Jeffrey Chilberto

Business Rules Engine: Samples

2009-11-21T07:18:55Z

I recently presented at the Auckland Connected Systems User Group. In order to prepare, I created several examples of methods of using the rules engine.

The sample solution is here.

Included in the sample solution are:

Various Vocabulary definitions including using a function in XPath Field (TotalCostOfOrder)
Various Rules definitions including retrieving from a database, using classes, adding XML Nodes to a document, forward chaining and priority
Examples implementations of IFactRetriever
BizTalk Application showing calling rules with both XML Document and objects
BizTalk Application showing an example of using the BRE for routing
A WCF Service and Client including an example tracking interceptor

The WCF Service was my focus of the talk and what I wanted to discuss most of all. I have found using the BRE very beneficial and now that I have spent time understanding it, I believe it helps to make more maintainable solutions by separating the business rules from the application framework. This allows for the business logic to more accessible and readable than embedded in BizTalk expressions, maps or custom classes.

Because of this I wanted to illustrate how a WCF service could be strongly typed but not require costly rebuilds when the facts or rules change. Though I believe strongly in SOA, I am not a fan of the un-typed XML in/XML out design. To achieve this, I took advantage of the KnownTypes attribute to allow WCF to serialize the payloads and to present metadata. In the example service, the specific classes are retrieved from the web.config, but there is not any reason why this could not have been retrieved from a more dynamic resource.

Cheers and feedback appreciated.

ACSUG Meeting Thursday November 19th - A Practical exploration of the Business Rules Engine in BizTalk Server

2009-11-16T07:42:00Z

I decided it was about time I presented something at the Auckland Connected User Group…

Recently I have been involved with two projects where the Business Rules Engine was exposed as a WCF web service. Though this is the basis for the presentation, most of the examples will be applicable to using the BRE within BizTalk or externally.

Slides and examples to follow when ready… I better get to work on them…

Movember

2009-11-16T07:32:17Z

I am growing a moustache for Movember to join the fight against prostate cancer and men's depression. Please give so that my Mo may grow: http://nz.movember.com/mospace/187663/

Force NTLM protocol instead of Kerberos in IIS

2009-07-03T04:06:00Z

Recently I deployed a WCF service using NTLM transport authentication.

An issue arose when calling the service from a remote server where Kerberos authentication was not working correctly.

After contacting the infrastructure team, I was informed to use NTLM and not Kerberos as the authentication provider.

By default, IIS uses Kerberos and NTLM for network authentication. It will choose to use Kerberos if the client is thought to support Kerberos as in this case. To force IIS to use NTLM only is relatively simple.

First, open IIS to determine the website identify. This can be done by navigating to the Websites folder:

The adsutil.vbs admin script can be used to set the authentication provider. The screenshot below illustrates:

Also, different virtual folders can have different NTAuthenticationProviders settings. For example, if the virtual folder “HelloWorld” was to be modified only, the command would be:

C:\>cscript c:\inetpub\adminscripts\adsutil.vbs //nologo SET "W3SVC/1/Root/HelloWorld/NTAuthenticationProviders" "NTLM"

Notes:

Related Knowledge Base: http://support.microsoft.com/kb/215383

WCF Test Harness: Release 1.2.2.0

2009-04-01T07:55:51Z

A new version of the test harness has been added to codeplex.

This release corrects some bugs and adds some new usability features (e.g., the ability to edit the client configuration via a menu item).

Feedback appreciated.

WCF Test Harness: Release 1.1.1.1

2009-02-24T03:11:21Z

The WCF Test Harness has been updated.

The previous version would fail to build the dynamic proxy for some services. In particular BizTalk isolated host service endpoints. The new version will still attempt to build the proxy even if warnings are encountered when downloading the metadata.
This version uses an untyped channel to send the requests and responses (see below). This allows for the messages to be more easily constructed and for more flexibility (or leniency) on the receive. As a test tool this is useful for debugging and determining when an application does not conform to the metadata. For example, the subtle differences between Java and .Net base soap solutions.
DataContractSerialization is now used for building the input message.

