsoft.Extensions.DependencyInjection.ServiceCollectionExtensions.CreateDatabase(IServiceProvider serviceProvider) --- End of inner exception stack trace ---

Have you considered options other than persisting a sqlite file to disk for the healthcheckdb?

We've been moving toward an immutable runtime environment model (Azure App Service using "Run From Package"), which means writing that file to disk is going to fail.

It looks like I can forgo instantiating the UI, and just query /healthchecks-api for check results, which may suffice for our needs. But, a UI is always nice.

Currently the project is creating a new and disposing of HttpClient per health check. Microsoft introduced IHttpClientFactory to help abstract the lifetime management of HttpClient as well as some other benefits in their original announcement. Managing HttpClient is also covered in a stack overflow topic: Do HttpClient and HttpClientHandler have to be disposed?

Could you please note that you need to get the health check packages from my get (it wasn't to obvious at least for me :) )

the class MaximumValueHealthCheck has the wfollowing property

        private readonly T maximunValue;`

the wording should be fixed here and in some other place
maximunValue => maximumValue

impact using UIResponseWriter and MemoryHealthCheck:

"virtualmemory":{"data":{},"description":"Maximun=2147483648, Current=.....

regards
Werner

Hi I was just wonder about how you'll integrate Prometheus in the port.

Do you have plans to port the PrometheusTracker? I was also kind of wondering if would be of interest to have a Prometheus"ResponseWriter" so that you could expose the metrics for the HealthCheck as straight Prometheus metrics.

Hi,

the response of GET requests produces an error response when a services is unavailable which is then unusable, e.g. for a health check ui service. First I used the healt check support for AzureStorageEmulator and I thought this is an emulator specific issue, but I've also tried the provided health check support for Identity Server and Uri and everything is working well as long as the services are available - when it is not, a self referencing loop was detected.

{
  "error": {
    "code": "500",
    "message": "Self referencing loop detected for property 'exception' with type 'Microsoft.WindowsAzure.Storage.StorageException'. Path  'entries.azureblob.exception.requestInformation'."
  }
}

The ConfigureServices method contains this:

services.AddHealthChecks().AddAzureBlobStorage(configuration.GetConnectionString("AzureStorage"));

TheConfigure` method looks like this:

app.UseHealthChecks("/api/health", new HealthCheckOptions()
{
    Predicate = _ => true,
    ResponseWriter = UIResponseWriter.WriteHealthCheckUIResponse
});

RedisCheck should switch to new StackExchange.Redis library as oppose to old 1.2.6 version; this creates reference confusion for VS.

We have environments with unencrypted SMTP connections and blank password. Yeah, it's bad, but it's the way it is.

When using a non TLS/SSL connection, thus having SmtpConnectionType.AUTO, this method raises an exception.

My temporary dirty fix that actually works:

ConnectionType = (SmtpConnectionType)int.MaxValue;

With that one "fixed" my next problem is the blank password which raises an exception:

Hope support for unencrypted connections and blank password can be added. Thanks!

Package AspNetCore.HealthChecks.IdSvr is listed in the readme file but it does not exist on the NuGet.org feed

.Net Core 2.2

http://localhost:5000/healthchecks-api

ConfigureServices

services.AddHealthChecks().AddSqlServer(_ConnectionString,name:"database").AddMQTTChatHealthChecks();
services.AddHealthChecksUI();

        private static void AddMQTTChatHealthChecks(this IHealthChecksBuilder builder)
        {
            builder.AddPrivateMemoryHealthCheck(1024 * 1024 * 1024,"privatememory")
             .AddDiskStorageHealthCheck(setup =>
             {
                 DriveInfo.GetDrives().ToList().ForEach(di =>
                 {
                     setup.AddDrive(di.Name, 1024);
                 });
             });
        }

Configure(IApplicationBuilder app, IHostingEnvironment env,------)
app.UseHealthChecksUI();

Allow to provide the name of the vault and credentials, and check that the key vault is reachable.

Does that make sense?

