The CEBIT trade show is an unbelievably vast series of more than 23 airplane hanger-sized halls, with thousands of vendors taking up booth space and a seemingly infinite number of products on display to the attendees moving from one exhibit to another. Trying to absorb it all is overwhelming - not unlike the challenges facing enterprise IT departments grappling with so-called "big data."
Though the official theme of CEBIT 2012 is "managing trust", big data was undisputedly the hot topic on the mouths of vendors, keynote speakers and conference sessions that have been taking place here over the past week. Though they sometimes differ on the particulars, most IT industry professionals broadly define big data as the unstructured information that's accumulating inside and outside organizations due to the proliferation of Internet-enabled mobile devices, the surge in social media-generated content and the increasing speed of transactions that take place both online and off.
Software AG used CEBIT 2012 to officially launch its foray into the big data market, holding a press conference for international media and including an area about it in the innovation corner of its booth. Last year Software AG bought Terracotta, which offered an in-memory database that competes with the likes of Teradata and SAP's HANA called Big Memory. By the end of this year, Software AG said it will combine Big Memory with its complex event processing (CEP) engine, WebMethods Business Events.
The company will also add a dashboard on top of the system which will offer up-to-the-minute information on critical customer areas, said Wolfram Jost, Software AG's CTO. "It's not going to be necessary to distinguish between transactions and analytics. In-memory can handle both," he said."Like the cloud, we see big data as hugely disruptive, and it will be less costly than acquiring more database licences."
Jost sees the combination as a competitive differentiator among the other range of big data players. "With complex event processing and in-memory you get real-time BI," he said, adding Software AG will be able to integrate Hadoop clusters. "To the best of my knowledge, SAP is concentrating on analytics."
Software AG will be working within SAP environments, however, and has established an SAP competency centre to learn more about HANA.
In his keynote speech on CEBIT's first day, meanwhile, Microsoft COO Kevin Turnersaid the Microsoft approach to big data would be based on using Windows Server and its System Center management software to handle processing workloads, then connecting to what he called "ublic data sets" in its Azure marketplace and offering the results through common productivity tools like Microsoft Office and SharePoint.
"We will be open and flexible using Windows Azure and SQL Server," he said. "Everything will be compatible with Hadoop connects to SQL Server."
Big data was also the focus of a panel discussion at CEBIT, where various firms talked about how they are starting to use big data in interesting ways. At Vodafone Germany, for example, the company has 350 million customers,37 million of which are mobile customers, according to its chief commercial officer for the enterprise business unit, Jan Geldmacher. Vodafone is now working with TomTom, a GPS system provider, to produce predictions on potential traffic congestion based on the movement of mobile customers, he said.
IBM, meanwhile, is using big data in some parts of South America to forecast major rainfall before it hits cities. "People often talk about real time, but real time is sometimes too late," said José Carlos Duarte Gonçalves, Big Blue's CTO from Brazil. "Many times we need information in advance."
The level of data mining that goes into such projects can raise concerns around data privacy, of course, but much depends on how granular the analysis becomes, said Yves de Talhouët, senior vice-president and managing director of HP's EMEA region.
"Most of the big data analysis is about one individual but about trending," he said, using health-care as an example where large data sets could unlock helpful insights on treatment and providing better care. "Of course those data are extremely sensitive and I don't want my personal data to be seen by anyone else. But within a group compared with other groups, I may not mind."
Information doesn't have the same value all the time, de Talhouët added. "When we talk about privacy/security, have to think about information aging and when does it have value and when does it stop having value? As we go into more real-time interaction, information starts having an extremely short life-cycle. Some information beyond 10 seconds has no value." That influences security thinking, he said.
Beyond trending, however, many firms are starting to eye big data as a means to get a better understanding of a particular customer's likes and dislikes so they can tailor products and services accordingly, said Steve Garnett, chairman and president of Salesforce.com in Europe. "The ultimate idea is of one-to-one marketing," he said, adding that companies are also beginning to invest a lot more in monitoring social media conversations to speak directly with customers. "The days, I think, are diminishing where you buy a product or service and something goes wrong where you call the company. You go to Facebook, Google, Twitter and find out what other people are saying."
As companies figure out a big data strategy, they will likely need to bring in or develop new skill sets. Kaspersky Lab CTO Nikolay Grebennikov said his firm is one of them. As part of its security research, Kaspersky creates some 70,000 malware records a day, most of them by robots, but the company is putting an increased emphasis on hiring staff with expertise in data mining and mathematical models.
"When we're talking to companies . . . they are not just interested in installing anti-virus and gateway (products)," he said. "They want people who can analyze trends in the network. That provides the best value."
The panel discussion was moderated by Amazon CTO Werner Vogels, who also gave a keynote speech where he discussed how the online retailer has used big data to power areas like its recommendation engine for related products to customers. When people get bad recommendations, Vogels said, that may be because there wasn't enough data to draw upon. Sometimes, he said, the challenge is not simply in analyzing data but making sure you can properly manage its lifecycle.
"Big data is not only about analytics, it's about the whole pipeline," he said. "So when you think about big data solutions you have to think about all the different steps: collect, store, organize, analyze and share."
On the CEBIT show floor, vendors capitalizing on the big data trend included open source BI provider Pentaho, regular suspects such as SAP and Actian, a company with headquarters in Redwood City, Calif. Actian offers Vectorwise, database management software that companies are using to analyze Web traffic patterns, cellular data records or stock purchases, according to the director of business development at its Central European office, Olaf Laber.
While a lot of hype is coming from the vendor side, IT professionals on the customer side appear to be just as interested in big data. "I did a talk about it this morning on the show floor," Laber said. "It was full."