Stephen Myers

WCF Service Message Timeouts, Size Limits, Tips and Tricks

2011-10-05T16:23:21-05:00:00

Is your WFC Service hitting the Maximum Limits set on the default service configurations? Open up the artificial constraints of the WCF Service Debugging WCF Services and message sizes. Caveat emptor: If you need to increase the size of the of the contracts, it probably indicates that you should reexamine the design of your data contract. In any case, to “Open Up” the service, you will need to set on both the client and server side configurations:

1. Timeouts

2. Message Sizes

3. Max Objects In Graph

Service side configurations

<system.serviceModel>

<serviceBehaviors>

<behaviorname="Server.WcfService.MyServiceBehavior">

<serviceMetadata httpGetEnabled="true" httpsGetEnabled="false" />

<serviceDebug includeExceptionDetailInFaults="true" />

<dataContractSerializer maxItemsInObjectGraph="2147483647" />

</behavior>

</serviceBehaviors>

</behaviors>

<endpoint address="http://MyServer:1260/MyService.svc"
binding="wsHttpBinding"

bindingConfiguration="MyServiceBinding"
contract="Server.WcfService.IMyService" />

</service>

</services>

<binding name="MyServiceBinding"

maxReceivedMessageSize="2147483647"

openTimeout="00:10:00"

receiveTimeout="00:10:00"

sendTimeout="00:10:00"

closeTimeout="00:10:00"

maxBufferPoolSize="2147483647" >

</binding>

</wsHttpBinding>

</bindings>

</system.serviceModel>

Client side configurations

<system.serviceModel>

<readerQuotas maxDepth="32"

maxStringContentLength="2147483647"
maxArrayLength="2147483647"
maxBytesPerRead="2147483647"

maxNameTableCharCount="2147483647" />

<reliableSessionordered="true"
inactivityTimeout="00:10:00"
enabled="false"/>

</binding>

</wsHttpBinding>

</bindings>

<endpoint
address=http://MyServer/MyService.svc
binding="wsHttpBinding"

bindingConfiguration="WSHttpBinding_IMyService"
contract="MyServiceReference.IMyService"
name="WSHttpBinding_IMyService"

behaviorConfiguration="MyServiceBehavior"/>

</client>

<endpointBehaviors>

<behavior name="MyServiceBehavior">

<dataContractSerializer
maxItemsInObjectGraph="2147483647" />

</behavior>

</endpointBehaviors>

</behaviors>

</system.serviceModel>

What is this 2,147,483,647business? It is equal to 2³¹ – 1, the max value of a 32 bit signed integer.

When setting these tags, you should save yourself some headaches and use the Configuration Editor (SvcConfigEditor.exe):
http://msdn.microsoft.com/en-us/library/ms732009.aspx

To test out your developed client, you can call the WCF service directly without writing your own client. Microsoft has provided one:
http://msdn.microsoft.com/en-us/library/bb552364.aspx

Security: use it. Make sure your client and service side tags match. You may have noticed that it is absent from the above example. That means that it is going to use the default which is Message for this particular binding type.
http://msdn.microsoft.com/en-us/library/ms731059.aspx

When the errors you are receiving don’t seem to make any sense, you can add robust diagnostics by adding the following tag (make sure the service has write access to the location specified in the InitializeData):

<system.diagnostics>

<source name="System.ServiceModel"

switchValue="Information, ActivityTracing"

propagateActivity="true">

<add name="sdt"

type="System.Diagnostics.XmlWriterTraceListener"

initializeData= "serviceLog.svcLog" />

</listeners>

</source>

</sources>

</system.diagnostics>

Call the WCF Service and then view the output file with the SVC Trace viewer tool:
http://msdn.microsoft.com/en-us/library/ms732023.aspx

Salient Qualities of Effective SCRUM

2011-10-05T13:59:31-05:00:00

1. The SCRUM must be short and NOT exceed 15 minutes daily. Achieve this with a Timekeeper. The Timekeeper role should rotate daily and the timekeeper should be fined $1 if the SCRUM does not start on time by opening the call in line and closing the door or if the SCRUM exceeds 15 minutes.

2. Team members must be on time for the SCRUM or be fined $1. To have productive developers, the SCRUM should be run at a reasonable time – a morning stand up should be a good way to share information and start a productive day, not a wake-up call to get Asses in seats or chastise developers who are not early risers.

3. Use a SCRUM Board. The board should consist of a list of each: milestones, issues, and risks with due dates as they apply to the team. There is also a team member component where each team member’s name is on one axis and two weeks’ worth of dates forward on the other axis. In each box, team members should put any Time off or what they plan on working on that day. This should be completely updated for the current week going forward before SCRUM on Monday morning or team members should be fined $1.

