I'm trying to get IIS6 to work reliably with a WCF service I have hosted in a separate Windows Service application on the same machine. Users connect to IIS via some HTTP exposed services, which is working fine, and then IIS needs to get some information from the Windows service to put in the HTTP response. I also need a callback channel between the Windows Service and IIS.
After a lot of effort I got it working with a netTcpBinding and everything would be rosey for 5 or 10 minutes but after that IIS would report the WCF channel as being faulted and then clam up and stop processing any requests until the worker process got recycled and the whole thing repeated.
I've been trying to swap to a netNamedPipeBinding but IIS refuses or is refused access to the pipe with an "There was no endpoint listening at net.pipe://localhost/mypipename" error. I can connect to the pipe fine from a console app.
So my question is has anyone got either of those two bindings working with IIS as a client or have any other approaches?
We are using IIS 7 hosting about 20 services with the net.tcp and net.pipe bindings and it's working fine.
Your problem with the pipe looks like a misconfiguration to me. If it helps, this is how we have them configured:
Server:
<endpoint address ="" binding="fooBinding"
contract="Bla.IBlaAPI"
bindingConfiguration="BlaAPI.BindingConfig">
Binding config:
<binding name="BlaAPI.BindingConfig"
receiveTimeout = "10:50:00"
sendTimeout = "10:50:00"
maxReceivedMessageSize="2147483647"
maxBufferSize="2147483647"
maxBufferPoolSize="2147483647"
transactionFlow="false">
<readerQuotas maxDepth="32"
maxStringContentLength="2147483647"
maxArrayLength="2147483647"
maxBytesPerRead="8192"
maxNameTableCharCount="2147483647" />
<security mode="None"/>
</binding>
Note that we are using long timeouts and have really high quotas for message size and etc. because we are passing some big chunks of data through this service. You can adjust for your own needs. We have the security set to "none" because the service is only being contacted from the local machine which is secured. Again, your mileage may vary.
Client:
<endpoint name="Bla.Bindings.BlaAPI" address="net.pipe://localhost/bla/IBlaAPI.svc"
behaviorConfiguration="BlaAPI.ServiceBehavior"
binding="netNamedPipeBinding" bindingConfiguration="BlaAPI.BindingConfig"
contract="Bla.IBlaAPI" />
About the Faulted state problem, please note that if an unhandled exception occurs during execution of the service code, the service instance will remain in Faulted state until it is closed properly. To avoid this, either handle exceptions at service top-level or use, for example, Enterprise Library Excexption Handling blocks.
Re NetNamedPipeBinding and "There was no endpoint listening at net.pipe://localhost/mypipename"
Is your web application impersonating its users? The above error is what you get if you try to access a WCF service via the named pipe binding, in a security context whose logon token has membership of NETWORK USERS. The WCF client-side channel stack doesn't distinguish between access denied errors and "not found" errors, when it attempts to read the shared memory object created by the service to publish the name of the pipe in use. (See http://blogs.charteris.com/blogs/chrisdi/archive/2008/05.aspx etc)
Impersonation tokens in an IIS application will always have NETWORK USERS membership.
Can you show me the code you use to dispose of the wcf client proxy?
Never use 'using' on a wcf proxy, as it will not dispose correctly every time. This can possibly lead to the faulted state.
Related
I host a WCF Service on IIS and have the following binding in web.config:
<bindings>
<wsHttpBinding>
<binding name="transactionalBinding"
transactionFlow = "true"
sendTimeout = "00:00:01"
receiveTimeout = "00:00:01"
openTimeout = "00:00:01"
closeTimeout = "00:00:01">
<security mode="Transport">
<transport clientCredentialType="None" proxyCredentialType="None" realm=""/>
</security>
</binding>
</wsHttpBinding>
</bindings>
In my service method I sleep for 10 seconds. I do not get a timeout exception when calling my service method from a client.
Is there any meaning in defining timeouts in server side bindings?
I do not get a timeout exception when calling my service method from a client.
TL;DR: because WCF timeouts by default are one minute so naturally a server operation that only takes 10 seconds isn't going to timeout. The timeouts you have specified on the server would only affect transmission not execution of your method. (you aren't calling anything else)
You are specifying the timeouts in the server config. What you need to do is specify the timeouts in the client's config file, specifically SendTimeout. Essentially whatever end is making the call, needs to specify the operation timeout. Probably not relevant in your case but if your "server" in turn made another WCF call to another service, you would want your own timeout there too.
