I want to send messages from the server (from a class, not a controller) via the SignalR Hub.
The hub works for messages originating from the client but not for messages from the server.
I've tried multiple methods of which non seem to work. For example, I tried retrieving the hub context using:
GlobalHost.ConnectionManager.GetHubContext<MyHub>()
with no success.
What is the best and up-to-date method of doing this in .NET Core?
Temporary Solution:
Having a websocket client inside the host api. Then making it connect to itself. This is not an ideal solution but works as a temporary fix.
You can inject the context in your class as service. Your class must be initialized via DI and added as a service. There is no difference between class or controller.
public class SomeClass
{
public IHubContext<ChatHub, IChatClient> _strongChatHubContext { get; }
public SomeClass(IHubContext<ChatHub, IChatClient> chatHubContext)
{
_strongChatHubContext = chatHubContext;
}
public async Task SendMessage(string message)
{
await _strongChatHubContext.Clients.All.ReceiveMessage(message);
}
}
You can also get service like following by injecting IHttpContextAccessor
var _strongChatHubContext = httpContextAccessor.HttpContext.RequestServices.GetRequiredService<IHubContext<ChatHub, IChatClient>>()
reference:
https://learn.microsoft.com/en-us/aspnet/core/signalr/hubcontext?view=aspnetcore-2.1
First you need to somewhere instantiate your hub (normally when your app is bootstrapping).
MyHub myHub = new MyHub();
Then on your class inject the context:
private readonly IHubContext<NotifyHub, ITypedHubClient> hubContext;
And in your class method just call the hub:
hubContext.Clients.All.yourHubMethod(yourPayload);
Related
Someone out there must have run into this already...
I created a WebApi solution with swagger implemented, full documentation, the whole 9 yards!
When I run my web api solution, see the swagger output (and I've tested the endpoints, all working fine)
I can see the swagger definition: https://localhost:5001/swagger/v1/swagger.json
Now, I want to consume this Api as a connected service on my web app.
So following every single tutorial online:
I go to my webapp
right click on Connected Services
Add Connected Service
Add Service Reference > OpenApi > add Url, namespace & class name
That generates a partial class in my solution (MyTestApiClient)
public parial class MyTestApiClient
{
// auto generated code
}
Next step, inject the service in Startup.cs
services.AddTransient(x =>
{
var client = new MyTestApiClient("https://localhost:5001", new HttpClient());
return client;
});
Then, inject the class into some class where it's consumed and this all works
public class TestService
{
private readonly MyTestApiClient _client; // this is class, not an interface -> my problem
public TestService(MyTestApiClient client)
{
_client = client;
}
public async Task<int> GetCountAsync()
{
return _client.GetCountAsync();
}
}
So everything up to here works. BUT, this generated OpenApi client doesn't have an interface which sucks for the purposes of DI and Unit Testing.
I got around this by creating a local interface IMyTestApiClient, added to the generated class (MyTestApiClient). I only have 1 endpoint in my WebApi so have to declare that on my interface.
public parial class MyTestApiClient : IMyTestApiClient
{
// auto generated code
}
public interface IMyTestApiClient
{
// implemented in generated MyTestApiClient class
Task<int> GetCountAsync();
}
services.AddTransient<IMyTestApiClient, MyTestApiClient>(x =>
{
IMyTestApiClient client = new MyTestApiClient("https://localhost:5001", new HttpClient());
return client;
});
public class TestService
{
private readonly IMyTestApiClient _client; // now injecting local interface instead of the generated class - great success
public TestService(IMyTestApiClient client)
{
_client = client;
}
public async Task<int> GetCountAsync()
{
return _client.GetCountAsync();
}
}
But this is a bad approach because it makes me manually create an interface and explicitly declare the methods I want to consume. Furthermore, every time my Api gets updated, I will have to tweak my local interface.
So question time:
How can I add an OpenApi Service Reference that automagically also generates an interface as well?
Thanks in advance for any help getting to a viable solution.
You may have already found the answer but I had the same issue and managed to resolve it by adding /GenerateClientInterfaces:true in the Options section for the OpenAPI reference in my .csproj:
<OpenApiReference Include="api.json" CodeGenerator="NSwagCSharp" Namespace="MyNamespace" ClassName="MyClassName">
<SourceUri>https://localhost:7040/swagger/v1/swagger.json</SourceUri>
<OutputPath>MyClient.cs</OutputPath>
<Options>/GenerateClientInterfaces:true</Options>
</OpenApiReference>
My web api looks like:
[Route("api/[controller]")]
public class ValuesController : Controller
{
private readonly ApplicationDbContext _context;
public ValuesController(ApplicationDbContext context)
{
_context = context;
}
[HttpGet]
public string Get()
{
// get random user
var user = _context.Users.SingleOrDefault();
return user?.Email ?? "";
}
}
Trying to call it via jquery:
$.get('/api/values', function (email) {
console.log(email)
})
I've 2 questions:
1/. Why didn't console.log(email) work although I was getting the response successful?
