Good Data Saves Lives

This guy is no doubt related to me.

This guy is no doubt related to me.

I was posed with the question of how we use SQL Server in our workplace by that irrepressible scamp Grant Fritchey, who was previously tapped gently on the noggin by David Taylor; the instigator in this latest SQL-centric meme.  It seem appropriate for me to first tell you a bit about the environment I support before I answer that question.  I am part of a medical care system (some may call it a collection of hospitals, and that would be fitting as well) that encompasses the Western side of Michigan’s lower peninsula from Grand Rapids all the way North to the Mackinac bridge.  This is the part where I hold my hand up like a mitten and point to the area of the state I’m speaking of.  Being from The Great Lakes State that is how we do things here when not <stereotypes> fishing, tying dead deer to things, dining on fast food, watching hockey, and cooking meth.  </stereotypes> 

We do all those things a typical hospital does: treat patients, bill insurance companies, submit for Medicare reimbursements, take part in clinical trials, budget for new equipment, research cures for the most vile of blights imaginable and so forth.  Each of those actions and the other I’ve not named has data associated with it.  That data is stored in a variety of relational database management systems such as SQL Server, and Oracle.  To a lesser extent we also have a smattering of Sybase, MySQL, Access, Excel, Word, flat files, and even Caché.

Without divulging too much I can tell you that we, like Grant in his response, use Oracle for our main data warehouse.  We also use it for many of the systems supporting our electronic medical record system.  It’s quite funny here at the brink of the new decade we are just now hotly debating the digitization of our medical record systems when my company have been doing it for almost 10 years.  I digress though…  back to what I was saying, which was that we use Oracle to support those initiatives and others which I will not dwell on.  Interestingly though, we also use SQL Server for components of those same systems.  Some data even does that awkward dance of Oracle-to-SQL-to-Oracle just to get into the data warehouse.  Reminds me of  a bit of Senior Prom back in 1986.  White tux, date 4″ taller than I; surrounded by equally gawky white kids trying to Cabbage Patch to Journey; not a pretty sight

Many smaller components of the electronic medical record are hosted in SQL and feed the same application(s) that tap into Oracle as well.  Fitting that a hospital performs data surgeries on par with those Dr. Frankenstéin did in his little biological data center in Transylvania.  We even use SQL to support our Oracle systems in case of widespread power or network failure.  One of these solutions was presented in the January 2008 issue of SQL Server Magazine(with your friendly neighborhood SQLAgentMan on the cover.)  In that issue I detailed how I used SQL Server and (then) DTS to create a downtime standalone instance of specific tables used to stage surgeries at our various surgical centers. 

Enough about how we use SQL Server as a supporting system for Oracle.  SQL Server has a rightful place in the lead role of many of our most important systems.  We use SQL Server as our database for MOSS and Sharepoint Portal Services, which is used for document management and as the content repository for our inter/intranet sites.  We use it to support SCOM, MOM, ILM, OCS, and all those other Microsoft-Acronym-happy products that allow us to function by monitoring our environment, allowing us to communicate, and facilitating security in the domain.  We use SQL Server for clinical systems that support the full spectrum of treatments for patients both young (we’re home to one of the best children’s hospitals in the nation) and old alike.  SQL Server is pervasive in our environment and has had phenomenal growth since I started working as their first SQL Server DBA.  (In less than a week from now I celebrate my 10th anniversary.)  We went from a single SQL Server instance with 13 databases when I took the reigns as DBA to a high of 80+ instances and almost 2,000 databases at our high-water mark before I was able to consolidate to our current 40+ instances and 1,000 databases.  These databases were originally catering to departmental applications that services a small pool of users.  Now we see that the RDBMS of choice for most of our enterprise applications requiring databases is, in fact, Microsoft SQL Server.  We have shared clusters, standalone instances, and dedicated clusters to support all sizes of databases from those in the low megabyte range to those pushing a terabyte. 

 I wish I could provide more detail into what we do use SQL Server for, but I fear I may have already said too much.  So the question posed was what do we use SQL Server for in our organization?  Well, just about anything we choose to.  We do this because, to quote our Oracle DBA: “Good Data Saves Lives.”


I’m not sure who has and has not been tagged in this meme, so I’ll take a stab and tag Andy Leonard because I am curious about how SSIS is employed in his company and because he is probably the most gentlemanly geek I’ve ever met , and I’ll also tag Jason Massie, because I’ve not heard much from his blog since he ported over to WordPress a while back.