It all started about a month ago, I got my first big assignment as a junior. The assignment was quite simple, design and implement a datawarehouse from scratch from a production system of an enterprise. Although the description of my assignment was quite simple, putting this assignment into practice proved to be a lot of work.
When I was at one of our customers a few weeks ago, they had some questions about the rebuild index process. On some servers it caused big TLog growths, and on other servers, it didn’t. So I did a quick investigation on the differences in versions and options.
A data warehouse is a relational database just like any other relational database. However, the database model is designed in a different way. It should facilitate business users querying large amounts of data, but also let them explore the data from different points-of-view by means of BI applications. Six steps will guide you through the design process and address some key fundamentals in data warehouse modeling. Finally, five additional tips will help you save time and money with future enhancements of your data warehouse solution.
What is better? Using GUIDs or Integer values? This is been an age long (religious) debate and there are advocates in both camps stressing on the disadvantages of the other. Both implementations have their advantages and disadvantages. In this blogpost, we'll see that the ultimate SQL Server answer is also valid for this debate: it all depends!
This post might sound a bit like a game of Cluedo, but it happened on a production SQL Server… The issue showed itself on a hosted production VM with 64GB of Ram, where SQL could allocate memory from 32GB to 55GB. The agreements with the hosting partner clearly mentioned that while the CPU could be overcommitted, memory couldn’t. All my alerts suddenly started to show that the machine was using all the available and became completely unresponsive (I couldn’t even open the Event Viewer to unload the logs). Remote connection with management studio gave me this worrisome error:
SQL Server 2014 has been released since the 1st of April and there are some really cool features that might help you and your organization and can definitely justify an upgrade. In my job as a SQL consultant, I have a lot of different customers and none of them is currently working on SQL Server 2014 or is planning to upgrade to this new version. I find it really hard to convince my clients to upgrade or even install it for a brand new installation. Why is it so hard to convince people to start using SQL Server 2014…? After questioning my customers I figured out these 5 reasons?