Why Big Data? The post-Cloud era has spawned the latest technology hotspots

Are you probably as surprised as I am that big data suddenly pops up to be the latest tech hotspot, like cloud computing in previous years? Http://www.aliyun.com/zixun/aggregation/13568.html "> Large data technology seems to exist a few years ago, in the history of databases, data warehousing, data marts and other information management areas of technology, To a large extent also to solve the problem of large-scale data, why they can not meet the requirements and make Hadoop as the core of the big data technology to become a hotspot? Chatting with a few friends, I made a list of 3 main points for you to correct.

1. Unstructured data is growing too fast and there is an urgent need for appropriate technology to deal with and discover value.

13 years ago in 1999, when I was in Microsoft for SQL Server technical support, a customer sent his unit to the suspect database to me to repair, and then called me: Our database is very large, there are 50MB! 50MB Ah, look at now, the amount of information around the world every two years to more than double the speed of growth, 2011 produced and replicated 1.8ZB of massive data, the growth rate even more than Moore's law. What is the concept of 1.8ZB, which requires every person in China to send 3 tweets per minute and uninterrupted for 6,217 consecutive years. Ubiquitous mobile devices, RFID, wireless sensors every minute in the production of data, hundreds of millions of users of Internet services at all times in the production of a huge amount of interaction ... The amount of data to be processed is too large and growing too fast. Worse, 85% of the data in the enterprise is unstructured data (email, documents, pictures ...). , which makes our relational database very emotional! We all know that these data are very likely to contain gold, but how to deal with this data and how to find value from these large data are all calling for big data technology.

2. How do we deal with obsolete data, delete it? In the past, we processed obsolete database data, the first is backup, and then removed from the current system. The second is to create a table partition, and then move the old data partition to the slow hard drive. But these initiatives have been powerless to deal with structured data, not to mention unstructured data that is more data-heavy. When you need some data in the backup disk, if it is me, I would like to eat them directly. Ideally, both outdated data and new data can be queried at any time when needed, and performance is unaffected. You don't have to dwell on how to handle the growing data and the gradual degradation of system performance, which requires big data technology.

3. The rapid decline in hardware costs, especially the progress of storage equipment, makes it possible for large data technology to move to business. Enterprises can use cheap PC server to build large data platform, and can be extended at any time. This can be done simply by applying Microsoft's Windows platform. and Windows Azure offers infinitely scalable storage and computing capabilities, and companies can use the hybrid cloud technology to provide greater power to the enterprise's development by using large data technology at will.

There should be other factors driving the big data into a hot spot, hoping to hear a better view