다음을 통해 공유

StreamInsight Goes Public!

  • 아티클

During the past few years we have heard from some of you on the need for incorporating “events” and “event streams” into your applications. In our conversations with all of you it emerged that some of you looked upon these applications as “Complex Event Processing (CEP) Applications” while some of you simply referred to them as low latency monitoring applications. These applications were referred to by different names in different domains ranging from manufacturing, utilities, oil & gas, financial services and IT monitoring to web analytics. Still, they all shared some common characteristics – the need to process large volumes of data in the form of event streams with minimal latency. In some applications the emphasis was on the ability to process events with very rich payloads (complex events) and in others the emphasis was on the ability to compute with minimum latency over very large event rates.

The key characteristics that emerged in all these cases was the necessity to have a rich model for expressing queries and an implementation of that query model that will facilitate incremental computation to incorporate newly arriving events into the query results with minimal latency. Both of these are key characteristics to fulfill the needs of the rich applications that address the business needs.

In conjunction with the database research team we worked on extending the relational algebra to incorporate the  extensions needed for processing event streams and developed an efficient implementation for it. Microsoft SQL Server StreamInsight is the result of these efforts. In order to ensure a very good developer experience we focused on integrating it with features offered in our development environment – a LINQ  provider for writing the queries and an event flow debugger. 

Some of the offerings that we have in this CTP are in the alpha stages while others are in internal production employment. We believe that we have reached a juncture where it would be extremely important for us to share with you what we have along with our plans for the future and solicit your feedback.  We value your feedback and look forward to working with all of you to develop the product that can address your needs.

Please download and try the CTP and give us your feedback. The best way to communicate your feedback is to report any bugs or feature requests at SQL Server Connect. Furthermore, the StreamInsight Forum is closely monitored by the development team and provides you with a basic level of support

Thanks,
The StreamInsight Team

Comments

  • Anonymous
    August 19, 2009
    Thats geat ! I'm using Microsoft Robotics Developer Studio for event processing. Its based on CCR and DSS frameworks. and based on message passing. Both StreamInsight and MS RDS has strong and weak sides. Thanks.

  • Anonymous
    August 20, 2009
    Is it going to be available in all editions of SQL Servers?

  • Anonymous
    August 20, 2009
    We are still figuring out licensing details. Availability of StreamInsight with SQL Server editions is part of that ongoing discussion. Roman

  • Anonymous
    September 14, 2009
    How does streaminsight is differrent from streambase? Is there any comparison report on this please?

  • Anonymous
    December 06, 2009
    Hi, When the output is a combination of low latency real-time streaming events combined with dimensional queires (Analysis Services) and historic parked data (SQL Server) What would be the ideal output format for any StreamInsight CE process. I thought MS Reporting Services at first but, that would render the whole object of having near real-time data streams irrelevant. What I am looking for is a dynamic spreadsheet!? that can handle the near real- time streaming data aspect of the Analysis or even better a browser based solution that can handle the equivalant. Is there anything I can use from Microsoft readily available today or in the pipeline we can use in bench-testing? Many thanks in advance Alp Hudson