See dotnet/extensions#639

I wanted to show health check result of different projects ( Micro Services) in one project.
I set up configuration like HealthChecks.UIAndApiCustomization sample and added my other project health url like this :

"HealthChecks-UI": {
"HealthChecks": [
{
"Name": "Http and UI on single project with customizations",
"Uri": "http://localhost:8001/custom/healthz"
},
{
"Name": "Payments",
"Uri": "https://localhost:44386/health"
}

],

but the result from other project is not been showed in the HealthChecksUI. It just shows the first project health result.
Is it the correct way to do this? Do I miss something ?

Thanks in advanced.
Saeideh

Currently some HealthChecks can throw, but they should always return a HealthCheckResult.

For example in the AzureBlobStorage

Throws:

FormatException: Settings must be of the form "name=value".
Microsoft.WindowsAzure.Storage.CloudStorageAccount+<>c.<Parse>b__97_0(string err) in CloudStorageAccount.cs, line 529

The CloudStorageAccount.Parse(connectionString) throws if the connectionstring is invalid. A solution for this would be to take in the connectionstring in the constructor as we do now, but parse the string in the CheckHealthAsync function. Which will catch the exception and return a HealthCheckResult

Hi Everyone

How can i use HealthCheck for REST Api Service Healthy ?

Thanks for your suggest.

Following the steps in the Scott Hanselman post and the ASP.NET 2.2 video, I've added the following to ConfigureServices

	// Register health check services
	services.AddHealthChecks()				
		.AddDbContextCheck<ApplicationDbContext>();
	services.AddHealthChecksUI();

In Configure added the following:

	// Healthcheck endpoint
	app.UseHealthChecks("/healthcheck", new HealthCheckOptions() {
		Predicate = _ => true,
		ResponseWriter = UIResponseWriter.WriteHealthCheckUIResponse
	});
	app.UseHealthChecksUI(setup => { setup.ApiPath = "/healthcheck"; setup.UIPath = "/healthcheckui"; });

And in appsettings

	"HealthChecks-UI": {
		"HealthChecks": [
			{
				"Name": "HTTP-Api-Basic",
				"Uri": "http://localhost:44393/healthcheck"
			}
		],
		"Webhooks": [
			{
				"Name": "Failhook",
				"Uri": "",
				"Payload": "",
				"RestoredPayload": ""
			}
		],
		"EvaluationTimeOnSeconds": 15,
		"MinimumSecondsBetweenFailureNotifications": 60
	}

When I navigate to /healthcheck I get the following JSON

{"status":"Healthy","totalDuration":"00:00:00.0029078","entries":{"applicationDbContext":{"data":{},"duration":"00:00:00.0026669","status":"Healthy"}}}

But when I navigate to /healthcheckui I get the following message "Could not retrieve health checks data"

I can see it's reading the appsettings as the Webhook name changes.

What am I doing wrong?

My (somewhat limited!) understanding is that for a Service Bus QueueClient for a particular queue should be a Singleton across an application.

Azure/azure-service-bus-dotnet#130

Now this issue is quite old now and I know there was some larger changes for ServiceBus in 3.X.X but I think this still applies as "best practice".

Currently the Healthchecks create a new client per check for ServiceBus. Should the interface instead be passed a QueueClient to use continously or create one for it's own uses instead?

Hello,

Elasticsearch behind X-Pack Shield is protected with a basic auth.
Here a sample to create an ElasticClient with basic auth:

var esUri = new Uri(Configuration.Uri);
var settings = new ConnectionSettings(esUri)
     .BasicAuthentication(Configuration.UserName, Configuration.Password);
var lowlevelClient = new ElasticClient(settings);

Can we add this to the current ElasticsearchHealthCheck?

Currently , I am using Healtchecks UI in my application which is running inside a docker container.
When i use
FROM microsoft/dotnet:2.2-aspnetcore-runtime-alpine
as base image , the container unexpectedly exits with code 139 with no log ( not that i could see ).
However when i have changed the base image to
FROM microsoft/dotnet:2.2-aspnetcore-runtime
it works fine.
Is there any way to make it work with alpine runtime image?

