Tuesday, 24 December 2013

Business Intelligence and Santa

As its Christmas Eve,  I thought I would post a link to a Microsoft Power Bi dashboard which shows NORAD's Santa tracking data.

Tracking Santa

Merry Christmas !

Tuesday, 17 December 2013

Understanding SAP tables

I haven't been working with SAP for that long and I often find it a challenge to understand if I have the right table.  The names are criptic and the same data is often repreated in many different tables, so what seems right, sometimes isn't.

I came across a good website yesterday for understanding this.  I thought i'd share it here for others:


I hope this helps you !

Wednesday, 4 December 2013

Agile BI road blocks

We have been trying to run out BI team in a more agile fashion over the last year with limited success.  There are a few main areas we have hit problems in:

Source System knowledge:  Due to the fact that we use big ERP/SAAS solutions like SAP and SFDC, out business don't have the technical knowledge of where their data comes from in the underlying systems.  This means we have needed to build this knowledge up in our team, which takes a lot of time (We don't have the budget to bring in experts).

Getting to the data - 2 main areas this hits us - Cloud solutions that don't provide adequate interfaces for getting to their data for BI & SAP ERP, which prohibits getting data more than once a day for some key areas (unless you buy Hana for $$$$).  Its almost as if these vendors want you to be locked-in to their solution !

Bad data - a classic, but still an issue. 

Design - I'm starting to think Star schema's and particularly slowly changing dimensions are agile handbrakes.. The amount of work that goes into adding fields to SCD2 tables and not having it generate a new generation requires a lot of extra coding.

These are only a few, but hopefully someone has a solution - and I don't mean a new tool - I am really trying to find a better process, if there is one !

Look forward to hearing your thoughts

Tuesday, 3 December 2013

Prioritising the Data warehouse

One area which is always contentious is determining which problems to solve first.  I found some great insides on this in David Stein's post titles "If We Build It, Will They Come?".  The full link is below, but I particularly like the simple Gartner-like Matrix below:

Thanks David !

If We Build It, Will They Come?

Tuesday, 8 October 2013

Work life balance

As I write this it is Tuesday.  Yesterday was a public holiday here so we got a log weekend.  I have a great time with my family and feel refreshed for the first time in months.  So why can't I do that every week if I want to ?

To me, I don't do 40+ hours a week 48 weeks a year very well.  It burns me out.  For some people that may be perfect, but not me.  One day I hope to find an employer that lets me work 4 days a week or 46 weeks a year (or both).  I think that would be about perfect. 

Has anyone found the perfect balance ?  Are there any employers out their who let their employees pick the conditions that suit them ?

Let me know if you've found the perfect balance !

Thursday, 15 August 2013

Moving data services repositories

We recently moved out repositories from one SQL server to another.  We have no end of problems and are still finding little issues.  So I thought it would help others if I posted a few tips:

  1. Before starting, make sure everything is checked into the central repository.
  2. For local repositories, you are better off to create new one's on the destination server, and remove the old ones.
  3. Ensure that you have everything correct in IPS/BO before worrying about the job server designer.
  4. If you get errors trying to add repositories into the job server, you will need to change the job server name.  One the repos are added, just change it straight back.  This seems to be a bug in the job server manager.

I hope these help !   If I knew this before we stared, I would have had a much easier few weeks !

Tuesday, 12 February 2013

Another Data Services link..


A few notes to supplement the wiki..

Friday, 11 January 2013

SQL Server 2012: Tabular & Multidimensional models on the same machine

Wow, 3 posts in a day - must be a record for me !

I have been trying to compare the pros and cons of moving to tabular for my company.  I wanted to be able to look at them side my side, but it wasn't obvious how to.  Luckly someone else worked it out for me:


Big thanks to James Serra for working this out !

Space issues when installing Data Services patches and upgrades

I thought I'd pass on an issue I had installing the latest Data Services patch.

The patches are compressed with WinRAR.  When you double click on the patch, it decompresses it to your temporary directory.  In our case that was on the C: drive, but the server in question did not have enough room on the C: drive.  This resulted in a corrupted install of data services - not good !

To fix this, I did the following:

Open a command prompt as administrator.

Manually set the TMP and TEMP variables to point to another location (in out case f::\temp).  I then ran the patch executable from this command prompt session.  This successfully uncompressed.  I then did a 'repair' of my data services installation which fixed the problem.

I hope this saves you from having the same issue !

Happy New Year !

I'm back at work and getting into a few new things.  SQL Server 2012, specifically Tabular models.  Also upgrading SAP Data Services to the latest patch set.

What will 2013 bring ?  Will the big data hype settle down.  Will the NO-SQL crowd be subsumed by the NEW-SQL crowd ?  Will the industry come up with some more buz-words (That one's a definite !)

Anyway, happy new year to my reader(s) !