Scaleout in SignalR (share server connection) - connection

I found some topics on MSDN about SignalR scaleout.
But these approaches don't fit me.
I have a windows app that host a SignalR server. But I should find a way to share the server connection between many application on the same PC because this app uses on terminal server and each client should have his own app with own server, but use the same port.
Some code:
using (SignalRConnection connection = new SignalRConnection("https://localhost:9090"))
{
Application.Run();
}
public class SignalRConnection : IDisposable
{
private IDisposable SignalR { get; set; }
public SignalRConnection(string serverURI)
{
SignalR = WebApp.Start<Startup>(serverURI);
}
public void Dispose()
{
SignalR.Dispose();
}
}
class Startup
{
public void Configuration(IAppBuilder app)
{
app.UseCors(CorsOptions.AllowAll);
app.MapSignalR(new HubConfiguration()
{
EnableDetailedErrors = true,
EnableJSONP = true
});
var listener = (HttpListener)app.Properties["System.Net.HttpListener"];
listener.AuthenticationSchemes = AuthenticationSchemes.Anonymous;
}
}

Related

Know where a connection to my server (domain) comes from in ASP.NET Core

I have a service ASP.NET Core in a GNU/Linux server with the following start settings:
using System.IO;
using Microsoft.AspNetCore.Hosting;
using Proyecta.PPlus.Web.Helpers;
namespace Proyecta.PPlus.Web.Startup
{
public class Program
{
public static void Main(string[] args)
{
CurrentDirectoryHelpers.SetCurrentDirectory();
CreateWebHostBuilder(args).Build().Run();
}
public static IWebHostBuilder CreateWebHostBuilder(string[] args)
{
return new WebHostBuilder()
.UseKestrel(opt => opt.AddServerHeader = false)
.UseContentRoot(Directory.GetCurrentDirectory())
.UseIIS()
.UseIISIntegration()
.UseUrls("https://*:5001")
.UseStartup<Startup>();
}
}
}
In CreateWebHostBuilder will step in .UseUrls) to define the URLs accepted as connection ('.UseUrls (https://*:5001) accepts connections from any IP provided the port and SSL are fulfilled`'. UseUrls("https://domain1:5001") ' I would accept only connections that come connect to that domain...).
Ok, this service will have a number of domains through which customers will connect and, depending on that domain, some or other options will be made.
.UseUrls("https://domain1:5001")
.UseUrls("https://domain2:5002")
.UseUrls("https://domain3:5003")
.UseUrls("https://localhost:8080")
My doubt is how do I know through where they connect so I can configure these actions?. If connected to 'domain3`or 'localhost'...
Finally I solved this incidence with the information stored in the context request:
HttpContext.Request.Host return the domain like domain1:5001.
If I only want the domain name or the port, I can do this:
HttpContext.Request.Host.Host for the domain name, it returns domain1
HttpContext.Request.Host.Port for only the port, it returns 5001.
I added it to the login controller to perform the relevant actions:
public virtual async Task<JsonResult> Login(LoginViewModel loginModel, string returnUrl = "", string returnUrlHash = "", string ss = "")
{
[...]
if (HttpContext.Request.Host.Host.ToString().Equals("domain1"))
{
[...]
}
else if (HttpContext.Request.Host.ToString().Equals("127.0.0.1:9229"))
{
[...]
}
else
{
[...]
}
return Json(new AjaxResponse { TargetUrl = returnUrl });
}