There was nothing in the Console tab.
2/. When I made a request to server (/api/values), the breakpoint had been caught:
My question: where had ValuesController been called within context? I'd sent a request from client, then the constructor was hit (I'm sure that I didn't send something like ApplicationDbContext from client to server :v)
UPDATE: By changing $.get('/api/values', {}, function (email) {} to $.get('/api/values', function (email) {}. I've fixed the first problem. It's my bad. Sorry about that.
Shorter Answer
My question: where had ValuesController been called within context? I'd sent a request from client, then the constructor was hit.
The HTTP Request arrived. ASP.NET MVC Routing told the application to use the ValuesController to handle the request. The application constructed the ValuesController, supplying an instance of ApplicationDbContext via dependency injection.
Longer Answer
The Startup.Configure method is used to specify how the ASP.NET application will respond to individual HTTP requests. Since you are using Web API, you have configured app.UseMvc(). Result: when an HTTP Request arrives, MVC Routing tells the application to use the appropriate controller.
The Startup.ConfigureServices method is used to specify services that are available via dependency injection. Since you are injecting an ApplicationDbContext into your constructor, you have configured services.AddDbContext<ApplicationDbContext>(). Result: when the application constructs a ValuesController, ASP.NET Dependency Injection will provide an instance of the ApplicationDbContext.
I am building a Spring Cloud project (Brixton.M4 with Spring Boot 1.3.1) with Eureka, Zuul and FeignClient where I am trying to add multi tenancy support (Tenants are identified by subdomain : tenant1.myservice.com). To do so, I would like to somehow pass the original subdomain along requests that are forwarded from a service to the other via Feign but I can't seem to be able to find the right way to do it.
What I have is a client that exposes a #RestController which calls a #FeignClient to communicate with my backend which exposes server operations to the client through its own #RestController.
The #FeignClient using same interface as my #RestController on the server :
#FeignClient(name = "product")
public interface ProductService extends IProductService {
}
What I am currently trying to do is set a header in a RequestInterceptor :
#Component
public class MultiTenancyRequestInterceptor implements RequestInterceptor {
private CurrentTenantProvider currentTenantProvider;
#Autowired
public MultiTenancyRequestInterceptor(CurrentTenantProvider currentTenantProvider) {
this.currentTenantProvider = currentTenantProvider;
}
#Override
public void apply(RequestTemplate template) {
try {
template.header("TENANT", currentTenantProvider.getTenant());
} catch (Exception e) {
// "oops"
}
}
}
My provider class is a simple component where I'm trying to inject a request / session scope bean :
#Component
public class CurrentTenantProvider {
#Autowired
private CurrentTenant currentTenant;
//...
}
The bean (I tried both session and request scope) :
#Bean
#Scope(value = WebApplicationContext.SCOPE_SESSION, proxyMode = ScopedProxyMode.TARGET_CLASS)
public CurrentTenant currentTenant() {
return new CurrentTenant();
}
On the server, I use Hibernate multitenant provider that is supposed to catch the header value and use it to define which DB to connect to :
#Autowired
private HttpServletRequest httpRequest;
#Override
public String resolveCurrentTenantIdentifier() {
return httpRequest.getHeader("TENANT");
}
It seems the Feign call to the server is done in another thread and out of the incoming request scope, so i'm not sure how to pass that value along.
It all works fine when I hardcode the tenant value in the RequestInterceptor so I know the rest is working properly.
I have also looked at many other posts about Zuul "X-Forwaded-For" header and cannot find it in the request received on the server. I have also tried adding a ZuulFilter to pass host name to next request but what I see is that original request to the Client is picked up by the ZuulFilter and I can add but not when the Feign request is sent to the backend service even if I map it in zuul (i guess that is intended ?).
I am not really sure what's the next step and would appreciate some suggestions.
Hope that it's of any use for you but we're doing sth similar in Spring-Cloud-Sleuth but we're using a ThreadLocal to pass span between different libraries and approaches (including Feign + Hystrix).
Here is an example with the highlighted line where we retrieve the Span from the thread local: https://github.com/spring-cloud/spring-cloud-sleuth/blob/master/spring-cloud-sleuth-core/src/main/java/org/springframework/cloud/sleuth/instrument/web/client/TraceFeignClientAutoConfiguration.java#L123
I'm trying to get the context for a hub using the following:
var hubContext = GlobalHost.ConnectionManager.GetHubContext<SomeHub>();
The problem is that GlobalHost is not defined. I see it is part of the SignalR.Core dll. At the moment, I have the following in my project .json file, under dependencies:
"Microsoft.AspNet.SignalR.Server": "3.0.0-*"
If I add the latest available version of Core:
"Microsoft.AspNet.SignalR.Server": "3.0.0-*",
"Microsoft.AspNet.SignalR.Core" : "2.1.2"
I get a whole bunch of errors because server and core are conflicting. If I change them to both use version "3.0.0-*", all the conflicts go away, but GlobalHost cannot be found. If I remove Server, and just user Core version 2.1.2 then GlobalHost works, but all the other things needing Server, obviously do not.
