Retrieving Large Volumes of Data

In the News:

SELECT Journal – Andrei Dzianisau, edited by Ian Abramson

Today’s database systems are being asked to store more information than ever before. According to research firm IDC, the size of data — which it calls the “digital universe” — will grow to 2.7 zetabytes in 2012, up 48 percent from 2011’s record year. This growth is unprecedented, and we must be able to manage databases that contain significantly more information that needs to be retrieved in less and less time. It is more important than ever to understand the options Oracle provides to users within the database that can truly empower your applications regardless of the size of your data sets.

Hardware appliances such as Oracle’s Exadata have helped significantly, but keeping the laws of physics in mind, there is a limit. Whether you use the faster memory, solid state disks or read technology, sometimes even very powerful hardware doesn’t help. For instance, it’s not possible to read 60 GB of data off from disk in just a few seconds. There are limits to how powerful a storage system can be. As a result, we need to understand how to take advantage of Oracle and the features and facilities available to us. Not everyone can afford to purchase an Exadata machine to solve their data volume issues, so this article investigates some of these capabilities that all Oracle databases can take advantage of.

It is a very common situation that a business asks for information to answer important business questions on daily/hourly basis. It then demands that these questions are answered within seconds (or, at very most, a matter of minutes). It is also very common to expect business intelligence (BI) systems to create reports that require vast amounts of data to support complex reporting needs and be able to use that data quickly and efficiently. The tools today try to optimize how they retrieve data, but, sometimes, with poorly deployed data warehouses, this is not always possible. What are the things you can do today with a basic Oracle installation? This article will discuss numerous strategies and approaches we at EPAM Systems have seen to be very successful in optimizing how we use and deploy data solutions. The top concepts we will discuss include the following:

  • Narrowing down amount of data to access
  • Using aggregates to optimize data retrieve and minimize calculations
  • Changing the design of your solutions
  • Using some of the built-in functions supplied by Oracle
  • Combining processes to streamline processing

These are just a few of the approaches one could take to support and ensure that your data warehouse performs well and consistently. Each approach we will discuss may appear basic on the surface, but the value each can provide can significantly impact your overall performance and save your organization from investing in unneeded hardware.
 

The full version of the article is available here, page 25.