MSDN:
SendTimeout – used to initialize the OperationTimeout, which governs the whole process of sending a message, including receiving a reply message for a request/reply service operation. This timeout also applies when sending reply messages from a callback contract method.
Generally, WCF client and server configs should match one another and unless you are using Add Service Reference/Refresh Service Reference each time the server contracts and/or config change, the client won't know about it. By the way, avoid the latter because it duplicates your model and can lead to runtime errors if they are out of sync. Not to mention service contracts can get out of sync.
A passing thought
And this brings up one of the problems of WCF configuration via config files, they are subject to runtime errors impossible to find at compile time.
A better practice is to do away with config files completely and do programatic configuration via a common assembly that both your client and server use. Specify bindings in code along with your timeouts.
That way both server and client are always in sync with regards to WCF configuration.
With both client and server agreeing on timeouts would have addressed some issues.
Tell me more
WCF the Manual Way… the Right Way
I have a WCF service in my application.
When the amount of data that passes from server to client is small I do not have problems.
The problems come when there's a big amount of data.
The proxy method runs a stored procedure that retrives data in an acceptable time.
The method server side return correctly,
BUT it takes more that two minutes to return properly to the client and continue the execution client side.
During this period if I stop the application it seems that no actions are in execution.
I don't know if it's an IIS problem or a configuration of the net.tcp binding.
The only thing I'm able to do is to execute the proxy in an external application and it that case I no have problems.
EDIT***
This is the binding configuration:
<binding name="NetTcpBinding_INService" closeTimeout="24.00:00:00" maxReceivedMessageSize="2147483647"
openTimeout="24.00:00:00" receiveTimeout="24.00:00:00" sendTimeout="24.00:00:00">
<security mode="None" />
</binding>
We have a logging service that works by pushing into an MSMQ queue and then a persistence service that pulls out of that queue to drop into a database. Both services are WCF. Recently, though the logging service has continued to work without issue, the persistence service has stopped working, and it throws the following error when trying to open:
An error occurred while opening the queue:Generic error code. (-1072824319, 0xc00e0001). The message cannot be sent or received from the queue. Ensure that MSMQ is installed and running. Also ensure that the queue is available to open with the required access mode and authorization.
Obviously, this error is less than helpful. I've tried switching the persistence service to use the same app pool as the logging service, creating a new app for the service, restarting MSMQ, and restarting the servers - but no dice. I'm out of ideas, so any suggestions would be greatly appreciated.
Service Binding
Here's the binding on the services:
<msmqIntegrationBinding>
<binding exactlyOnce="true" durable="true" serializationFormat="Xml" maxReceivedMessageSize="2147483647" closeTimeout="00:00:30" sendTimeout="00:00:30" receiveTimeout="00:00:30" timeToLive="24:00:00" receiveRetryCount="1" maxRetryCycles="1" retryCycleDelay="00:10:00" receiveErrorHandling="Move">
<security mode="Transport"/>
</binding>
</msmqIntegrationBinding>
It hasn't changed since deployment many moons ago, but it looks like the service hasn't been working for just over a month.
EDIT: Well, reinstalling MSMQ on all servers didn't seem to help (in fact, just seemed to mess up my queue access even more). Groovy.
EDIT2: Even if I switch to netMsmqBinding in place of msmqIntegrationBinding, I get the same error.
EDIT3: If I create the queue local to the server instead of on another server, it works. So something about going to the other server is the issue. Restarting DTC doesn't resolve.
EDIT4: I can use a queue on another server, just not the MSMQ server. So it must be something with the connection to the MSMQ server (or the MSMQ server itself).
EDIT5: Ports 1801, 135, 2103, and 2105 are all open between the service server and the queue server.
EDIT6: DTCPing comes back with successes.
I want to test out the possibility of queuing message on remote clients who may or may not be connected, those clients when connected will push the messages sent to an msmq over the internet that is hosted in IIS 6.
Now, I setup MSMQ on the win server2003 hosting IIS. After I did this "MSMQ" shows up in the IIS default web site.
Ok, then I added a new transactional private queue through computer management-> message queuing.
From there all I want to do is see messages stack up, I'll deal with those after this works.