4. The format should be that each team member in the circle tells briefly what they are working on and any issues or risks that are going to cause the work to stall or miss deadlines. The timekeeper must keep the SCRUM moving by deciding when a topic is getting too deep and should be moved to a breakout session with the appropriate parties. The Team Leaders should go over the issues and risks to make sure that they have owners.

5. Fine Money can be used to purchase beer or something else for the team.

Connecting MVC to a SunOne LDAP store by implementing a custom Membership Provider

2011-06-02T16:22:37-05:00:00

LDAP is not Active Directory, though Active Directory is LDAP. As someone who drinks the Microsoft Kool-Aid, I found myself using LDAP for something other than Active Directory and I thought I should share what I have found. The problem domain was to connect a new MVC application to an existing Sun One LDAP Store.

First off, authenticating an MVC application using forms mode authentication and the Membership providers is straight forward. Start with the ASP.NET MVC 2 Web Application Template that comes out of the box with Visual Studio 2010 when you create a new project and it will have the code to use forms authentication with the membership provider model built in. In the web.config, add the tag:

<authenticationmode="Forms">

</authentication>

This points to the \Controllers\AccountController.cs class which already contains the code that you need to perform the authentication.

Now for the Membership provider… If you were using Active Directory, you could just use the Active Directory Membership provider, and be done without writing a lick of C# code:

http://msdn.microsoft.com/en-us/library/system.web.security.activedirectorymembershipprovider.aspx

<membershipdefaultProvider="AspNetActiveDirectoryMembershipProvider">

<addname="AspNetActiveDirectoryMembershipProvider"

type="System.Web.Security.ActiveDirectoryMembershipProvider,

System.Web, Version=2.0.3600, Culture=neutral,

PublicKeyToken=b03f5f7f11d50a3a" />

</providers>

</membership>

However, since we are looking at SunOne, we need to use the Generic LDAP Membership provider. Unfortunately, this provider is only included with SharePoint. If you are working in an environment where you have access to this, check out for another code-free solution:
http://msdn.microsoft.com/en-us/library/ms517148.aspx

Otherwise, all that you need to do is implement a membership provider that speaks vanilla LDAP so it can connect to the SunOne store and there you go. Please keep in mind the scope of this blog is only to authenticate to the SunOne store, not to be able to manage it, though carrying this implementation through to that level if needed should be pretty straight forward. The general strategy for implementing the authentication check is a little bit different than I would have expected. I would have expected to authenticate to the LDAP provider using some sort of administrative credential and then calling a “Check Password” function of some kind. The actual way to do it is bind to the LDAP server using the credentials passed in for authentication, and if the bind is successful, then Authentication = true, else, especially in the case of exception is false. I don’t like programming to exceptions so if someone has a better approach, please reply.

In order to do this, create a new project with an output type of class library. Create a public class which will match your configuration file and have it implement MembershipProvider. You can then right click on MemberShipProvider and select “Implement Interface” in order to generate method stubs. Then create a helper class to do the actual work. The AspNetSunOneMemebershipProvider class listing below has been truncated to exclude the methods which I did not implement. In order to compile this you will need to implement them.

File: AspNetSunOneMembershipProvider.cs

using System;

using System.Collections.Generic;

using System.Text;

using System.Web.Security;

using System.DirectoryServices;

namespace SunOne

{

public class AspNetSunOneMembershipProvider : MembershipProvider

{

// Password to SunOneMembershipProviderKey.pfx is SunOne

private static SunOneHelper _helper;

private string _applicationName;

public override void Initialize(string name, System.Collections.Specialized.NameValueCollection config)

{

// Overriding the Initialize method allows us access to the NameValueCollection that
// is read from the XML attributes

// in the web.config file.

string ldapServer = config["LDAPServer"];

string userNameConnectString = config["UserNameConnectString"];

// Scott Gu says always set the app name

// http://weblogs.asp.net/scottgu/archive/2006/04/22/Always-set-the-_2200_applicationName_2200_-property-when-configuring-ASP.NET-2.0-Membership-and-other-Providers.aspx

ApplicationName = config["ApplicationName"];

_helper = new SunOneHelper(ldapServer, userNameConnectString);

base.Initialize(name, config);

}

public override bool ValidateUser(string username, string password)

{

return _helper.ValidateUser(username, password);

}

public override string ApplicationName

{

get

{

return _applicationName;

