SAP HANA

SAP HANA is high performance in-memory, column-oriented, relational database management system developed by SAP SE. It’s a hybrid database that has as primary function  – to store and retrieve data for applications. Data is stored in terms of rows or in terms of columns. SAP HANA can read data in 5 milliseconds and thus why it is known to stand for High Performance Analytical Appliance. SAP HANA allows you to create boundless intelligent solutions for real time decision making. It’s advanced analytics features lets you support next generation transactional processing.

• Real Time Analytics – More integrated, intuitive and real time analytics on huge data volume for decisions support and operations
• Opportunity for new business models – Integrating real time data with speed means ability to redefine process
• Superior Performance on Large Data Volume – In-memory computing power to analyze huge data without need for building complex data models
• TCO Reduction – Data compressions, reduce data modeling needs, eliminating separate database for DWH(long run)

HANA allows us to store data in both methods to leverage the benefits of both storage approaches. Database storage structures are based on application needs.

• OLTP systems(Row storage)
  • High amount of write and update operations
  • Typically needs to access the complete record(i.e. require attributes stored side by side
  • Process only a single record at a time
  • Allows reading of large number of attributes against a single key
  • High amount of write and update operations
  • Eg. Retrieve all customer details using CUSTOMER_ID

• OLAP systems(Column storage)
  • High amount of read operations
  • Calculations are executed in a single column or a few columns
  • Allows reading and processing vast amount of rows for select attributes
  • Eg. Total sales for a given period

• Vertical Partitioning
  • It is rearranging of individual database columns by splitting columns of database table into multiple column sets
  • Each column set can be distributed on individual database servers
  • Used to achieve better search performance while guaranteeing high-availability of data
  • Need in depth understanding of application’s data access patterns
  • Attributes that are accessed in the same query should rely in the same partition since locating and joining additional partitions may degrade overall performance

• Horizontal Partitioning
  • Used for dividing large database tables into smaller pieces of data
  • Each piece of the database table contains a subset of the complete data with table
  • Splitting data into equivalent long horizontal partitions is used to support search operations and better scalability

• Multiple physical machines contribute to one virtual machine, giving the user the impression of working on a large server
• Multiple queries can be handled by multi-threaded applications, where the application does not stall when dealing with more than one query
• Each query is mapped to a single core, thus optimizing the system’s response time
• Query processing also involves data processing i.e. the database needs to be queried in parallel too
• From the database perspective, the partitioning of data sets enables parallelization since multiple cores across servers can be involved for data processing