In Memory Management – Focus on SAP HANA
Before jumping into in-memory and other fancy words, let us
first understand the information world we live in. Information is everywhere.
It is the relevancy that matters most for individuals and organizations. May be
it is an individual’s bank transactions, airline reservations , product reviews,
weather updates, global scenarios or an organization’s market penetration,
competitors, economic climate, key performance indicators, customer
preferences, what not. Everywhere it is the information which gives answers to
questions like, what, when, how
and why. These questions are a direct result of
individual or organization’s thrust for more knowledge to become stronger and
stronger for sustainability and advancement.
Having said that, let us focus on three points: Information availability, Information relevancy and effective Information usage. We are going thru an Information Explosion. Massive amounts of data is being created
or generated on a daily basis around us. It could be human contribution and/or
machine generated. A generic example of human contribution is social network data
originating thru our popular friends like Facebook, twitter and YouTubes. Apart
from this, organizations do generate transactional data as part of their business
functions, contributions from academia and media. While capturing machine data,
has been at a low key all these years, but have risen to a level of very important
ingredient in the current day business intelligence for quicker decisions.
Talking about Information availability, how do organizations
collect, organize these huge volumes of data. Over the years the storage costs
have come down significantly resulting in less expensive storage solutions.
However, there are two tricky questions which drive the other two points from
above, Information relevancy and effective usage. Relevancy derives the
selective data collection or procurement for the specific business functions
the organizations looking for. Corporate spending always needs to answer the return
on investment question. Whatever data is captured, maintained, organized how it
is providing the Business intelligence value for winning over the competitors,
customer satisfaction and business growth.
Effective usage of information revolves round the ease and
timely availability of information. How to reduce the latency between
information requirement and information availability? One answer for this is
having a scenario where the information is available in real-time fashion. What
does that mean? Information is available
for decision making as it happens. Sounds amazing. But how? It needs systems
with robust storage, retrieval and analytical capabilities. We did talk about inexpensive
storage options; however, effective retrieval is always constrained with speed
of disc operations. This lag can be addressed using in-memory storage for
faster retrieval. The last piece in the puzzle is how to consume information in
a user friendly manner. This needs a set of user interface applications with flexible
and rich visualization capabilities.
Several vendors are in the market place for providing the
applications and tools for fulfilling these current day organizational needs.
SAP has come up with a solution called HANA (High
Performance Analytical Appliance) which is built on in-memory capabilities
supplemented with a set of peripheral tools for ETL (Extraction,
Transformation, Loading), data replication and analytics visualization.
What is
HANA? It is an application platform comprising of in-memory
database and flexible development environment.
For a basic level understanding, it is a replacement to the
traditional relational databases (RDBMS) such as Oracle, DB2 and SQL Server. All
the data will be residing in-memory which provides high performance data
retrieval compared to traditional I/O from the disk based RDBMS. This results
in a significant improvement in the transactions happening in the application.
Next level of understanding is applications logic running in
HANA environment, means pushing the application logic to memory will enhance
the performance on the overall application processes.
Third level of understanding is using HANA for creating
database type of applications using custom tables and processing logic. This
opens gates for marriage between enterprise applications data and non-enterprise/custom
data on the HANA side, which has been the biggest challenge over the years with
large organizations. This is the scenario where enterprise data can be combined
with other sources of data such social
networks, machine not jeopardizing the relevancy.
Finally, SAP has opened the HANA as an application
development environment thru XS engine, where organizations can build their own
applications using the rich UI5 and Java capabilities. In this scenario,
applications can leverage enterprise as well as non-enterprise data available
in the system.
Globalization, tough market and economic conditions
naturally drive organizations to keep on finding better ways to strengthen
their business intelligence competitive advantage. SAP HANA can help in
bringing together the information explosion and faster and flexible information
availability for building insights, so that they are equipped for taking
advantage of favorable market conditions, customer trends, price fluctuations
and other factors that directly influence the bottom line.
----Sundar
Dittakavi