Monday, November 29, 2010

Working On Presentations

Happy Thanksgiving!!!!   Just wanted to take a second to write a quick update, I'm working on some presentations for an upcoming SQL Saturday in Tampa.  I've submitted a couple sessions and we'll see if any get picked up.  Regardless I'm looking forward to attending the event and catching up with some friends and fellow DBA's.

Monday, November 15, 2010

Master Data Services, Step 1 What is it?

So I’m taking a look at Master Data Services, which is a new addition to SQL 2008 R2.  SQL 2008 R2 added a lot of new features to SQL Server, a lot of them will push DBA’s outside of the normal box we sit in.  That box has become a lot smaller over the last couple years as certifications have pushed us to be Database Administrators, Database Developers, or Business Intelligence Developers.  But all three professions had area’s that overlapped one another, and more often than not the way a work place will define the title depends on what they want you to do.
Strictly speaking, and going by the area of study covered by the certifications in each area, Master Data Management would seem to be a concept covered by Data Architects and maybe Database Developers.  However, Master Data Services ships with SQL 2008 R2 and if you company bought it you may be asked at some point to install it, or even what it does, And in that case here is a bit of an overview.
1st if you are using SQL 2008 R2 or looking at it I would encourage you to go to, http://blogs.msdn.com/b/microsoft_press/archive/2010/04/14/free-ebook-introducing-microsoft-sql-server-2008-r2.aspx?wa=wsignin1.0 , and download the free e-book Introducing Microsoft SQL Server 2008 R2 by Ross Mistry and  Stacia Misner.  Ross is an expert with Microsoft and is currently working implementing SQL Server solutions out of the Microsoft Technology Center in Silicon Valley and Stacia is a BI expert that currently works for the amazing folks at SQLSkills.com.   The E-book is free, which is always a great bargain, and is chocked full of tasty knowledge nuggets……. mmmmmm nuggets.
So what is Master Data Services?  The 50,000 high view is a location where you can import multiple sources of data, organize the data, present a master view of the data, that can be exported to source databases, but will not be in and of itself the Gold Master Database.  Master Data Services uses a web API, and is not exposed via SSMS, SQL Server Management Studio.  It can integrate with SQL Server Sharepoint and have work flows and business rules written against it.  It exposes the Windows API and can have custom C# components coded for it.
You can set it up as a multi server deployment, or on one server.  However it will require x64 hardware and a Web Server.  For an overview of how to set up/install MDS take a look at the Master Data Management Team Blog, http://sqlblog.com/blogs/mds_team/archive/2009/12/10/installing-and-configuring-master-data-services-2008-r2-november-ctp.aspx . 
So reading over that description you may say, “this is a DBA/database technology?!?”.  Microsoft has decided that it will bundle it’s data related components with SQL Server.  This is good and bad in that normally you would do a point and click install for SQL Server, there are much more complex installs, but normally you have a Database Server, a set of drives, and your application lives on this server (be it virtual or physical).
You have applications that use your databases and data cubes, you may have dedicated SSRS and SSIS servers.  But at the end of the day we are a portion of an application, not typically the whole of an enterprise level application.  Master Data Services is, IMHO, an Enterprise level application.  It is a much deeper dive than just point click, optimize, and maintain.
I’ll be taking a deeper look as I go, but I just wanted to offer an up front, that this will not be a normal DBA road trip.
Thanks,

Brad

Denali

Denali CTP1 is available for download, http://www.microsoft.com/sqlserver/en/us/product-info/future-editions.aspx, This will be the latest major Version release since SQL 2008, and will build upon the minor version release from SQL 2008 R2.  There look like some really interesting features that will need to be flushed out more.  Brent Ozar has blogged about Hadron here, http://www.brentozar.com/archive/2010/11/sql-server-denali-database-mirroring-rocks/ .

Hadron will be where mirroring is going for SQL 2011, some new features will be that it is built on top of Windows Clustering, you can have up to 4 mirrors (only 1 in CTP 1), and you can execute read only queries against the mirrors, no need for snapshots. For more info see here, http://msdn.microsoft.com/en-us/library/ff877884(v=SQL.110).aspx .
Columnar level indexes, this was teased during the keynote for PASS opening night.  This would be big for Data Analytics for SQL Server, this will allow SQL to scream and compete with Oracle, DB2, TeraData, and Informix.  The TCP benchmarks on this will be unlike anything SQL has seen before.
This will be a really interesting ride.  More to come.
Thanks,

Brad