}

set

{

_applicationName = value;

}

public override MembershipUser GetUser(object providerUserKey, bool userIsOnline)

{

return GetUser(providerUserKey.ToString(), userIsOnline);

}

File SunOneHelper.cs

using System;

using System.Collections.Generic;

using System.Configuration;

using System.DirectoryServices.ActiveDirectory;

using System.DirectoryServices.Protocols;

using System.DirectoryServices;

using System.Security.Cryptography.X509Certificates;

using System.Net;

using System.Text;

namespace SunOne

{

public class SunOneHelper : IDisposable

{

#region Member Variables

public string Server { get; set; }

public string UserNameConnectString { get; set; }

private LdapConnection _con;

#endregion

#region Constructor

///<summary>

/// construcor

///</summary>

///<remarks>

/// Standard LDAP ports:

/// LDAP = 389

/// LDAP SSL = 636

///</remarks>

public SunOneHelper(string server, string userNameConnectString)

{

Server = server;

UserNameConnectString = userNameConnectString;

}

#endregion

#region Public Methods

///<summary>

/// Cleanup connection if we have one

///</summary>

public void Dispose()

{

try

{

if (_con != null)

{

_con.Dispose();

}

catch

{

}

///<summary>

/// Validate an LDAP user and password by attempting to log on.

///</summary>

///<remarks>

/// This method catchs log on exceptions,

/// preferably in DirectoryServicesCOMException but always in the generic

/// exception below and returns false to indicate that the logon failed.

///</remarks>

///<returns>true if logon is successful; otherwise false</returns>

public bool ValidateUser(string username, string password)

{

bool authenticated = false;

try

{

if (_con != null)

{

_con.Dispose();

_con = null;

}

string connectString = UserNameConnectString.Replace("{0}", username);

_con = new LdapConnection(new LdapDirectoryIdentifier(Server));

_con.AutoBind = false;

_con.SessionOptions.AutoReconnect = false;

// no ssl, setup the certficate callbacks

_con.SessionOptions.SecureSocketLayer = false;

_con.SessionOptions.VerifyServerCertificate = null;

_con.SessionOptions.QueryClientCertificate = null;

_con.AuthType = AuthType.Basic;

// try the connection

_con.Bind(new NetworkCredential(connectString, password));

_con.Dispose();

authenticated = true;

}

catch (DirectoryServicesCOMException cex)

{

//not authenticated; reason why is in cex

authenticated = false;

}

catch (Exception ex)

{

//not authenticated due to some other exception

authenticated = false;

}

finally

{

_con.Dispose();

_con = null;

}

return authenticated;

}

#endregion

}

You can then either deploy this assembly locally to the MVC application, or preferably, install it in the GAC. If you want to install it in the GAC of the Web Server, sign the assembly:
http://msdn.microsoft.com/en-us/library/ms247123(v=vs.80).aspx

And then install it in the GAC of the web server:

http://msdn.microsoft.com/en-us/library/ex0ss12c(v=vs.80).aspx

Now all that you have to do is use this configuration (note that you will need to update the Public Key Token to match your own signed assembly):

<membershipdefaultProvider="AspNetSunOneMembershipProvider">

<addname="AspNetSunOneMembershipProvider"

type="SunOne.AspNetSunOneMembershipProvider,

SunOne, Version=1.0.0.0, Culture=neutral, PublicKeyToken=xxxxxxxxxxxxxxxx"

ApplicationName="/myApplication"

LDAPServer="SunOne.Domain.com"

UserNameConnectString="uid={0}, ou=People, ou=myApplication, dc=myContext,dc=com"/>

</providers>

</membership>

When determining your connection string, the LDP tool is invaluable when testing your LDAP Bindings. Check it out

http://technet.microsoft.com/en-us/library/cc772839(WS.10).aspx

It is contained in the Windows XP Service Pack 2 Support tools:

http://www.microsoft.com/downloads/en/details.aspx?FamilyID=49ae8576-9bb9-4126-9761-ba8011fabf38

One joy of uniqueidentifier PK's

2009-02-23T09:09:58-06:00:00

When using uniqueidentifiers as Primary keys, you get the bonus of being able to write a mutli-function Read stored procedure. Lets take a look at a table used for storing client Addresses:

-- ================================================
CREATE TABLE [Address](
[AddressId] [uniqueidentifier] NOT NULL DEFAULT NEWID(),
[ClientId] [uniqueidentifier] NOT NULL,
[AddressTypeId] [uniqueidentifier] NOT NULL,
[StreetLine1] [varchar](50),
[StreetLine2] [varchar](50),
[ZipCode] [int],
[LastRevisedDate] [datetime] NOT NULL DEFAULT GETDATE(),
[LastRevisedUser] [uniqueidentifier] NOT NULL
CONSTRAINT [PK_Address] PRIMARY KEY NONCLUSTERED ([AddressId] ASC)WITH (PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, IGNORE_DUP_KEY = OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON, FILLFACTOR = 90) ON [PRIMARY]) ON [PRIMARY]GO
-- ================================================

Assume that the ClientId is a link to the client table, the adressTypeID is a link to a type table (think Home, Business, etc) and the LastRevisedUser is a link to a users table to record who last updated this record. Instead of writting four seperate procedures to get Addresses by each of these ID's, since you are using uniqueidentifiers, you can write a single stored procedure that would retrieve addresses by any of these criteria.

-- ================================================
IF OBJECT_ID ( 'Address_GetById', 'P' ) IS NOT NULL
DROP PROCEDURE Address_GetById;
GO
CREATE PROCEDURE Address_GetById
@Id uniqueidentifier
AS
BEGIN
-- =======================================================================
-- CRUD Code Stored Procedure - [Address_GetById]
-- =======================================================================
SELECT A.[AddressId]
,A.[ClientId]
,A.[AddressTypeId]
,C.[Description] AS AddressDescription
,A.[StreetLine1]
,A.[StreetLine2]
,A.[ZipCode]
,A.[LastRevisedDate]
,A.[LastRevisedUser]
FROM [Address] A
LEFT OUTER JOIN Code C ON A.AddressTypeId = C.CodeId
WHERE
AddressId = @Id OR ClientId = @Id OR A.[AddressTypeId] = @Id OR A.[LastRevisedUser] = @Id
END-- ================================================

This will save you not only on the stored procedure side of things, but also cut out boilerplate code when you are writing the data layer methods to call stored procedures to get the address or addresses by these different criteria. In this case you turn 4 CRUD Read stored procedures into 1, and cut out the all of the associated cut & paste in the data layer. You could also let a NULL parameter mean "Give me all the Addresses" if you add another WHERE clause "OR @Id IS NULL"

FxCop Custom Dictionary

2009-01-12T07:14:46-06:00:00

FxCop is a great tool, and I would hesitate to turn off the spell checker in projects since spelling has never been my strongest suit. However, the are occasions when you may want to modify the words that FxCop accepts or rejects. Such is the case with 3 letter abbreviations. If you put a three letter abbreviation in as all caps, for example, EIO, FxCop will give you a casing error. If you correct the casing to Eio, it then gives you a spelling error. You can explicitly add or remove words from the custom dictionary file, which is in located in the installation directory for FxCop, most likely C:\Program Files\Microsoft FxCop 1.36\ or C:\Program Files\Microsoft FxCop 1.35\. The name of the file is "CustomDictionary.XML" and then add
<Word>eio</Word>
inside the
<Unrecognized>
tag to make that word cause an error, or more likely inside the
<Recognized>
tag to make FxCop allow the word.

Note: If you are using Visual Studio Team System 2008, you can add a custom dictionary to the project: http://blogs.msdn.com/fxcop/archive/2007/08/20/new-for-visual-studio-2008-custom-dictionaries.aspx

SSRS: Unable to load client print control

2008-10-30T10:13:49-05:00:00

Here is a frustrating little error. The problem occurs when trying to print reports from a report server that is not properly updated on a client machine which has been. Updating the SQL Server installation using Microsoft Update (Not just Windows update) fixed this on my local development box, however I had to dig a little deeper to get it to work on our production box which is 64 bit. The root of the problem is this: If you are using SSRS to print the reports, the way it works is that an Active X control is downloaded to the client machine and is used to print the report. However, there was a security flaw in Active X controls which Microsoft recently patched by adding "Kill Bits": http://www.microsoft.com/technet/security/advisory/956391.mspx This patch was made on October 14th (conveniently in the middle of the release cycle for this project) and has the unfortunate side effect of killing the RS Client Print Class which is the actual Active X client print control responsible for printing SSRS reports. Thankfully, Microsoft has released SQL Server updates which when run on the machine which houses the SSRS instance, also fix the Active X control which is seamlessly downloaded to client machines. On a particular server (our clients production server in this case) which was running the 64 bit version of SQL 2005, even though we ran the Microsoft updates, the correct patch was not applied. If you are having this problem, make sure that the SQL Server Reporting Services is updated to at least build 9.0.3073 and you will be good to go! If you have patched the server and the clients are still having problems, have them uninstall the control by selecting Tools-->Internet Options--> and then Manage Addons from the Programs tab. If you are unable to remove it from there, check the C:\WINDOWS\Downloaded Program Files directory of the client machine. Once it has been successfully removed, it will reinstall when a report is printed from a properly configured SSRS Server. Here is a thread that you can follow for more information about this problem, http://forums.microsoft.com/MSDN/ShowPost.aspx?PostID=4038983&SiteID=1 However, in the end in order to upgrade my production x64 SQL Server past build 9.0.3069, we needed to contact Microsoft, but I expect this will not be the case soon.