What often comes up is that AppSettings are not specified correctly, however I don't believe in the idea that the application should not StartUp if the AppSettings are invalid. I'd rather let the application start up and afterwards check the Settings using a Health Check.

API Proprosal:

services.AddAppSettingsHealthCheck<AppSettings,AppSettingsValidator>()

AppSettings

    public class AppSettings
    {
        public string MySetting1 { get; set; }
        public int MySetting2 { get; set; }
    }

AppSettings Validator (FluentValidation)

public class AppSettingsValidator : AbstractValidator<AppSettings>
    {
        public AppSettingsValidator()
        {
            RuleFor(x => x.MySetting1).IsValidSQLConnectionString();
            RuleFor(x => x.MySetting2).NotEmpty();
        }
    }

HealthCheck

• No Generics yet, but you'll get the point
• Using FluentValidation to check for the settings;

public class AppSettingsHealthCheck : IHealthCheck
    {
        private readonly AppSettings _settings;
        private readonly IValidator<AppSettings> _validator;

        public AppSettingsHealthCheck(IOptions<AppSettings>, IValidator<AppSettings> validator)
        {
            _settings = settings.Value;
            _validator = validator;
        }
        public async Task<HealthCheckResult> CheckHealthAsync(HealthCheckContext context, CancellationToken cancellationToken = default(CancellationToken))
        {
            var result = await _validator.ValidateAsync(_settings);
            if (result.IsValid)
                return HealthCheckResult.Healthy("AppSettings are valid.");

            return HealthCheckResult.Unhealthy("AppSettings are invalid", new ValidationException("AppSettings Validation Failed", result.Errors));
        }
    }

Result (as an example)

• MySetting1 is not a valid ConnectionString
• MySetting2 is cannot be empty

The exception on this line is never logged since the exception object is interpreted as an arg of the message.

catch (Exception ex)
{
    _logger.LogError("HealthCheck collector HostedService throw a error:", ex);
}

should be (spelling mistake fixed as well):

catch (Exception ex)
{
    _logger.LogError(ex, "HealthCheck collector HostedService threw an error");
}

Hi,

Is it possible to provide a timeout for when checking a URL? If nothing responds in 1 minute then consider the URL Degraded. I'm currently doing the following:

AddUrlGroup(new Uri("http://anotherurl/"), "Description", Microsoft.Extensions.Diagnostics.HealthChecks.HealthStatus.Degraded)

Help will be much appreciated!

Is it any way how to expand all rows by default after start?

Thanks

Just so you are aware, using the Client.UI with InProcess hosting will throw an error that it cannot read the SQLite database. The issue seems to be caused by the following note found here:

Directory.GetCurrentDirectory() returns the worker directory of the process started by IIS rather than the application directory (for example, C:\Windows\System32\inetsrv for w3wp.exe).

There is a more detailed discussion of the issue and possible work arounds here.

I think it is a little bit confusing having the name AspNetCore.Diagnostics.HealthChecks when comparing it to Microsoft.AspNetCore.Diagnostics.HealthChecks. Maybe we should move to another name like BeatPulse.AspNetCore.HealthChecks . This will also help newcomers to see this is the new repository because the old BeatPulse repository is still growing more than this.

This is just an idea. What do you think?.

I'm working on a Hangfire HealtCheck package.

Is there anything that i should be aware of before creating a PR? I'm new to open source contributions :)

See my work here:
https://github.com/pajzo/AspNetCore.Diagnostics.HealthChecks/tree/add-hangfire

I was studying to perform the migration of the old beatpulse to this new implementation.
I missed the individual path verification.

This is in paper for implementation or it will not be possible to be done in this new version.

My API application is configured to run below the root web app. e.g.: https://MyCoolApp.com/API
When I try and view the UI the page returns 404 for all of the resources because the page is trying to load the resources using the root web application.
https://MyCoolApp.com/ui/resources/bootstrap-min.css
instead of https://MyCoolApp.com/api/ui/resources/bootstrap-min.css

I have tried to update the settings / configuration to no avail.
How do I configure the UI to reference the correct url?