Untyped Channel

After a TestCase has been created that indicates the contract to call and the input, the InvokeProxyMethod method is used to call the service using an untyped channel. By untyped, I mean conforming to the following signature:

[ServiceContract]public interface IUntypedChannel{    [OperationContract(Action = "*", ReplyAction = "*")]    Message Post(Message content);}

The following code is used to create a ContractChannelFactory<T> using reflection. Reflection is necessary as the contract type (T) is not known until runtime. The ContractChannelFactory is based on the ClientChannelFactory<T> from the Microsoft WCF Samples. The main difference is it can be used to create a IUntypedChannel channel. This can be used to post a Message based on the contract type's message properties.

public string InvokeProxyMethod(TestCase test){    string ret = "Failed to call test case...";     // determine the contract type    ServiceClientContract contract = _contracts[test.ContractName];    ClientEndpoint endpoint = contract.Endpoints[test.EndpointName];    Type contractType = Assemblies[test.Assembly].GetType(endpoint.Contract);        // create the factory to generate the channel    Type factoryType = typeof(ContractChannelFactory<>).MakeGenericType(contractType);    Object factory = Activator.CreateInstance(factoryType, new object[] {test.EndpointName, contract.FileName});       // create the channel    IUntypedChannel channel = (IUntypedChannel)factoryType.InvokeMember("CreateUntypedChannel", BindingFlags.Public | BindingFlags.InvokeMethod | BindingFlags.Instance, null, factory, new object[] { test.URL });        // determine the action    MemberInfo s = contractType.GetMethod(test.Method);    object[] attributes = s.GetCustomAttributes(typeof(OperationContractAttribute), true);    if (attributes.Length == 1)    {        using (StringReader sr = new StringReader(test.Input))        {            // build an untyped message with the same message version, input and action            XmlReader xr = XmlTextReader.Create(sr);            OperationContractAttribute attribute = (OperationContractAttribute)attributes[0];            MessageVersion version = (MessageVersion)factoryType.InvokeMember("GetMessageVersion", BindingFlags.Public | BindingFlags.InvokeMethod | BindingFlags.Instance, null, factory, new object[] { });            Message request = Message.CreateMessage(version, attribute.Action, xr);             // call the service            Message response = channel.Post(request);             ret = response.ToString();             xr.Close();            sr.Close();        }    }     return ret;}

ContractChannelFactory<T>

The ContractChannelFactory<T> is used to build the ChannelFactory<T> using a supplied configuration file. The new methods, CreateUntypedChanee() and GetMessageVersion() are used to create a channel that will send a untyped message.

public class ContractChannelFactory<T> : ChannelFactory<T>{    // contructor and methods to build channelfactory based on supplied configuration file    ...    public IUntypedChannel CreateUntypedChannel(string overrideUrl)    {        EndpointAddress address = this.Endpoint.Address;                  // if an override then take the new address        if(overrideUrl!=string.Empty)            address = new EndpointAddress(new Uri(overrideUrl), address.Identity, address.Headers, address.GetReaderAtMetadata(), address.GetReaderAtExtensions());         ChannelFactory<IUntypedChannel> factory = new ChannelFactory<IUntypedChannel>(this.Endpoint.Binding, address);                    return factory.CreateChannel();    }     public MessageVersion GetMessageVersion()    {        return this.Endpoint.Binding.MessageVersion;    }  }

Enterprise Library: Filter on FormattedEventLogTraceListener does not work

2009-02-23T04:13:00Z

As the project I am working nears completion, attention is drawn to reporting. In my scenario, I want all logging to be sent to the database and only warning+ to be logged to the event log. No problem, as the reporting is based on Enterprise Library 4.0, this is just configuration.

Unfortunately there is a defect in 3.0 and 4.0 where the Filter on the FormattedEventLogTraceListener does not influence the logging. Fortunately, using the System.Diagnostics.EventLogTraceListener works a treat.

http://www.codeplex.com/entlib/Thread/View.aspx?ThreadId=30645

This appears to be fixed in Enterprise Library 4.1, but I have not confirmed this yet.

BizTalk: CreateBTSMessage revisited

2009-02-04T21:41:00Z

This post is a follow-up to the previous BizTalk: Create Message based on deployed schema post.

In the current solution I am working on, we are implementing the ability to run versions of the BizTalk application side-by-side. This will allow us to deploy early and verify the solution before any real messages are sent through. This will also allow us to finish the processing of dehydrated orchestrations. Aside: if anyone has some good best practices/techniques around this, please let me know!