Background Thread that uses ApplicaitonDBContext

I am trying to wire up a background thread that will update the database once an hour from Active Directory. I am not sure how to pass the current
public void ConfigureServices(IServiceCollection services)
{
services.Configure<CookiePolicyOptions>(options =>
{
// This lambda determines whether user consent for non-essential cookies is needed for a given request.
options.CheckConsentNeeded = context => false;
options.MinimumSameSitePolicy = SameSiteMode.None;
});
// Add framework services.
services.AddDbContext<ApplicationDbContext>(options =>
options.UseSqlServer(Configuration.GetConnectionString("Connection")));
services.AddMvc().SetCompatibilityVersion(CompatibilityVersion.Version_2_1).AddSessionStateTempDataProvider();
services.AddSession();
services.AddHttpContextAccessor();
services.AddSingleton(Configuration);
services.AddScoped<IAppDbRepository, AppDbRepository>();
services.AddScoped<IActiveDirectoryUtility, ActiveDirectoryUtility>();
services.AddScoped<IActiveDirectoryManager, ActiveDirectoryManager>();
services.AddHostedService<LdapManager>();
services.AddScoped<ILdapManager, LdapManager>();
}
In the LdapManager class I would like to call the UpdateUsers method every hour:
public class LdapManager : ILdapManager, IHostedService
{
private IConfiguration _configuration = null;
private Logging _logger;
private List<string> ldapConnectorForDirectoryEntries = new List<string>();
public LdapManager(IConfiguration configuration)
{
_configuration = configuration;
UpdateUsers();
SyncActiveDirectoryUsers();
}
public void SyncActiveDirectoryUsers()
{
try
{
using (var waitHandle = new AutoResetEvent(false))
{
ThreadPool.RegisterWaitForSingleObject(waitHandle, (state, timeout) => { UpdateUsers(); }, null, TimeSpan.FromHours(1), false);
}
}
catch
{
throw;
}
}
}
The UpdateUsers() method should be able to call the applicationDBContext.SaveChanges() method.
How can I ensure that the LDAP manger class can use the Application DB context?
You probably want class LdapManager : BackgroundService, ILdapManager
BackgroundService is .NET Core 2.1, there is a code sample available for core 2.0
Inject IServiceScopeFactory and override Task ExecuteAsync( ), run a while loop there.
while(!stoppingToken.IsCancellationRequested)
{
using (var scope = _serviceScopeFactory.CreateScope())
{
var context = scope.ServiceProvider.GetRequiredService<ApplicationDbContext>();
...; // do your stuff
}
await Task.Delay(myConfig.BackgroundDelay, stoppingToken);
}
And here is a good read about this on MSDN, including the code sample for 2.0
For accessing ApplicationDbContext from HostedService.
DbHostedService
public class DbHostedService : IHostedService
{
private readonly ILogger _logger;
public DbHostedService(IServiceProvider services,
ILogger<DbHostedService> logger)
{
Services = services;
_logger = logger;
}
public IServiceProvider Services { get; }
public Task StartAsync(CancellationToken cancellationToken)
{
_logger.LogInformation("Consume Scoped Service Hosted Service is starting.");
DoWork();
return Task.CompletedTask;
}
private void DoWork()
{
_logger.LogInformation("Consume Scoped Service Hosted Service is working.");
using (var scope = Services.CreateScope())
{
var context = scope.ServiceProvider.GetRequiredService<ApplicationDbContext>();
var user = context.Users.LastOrDefault();
_logger.LogInformation(user?.UserName);
}
}
public Task StopAsync(CancellationToken cancellationToken)
{
_logger.LogInformation("Consume Scoped Service Hosted Service is stopping.");
return Task.CompletedTask;
}
}
Register DbHostedService
services.AddHostedService<DbHostedService>();

