Day 327: From Prototype to Production

  • Simon Doubt

    Mr or Mrs. 500

    Points: 565

    Comments posted to this topic are about the item Day 327: From Prototype to Production

  • Greg Edwards-268690

    SSC-Insane

    Points: 20528

    Great article! Coming from the BI side, I can see now you understand just how different it can be from the old SQL world. Some of it requires more discipline, and cleaner data, as you found out.

    Yes, the speed is amazing. I kind of compare to an Excel pivot table on steroids.

    Think back about the timeframe - first part in a week, then the journey.

    Then consider what it must mean when a person comes in an tells you they can give you a data warehouse in 2 or 3 weeks.

    Overpromising? Or likely just roughed in?

    Once again - great job!

  • Simon E Doubt

    SSCarpal Tunnel

    Points: 4191

    Greg Edwards-268690 (10/10/2013)


    Great article! Coming from the BI side, I can see now you understand just how different it can be from the old SQL world. Some of it requires more discipline, and cleaner data, as you found out.

    Yes, the speed is amazing. I kind of compare to an Excel pivot table on steroids.

    Think back about the timeframe - first part in a week, then the journey.

    Then consider what it must mean when a person comes in an tells you they can give you a data warehouse in 2 or 3 weeks.

    Overpromising? Or likely just roughed in?

    Once again - great job!

    Thanks for the feedback, Greg!

    When I look back on the project, it is definitely a tale of two timelines.

    The prototype development (my OLAP Sprint) was an intense, non-structured, just-get-it-working experience, with no real regard for sustainability. Just a prototype and learning experience. The year that followed involved so much more planning - much of it on the business side to make sure we managed the risk and opportunity that the project entailed, and then of course all the layers of prep and discussion and development. For each layer of the project, we tried to research and make informed decisions - server hardware, software edition, etl, data warehouse design, cube design, reporting goals, etc. I can't imagine doing the work that we did in 2 or 3 weeks!

  • Greg Edwards-268690

    SSC-Insane

    Points: 20528

    I went from reporting off an ERP system, to working on a DW team which built a base warehouse and cube off this system. Most people don't truly understand how different a pure transactional db differs from a good base warehouse. Very enlightening when someone sees 'clean data' generating errors.

    Planning and mapping out are critical, and also need to take into account more of users and their expectations.

    We did see slowing down some, and letting some of the thoughts and questions settle for just a bit, then a fresh review before finalizing what we were going to do, worked very well for us.

    I remember 10 years ago, being in SQL events, and the question would come up 'who uses a cube?', and I was usually one of the only ones raising my hand. Seems like once someone is exposed to the power of SSAS, they get pretty excited. Whole new set of possibilities.

    Some of the tools have come a long ways in the last few years for the users to access / view data stored in a cube. As you could see, MDX natively, isn't exactly real user friendly and intuitive.

    Your relief might be short lived, although in a good way. Once users get a handle on things, usually it will spark a bunch of can we add this requests. If you haven't done much with drill to detail yet, you might want to get familiar with it.

  • Jeff Atherton

    SSCrazy

    Points: 2663

    Great Article. Thanks for sharing your experience. I've been thinking about getting into BI but it sounds like I'll really need to prepare myself if I decide to make the transition.

  • Simon E Doubt

    SSCarpal Tunnel

    Points: 4191

    Jeff Atherton (10/10/2013)


    Great Article. Thanks for sharing your experience. I've been thinking about getting into BI but it sounds like I'll really need to prepare myself if I decide to make the transition.

    Thanks very much for the feedback, Jeff. I'm glad you enjoyed the article.

    I had similar reservations as you are describing - wanting to get into BI, but not knowing how to make the transition from relational/TSQL.

    In the end, I closed my eyes and jumped. I used last year's PASS Summit as my own 1-man, 7-day hackathon, with great results.

    Have a look at my original blog series - http://OLAPsprint.com - all I had was a laptop, a book, 7-days, and the PASS Summit.

    It was a great experience, and a relatively low-cost way to enter the BI space.

    - Simon

  • thewoodymax

    SSC Enthusiast

    Points: 148

    Simon,

    Thank you for sharing this insightful and inspirational article.

    Woody Torres

  • Jeff Atherton

    SSCrazy

    Points: 2663

    Simon E Doubt (10/10/2013)


    Have a look at my original blog series - http://OLAPsprint.com - all I had was a laptop, a book, 7-days, and the PASS Summit.

    It was a great experience, and a relatively low-cost way to enter the BI space.

    Awesome, Simon. Thanks. I'll check out that website for more of the story.

  • Simon E Doubt

    SSCarpal Tunnel

    Points: 4191

    Jeff Torres (10/10/2013)


    Simon,

    Thank you for sharing this insightful and inspirational article.

    Woody Torres

    My pleasure, Woody. I have learned and gained much from the SQL community - hopefully someone finds this interesting and begins their own journey into BI.

    - Simon

Viewing 9 posts - 1 through 9 (of 9 total)

You must be logged in to reply to this topic. Login to reply