Thanks

Hi,

Thanks for the great library. If using Kubernetes Service Discovery is there any way to access the host if the health app is running on a pod in the same cluster without passing a token? Also what would be the host url in that case?

Thanks

In the docs it says the default path for the health UI is: /health-ui but the default value of UIPath in Options is actually /healthchecks-ui.

I am trying to use the HealthChecks.UI.Client package to pull health checks from an API end point and there seems to be some disconnect between the client application and the end point that it is trying to communicate with. Is there something that I have missed in the configuration for this? It appears that the UI application is not finding and registering the end points but I am not sure why that it.

Result from healthchecks-api endpoint -
[{"id":1,"status":"Unhealthy","onStateFrom":"2018-12-14T18:33:10.6624959","lastExecuted":"2018-12-14T18:33:21.7626925","uri":"https://localhost:6001/health-ui","name":"Target API 1","discoveryService":null,"entries":[],"history":[]}]

Appsettings.json for UI application -
"HealthChecks-UI": { "HealthChecks": [ { "Name": "Target API 1", "Uri": "https://localhost:6001/health-ui" } ], "Webhooks": [ { "Name": "", "Uri": "", "Payload": "", "RestoredPayload": "" } ], "EvaluationTimeOnSeconds": 10, "MinimumSecondsBetweenFailureNotifications": 60 }

Result from target end point -
{"status":"Healthy","totalDuration":"00:00:00.0106605","entries":{"heartBeat":{"data":{},"duration":"00:00:00.0017263","status":"Healthy"}}}

Target API Startup.cs -
`public IServiceProvider ConfigureServices(IServiceCollection services)
{
...

        services.AddHealthChecks()
            .AddCheck<HeartBeat>("HeartBeat");

        services.AddHealthChecksUI();

}

public void Configure(IApplicationBuilder app, IHostingEnvironment env, ILoggerFactory loggerFactory)
{
...

        app.UseHealthChecks("/health", new HealthCheckOptions()
        {
            Predicate = _ => true,
        });

        app.UseHealthChecks("/health-ui", new HealthCheckOptions()
        {
            Predicate = _ => true,
            ResponseWriter = UIResponseWriter.WriteHealthCheckUIResponse
        });

        app.UseHealthChecksUI();

}`

Kafka health check is not working in the case where kafka is configured by default to create compacted topics. The reason is that compacted topics require key, but the health-check is without such. In this case we get: ErrorCode: InvalidMsg.
It is fairly easy to add key for the check.

Hi all

We are planning to create a new milestone for HealthChecks UI with some improvements, like:

• Better UI
• Enable Style customization
• Enable Company Branding
• More features in the UI
  • show some old states
  • create failure notifiers from the ui

We are delighted to receive your suggestions!

SmtpClient has been deprecated with MailKit recommended as its replacement.

The output formatter returned by UIResponseWriter.WriteHealthCheckUIResponse seems to format the json document in a way the UI does not understand. I have followed the documentation and I can't get the UI up running.

Hi guys,
I tried the health checks locally on my dev machine and everything worked great.
I'm deploying this to my staging server and the app is trying to create the Sql Lite DB on the root of c:
This is what I find on the app log:

info: Microsoft.EntityFrameworkCore.Infrastructure[10403]
Entity Framework Core 2.2.0-rtm-35687 initialized 'HealthChecksDb' using provider 'Microsoft.EntityFrameworkCore.Sqlite' with options: None
fail: Microsoft.EntityFrameworkCore.Database.Connection[20004]
An error occurred using the connection to database 'main' on server 'C:\healthchecksdb'.
Microsoft.Data.Sqlite.SqliteException (0x80004005): SQLite Error 14: 'unable to open database file'.
at Microsoft.Data.Sqlite.SqliteException.ThrowExceptionForRC(Int32 rc, sqlite3 db)
at Microsoft.Data.Sqlite.SqliteConnection.Open()

Any idea on why it is trying to create it outside the app folder?
Is there any way I can change this?
Thanks!