Now, I made a client app that has the following code:
using (var contract = new HttpMsmqBridgeProxy())
{
var valueToSend = 2456;
contract.TestFunction(valueToSend);
Console.WriteLine("value sent: " + valueToSend + "\r\n");
}
Here's the app.config of this client:
<configuration>
<system.serviceModel>
<client>
<endpoint
address="net.msmq://**.**.***.228/private/MarksTestHttpQueue"
binding="netMsmqBinding"
bindingConfiguration="srmpBinding"
contract="HttpMsmqBridgeLibrary.IHttpMsmqBridgeContract">
</endpoint>
</client>
<bindings>
<netMsmqBinding>
<binding name="srmpBinding"
queueTransferProtocol="Srmp">
<security mode="None"/>
</binding>
</netMsmqBinding>
</bindings>
</system.serviceModel>
</configuration>
The IP is my public facing IP that works, I can host a wcf service or webpage just fine. I followed this guide somewhat for using srmpBinding.
http://msdn.microsoft.com/en-us/library/aa395217.aspx
So, in short what happens when I run the app is it succeeds, tells me it was sent, I go into Message Queue of my client and see that a new queue has shown up in Outgoing folder called:
Direct:http://..*.228/msmq/private$/MarksTestHttpQueue
there is no outgoing messages waiting in this queue so I assume the message was sent.
When I look at my msmq now on the winserver2003 there are no arrived queued messages waiting.
ETA: I can send messages to a non-transactional queue using the classic MessageQueue implimintation:
var queue = new MessageQueue("FormatName:DIRECT=http://**.**.***.228/msmq/private$/nonTransQueue");
System.Messaging.Message Msg;
Msg = new System.Messaging.Message();
Msg.Formatter = new ActiveXMessageFormatter();
Msg.Body = "Testing";
queue.Send(Msg);
The messages show up (after altering the mapping file in the system32/msmq/mapping directory) just fine. I'm wondering if because it's IIS6 I won't be able to send using the net.msmq binding.
You are correct in that your WCF service hosted in IIS6 won't be able to process the messages. This is because IIS6 doesn't use WAS which instantiates processes for non-http requests. But I think that this comes after everything you're doing in the workflow. I would expect
you run your client, pushing the message to the remote queue
the message appears in the remote queue
your WCF service does not pickup the message because it's hosted in IIS6, so you are left with a message in the remote queue.
I don't believe that IIS is involved at all up until the point where it wouldn't be working anyway.
A simple test for this is to self host your service on the server, e.g. run it in a console app. It will be able to accept MSMQ messages just as IIS7 would, and will remove that as a potential problem from your rig.
You might also want to test whether you can push a message directly from the client to a transactional queue on the server. If you're having problems sending messages to transactional queues on other machines then you can possibly check the MSDTC log. I don't envy having to delve into there.
I am building a WCF service interface for an existing Windows service process. The purpose of the WCF interface is to provide an "Command Channel" to implement an administrative capability for the Windows Service. There are several OperationContract methods defined that are intended to extract information from and control the behaviour of the Windows service far beyond the Start/Stop/Pause capability of the Services applet.
This WCF service is intended to be part of the existing process. As such, running the WCF service in IIS or ServiceHost is not an option.
My problem is that although the ServiceHost does not throw an error on Open(), I cannot get "WCF Test Client" (or anyting else) to find the service.
This is my first WCF Service, and have had trouble finding examples or patterns that fit what I am trying to do. So I have no illusions and would not be suprised if I did many things wrong. Also, not that I have 'portSharingBinding=false'. I did have that on but it was throwing an error that pointed to another service that I do not wish to run.
Is port sharing required?
Config information:
<system.serviceModel>
<bindings>
<netTcpBinding>
<binding name="PortBinding" portSharingEnabled="false" />
</netTcpBinding>
</bindings>
<services>
<service name="NameChanged.ServiceManager.CommandService">
<endpoint address="net.tcp://localhost"
binding="netTcpBinding"
bindingConfiguration="PortBinding"
name="ServiceManagerCommandChannel"
contract="NameChanged.ServiceManager.ICommandService" />
</service>
</services>
</system.serviceModel>
I also tried the no config route using the following code:
ServiceHost host = new ServiceHost(typeof(CommandService)))
host.AddServiceEndpoint(typeof(ICommandService),
new NetTcpBinding(), "net.tcp://localhost:8000");
host.Open();
Also, no error on the Open(). But, no success connecting to the service.
Thanks for your time,
Jim
I can only speak to the WCF Test Client, but it is looking for the metadata for your service so it can generate a proxy for it. From the above configuration, it does not appear that you are exposing a metadata exchange endpoint. Take a look at this link for more info:
http://weblogs.asp.net/fabio/archive/2009/02/28/net-tcp-mex-endpoints-and-portsharing-in-wcf.aspx
You can access your service without using exposed metatdata to generate a proxy, but it will require you to manually create channels to do so:
http://msdn.microsoft.com/en-us/library/ms734681.aspx