Skip to Content
Author's profile photo Bala Prabahar

01001000 01000001 01001110 01000001

01001000 01000001 01001110 01000001 – Binary representation for HANA.

HANA is not simple technology. HANA is not technology that would just automagically convert data into business-consumable format and help the customers improve their bottom line. One would require skills plus invest time and efforts  to decipher the binary representation, 01001000 01000001 01001110 01000001. Similarly HANA appliance or technology would require skills, time and efforts to make it useful for the businesses.

In the last two months or so, I’ve read several blogs describing how HANA is going to be useful to the businesses. A few of them implied the death of SDLC. Is SDLC or good design philosophies just going to die with the advent of HANA? I don’t think so.

IMHO, one of the reasons why we need HANA to solve our today’s problems is because we didn’t use what was available yesterday. We didn’t have discipline, commitment, energy, interest, education, training, dedication, devotion, steadfastness to do the right thing in the past.

SAP played a role in developing that culture  by providing wizards to develop data models. Those data models are strongly coupled to ABAP dictionary but loosely coupled to the principles. As a result, anyone with ABAP background could develop BW data models which may or may not follow DW principles.

Recently I heard one of SAP customers wanted to analyze real time data. And the number of records was 465+ billion records. Were all of them real time? I would be surprised if it was. Even if everyone in the globe(including just born, infants etc) was a customer of SAP’s customer, they should be on an average generating 60-70 transactions(minus master data records?) to generate 465+ billion records. This in my opinion is not likely. So let us assume all 465+ billion records was not real time.
 
Did SAP’s customer implement good partitioning scheme to segregate real time data from non-real time data? Was BW used for non-real time data and remote cube(or some other technology) for real time data before HANA was implemented? 
 
And who was interested in real time data? I don’t believe CEO or anyone in top management would be interested in real time data. I don’t know the organization structure of that customer so let us assume that customer has 10 hierarchy levels. Let us presume  anyone beyond 5th or 6th level of that organization structure would not be interested in real time data. My point here is that data can be partitioned in such a way that person X accesses only partitions(amongst real time data partitions) that are relevant to him/her. Even if CEO is interested in real time data, this scheme would still support that requirement. This feature could’ve made the database logically smaller.
 
SAP couldn’t keep up with the advances made in partitioning due to the complexity in implementing it. In SAP world, the partitioning-if one wants to implement-should be implemented at the DB level except in a few cases(in case of Oracle). This is not easy task. This would require skills, data knowledge, time, efforts, outage and coordination between several groups. Recently SAP delivered a program to support partitioning. This program is too little, too late. Partitioning has been available for 10+ years.
             
BW wizards(RSA1) and lack of progress in implementing the latest DB technologies played a role in not realizing the benefits of BW systems for a good percentage of the customers. Would HANA add value to them? I doubt it. Would HANA be useful to any customer at all? Absolutely yes. HANA would be useful and beneficial to very large customers. 
                    
What I see happening with HANA is an attempt to move challenges and chaos  from point A to point B.  As a consultant, I’m really excited. What I’m not excited about is why the culture of doing the right thing is disappearing(if not completely disappeared already). SAP appears to be operating on the premise that the  amount of available resources(memory, CPU and other resources) are infinite. IMHO the availability of resources is not infinite. 
     

Assigned Tags

      7 Comments
      You must be Logged on to comment or reply to a post.
      Author's profile photo Bikas Tarway
      Bikas Tarway
      Hi Bala,

      I appreciate your way of thinking.

      Regards
      Bikas

      Author's profile photo Bala Prabahar
      Bala Prabahar
      Blog Post Author
      Thank you, Bikas.

      Regards,
      Bala

      Author's profile photo Former Member
      Former Member
      Hey Bala!

      It was nice meeting you at TechEd. 

      I should hope that HANA doesn't mean the death of SDLC.  That's just normal quality development.  MMmmmm...  I missed reading that somewhere.  But I am just catching up on reading blogs.

      I agree customers do not have infinte resources.   However, hopefully the cost continues to go down for memory, CPU...  HANA would allow us to do MORE with our programs.  While keeping our system in good shape.

      BUT we would always use good SDLC practices.  Otherwise - well otherwise we would only have "junk" to support.

      Just my thoughts. (I think they mirror your thoughts to some degree.)

      Michelle

      It will be interesting to watch the roll out of the technology, don't you think?

      Author's profile photo Bala Prabahar
      Bala Prabahar
      Blog Post Author
      Michelle,
      I enjoyed meeting you and others (notably Marilyn, Jon, Arun V, Vijay V,  Vitaliy) in LV. I had a great time. Great people, great thoughts!.  My response below is very long. Sorry!
      Yes, I'm looking forward to working on HANA.
      It seems our database size is getting larger and larger because we don't spend time to control the size. A lot of people think disk, memory, CPU, network bandwidth etc. are getting cheaper so no need to spend time and efforts on controlling the DB size.
      I've a slightly different viewpoint: Cheaper resources (compared to a few years ago) don't justify keeping "junk" in the database. For example, I’ve seen customers not deleting the change logs for 7+ years. Basis team brought up the issue because change logs were using more than 1/3rd of the production database. BW team - instead of analyzing the change log requirements - consulted an SAP expert. He mentioned he had never seen any customer needing change logs older than a year. BW team, despite SAP expert’s opinion, insisted their situation is unique (no document to support) and decided to keep change logs for all 7+ years. Deletion of change logs is a part of SDLC, isn't it?
      Second topic is archiving. How many customers have implemented archiving? The issue of not implementing archiving makes the DB size larger artificially. Archiving again is a part of SDLC, isn't it?
      Third topic is BW design itself. Unfortunately there are a very few people who know both SAP BW and Data Warehousing principles. Anyone with SAP BW knowledge can easily design, develop and implement BW solutions. Unfortunately BW implementations from my own experiences and from reading other blogs in sdn imply they don't follow DW principles.
      These three are just a few challenges of the current environment(s). There are other indirect challenges as well: Instead of trying to understand and address the issues on hand, I've heard customers/consultants asking: What is the best practice? or what does SAP recommend? Is the solution one of team members offers in line with SAP's recommendations? They would like to have SAP's recommendation for every unique situation they run into.  Why not use their own experience and product knowledge to come up with a solution meeting their unique requirement?
      Would HANA address those challenges? Let us wait and see.

      Regards,
      Bala

      Author's profile photo Former Member
      Former Member
      Love your long reply!!!

      For the most part I agree.   My exception would be, I don't think HANA is meant to solve the problems.  It's just another tool to use.  "We" have to do better with design, database, etc.  I'm not a BW expert.  However, I think the same holds true no matter what we are doing.

      Archive plans always help! 

      Michelle

      Author's profile photo Marilyn Pratt
      Marilyn Pratt
      And your blog title is quite original.  🙂
      Good to see you sharing your ideas and challenging thinking (and marketing 🙂 )
      Author's profile photo Bala Prabahar
      Bala Prabahar
      Blog Post Author
      Marilyn,

      Thanks for your feedback! I love 'openness' of sdn. In sdn, I don't need to worry too much about "political correctness". This makes blogging a bit easier.

      Regards,
      Bala