One bit of functionality that has cause a little grief is using the helper utility described in the previous post. Fortunately, this was overcome with a minor code change to filter by the assembly full name:

public static XmlDocument CreateMessage(string schemaName, string rootNode, string assembly)
{
    XmlDocument doc = new XmlDocument();
    DocumentSpec docSpec;
 
    BtsCatalogExplorer catalog = new BtsCatalogExplorer();
    catalog.ConnectionString = System.Configuration.ConfigurationManager.ConnectionStrings[BTS_CONNECTION_STRING].ConnectionString;
            
    Schema schema = null;
 
    if (assembly == string.Empty)
        schema = catalog.Schemas[schemaName];
    else
        schema = catalog.Schemas[schemaName, assembly];
 
    if (schema == null)
        throw new BTSUtilitiesException(string.Format("Failed to retrieve the {0} schema from the management database.", schemaName));
    else
    {             
        docSpec = new DocumentSpec(rootNode, schema.BtsAssembly.DisplayName);
                
        if (docSpec == null)
            throw new BTSUtilitiesException(string.Format("Failed to build the document {0}.", rootNode));
        else
        {
            StringBuilder sb = new StringBuilder();
            StringWriter sw = new StringWriter(sb);
 
            try
            {
                doc.Load(docSpec.CreateXmlInstance(sw));
            }
            finally
            {
                sw.Dispose();
            }                    
        }
    }
    return doc;
 }

WCF Test Harness: DynamicProxyFactory

2009-01-19T22:27:49Z

This is in response to a post on CodePlex in regards to how to retrieve the metadata of a service. I based this on a WCF Sample create by Vipul Modi.

DynamicProxyFactory

In the WCF Test Harness solution, the DynamicProxyFactory class is used to generate a proxy class based on a metadata enabled service. The class performs the following:

Downloads Metadata - uses the System.Web.Services.Discovery.DiscoveryClientProtocol to retrieve the metadata
Imports Metadata - uses the System.ServiceModel.Description.WsdlImporter to extract the bindings, contracts and endpoints
Creates Proxy - generates the proxy code
Writes Proxy - writes the code into memory
Compiles - compiles the proxy code into a temporary assembly

If you are interested in exploring this, I suggest retrieving the source code and investigating the classes located in the Dynamic folder of the Spike.ContractBuilder library.

Cheers, Jeff

WCF Test Harness: Creating a Stub Service

2008-12-09T00:03:00Z

The following steps can be performed to create a stub service based on the service contract downloaded and compiled in Add a New Proxy via URL:

Select the menu option Show Hosting Window on the Host menu. This will bring up a non-modal dialog that can be used to manage stubbed services.
Select the Add Service Contract menu option on the Host menu of the Hosting Window:
In the dialog, enter in a unique name for the service, select a service contract, and provide an address for the service:
Next, click Create Stub to generate an implementation of the service contract:
This will generate a metadata enabled service endpoint at the address specified. Click the Start button to create and open the host:
If the TestHost.exe is currently running, you will get a failure as the Uri has already been allocated:
Stop the TestHost by pressing return in the consol application window:
Click the Start button again. The host is now running...

Overview

WCF Test Harness: New Release

2008-12-08T23:52:00Z

I've created a new version of the WCF Test Harness including some bug fixes and the ability to create a mock host. This is a stubbed service based on a loaded service contract. This is very limited and its main purpose is to illustrate what is possible. If this proves to be useful, I'll flush it out and work on its limitations and spend some time on the GUI.

Note: This replaces the previous version and is not compatible with the previous version (including the SampleHost.exe).

With this new release it is possible to perform the following:

First create a new client configuration and proxy by following: Add a New Proxy via Url.
Add a new test case and verify the service is valid by following: Creating a New Test Case.
Create a stubbed service based on the service contract downloaded in Step 1: Creating a Stub Service.
After stopping the TestHost service running, run the test created in Step 2 again.

This does have potential for being really useful. The next step is to apply the test harness against some real-world services to see where the holes are...

As always, any feedback or suggestions appreciated.

For Standard and Branch editions, joining a BizTalk group is not allowed.

2008-11-25T04:12:45Z

The "For Standard and Branch editions, joining a BizTalk group is not allowed." failure was recently encountered by a colleague while configuring a new implementation of BizTalk. After trudging through countless posts, he came across the solution to this issue.

When encountering this error, just double check the name of the server is 15 characters or less (both NetBios and windows server name).

WCF Test Harness: Sample Contract

2008-11-21T01:04:00Z

To illustrate the WCF Test Harness, I create a service contract with a variety of operations defined.

Also a service host was created to host the contract at http://localhost:9991/CustomerService:

Overview

WCF Test Harness Guide updated

2008-11-16T21:06:17Z

I've added a new page to the Guide: Adding a New Proxy via Assembly.

First CodePlex Project: WCF Test Harness

2008-11-14T02:26:48Z

I've just published my first CodePlex project: WCF Test Harness. This is a testing utility I have been working on for a while now and I thought it could be helpful for others.

I wrote a short Overview and I plan on adding additional information depending on interest.