Image courtesy of Shutterstock
With the speed of business constantly increasing, the demand for real-time data has increased along with it. As business-driven demands on data speed increase, however, some organizations are still inhibited by slow moving data that can’t keep up. New strategies must be created to process data at the speed that is expected from today.
A recent DBTA webcast featuring Maria Colgan, master product manager with Oracle, and Unisphere Research analyst Joe McKendrick, examined the technologies available today to address these changing requirements for data delivery.
The two biggest factors that are hindering companies from becoming adept at using data in their decision making are lack of complete information and delays in the availability of required information, according to a new Unisphere Research survey conducted among 303 members of the Independent Oracle Users Group (IOUG) and sponsored by Oracle.
Fifty-five percent of respondents claimed a lack of complete information, while 51% claimed delays in the availability of required information. “Thus, the two major inhibitors for an organization becoming a data-driven organization are that the information still remains bottled away and simply takes too long to extract and deliver to decision makers,” noted McKendrick. Data is the most important factor when it comes to decision making. Making a sound decision using data is clearly the better option than making a decision on a hunch. The Unisphere survey discovered that only 24% of respondents’ organizations are using data in their decision making all of the time. The business world is increasingly demanding real world insights but challenges persist that inhibit the process. Enterprises are also slowed by poor configurations of their data management tools, as well.
“Oracle’s primary goals for in-memory are to really allow our customers to ask those business-driving questions anywhere in the lifecycle of the data - to give folks the ability run real-time analytics,” stated Colgan. Traditionally, there have been two types of data formats: row and column. When Oracle, began it chose to be a row format database. Many vendors have forced users to select one format or the other, but Oracle Database In-Memory is looking to change that by offering the best of both worlds. It does this by reading the data that a user enters into the database and determines which format would work better, said Colgan.
An advantage of the product being a part of the Oracle family is the features from other applications that can be used for Oracle Database In-memory. Joins is an example, said Colgan. Oracle Database In-Memory is able to convert star joins into 10 times faster column scans. Traditionally, Oracle has offered two types of aggregation, but is now offering a third - in-memory aggregation. This creates an in-memory report outline that allows the user to keep a running total of the calculation as the data is scanned and filtered. Oracle Database In-Memory also allows multiple ways to scale. It allows scaling out across servers to grow memory and CPUs. Users are also able to scale up on large SMPs. Finally, users can combine flash and disk. This allows for the data to be placed on the most cost effective tier.
To watch a replay of this webinar go here.
To access the executive summary of the IOUG/Oracle report titled “Moving Data at the Speed of Business: 2016 IOUG Survey on Data Delivery Strategies,” go here.
The full survey report is available to IOUG members.