Skip to Content
Author's profile photo Former Member

BI Redux

Upto couple of years back larger size of BI system gave a BI team bragging rights. And the threshold to join the club kept getting bigger (you have a size of 6 only? we are approaching 11 (TeraBytes ofcourse)).  

Well, it may still be so. But the novelty is giving way to weary realization that there are no easy tools to get the genie back in the bottle.

The DB size may be adding a few TBs every year, multiplying costs through increased hardware and resources to manage it.

Organizations have looked at solutions like BIA (and overlooked the budget increases for hardware and related services) to continue to derive incremental value out of their BI systems. Questions not asked often enough are “is all that disksize containing real meaningful data”, “is this data what users are using, or, going to use in future”, “are there holes in my dataset – placeholders that are never used but eating up diskspace”, “are there orphan fields in my schema” and so on.

It is good to explore and find the best combination of optimizing techniques such as aggregates however it may not be enough anymore. With many organizations reaching a critical DB size after which the incremental cost with increasing size of DB will rise significantly due to the complexity involved in normal maintenance tasks (eg backup, system copy, DB reorg etc) it is about time to look at what more a BI team shall be exploring.

My opinion is, now that the BI teams are confident and comfortable managing their BI system, they need to proactively identify and implement methods that reduce the growth of DB (or in best case, reduce the DB size followed by a minimal growth rate). This would involve analyzing existing data models, remodelling, archiving, discarding (yes, this is what it ultimately will need to come to, archiving offline is what it will mean in real terms) data and so on. Sooner organizations get to this mindset (of managing the DB size) in addition to applying other emerging techniques and solutions in SAP BI, better it will be long term.

It would be a welcome change to see bragging rights (and bonuses) linked to SAP BI systems growth (ie lack thereof).

Assigned tags

      You must be Logged on to comment or reply to a post.
      Author's profile photo Nathan Genez
      Nathan Genez
      Good blog Ajay.  Customers/Consultants are often times too technology focused when they're looking for solutions.  Performance is a problem?  Go buy BIA.  Need some formatting reporting?  Let's go get a license for Crystal. 

      But sometimes we all need to be more disciplined in where we look for answers.

      Author's profile photo Witalij Rudnicki
      Witalij Rudnicki
      Many people look at growing volume thru perspective of storage space getting cheaper, but forgetting other cost aspects like managment and maintainance of this volume as well as cost of retriving the meaningful data out of this (in terms of both: finding correct data and speed of access to it).

      But you sound a bit sceptical about BIA. Why? It is causing redundant storage of the data today, but no doubt long term it will eliminate duplicate storage of reporting data in both BIA and DB.


      Author's profile photo Former Member
      Former Member
      This is an interesting paradox. On one hand you have several TB of data, that probably is sub-optimal. Let us say it costs $ XX a year to maintain it (hardware, staff etc). If you commission a project to re-design it - most probably it will cost a lot of money too (probablt more than XX since you might have to hire external help, additional hard disk for migration etc, user's time to test and so on.

      After spending two years and say you succesfuly changed everything, and reduced the size of the BW system by some factor - does it provide significant business value? are the reports any faster than what you would have got with BIA? and is that speed even a big deal to the business? By the time you are ready to put your feet up - most probably you will need to start re-designing again.

      Author's profile photo Former Member
      Former Member
      Blog Post Author
      I am not proposing an either-or scenario. Keeping an eye on size and complexity and proactively reducing it where appropriate is not a one-off project rather an integral part of the process (or should be so).

      BI teams can (and shall) use tools like BIA to bring additional value to the customer. It doesn't obviate their duty to be an aggressive keeper of a BI system that is "healthy". If BI teams see their role inclusive of this goal, it can only help (and by extension if they forget it, it will hurt sooner or later).
      We probably disagree on the incremental cost/value that such an approach can bring.

      Author's profile photo Former Member
      Former Member
      I do see your point - and it is logical. However, I am still debating in my mind if it is possible to somehow measure the incremental cost/value to make a solid case. The reason is - most clients probably will not fund such an excercise unless they see a cost-benefit analysis that supports the investment.