|
>
 Friday, April 25, 2008
 |
|
 |
|
|
|
|
|
Here is my usual post-conference post with updated code samples related to the topics I presented on. I did 2 full day tutorials, and 4 sessions...enjoy! Many of the demos come from my book, Learning WCF. Since there is setup required for most of the samples that illustrate security or rely on a database, it is best you download the entire package of samples and follow the setup instructions provided in the appendix. Here's the link: http://www.thatindigogirl.com/LearningWCFCode.aspx TUTORIAL: Improve Your SOA: Designing a Secure, Reliable and Scalable System with WCF TUTORIAL: .NET Roadmap SESSION: ADFS and ASP.NET: Supporting Single Sign-On in your Web Applications SESSION: Building a Router for your Applications SESSION: Going Federated with WCF SESSION: Load Balancing Considerations for WCF - Samples for this instancing and throttling come from my book (see above).
- I have additional samples related to proxies here, including a proxy wrapper to address timeouts and uncaught exceptions that fault the channel: http://www.dasblonde.net/downloads/Proxies.zip
|
|
|
 |
|
 |
Thursday, April 24, 2008
|
|
|
|
|
|
|
|
|
A few weeks ago I was presenting a code sample that I created last year and discovered a mysterious problem. First, let me explain the sample. It is a proxy wrapper for WCF clients that illustrates how to swallow timeout exceptions and recreate channels automatically when there is an exception that faults the channel. The idea is this: - When a channel with a transport session times out, does the user need to see an error? Not really. But, the exception won't show up until you try to call the service, so my wrapper catches communication exceptions and if they are not faults it creates a new channel and retries the same call to the service once more. The theory is, if it fails again, we probably have a bigger problem. Otherwise, we will have successfully allowed the client to continue working without seeing an error.
- When an uncaught exception from the service faults the channel, the client channel will also be faulted if the call is not one-way. The user should see the error message, even if it is an uncaught exception, but the next time they use the proxy they should get a new channel so they can continue to work. So, the proxy wrapper creates the channel again if the channel is faulted, before making the next call.
I explained this in my ASP.NET Pro article on the subject. The latest code for this is here: http://www.dasblonde.net/downloads/Proxies.zip Well, the problem that In encountered is that all of a sudden my logic for checking if the channel was faulted after a timeout, was failing!!!! It was strange - from one stack frame to another, the channel went from Faulted to Created. But none of my code affected that change! So, of course I thought that something had changed in .NET 3.5 related to channel factory caching that might have had a side-effect of my code...and I didn't have time to investigate further until today while I was talking with my colleague Brian Noyes about the subject. Long story short, he ran the code, reproduced the problem, and remembered that there were some strange behaviors with SUO files for a solution that could cause this. He deleted the SUO and then the sample worked like it originally did!!!!! I did the same, and found the same result. This is really messed up! Apparently this is a common problem, but I have never heard about it before. The real annoyance is the time I spent troubleshooting this before I talked to Brian, and the doubts it put in my mind about new features of WCF and possible regressions....and yet I was wrong...it was the stupid SUO file. What the? Holy? This is an unacceptable bug. Who knows what kind of misleading issues this could cause developers in their day-to-day work. Technorati Tags: WCF, VS 2008 |
|
|
|
|
|
|
Wednesday, April 09, 2008
Wednesday, November 14, 2007
|
|
|
|
|
|
|
|
|
Thanks to all that attended the full-day tutorial at Dev Connections last Monday - Improve Your SOA: Designing a Secure, Reliable and Scalable System. It was certainly an avalanche of rich topics related to SOA and WCF. At last I have compiled the long list of resources from the day, including references to some getting started resources for those new to WCF. Getting Started Demos Code from the book is organized by subject matter. I specifically illustrates samples from these subdirectories: \Exceptions, \Security, \Security\ClaimsBased, \Instancing, \Concurrency, \Bindings, \QueuedMessages, \Transactions, \ReliableSessions. If there are other resources you are looking for specifically, please drop me an email and I'll add to this post. Thanks! Technorati Tags: Dev Connections, WCF, SOA |
|
|
|
|
|
|
Monday, October 22, 2007
Saturday, September 15, 2007
|
|
|
|
|
|
|
|
|
