Large data or traditional data, just find the right tool

At the recent Enterprise data World Conference, participants have expressed their views on the topic of "Big Data". Karen Lopez, chief consultant from Infoadvisors, said the big numbers weren't as reliable as they seemed.





Lopez said: "From the most basic question, what is the big data?" I believe that no one can really speak clearly. Lopez The Wikipedia explanation, which says that big data is data sets that have grown so much that they are hard to handle. Lopez said: "What kind of definition is this?" "


, a very well-known consultant on Twitter, talked about http://www.aliyun.com/zixun/aggregation/14417.html ">apache Hadoop and other related technologies, She says the great thing about Hadoop is that all of its related transactions are named after Hadoop, such as Hadoop column,hadoop Distributed file Systems, Hadoop MapReduce, and so on. Why didn't the yellow elephant mascot call Harry or Harvey, and that was the strange name of Hadoop?





Lopez also lists techniques that are not named after Hadoop, including Hive, Pig, zookeeper, and mahout (the meaning of the Elephant Rider in Hindi). "Projects that use these technologies must not be completed on time, because the elephant riders work with pigs (Pig) and breeder (zookeeper)," she said with a hint of irony. "





In addition, Lopez to make their own judgments about the unpredictability of large data. "You can't know in advance what type of data you're going to get, it's a lack of schema," she said. The problem now is that you don't know how to design a schema until the data arrives and unfolds. In other words, when you are modeling your data, you have to do it at a sprint speed. My so-called Sprint speed should be near the speed of light, so I hope you can start practicing now. "





Of course it was just a joke, but Lopez explained why she had to throw cold water on the user. She believes that big data is a big part of the hype from the manufacturers, in her view, large data technology and any other technology is the same, is not a "cure-all" "Panacea." "There is no reason to think that big data is antithetical to traditional relational data or structured data, and we are already very familiar with the latter," says Lopez. We want the right tools to deal with the right job, big or small, that doesn't prevent us from finding the tool. "





also has a number of users who have expressed different insights from the Phasic Bae, CEO of Geoffrey P. Malafsky, who believes that relational database technology is obsolete and discarded by the torrent of technology. Big data is not popular for no reason. No schema data is present, as long as the data is stored in the system, it has a schema.





Amir Halfon, a financial services technology director from Oracle, did not refute Lopez as categorically as Dr Malafsky, but he found it inappropriate to Lopez the irony of Hadoop. Halfon is also half joking that the author of Hadoop is a technical Daniel, he also launched a number of other technical projects, it is not possible to give each technology a satisfactory name for her.

(Responsible editor: The good of the Legacy)