SSRS Subreports

2008-10-10T08:23:42-05:00:00

So today I ran across the dreaded "Error: Subreport could not be shown." The odd thing was that the report and subreport executed fine in business intelligence studio, yet this problem arose when the report was deployed to the SSRS server. The key to broken subreports that execute fine in stand-alone mode are the report parameters passed in from the root report to the subreport. A little background: The subreport was being called iteratively in a table where each row in the table contained the subreport based on the result set of a SQL query. The subreport took two parameters, one that resolved using the Fields! of the SQL query for the table itself, and the other was one of the parameters of the root report, just passed down to the subreport. It turns out that while the Fields! parameter resolved just fine, the report parameter did not resolve in the stage environment for some reason. Luckily for me, this parameter was useless in the case that the report was being run as a subreport, so I just removed it. This problem was further obfuscated by the caching performed by SSRS. That "Overwrite Existing" checkbox on the upload reports screen does not clear the database level caching that SSRS does, and even an IISreset and resetting the SSRS service itself didn't clear the cache. I found that deleting the report and then re-uploading it worked just fine to "clear the cache". Problem solved, good times :)

Unicode SQL Server Scripting Error

2008-10-01T08:56:10-05:00:00

Sometimes, when executing a new SQL script via osql or SQLCmd, generated by Visual Studio's "Add New", you will get an error due to a Unicode character. This character is usually at the beginning of the file. In order to get around this, you can open the file with notepad and select file -> save as… and then when saving the file, select ANSI in the encoding box. If this doesn’t solve your problem, you can change the file extension to .bin and open it with visual studio which will present you with a binary editor with which you can remove the culprit and then change the extension back to .sql.

Visual Studio Magic

2008-09-22T22:40:40-05:00:00

This is my favorite Visual Studio trick! By holding down Alt while selecting with the mouse, you are able to select a square block of text, right in the middle!

Welcome to my blog!

2008-09-12T09:02:31-05:00:00

It is my intention to discuss topics pertaining to my professional experience, specifically my current work with e.magination where I often get the opportunity to implement business solutions using the latest Microsoft technologies. It is my goal to give back to the community and help others out who may run across the same types of challenges I have run into and solved and to foster discussions pertaining to the viability of different approaches to solving technical problems.

Since this is my first post, I will start out by telling you a little about myself. I am what you might call a "Classically Trained" programmer; I received my Bachelor of Science in Computer Science from the University of Maryland, College Park in 2001. Immediately after college, I started my career as a Software Engineer with a flight test instrumentation company named ACRA Control . This was an amazing experience as I was able to travel to Dublin Ireland extensively, gained great experience interfacing between US clients and Irish developers, the opportunity to design and build a light ground station application in C++ 6.0, and make some great, lasting friendships. I also got to travel to some very interesting locations in the US to meet with clients. Next I was hungry to get my hands into the recently released .Net framework and moved on to develop a web based timesheet application with predictive scoring in VB.Net for Metier Ltd. We worked from a penthouse in Georgetown and I got to work with another team of very talented individuals. In addition to working on smaller features and bug fixes, I got to lead a small team on a very interesting project - an interface with Microsoft Exchange that polled the exchange server, digested calendar items, and injected them into the timesheet application using web services. I left DC to move to Baltimore, and changed my job again – this time to work for MICROS where I was part of the lesuire and entertainment division which serviced large installations like the Wynn casino and the Baltimore Ravens stadium with networked point of sale systems. In addition to small feature implementaion and bug fixing, I was able to apply my .Net experience there by creating a Web service for 3^rd parties to create kiosk integration with the system. Whenever you charge up a Powercard at Dave and Busters you are running my code

I was attracted to e.magination because of their reputation for staying on the cutting edge of technology and their Microsoft Gold Partner status. I currently hold the position of Senior Developer I, and my key interests lie in staying on top of the latest Microsoft technologies by applying them appropriately to solve real world business problems elegantly, efficiently and cost effectively.