What is SAP's approach to managing big data?
The big data hype machine is in full swing and buzzwords are flying left and right. Big data has always existed, but its definition has changed over time as computing power has improved. Today, if we're talking about big data, we're probably talking about data sets in the petabyte (1,000 terabytes, 1,000,000 gigabytes) order of magnitude.
SAP's HANA project parted ways with the rest of the industry by keeping relatively large data sets (terabytes) in memory. Over the last couple of years, the industry has come around on SAP's approach. But a terabyte is no longer so big. While it may be technically possible (SAP has built a petabyte HANA cluster), most analytics scenarios won't justify the cost of keeping petabytes of data in RAM on HANA.
SAP's big data strategy centers around HANA but farms out really big data to SAP's Sybase IQ product (recently renamed as SAP IQ) or Hadoop clusters (via Hive) -- technologies designed to store and query very large data sets cost-effectively. The data appears to be in HANA, but is actually stored in IQ or Hive. IQ is also an excellent and widely used option standalone (without HANA) for analytics on very large data.
This strategy to manage big data on a fit-for-purpose database promises to allow customers to balance speed against cost. Customers should stay up to speed on the options available to them so that they can balance this equation effectively.
What are the data warehousing options for SAP landscapes?
The new NLS application is changing the SAP data management equation
Top tips for SAP BW
Dig Deeper on SAP data management
Related Q&A from Ethan Jewett
The design rule of thumb is to keep the Launchpad tile count under 20. But what about power users who need to launch more applications? Here are ... Continue Reading
The way Launchpad versions relate to each other and to SAPUI5 versions isn't intuitive, and both affect app compatibility and access to newer ... Continue Reading
If you want to get more serious about developing UI5 apps, you need to know how the libraries fit into the broader architecture, as well as their ... Continue Reading