Traditionally a query was processed sequentially. As the need for trend analysis grew, it is required to query tables with large volumes of data. As a result these queries took longer time. With the use of the various optimization techniques, the query processing time was reduced. But by the time we rejoice the benefits of this improvement, volume of data to processed increases and we are back to square one.
A typical Data warehouse contains large volumes of Data. Traditional oracle database could not handle this data. Oracle 8i, 9i, 10g etc. are designed for better performance of transactional system. I see that off late companies started using Teradata, Data Allegro, Netezza etc. for implementing their BI solution.
What makes Teradata different? Parallel query processing, fallover mechanism, distribution of data across AMPs makes it process query parallelly and return results in less time.
Is this the end? Reducing the query process time is always a good thing to cheer. Oracle’s claim of The HP Oracle Database Server being fastest on earth is a good news to big organizations with large BI environments. The system consists of
How is the query time reduced? Data is stored in the form of blocks on the hardware. When database processes a query, it fetches all the data blocks from the hardware and then processes the query. When the size of the table is huge (which is typically the case with BI solutions) lot of time is taken in transferring the blocks from hardware to database. Oracle bypassed this part and instead took the query to hardware and processed it at the hardware level. The end result is then passed to database. This results in saving the time for transferring data blocks from hardware to Database. This makes Oracle Exadata database as the fastest database in the world.
Architecture of The HP Oracle Database server: There will be three grids, grid of Database, grid of Fusion middleware and grid of storage servers. The connection between the storage server and DB is infiniband pipe as that can practically transfer data at 1 GB/sec.