As some of you may know, several of us at IDesign (Juval, Brian and myself) are in the midst of a two-week .NET 3.5 Roadshow - six cities in two weeks where we collectively cover WCF, WF, WPF, CardSpace, federated and claims-based security concepts, and some key aspects of .NET 3.5 such as new C# 3.0 language features and ADO.NET 3.5 including LINQ and the Entity Framework. I'm personally covering WCF security, federated and claims-based security, C# 3.0 and ADO.NET 3.5. For those of you attending (or, not) here are links to the code samples I'm presenting: VS 2005 samples Download VS 2008 Samples (UPDATED 10/11/07) This download includes all samples referenced above, in addition to .NET 3.5 samples for C# 3.0 and LINQ, and IDesign's declarative security model including a recent version of our ServiceModelEx library. Other relevant resources discussed: Any questions? Email me. -Michele Technorati Tags: CardSpace, WCF, LINQ, C# 3.0 |
|
|
|
|
|
|
Wednesday, August 22, 2007
Thursday, August 16, 2007
|
|
|
|
|
|
|
|
|
Today I completed a webcast as part of a 15 part series - today's subject concurrency, throughput and throttling. I received some questions about callback and Windows client applications that I thought I would elaborate on here. In fact, I went a little overboard and created a bunch of samples that would illustrate the behavior of services and clients when you have a Windows client, a service with a callback contract (thus, two-way communication) and various WCF settings at the client and service that relate to concurrency, multithreading, synchronization with the UI thread, and so on. The following table summarizes various settings at the client, service and callback and the resulting behavior at runtime. Here's the breakdown for each column: - Callback Sync Context - refers to the UseSynchronizationContext setting for the CallbackBehaviorAttribute on the client callback object.
- Callback Concurrency Mode - refers to the ConcurrencyMode setting for the CallbackBehaviorAttribute on the client callback object.
- Service Operation - Indicates if the service contract operations are one-way or two-way.
- Callback Operation - Indicates if the callback contract operations are one-way or two-way.
- Service Concurrency Mode - refers to the ConcurrencyMode setting for the ServiceBehaviorAttribute on the service type.
- Resulting Behavior - when the corresponding sample is run, what happens?
I have also uploaded sample code for each of these scenarios, numbered in order of table row description below. Get the code here. | Callback Sync Context | Callback Concurrency Mode | Service Operation | Callback Operation | Service Concurrency Mode | Resulting Behavior | | True/ False | Single/ Reentrant/ Multiple | One-way/ Two-way | Two-way | Single | These are defaults. InvalidOperationException at the service. Operation would deadlock calling the client application since the callback is two-way and reentrancy isn’t possible. | | True | Single/ Reentrant/ Multiple | Two-way | Two-way | Reentrant/ Multiple | Service able to call client, but client is blocked because of outgoing call. | | False | Single/ Reentrant/ Multiple | Two-way | Two-way | Reentrant/ Multiple | Service able to call client, client callback object is not blocked. If client callback object tries to communicate with UI thread (i.e., setting properties of controls) it will block. | | False | Single | One-way | Two-way | Reentrant/ Multiple | Service able to call client, client callback object is not blocked. Client callback object can communicate with UI thread using Invoke() from callback thread. Client can issue multiple calls to service but only one callback can be processed at a time. | | False | Multiple | One-way | Two-way | Reentrant/ Multiple | Service able to call client, client callback object is not blocked. Client callback object can communicate with UI thread using Invoke() from callback thread. Client can issue multiple calls and multiple callbacks can be processed at a time. | | False | Multiple | One-way | One-way | Single/ Reentrant/ Multiple | Service able to call client, client callback object is not blocked. Client callback object can communicate with UI thread using Invoke() from callback thread. Client can issue multiple calls and multiple callbacks can be processed at a time. | | False | Multiple | Two-way/ Multi-threaded Client | Two-way | Reentrant/ Multiple | Service able to call client, client callback object is not blocked. Client callback object can communicate with UI thread using Invoke() from callback thread. Multiple calls from client can be processed at the service, and multiple callbacks can be processed at the client. | | False | Multiple | Two-way/ Multi-threaded Client | Two-way | Reentrant/ Multiple | Service able to call client, client callback object is not blocked. Client callback can communicate with UI thread using Invoke() from callback thread. Multiple calls from client can be processed at the service, and multiple callbacks can be processed at the client. | Here is a summary of considerations. - Contract design:
- Make callback operations one-way. If this isn’t possible, make the service reentrant. Prefer one-way callbacks since most likely the service doesn’t care about responses.
- If service operations will notify clients with a callback, make them one-way as well.
- Exception handling:
- Remember that uncaught exceptions invalidate service, client and callback channels where sessions are involved. Two-way communication can only happen over duplex, thus there is a transport session (TCP, named pipes) or HTTP session (using reliable sessions over WSDualHttpBinding).
