SQLServerCentral Article

Day 2 of the OLAP Sprint


Day 2: 9:35am

Dawn of Day 2 – 6 days of dev to go.

I’m pleased with my progress thus far: in something like 36 hours I’ve used SQL Server Data Tools to build a very rudimentary Mutlidimensional OLAP Cube, with one Measure and a handful of Dimensions. I spent some time exploring different sections of my data using the Cube Browser and was excited with the results and the myriad possibilities for analysis.

Today will be my last full day of unstructured time for development, so I want to take full advantage. Tomorrow PASS Summit 2012 kicks off across the street, and I’ve got two full-day seminars to look forward to on Monday and Tuesday, followed by three days of conference sessions before heading back to Toronto.

I’d to accomplish a few things today:

Stuff the Goose:

I’ve got just under 7 million rows in my lone Fact table.

Chump change.

To really understand the scaling potential, I’m going to fill that skinny Fact table with a lot more data. I’ve always been a fan of Red Gate‘s DBA tools, and will use theirSQL Data Generator to that end today.

Speak the Language:

Cube Browsing in SSDT was fun – seeing the results grid light-up instantaneously as I dragged and dropped Measures and Dimensions – so I’ll be venturing back into familiar territory with SSMS and trying to replicate the results using MDX. I’m already feeling quite saturated with the all of new OLAP nomenclature, so I’ll have to find some room in my head for the MDX syntax. I’m comforted by the sight of the familiar SELECT, FROM, and WHERE structures, but know that the experience is going to be different from writing TSQL.

Time for some breakfast, and then it’s back to battle.

Looks like the weather will cooperate with 6 more days of programming:

Programming Weather

Day 2: 11:47pm: Hoisted by my own Petard!

“Chump change”, I said.

“Stuff the Goose”, I said.

Well, I’m eating those words now.

It would seem that trying to cram 150 million rows into the fact table in my pseudo-data warehouse did not agree with my disk drive.

I started the data generation job, using Red Gate’s SQL Data Generator (which is a great tool; this is clearly the carpenter’s fault, to use the analogy), and then went off to the gym, had a shower, grabbed a bite to eat, etc. I sat back down to work, and wasn’t paying attention to growing size of my data and log files. I was making progress in SSMS writing some basic MDX, but was eventually harangued to the point of interruption by the low disk space warnings, and subsequent fallout from having no space on my C:\ drive. And so I had to abort my MDX fun and take some action. I spent more time than I would have liked to undoing the damage, but can’t help but laugh at my predicament.

Tomorrow the PASS Summit 2012 begins, and I’ve got a full-day session with Brian and Devin Knight on Building a SQL Server 2012 BI Platform, which should be great. I’ll find some time after all that to post my thoughts on MDX – assuming my laptop is still in working order and I haven’t blown everything up.

This wouldn’t be fun if it was all a piece of cake, right?

Reprinted with permission from http://OLAPsprint.com/

The Other Days

Follow along on this journey and read the other days of Simon's pointed learning journey.


You rated this post out of 5. Change rating




You rated this post out of 5. Change rating