Best Practices for Maximizing Scalability and Cost Effectiveness of Queue-Based Messaging Solutions on Windows Azure

This post has been moved to a new location. Please follow the link below. Thank you.

Comments (8)

  1. mschulz says:

    Thanks for a very informative article.  As an Azure solution architect, this is what I would describe as a fundamental piece of guidance regarding Azure solutions.  This is also very timely as I have just been asked to architect such a solution.  I look forward to the sample code that is forthcoming…and I would be happy to review the code as soon as possible.

  2. Valery M says:

    Michael, your comments are very much appreciated.

    We are currently going through each component in the referenced codebase to make sure that they carry the sufficient level of inline documentation. Once it's completed. we will update this write-up. Thanks and regards.

  3. Nikolai says:

    Hi, as Michael stated this is a great article and has some good lessons. How far off is the code sample publication? I would be very interested in reviewing this sample as I am looking at implementing a scatter gather/MapReduce pattern for parallel processing in a POC I am currently developing.

  4. Valery M says:

    Nikolai, we are almost there. It has taken a lot more time than anyone anticipated. Our last milestone is comprised of some final steps before a package can be made available, hopefully before the end of March.

  5. Mel Gerats says:

    Great read, this is what Best Practice articles should be like. I too am looking forward to the sample code.

  6. Valery M says:

    Folks, we have made the CTP (AKA community preview) version of a reference implementation available on our Code Gallery. This reference implementation makes very extensive use of the pattern discussed in this post with the exception of role instance scaling. The package provides all the required code artifacts to support elastic event-driven queue listeners in all kinds of Windows Azure solutions. Simply look for CloudQueueListenerExtension<T> and its dependencies to see how the entire best practices discussed in this post unfold themselves in the context of the reference implementation.

    Just to clarify – this is a CTP version and it’s out there for the purposes of earlier preview. We are still working on supplemental technical documentation to cover a detailed technical overview of the reference implementation is question as well its deployment scenario. What you get with the CTP is just the API documentation (Contoso.Cloud.Integration.chm).

    One more clarification – since this is a full-blown end-to-end reference implementation, there is a lot of source code in there. If you are primarily interested in the queue listener pattern, you obviously need to be prepared to extract bits & pieces from the solution purposefully and at your own pace and convenience.

  7. Jay says:

    The anticipation is over!  Great work.  Thank you so much.  I think I was first to download the CTP.

  8. Nikolai @ Adaptiv says:

    Thanks Valery (and the whole team), this is a great reference implementation and I look forward to many hours going over the details =)

Skip to main content