Hi, I’m Derek
I started this site mainly out of my passion for SQL Server. I began my career in 2000 working for activePDF as one of the early employees. Moving up, I defined most of the policies and procedures for the support department, along with creating custom reporting and applications. This is where my love of SQL Server first occurred. I remember looking at stored procedure syntax and being frustrated that it was so different from anything I have seen. Shortly thereafter, I created my first stored procedure. Working at activePDF gave me a foundation that no other job has. Working with my boss Tim (the CEO) was one of the most rewarding experiences of my career. I am very grateful to have worked with Tim. He helped me lay a strong foundation for understanding Windows, programming, PDFs, and also for life.
Moving along, I found my home at a subprime mortgage company named ResMae. I was luckily thrown into dataflow despite my extenisive knowledge in Visual Basic. Soon after, I created my first incremental (delta) refresh using a multi-tiered DTS package structure and fell in love with dataflow. The challenge of being able to move large amounts of changing data quickly across an organization sparked my intrests. At that point, I also doubled as a DBA, troubleshooting SQL Server slowdowns, investigating correlations between performance counters and other statistics. It was here that I gained confidence that I could diagnose and optimize any query to perform at it’s maximum speed. The biggest contribution I gave to this company is likely relieving their processor bottleneck by simply adding 6 indexes and rewriting 2 critical procedures. This relieved the company of long standing slowdowns that had been occuring for over 2 years. I had also started early on SQL Server 2005 at this point, learning the differences between the query optimizer of 2000 and 2005.
My next job consisted of working at a Collateral Debt Obligation Trustee. Some may say I jumped from the frying pan into the fire. First subprime, now CDO’s. That may have been the case, however it was worth it. Not only was this an excellent company, I was also placed in a position where I could make a difference. When I started I helped convert SQL 2000 to SQL 2005. After rewriting all the troubled procedures, we upgraded only to hit big walls of slowness. This was mainly due to a few things. First, we did not have enough RAM and were paging out. Second, the parallelism setting caused us to strain the disk subsystem. And last, we experienced process trimming which was common at the time of 64-bit OS’s.
My biggest contribution at this company was the rewriting of their most critical snapshot system. This system was the core of the business and boy was it a pig. There was approximately 15,000 lines of code written using udf’s; I completely rewrote using SET based methods. The results were impressive (approximately a 10X increase of throughput). This change along with rewriting a scalar udf that was defined as a computed column (??!!) I would say paid for my salary. The change of the computed column allowed the system to go from a maximum batch requests/ sec of 180 to hit peaks of 3000+. Shortly hereafter I got the title of Data Architect. We also experienced a lot of web slowdowns which occured because of classic ASP threading model being a single threaded apartment which I helped troubleshoot with some senior Microsoft Support personel.
The culmination of all my experience (and the falling in love with my wife Meghan) led me to my next and current job in Los Angeles. Here my biggest contribution has been the creation of a Rules Engine that utilizes CDC (change data capture). It operates near real-time and only runs the rules required to run against a subset of records every two minutes. This allows the statuses of the rules to be updated constantly. This is done by monitoring columns that change and registering them with the rules that depend on them. After implementing this I got promoted to Architect, which was an honor because I had only been there barely one year and it is a large company (40,000+ employees) and only has a handful of architects. Currently I continue to work on another rules engine which also utilizes CDC. This time, the engine will accommodate quick rule creation turn around times, and will also contain centralized logic for rules and a plug-in architecture to support minimal regression testing and code modification.
I hope you enjoy the site. I love to hear from all my visitors. I want to make the site as informative as possible and would like to accommodate all levels of developers.
To advertise with us, please contact advertising <at> SQLServerPlanet dot com. For all other inquiries please email derek <at> SQLServerPlanet dot com.
Thanks for Visiting!