Service Fabric with multiple endpoints and dependency injection

I'm currently working on a POC project and I'm trying to figure out how I can share a service dependency between different endpoints to control application state and handle all service requests (lets call it ControlService) - specifically when one of those endpoints is a KestrelCommunicationListener / HttpSysCommunicationListener and combined with a FabricTransportServiceRemotingListener (or any other type of custom listener)
Autofac looked promising but the examples don't show how to get a HTTP listener working when the container is built in startup rather than the main entry point - would I need to pass the container to MyFabricService so it can be passed into and added to by the startup registrations?
I've seen references to using container.Update() or adding registrations on the fly using container.BeginLifetimeScope() but they are all using a container built in main and then I'm not sure how I would add the APIs created by the HTTP listener to the original container.
I'm possibly not explaining it that well so in summary I'm looking to have something like the below service that can receive communications via n. different endpoints - process the message and then send messages out via n. clients (aka other service endpoints)
Happy to clarify if anything is unclear - perhaps even using another creative diagram :)
Updated:
From Program.Main()
ServiceRuntime.RegisterServiceAsync("ManagementServiceType",
context => new ManagementService(context)).GetAwaiter().GetResult();
Here is my fabric service
public ManagementService(StatefulServiceContext context)
: base(context)
{
//this does not work but is pretty much what I'm after
_managementService = ServiceProviderFactory.ServiceProvider.GetService(typeof(IManagementService)) as IManagementService;
}
protected override IEnumerable<ServiceReplicaListener> CreateServiceReplicaListeners() =>
new ServiceReplicaListener[]
{
//create external http listener
ServiceReplicaListenerFactory.CreateExternalListener(typeof(Startup), StateManager, (serviceContext, message) => ServiceEventSource.Current.ServiceMessage(serviceContext, message), "ServiceEndpoint"),
//create remoting listener with injected dependency
ServiceReplicaListenerFactory.CreateServiceReplicaListenerFor(() => new RemotingListenerService(_managementService), "ManagmentServiceRemotingEndpoint", "ManagementServiceListener")
};
ServiceReplicaListener
public static ServiceReplicaListener CreateExternalListener(Type startupType, IReliableStateManager stateManager, Action<StatefulServiceContext, string> loggingCallback, string endpointname)
{
return new ServiceReplicaListener(serviceContext =>
{
return new KestrelCommunicationListener(serviceContext, endpointname, (url, listener) =>
{
loggingCallback(serviceContext, $"Starting Kestrel on {url}");
return new WebHostBuilder().UseKestrel()
.ConfigureServices((hostingContext, services) =>
{
services.AddSingleton(serviceContext);
services.AddSingleton(stateManager);
services.AddApplicationInsightsTelemetry(hostingContext.Configuration);
services.AddSingleton<ITelemetryInitializer>((serviceProvider) => new FabricTelemetryInitializer(serviceContext));
})
.ConfigureAppConfiguration((hostingContext, config) =>
{
config.AddServiceFabricConfiguration(serviceContext);
})
.ConfigureLogging((hostingContext, logging) =>
{
logging.AddConfiguration(hostingContext.Configuration.GetSection("Logging"));
logging.AddDebug();
})
.UseContentRoot(Directory.GetCurrentDirectory())
.UseServiceFabricIntegration(listener, ServiceFabricIntegrationOptions.None)
.UseStartup(startupType)
.UseUrls(url)
.Build();
});
});
}
Startup
public class Startup
{
private const string apiTitle = "Management Service API";
private const string apiVersion = "v1";
private readonly IConfiguration configuration;
public Startup(IConfiguration configuration)
{
this.configuration = configuration;
}
public void ConfigureServices(IServiceCollection services)
{
var modules = new List<ICompositionModule>
{
new Composition.CompositionModule(),
new BusinessCompositionModule()
};
foreach (var module in modules)
{
module.AddServices(services, configuration);
}
services.AddSwashbuckle(configuration, apiTitle, apiVersion, "ManagementService.xml");
}
public void Configure(IApplicationBuilder app, IHostingEnvironment env, ILoggerFactory loggerFactory)
{
loggerFactory.AddApplicationInsights(app.ApplicationServices);
// app.UseAuthentication();
// app.UseSecurityContext();
if (env.IsDevelopment())
{
app.UseDeveloperExceptionPage();
// app.UseBrowserLink();
}
else
{
app.UseExceptionHandler("/Home/Error");
}
app.UseStaticFiles();
app.UseCors("CorsPolicy");
app.UseMvc();
app.UseSwagger(apiTitle, apiVersion);
//app.UseMvc(routes =>
//{
// routes.MapRoute(
// name: "default",
// template: "{controller=Home}/{action=Index}/{id?}");
//});
}
}
All the service dependencies are added in the CompositionModules using Microsoft.Extensions.DependencyInjection (not autofac) in startup.cs
This works great and creates my HTTP listener - I now just need a way of getting access to my services that were added to the container during startup of my http listener / webhost.
You can use Autofac.Integration.ServiceFabriŅ, an Autofac extension to support Service Fabric. You need to create a container in Program.cs
var builder = new ContainerBuilder();
builder.RegisterServiceFabricSupport();
builder.RegisterType<SomeService>().As<IManagementService>();
builder.RegisterStatelessService<ManagementService>("ManagementServiceType");
using (builder.Build())
{
// Prevents this host process from terminating so services keep running.
Thread.Sleep(Timeout.Infinite);
}
Then you can inject it to the constructor of your fabric service. You can find more information on this topic on https://alexmg.com/posts/introducing-the-autofac-integration-for-service-fabric
#Tim
Sorry for a late response. Currently I am working on library package that we use in our company for internal projects. This library simplifies configuration of Reliable Services. I think our recent enhancements can do what you need to do (hope I get the use case correctly).
All the information about the library can be found on project page on GitHub and NuGet package can be found here (please note that it is a pre-release version but we are planning to turn into complete release soon).
In case you have any questions or need more information feel free to contact me.
UPDATE
I have create a sample application. Please feel free to try it.
Here is a code example.
public interface IManagementService
{
string GetImportantValue();
}
public interface IMessageProvider
{
string GetMessage();
}
public class MessageProvider : IMessageProvider
{
public string GetMessage()
{
return "Value";
}
}
public class ManagementService : IManagementService
{
private readonly IMessageProvider provider;
public ManagementService(
IMessageProvider provider)
{
this.provider = provider;
}
public string GetImportantValue()
{
// Same instances should have the same hash
return this.provider.GetMessage() + $"Hash: {this.GetHashCode()}";
}
}
public interface IRemotingImplementation : IService
{
Task<string> RemotingGetImportantValue();
}
public class RemotingImplementation : IRemotingImplementation
{
private readonly IManagementService managementService;
public RemotingImplementation(
IManagementService managementService)
{
this.managementService = managementService;
}
public Task<string> RemotingGetImportantValue()
{
return Task.FromResult(this.managementService.GetImportantValue());
}
}
public class WebApiImplementationController : ControllerBase
{
private readonly IManagementService managementService;
public WebApiImplementationController(
IManagementService managementService)
{
this.managementService = managementService;
}
[HttpGet]
public Task<string> WebApiGetImportantValue()
{
return Task.FromResult(this.managementService.GetImportantValue());
}
}
public class WebApiStartup
{
private readonly IConfiguration configuration;
public WebApiStartup(
IConfiguration configuration)
{
this.configuration = configuration;
}
public void ConfigureServices(
IServiceCollection services)
{
services.AddMvc();
}
public void Configure(
IApplicationBuilder app,
IHostingEnvironment env,
ILoggerFactory loggerFactory)
{
app.UseMvcWithDefaultRoute();
}
}
internal static class Program
{
/// <summary>
/// This is the entry point of the service host process.
/// </summary>
private static void Main()
{
var host = new HostBuilder()
.ConfigureServices(
services =>
{
services.AddTransient<IMessageProvider, MessageProvider>();
services.AddSingleton<IManagementService, ManagementService>();
})
.ConfigureStatefulService(
serviceBuilder =>
{
serviceBuilder
.UseServiceType("StatefulServiceType")
.DefineAspNetCoreListener(
listenerBuilder =>
{
listenerBuilder
.UseEndpointName("ServiceEndpoint")
.UseKestrel()
.UseUniqueServiceUrlIntegration()
.ConfigureWebHost(
webHostBuilder =>
{
webHostBuilder.UseStartup<WebApiStartup>();
});
})
.DefineRemotingListener(
listenerBuilder =>
{
listenerBuilder
.UseEndpointName("ServiceEndpoint2")
.UseImplementation<RemotingImplementation>();
});
})
.Build()
.Run();
}
}

