How is it possible that
HANA is so fast? HANA is snappy in view of many reasons. The going with picture
1 depicts an outstandingly unraveled interpretation of what's inside the
In-memory Computing Engine.
Section
Storage While the standard databases store the social table one line after
another, IMCE stores tables in fragments. In a perfect world the going with
figure clears up the refinement between the two stockpiling segments viably .
Compression :
As the information
composed by each other is of same sort, there is no compelling reason to
compose similar values over and over. There are numerous pressure calculations
in HANA with the default being the lexicon calculation, which for instance maps
long strings to whole numbers.
Case of word reference
calculation: You have a Country section in your Customer table in your
database. Suppose you have 10 million clients from 100 nations. In the standard
line based capacity you will require 10 million string values put away in
memory.
Parallel Processing :
It is dependably
execution basic to make full utilization of the assets accessible. With the
present support in the quantity of CPUs, the more work they can do in parallel,
the better the execution.
The columnar
stockpiling empowers parallel preparing as various CPUs can take one section
each and do the required operations (conglomerations and so on ) in parallel.
On the other hand numerous CPUs can take a parceled segment and work in
parallel for quicker yield.



No comments:
Write comments