In-Depth

Q & A—Microsoft Details Reporting Services Features

Group product manager for SQL Server explains the added value of Microsoft's new RS

• By Stephen Swoyer
• 10/01/2003

Way back when, before a summer of turmoil rocked the BI industry—and the enterprise reporting landscape, in particular—Microsoft Corp. announced a new add-on for its SQL Server 2000 database, Reporting Services (RS). The software giant positions RS as a comprehensive platform for the creation, management and publishing of reports. There’s just one problem: Actuate Corp., Cognos Inc., Crystal Decisions, and other established vendors already market comprehensive reporting suites.

What distinctive value-add will Microsoft bring to the table with RS? To find out, we spoke last week with Tom Rizzo, group product manager for SQL Server with Microsoft.

Microsoft has quietly built an integrated BI tool stack into SQL Server 2000. Could you give us an overview of some of the BI features that you’ve built into the database?

We jump-started a trend when we integrated OLAP in SQL Server [7.0], and we’ve since added other business intelligence features. So we have things like data mining [Analysis Services] built in, which IBM and others have started to copy. We also have ETL [Data Transformation Services], now, too. We also have a BI Accelerator [which helps developers write applications to exploit these services], and we just released a number of Office Accelerators for corporate performance management … and other business scenarios that customers are interested in when it comes to BI.

With SQL Server, we can include a number of core features in the box that you have to pay for from other vendors. The other thing that we just recently talked about with regard to BI is [Enterprise] Reporting Services.

Is this something that you’re shipping now?

We have built in reporting services for SQL Server coming online soon. We’ll have a public beta soon. We believe we’re going to have this [available] by the end of the year.

Reporting is a big topic. What’s the scope of the Enterprise Reporting Services that you’re developing?

I would break out reporting into three key things, all of which [Enterprise Reporting Services] addresses. There’s authoring, where you actually go and author your reports. There’s management—if you’re the author of 60,000 reports, you’ve got to be able to manage them. Then there’s deploying them, getting them out to your users.

So if you take authoring, the key thing we did there was we built integrated development tools right into Visual Studio for report authoring. We added in drag-and-drop capabilities for creating tables, fitting charts, creating sub-reports, doing the standard sort of graphical development that you want to do, but you get the power of Visual Studio, so all you have to do is fire up Visual Basic or C# and you’ve got the ability to create these reports using tools and in a language that you’re comfortable with. So it doesn’t matter if you’re a report developer or not. You can embed this [reporting capability] in the applications that you build.

What about report management and deployment? Does [Enterprise] Reporting Services offer anything that makes it easier for an organization to manage and publish reports?

Well, once you build your report, you deploy it out to a server for reporting services. As an IT administrator, you can go in and manage your reports. Or you can do things like take any execution of my report and see a history. This is very important for something like Sarbanes-Oxley, where you need to know who ran a report, what they saw, how they saw it, and when then saw it. You may want to manage data sources and secure access to data sources. One of the neat things we did was separate reports out from data sources.

Or say for some reason, you have to change the name of the server, or the secure access method, or the IP address—or whatever—in 10,000 reports. Instead of going to all 10,000 of those reports and changing the text in those reports to reflect that new data source, you just designate one data source that these reports connect to and change it once.

As for deployment, that’s really getting it out to your users. Out of the box, we support publishing it to a Web site, so you can put your report up on a Web site or portal. We support putting it out via e-mail, so you can subscribe and be notified when a report changes automatically through a Web site that we have. That is a pull model for the end user. We also have the push model, which is when the report writer says, ‘Here are the people I’m going to e-mail it to every day,’ and it’s automatically pushed out to them.

What kinds of reporting does Enterprise Reporting Services support?

The reports can be dynamic, so you can have things like drill-down, and charting into reports, and they can be personalized, so you can generate a customized report for Tom, for Sarah, for Bob, and e-mail it to them so that they only get their information. It supports drill-down, so you may get sales data for how you’re doing for this quarter, which product is doing well, which isn’t doing well, you can drill down into the data as well.

Will it be a SQL Server-only play?

Reporting services works against Oracle and DB2, just as our data mining works against SQL Server and these other things. We made sure in real world customer environments that we can work against what’s out there.

Also, all of the management technologies are XML Web services-enabled, so if you wanted to generate a report in reporting services, say you wanted a new report generated and sent out from a Web-services enabled mainframe based on some new data, you could do it. In your COBOL code on your mainframe, when it sees the new data, it could call the Web services and generate a new report. [XML Web services support is] a big thing for interoperability with other platforms.

You’ve made the other BI add-ons for SQL Server available for free. Do you plan to do the same thing with the reporting services component?

Yes, we’re going to license it like we licensed our Analysis Services product, which is if you run it on the same server as SQL Server, you don’t pay anything extra, but if you run it on a different server, then you have to pay.