How to run IdentityServer3 as a Windows Service

I have some experience with Windows Services and am just getting my feet wet with IdentityServer3. My current solution works fine in IISExpress but I can't get it to work in IIS. So, I thought it might be easier to host it in a Windows Service, but I have not been able to find any samples to get up and running. Has anyone taken this approach?
My first though is I will need to instantiate my IS3 Startup class in my OnStart method and then call the Configuration method to create my authentication server, but that method takes an IAppBuilder parameter and I don't know how to create this. Any ideas would be appreciated.
I am doing something similar to this...
public partial class ServiceHost : ServiceBase
{
private IDisposable _service;
public ServiceHost()
{
InitializeComponent();
}
public void Start(string[] args) { OnStart(args); }
protected override void OnStart(string[] args)
{
var options = new StartOptions("https://localhost:44331/");
_service = WebApp.Start(options, Configuration);
}
protected override void OnStop()
{
_service?.Dispose();
}
private static void Configuration(IAppBuilder app)
{
var factory = new IdentityServerServiceFactory()
.UseInMemoryUsers(Users.Get())
.UseInMemoryClients(Clients.Get())
.UseInMemoryScopes(Scopes.Get());
var idsrvOptions = new IdentityServerOptions
{
Factory = factory,
SigningCertificate = Cert.Load(),
AuthenticationOptions = new AuthenticationOptions
{
// This is where we configure External Identity Providers
IdentityProviders = ConfigureIdentityProviders
}
};
app.UseIdentityServer(idsrvOptions);
}
private static void ConfigureIdentityProviders(IAppBuilder app, string signInAsType)
{
}
}