Any ideas?
IConnectionManager does not exist any more in SignalR for ASP.Net Core.
I've been using HubContext for getting access to a hub.
public class HomeController : Controller
{
private readonly IHubContext<LiveHub> _hubContext;
public HomeController(IHubContext<LiveHub> hubContext)
{
_hubContext = hubContext;
}
public void SendToAll(string message)
{
_hubContext.Clients.All.InvokeAsync("Send", message);
}
}
I'm using .net core 2.0.0 and SignalR 1.0.0-alpha1-final
Microsoft.AspNet.SignalR.Infrastructure.IConnectionManager is a DI injected service through which you can get the hub context...For example:
using Microsoft.AspNet.SignalR;
using Microsoft.AspNet.SignalR.Infrastructure;
using Microsoft.AspNet.Mvc;
public class TestController : Controller
{
private IHubContext testHub;
public TestController(IConnectionManager connectionManager)
{
testHub = connectionManager.GetHubContext<TestHub>();
}
.....
To use the hub in a backgroud service, in addition to controllers, you must use the IHostedService interface and get the hub by DI.
public class MyBackgroundService : IHostedService, IDisposable
{
public static IHubContext<NotifierHub> HubContext;
public MyBackgroundService(IHubContext<NotifierHub> hubContext)
{
HubContext = hubContext;
}
public Task StartAsync(CancellationToken cancellationToken)
{
//TODO: your start logic, some timers, singletons, etc
return Task.CompletedTask;
}
public Task StopAsync(CancellationToken cancellationToken)
{
//TODO: your stop logic
return Task.CompletedTask;
}
public void Dispose()
{
}
}
Then you can call your hub from anywhere in your code from HubContext static field:
MyBackgroundService.HubContext.Clients.All.SendAsync("UpdateData", myData).Wait();
Learn more about IHostedService:
https://learn.microsoft.com/en-us/aspnet/core/fundamentals/host/hosted-services?view=aspnetcore-2.1
You can create and start a timer in MyBackgroundService and call the hub in ElapsedEvent.
I needed to be able to access the Hub Context from outside the app request thread - because I was subscribing to NServicebus messages, and needed to be able to trigger a client function when I received a message.
Here's how I got it sorted:
public static IServiceProvider __serviceProvider;
then during startup configuration
app.UseServices(services =>
{
__serviceProvider = new ServiceCollection()
.BuildServiceProvider(CallContextServiceLocator.Locator.ServiceProvider);
});
Then anywhere else in the vNext asp.net application (any other thread)
var manager = Startup.__serviceProvider.GetRequiredService<IConnectionManager>();
var hub = manager.GetHubContext<ChatHub>();
Hope this helps!
I added some code to my Startup.cs to grab reference to the ConnectionManager which you can then use to do a GetHubContext at anytime from anywhere in your code. Similar to Nimo's answer but a little different, maybe simpler.
services.AddSignalR(options =>
{
options.Hubs.EnableDetailedErrors = true;
});
var provider = services.BuildServiceProvider();
//Hold on to the reference to the connectionManager
var connManager = provider.GetService(typeof(IConnectionManager)) as IConnectionManager;
//Use it somewhere else
var hub = connManager.GetHubContext<SignalHub>();
I'm looking at SignalR source code and it seems that IHubContext is registered as a singleton.
Which means you get the same instance whenever you access it.
Which means you can simply save it in a static var and use it from whatever.
public void Configure(IApplicationBuilder app, IWebHostEnvironment env, IHubContext<MyHub> hubContext)
{
_staticVar = hubContext;
}
But be warned - it's an anti-pattern.
I am using SignalR to broadcast messages to all my clients. I need to trigger the broadcasting outside of my hub class i.e. something like below:
var broadcast = new chatHub();
broadcast.Send("Admin","stop the chat");
I am getting error message as:
Using a Hub instance not created by the HubPipeline is unsupported.
You need to use GetHubContext:
var context = GlobalHost.ConnectionManager.GetHubContext<chatHub>();
context.Clients.All.Send("Admin", "stop the chat");
This is described in more detail at http://www.asp.net/signalr/overview/signalr-20/hubs-api/hubs-api-guide-server#callfromoutsidehub.
A small update for those who might be wondering where the GlobalHost has gone. SignalR has been completely rewritten for .net core. So if you are using the SignalR.Core package (Difference between SignalR versions), you get an instance of SignalR hub context by injecting it into your service:
public class MyNeedyService
{
private readonly IHubContext<MyHub> ctx;
public MyNeedyService(IHubContext<MyHub> ctx)
{
this.ctx = ctx;
}
public async Task MyMethod()
{
await this.ctx.All.SendAsync("clientCall");
}
}
And in Startup.cs:
services.AddSignalR()/*.AddAzureSignalR("...")*/;
Microsoft docu is here: Send SignalR messages from outside the hub.