Expand Hadoop Category

October may be the one months worth noting in the history of large data. Because in this month, we can redefine Hadoop. It can be a research framework of large data batch processing, and it can also be a high speed engine and interactive analysis product of structured and unstructured data large-scale parallel analysis data.

>

There are already companies trying to prove it. Recently, the industry held a special meeting to promote Hadoop itself Hadoop-plus-sql architecture, add advanced analysis function, through the image display and so on. Among them, three companies are fairly representative.

Birst:birst has transformed into BI and released a set of related products, and is now back on the role of a service provider based on Hadoop providing cloud computing and large data. Birst Big Data Services enables the storage of unstructured and semi-structured data, and on this basis enables users to implement analytical applications without the need for more mapreduce knowledge or other complex skills. Thus, a new connection service for structured relational data storage and a visualization tool analysis for new types of data are implemented.

Splice Machine:splice Machine is an innovative company in San Francisco. The Hadoop Distributed File system was built on top of the SQL database, and the first round of 4 million dollars of financing from Mohr Davidow Venture was announced in Wednesday. Like another innovative enterprise drawn to scale, Splice machine promises to provide services that provide SQL functions and transactions on a distributed basis HDFs and HBase. It sounds like a lovely story. If it can be implemented, a flexible architecture of unstructured data, and a lot of scalability, can achieve seamless integration with many businesses ' favorite SQL BI products.

Teradata:teradata has finally done something that everyone has long wanted to do, and has built a program named Big Analytics Appliance that will aster the data database with Hadoop. In fact, Teradata has acquired Aster data a few years ago and is the necessary investment in unstructured data as one of the core data areas, but it has not been compatible with Teradata's core data warehousing and analytics business. Aster data is a product of sql-mapreduce software that enables users to perform mapreduce jobs by using standard SQL. Linking Aster-hadoop and Teradata's flagship database is sql-h. This secret weapon allows users to access Hadoop data and connect to Aster data, which enables analysis.

The meaning of these unique products is impressive, but we have not yet seen the landing of any product. This is not impossible in the Hadoop ecosystem. At next week's O ' Reilly Strata Conference and Hadoop World, there will be more heavyweight appearances and share how they combine Hadoop technology with business needs more closely.