Is there a reason why there isn't a AddTcpHealthCheck extension method in NetworkHealthCheckBuilderExtensions? You have a TcpHealthCheck but it's not referenced.

System.ArgumentException: An item with the same key has already been added.
   at Microsoft.AspNetCore.Server.Kestrel.Core.Internal.Http.HttpHeaders.ThrowDuplicateKeyException()
   at Microsoft.AspNetCore.Server.Kestrel.Core.Internal.Http.HttpHeaders.System.Collections.Generic.IDictionary<System.String,Microsoft.Extensions.Primitives.StringValues>.Add(String key, StringValues value)
   at HealthChecks.UI.Core.UIResourcesMapper.<>c__DisplayClass2_0.<Map>b__3(HttpContext context)
   at Microsoft.AspNetCore.Builder.Extensions.MapMiddleware.<Invoke>d__3.MoveNext()

In my host the Cache-Control header is already set before processing the HealtChecks UI middleware.

In a shared hosting environment such as MLab, the MongoDB health check will fail if the user does not have permissions to ListDatabases().

mongodb: {
data: { },
description: "Command listDatabases failed: not authorized on admin to execute command { listDatabases: 1 }.",
duration: "00:00:00.1014917",
exception: "Command listDatabases failed: not authorized on admin to execute command { listDatabases: 1 }.",
status: "Unhealthy"
},

I could prepare a MongoDbSingleDatabaseHealthCheck that would call GetDatabase() and ListCollectionsAsync() to work around this issue.

I can see in the code you have a function to get the locale and format the date (livenessTable.tsx:41) but it does not seem to display the correct time. It is still showing UTC time for me here in Australia.
e.g. showing

You can see bottom right my local time being UTC+11.

Hi @unaizorrilla

Did you consider adding a healthcheck for AWS S3? Or is there a reasons why there isn't one atm ?

Cheers

Is there an interest of having a publisher for cloudwatch?

Not really an Azure customer and have my grafana plugged into my cloudwatch as backend so both AppInsights and Prometheus aren't of much use to me at the moment.

I was planning on adding one that would publish metrics and can PR that in if there is interest.

Hi could you publish this package also for 2.2.0-preview3-35497 at the moment its only available for 2.2.0-rmt-35613 ?

Cheers

Is there a way to set database location? Thanks in advance...

Minor issue, but this string should be an interpolated string.

As you may see below, serving address may change depends on environment.
Unfortunately adding multiple references and/or changing it constantly is not a pleased workaround.

It would be convenient to have an option to use relative address in Uri like we can in launchSettings.json's launchUrl property.

appsettings.json

"HealthChecks-UI": {
	"HealthChecks": [
		{
			"Name": "Self",
			"Uri": "http://localhost:47920/health/beatpulse"
		}
	]
}

launchSettings.json

{
	"iisSettings": {
		"windowsAuthentication": false,
		"anonymousAuthentication": true,
		"iisExpress": {
			"applicationUrl": "http://localhost:47920/",
			"sslPort": 0
		}
	},
	"$schema": "http://json.schemastore.org/launchsettings.json",
	"profiles": {
		"IIS Express": {
			"commandName": "IISExpress",
			"launchBrowser": true,
			"launchUrl": "health/beatpulse-ui",
			"environmentVariables": {
				"ASPNETCORE_ENVIRONMENT": "Development"
			}
		},
		"MyProject.Api": {
			"commandName": "Project",
			"launchBrowser": true,
			"launchUrl": "health/beatpulse-ui",
			"environmentVariables": {
				"ASPNETCORE_ENVIRONMENT": "Development"
			},
			"applicationUrl": "https://localhost:47923/;http://localhost:47922/"
		},
		"Docker": {
			"commandName": "Docker",
			"launchBrowser": true,
			"launchUrl": "{Scheme}://{ServiceHost}:{ServicePort}/api/values"
		  }
	}
}