- Throw faults for known exceptions unless you want the client to be forced to recreate the proxy and start a new session. That includes clients – who should not allow exceptions to flow back to the service during a callback unless they intend to destroy the session.
- To avoid destroying the session: if callback operations are one-way, swallow the exception after handling it at the client; if callback operations are two-way, report a fault to the service in response.
- Windows client issues:
- Always disable synchronization with the UI thread for callback objects, by setting UseSynchronizationContext to false in the CallbackBehaviorAttribute.
- Multithreaded clients should invoke services on another thread (can use asynchronous proxies) when callbacks or long running service operations are involved. This prevents UI thread from blocking during call, allows callbacks to communicate freely with UI.
- Use Invoke() to communicate with UI from the callback thread.
- Don’t forget to protect other client state if the callback object supports multiple concurrent callbacks. That means the ConcurrencyMode setting is Multiple in the CallbackBehaviorAttribute.
|
|
|
|
|
|
|
Monday, April 02, 2007
|
|
|
|
|
|
|
|
|
Once again a fantastic conference in Orlando. Dev Connections just keeps getting better and I always enjoy being part of it. Not to mention the weather in Orlando isn't bad! Here are links to my code samples for each talk I delivered. Enjoy! .NET Technology Roadmap Tutorial ASP.NET and WCF ASP.NET and CardSpace - Demonstrations in this talk can be found here:
WCF Federated Security - My claims-based samples can be found here:
- And, my STS sample here (NOTE: this sample will be updated shortly with an upcoming article, stay tuned!):
WCF Contracts and Versioning ASP.NET Performance (Updated 06/07/2007) Ok people, I had this in my Windows Live Writer to send a long time ago, and somehow it did not post...but since I haven't posted in a while I didn't notice. Many apologies for the delay. Does the "better late than never" statement apply here? I hope so... - In this talk I covered a lot of ground, and theory around performance including simpler performance tips, the progression of asynchronous handlers to component distribution, and the importance of performance counters for your SLAs.
- You can look at my ASP.NET Sandboxing articles and samples for more resources on component distribution.
- See also the following data demos for examples of data caching
- My LocalizedGallery globalization example (posted here) illustrates the use of complex output caching based on custom caching by browser, culture and profile
- Asynchronous HTTP handlers:
- Here is an example showing how to create custom performance counters:
- Here is an example illustrating some of the health monitoring configuration features of ASP.NET 2.0:
|
|
|
|
|
|
|
Friday, March 23, 2007
|
|
|
|
|
|
|
|
|
I just spent the last week at SD West in San Jose...this year the conference was bigger than ever before! Thanks to everone that attended my tutorials and sessions, I really enjoyed the questions and discussions...as I hope you did. As promised, here is a list of all my resources from the conference talks. Enjoy! Intro to Web Services (Tutorial) - Christian Gross and myself presented this one, discussing everything from POX, REST, RSS, SOAP/WSDL, WS* and SOA.
- Code I demonstrated in this tutorial is from the .NET Web Services tutorial next.
.NET Web Services TODAY (Tutorial) Microsoft Technology Avalanche (Tutorial) Top 10 Web Service Standards You Need To Know WCF Contracts and Versioning - Demonstrations in both of these talks are based on WCF code from my book here: http://www.thatindigogirl.com/LearningWCFCode.aspx
- See \Security, \ReliableSessions, \Transactions in particular for the WS* discussion
- See \DataContracts, \AdvancedSerialization, and \ServiceContracts for the contracts discussion
CardSpace - Demonstrations in this talk can be found here:
The Amazing World of Federated Security - This talk was mostly discussion of standards and architectures however the demonstrations I used to illustrate a few points are based on my claims-based samples here:
- And, my STS sample here (NOTE: this sample will be updated shortly with an upcoming article, stay tuned!):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
I created a presentation in November of last year with the goal of helping folks make sense of the vast number of Microsoft technologies from development tools, language, data, windows, web and SOA development. This blog post holds the latest links to resources and code for each section. System Requirements The links below use the following technology platforms: - Visual Studio 2005 and .NET 2.0
- .NET 3.0
- NET 3.0 Runtime (installed with Vista)
- Windows SDK for .NET 3.0
- Visual Studio 2005 Orcas Extensions for .NET 3.0:
- WCF&WPF (Nov 2006)
- WF (Nov 2006)
- ADO.NET and LINQ CTP for VS 2005 (May 2006)
- LINQ CTP
- ADO.NET vNext CTP
- ADO.NET vNext Entity Data Model Designer Prototype, CTP
- Visual Studio Orcas CTP (March 2007)
| | |