SAP HANA has been around for a little over 2 years now. Its new vision and approach, as with many things novel and paradigm shifting, can be met with resistance and disbelief. Change is not always easy and many who are comfortable with existing approaches try to evaluate HANA in the light of those older paradigms. Hence they often fail to see the bigger opportunity – the opportunity to re-think how we fundamentally approach data processing architecture.
In the days ahead, you will see a new feature here on saphana.com entitled “The HANA Difference” focused on SAP HANA’s defining capabilities. Within this feature you will see a variety of perspectives on these defining capabilities from SAP, customers, and from our ecosystem of experts. The discussions will focus on how HANA optimizes information processing,, backed up by examples, demos or code snippets. These examples will illustrate the impact of SAP HANA in simplifying application design and data management, provide deeper understanding on how SAP, customers and partners are applying them, and how these capabilities translate into a very different way of assessing value for those who deploy it.
This blog highlights few of SAP HANA’s defining capabilities that will be the focus of detailed blogs in the coming series, which we hope you will discuss along with us. As indicated, there is a HANA Difference. Let’s look at some of these unique HANA capabilities.
- SAP HANA’s optimized cache awareness. HANA’s optimizer knows the locality of the data in the CPU L1, L2 and L3 caches on the CPUs. It leverages this information along with knowledge of all the bandwidth and latencies between every cache, every CPU, and every node in its execution plans. This accelerates repetitive queries to mere milliseconds revolutionizing how information is digested by users.
- SAP HANA can run real-time OLTP and OLAP on a single copy of the data so there is no need for an OLTP and an OLAP copy. SAP HANA stores data only once for transactional and analytical applications –transforming how businesses build and consume applications. Significantly cutting down cost for permanent storage and, the time, cost and effort involved in ETL, backup and archiving.
- Multiple Parallel Processing (MPP) on a shared nothing architecture along with its support for Single instruction, multiple data (SIMD), is another unique capability. SAP HANA breaks the work process into sub components and distributes these combining the storage and query processing capacities of several servers in a cluster. This increase capacity and reduces response times enabling increased the use of the system by end users.
- Despite the fact that SAP HANA is supporting multiple domain-specific data processing capabilities, it has not sacrificed on ACID compliance or Reliability, Full High Availability, Disaster Recovery and Supportability. What we have done is to take the opportunity to re-think how these requirements can be further enhanced and simplified because of the HANA’s cache-memory aware architecture
- SAP HANA’s on-the-fly schema extension capability allows for flexible business model changes creating another strong differentiator and business value.
- Dynamic Data Tiering optimizes the balance between data processing and data storage. Data identified as frequently used is kept in memory. Any data that hasn’t been accessed recently is purged from memory but persisted on-disk with no write-back is needed to preserve it. And cold infrequently used data can be persisted in systems like SAP Sybase IQ or Hadoop but is still made accessible dynamically via data virtualization using SAP HANA smart data access. This dramatically reduces the cost of permanent storage and the time and effort for data movement and ETL.
- SAP HANA is more than just a database. It converges, platform, database, data processing capabilities, handles spatial and textual data analysis, and provides libraries for predictive, planning and business analytics. To do what you can do in one HANA appliance you would need several separate individually purchased, supported and maintained comments for competing systems. Now that is simplification and cost reduction.
- SAP HANA comes with the most comprehensive data provisioning allowing data to come from any source to benefit from the real-time performance of true in-memory computing. Synchronization of mobile and machine data, analysis of streaming or sensor data, data virtualization leveraging the unique processing capabilities of the source system, batch loading, real-time replication and in-memory massively parallel transforms.
- SAP HANA is open and agnostic and is available for any application, data and source, giving you the flexibility and adaptability you need. It runs on commodity x86 based hardware from 9 HW partners and 7 cloud infrastructure providers. It supports a wide variety of programming languages, integrate with certified 3rd party tools. Numerous ISV and StartUp apps are deployed on HANA without modification. It is data agnostic with support for structured, unstructured (i.e. text), spatial, document data and sparse data
- Extreme linear scalability is another unique SAP HANA capability that sets it apart. Let’s see someone else do this. Year over year trending report for top 100 customers over 5 years – 1,200 Billion rows in 3.1 seconds. How can real-time insight like this transform how you see and do business?
Of course, what is listed above is not the complete exhaustive list. There are many more topics important for technologists of all shades as well as for others in the ecosystem. We invite you to suggest additional topics that you feel it would be of value to discuss, or on which you feel more clarity would be of benefit. Anything “hot” and top-of-mind – as it pertains to true real-time in-memory platforms.
Remember, the discussion around SAP HANA and related matters is not just a discussion about databases but how to introduce your organization to the renaissance of computing, and drive real business success therefrom.