Calling a Client Method on a Windows Service

I have a SignalR client in a Windows Service that successfully calls a Server method in an MVC app. First the Server Code:
public class AlphaHub : Hub
{
public void Hello(string message)
{
// We got the string from the Windows Service
// using SignalR. Now need to send to the clients
Clients.All.addNewMessageToPage(message);
// Send message to Windows Service
}
and
public partial class Startup
{
public void Configuration(IAppBuilder app)
{
ConfigureAuth(app);
app.MapSignalR("/signalr", new HubConfiguration());
}
}
The Windows Service client is:
protected override async void OnStart(string[] args)
{
eventLog1.WriteEntry("In OnStart");
try
{
var hubConnection = new HubConnection("http://localhost.com/signalr", useDefaultUrl: false);
IHubProxy alphaProxy = hubConnection.CreateHubProxy("AlphaHub");
await hubConnection.Start();
await alphaProxy.Invoke("Hello", "Message from Service");
}
catch (Exception ex)
{
eventLog1.WriteEntry(ex.Message);
}
}
It sends a message to the MVC Server. Now I want to call the other way from server to client. The Client Programming Guide has the following code examples which will NOT work as this is not a desktop.
WinRT Client code for method called from server without parameters (see WPF and Silverlight examples later in this topic)
var hubConnection = new HubConnection("http://www.contoso.com/");
IHubProxy stockTickerHubProxy = hubConnection.CreateHubProxy("StockTickerHub");
stockTickerHub.On("notify", () =>
// Context is a reference to SynchronizationContext.Current
Context.Post(delegate
{
textBox.Text += "Notified!\n";
}, null)
);
await hubConnection.Start();
How can I call a method on the client?
The .NET client side code seems fine. You can simply get rid of Context.Post since your client is running inside of a Windows Service and doesn't need a SyncContext:
protected override async void OnStart(string[] args)
{
eventLog1.WriteEntry("In OnStart");
try
{
var hubConnection = new HubConnection("http://localhost.com/signalr", useDefaultUrl: false);
IHubProxy alphaProxy = hubConnection.CreateHubProxy("AlphaHub");
stockTickerHub.On("Notify", () => eventLog1.WriteEntry("Notified!"));
await hubConnection.Start();
await alphaProxy.Invoke("Hello", "Message from Service");
}
catch (Exception ex)
{
eventLog1.WriteEntry(ex.Message);
}
}
You can invoke the "Notify" callback from inside your AlphaHub on the server like so:
public class AlphaHub : Hub
{
public void Hello(string message)
{
// We got the string from the Windows Service
// using SignalR. Now need to send to the clients
Clients.All.addNewMessageToPage(message);
// Send message to the Windows Service
Clients.All.Notify();
}
Any client will be able to listen to these notifications since we are using Clients.All. If you want to avoid this, you need some way to authenticate your Windows Service and get its ConnectionId. Once you have that, you can send to the Windows Service specifically like so:
Clients.Client(serviceConnectionId).Notify();
Hope this helps.
Windows Service with self hosted SignalR
public partial class MyWindowsService : ServiceBase
{
IDisposable SignalR { get; set; }
public class SignalRStartup
{
public static IAppBuilder App = null;
public void Configuration(IAppBuilder app)
{
app.Map("/signalr", map =>
{
map.UseCors(CorsOptions.AllowAll);
var hubConfiguration = new HubConfiguration()
{
// EnableDetailedErrors = true
};
map.RunSignalR(hubConfiguration);
});
}
}
public MyWindowsService()
{
InitializeComponent();
}
protected override void OnStart(string[] args) { Start(); }
protected override void OnStop() { Stop(); }
public void Start()
{
SignalR = WebApp.Start<SignalRStartup>("http://localhost:8085/signalr");
CallToMvcJavascript();
}
public new void Stop()
{
SignalR.Dispose();
}
private void CallToMvcJavascript(){
GlobalHost.ConnectionManager.GetHubContext<MyHub>().Clients.All.addNotice(// object/data to send//);
}
}
The Hub in the Windows Service
public class MyHub : Hub
{
public void Send()
{
Clients.All.confirmSend("The service received the client message");
}
}
The Javascript
$.connection.hub.logging = true;
$.connection.hub.url = "http://localhost:8085/signalr";
var notices = $.connection.myHub;
notices.client.addNotice = function(notice) {
console.log(notice);
};
notices.client.confirmSend = function(msg) {
alert(msg);
};
$.connection.hub.start().done(function() {
$('#myTestBtn').on('click', function() {
notices.server.send();
});
});

Resources