During the last years SAP has entered the database industry with a brand new revolutionary database: the in-memory HANA database. SAP is now considered one of the leaders in the database industry.   This blog talks about the main design principles of HANA and it discusses if HANA really is revolutionary.

“HANA” is short for “High Performance Analytical Appliance”.   Yes, the marketing team must have worked hard on this abbreviation to make something nice of it… As the name indicates “analytics” is in focus.

The database is optimized for analytical workload. In my opinion there are two main design principles of HANA: 1) column base structure and 2) in-memory based.

Column based structure: while databases in the last 40 years have been designed to read and write to the DB as fast as possible for individual transactions (think about adding one new sales order with 50 field into the DB as one record), HANA is designed with key priority to read and process a vast amount of similar data for analytical purposes as fast as possible. Think about reading customer name and sales amount from the database for 10.000 sales orders to create an overview of total sales per customer. In HANA, instead of having to read 10.000 records, you only have to read two “records” (the record of customer name and the record of sales amount).   This makes reading and analyzing data extremely fast. Watch this video for short and great introduction to some more details on this principle.

In memory: The disadvantage of the a column based DB (as discussed in the previous paragraph) is that inserting and reading one complete sales order is a lot more work for the DB vs a rows oriented DB. Think about the sales order with 50 fields: adding this sales order in a column based DB means updating 50 records. However, by running the DB in memory (which speeds up “everything”), updating the 50 records can be done with “normal” speed. Reading 10.000 records to create a report will obvously become even faster when the data is kept in-memory!

As a result of 1) column base DB and 2) in-memory DB, transactional work (like adding a new sales order) can be expected to have similar speed in HANA as in traditional “disk based” databases. However, analytical work will be incredible faster!   These are activities like searching for data, producing dashboards, producing overviews and reporting, graphs, data mining, user playing with data, etc.

Designing and creating an in-memory database design is tremendously more complex than it sounds like. It is not just to “put the database in memory”. You have to redesign and re-code everything to fit another memory and hardware type. So HANA is a real innovation! One of my favorite proof point of this is the reaction of Oracle boss (and SAP competitor) Larry Ellison 3 years ago when he heard that SAP planned to enter the DB market with in-memory DB HANA: he called the plan “Wacko” and “extremely complex” and said that his company had not managed to do anything like this yet. Oracle has been one of the leading DB companies for the last decades. I strongly advise you to watch this 2:00 min video of the interview with Larry. It is good entertainment!

Further, Gartner (recognized independent IT research and advisory firm) now rates SAP as one of the leaders in database technology, proof point that SAP has really established itself with its HANA technology.

For business management systems, running on a HANA database can give tremendous advantages.  HANA gives an unprecedented opportunity to increase employee efficiency and for management of a business to obtain better insight into what is happening in their business for decision making.

SAP Business One solution is one of the SAP solutions take advantage of this new SAP technology and helps organizations with as few as 1 user up-to 500 users to be able to take their business to the next level leveraging HANA.  Watch this video for some of the cool “out of the box” features based on HANA.