Cloud computing era: the big data bubble is expanding infinitely)

In today's enterprises, 80% of the data is unstructured data, which increases by 60% every year. Big Data will challenge enterprises' Storage Architecture and Data center infrastructure. It will also trigger a chain reaction to applications such as data warehouse, data mining, business intelligence, and cloud computing. In the future, enterprises will use more TB-level (1 TB = 1024 GB) data sets for business intelligence and business analysis. By 2020, global data usage is expected to surge by 44 times, reaching 35.2zb (1zb = 1 billion TB ). Big Data is revolutionizing the IT world. In October, the actions of several tech giants made more people realize that the so-called hi-tech bubble-that is, "Big Data" is expanding infinitely.

Microsoft cooperates with hortonworks to develop hadoop

As early as February this year, Microsoft's HPC development team announced a distributed computing platform called Dryad. This also marks that Microsoft provides Windows HPC server users with tools for massive data processing. Microsoft launched Dryad to encourage developers to write large-scale parallel application programming on Windows or. NET platforms. At that time, it was also regarded as a powerful product that Microsoft had fought against hadoop in the big data field.

However, at the SQL pass 2011 summit held in Seattle on September 14, Microsoft announced that it would cooperate with hortonworks, which was split from Yahoo! To develop hadoop, the Windows azure and Windows Server platforms will be built on Apache hadoop. At the same time, the hadoop-based Windows server will also work with Microsoft's existing bi tools to process tasks.

The goal of the in-depth cooperation between Microsoft and hortonworks is to leverage the expertise of hortonworks in this field to help maximize the integration of hadoop into Microsoft products. At the same time, the cooperation between the two companies can help simplify the download, installation and configuration of several hadoop-related technologies. Including HDFS, hive, and pig. This will help enterprises expand their business through hadoop. Microsoft will also write a new ODBC driverProgramAnd extend your existing Query System to hive. In this way, you can directly execute hadoop queries from Excel and powerview.

Red monk analyst Stephen o'grady is also optimistic about the combination of windows and hadoop. He said it would be very attractive, which will attract a large number of Windows users. Microsoft is competitive in this field.

Joint efforts of Oracle hardware and software in the Big Data Field

Oracle, as the world's largest relational database provider, is also unwilling to feel lonely. It has joined a nosql database called "nosql database" in its product chain. Nosql database is an integral part of Oracle Big Data appliance announced at the 2011 Oracle global conference. Big Data appliance is a system that integrates hadoop, nosql database, Oracle Database hadoop adapter, Oracle Database hadoop loader, and r language.

Oracle's investment in the big data field is far more than that. They not only launched Oracle Big Data appliance at the software level, but also Oracle exalytics at the hardware level. Exalytics targets big data. In-memory computing launched by Oracle provides massive information in the big data era, including structured, semi-structured, dataset, and unstructured data analysis. At the same time, exalytics also supports hybrid data sources, including Oracle databases, teradata, Microsoft SQL Server, and standalone corner databases.

In addition, exalytics has powerful hardware and software configurations: 1 TB memory and 48 core processors; Support for obiee 11 GB; 200 GB/s wide TimesTen parallel memory database; support for memory parallel processing of the corner stone OLAP server; the new high-bandwidth analysis-oriented user interface and the fastest connection to the exadata InfiniBand connection.

In the past, Oracle has always been somewhat conservative in the cloud computing field, but with Oracle's launch of powerful products at both the hardware and software levels at the Conference. This marks an epoch of Oracle in the cloud computing field.

IBM combines DB2 with nosql Databases

Likewise, at the IOD October conference held by IBM in 2011, Curt Cotner, vice president of IBM database server department, announced that IBM will launch the DB2 flagship Database Management System with built-in nosql technology next year.

IBM has some experience in the nosql technology field. Its own rational jazz collaboration software delivery platform uses the "triplestore" technology, the "triplestore" technology is roughly the same as that involved in nosql databases. Triplestore technology allows you to retrieve metadata and other related information in a concise and fast manner.

However, the IBM Rational team eventually found that triple did not have the expected availability features, such as failover and horizontal scaling to multiple nodes. The IBM Rational team found that if it received a large number of triple in a short period of time, nosql storage indexes would lock the database. The rational team is actually from Open SourceCommunityUse nosql triplestore and modify it to embed it into the DB2 database, with such modifications, you can use DB2 indexes, logs, high-availability solutions, and all functions in the DB2 database.

Codecome indicates that the modified nosql function will run more than four times faster in DB2 databases than in the previous open-source products, while eliminating problems caused by availability and scalability. Today's nosql functions are still under development, but the rational team will be able to integrate more nosql functions for DB2 in the future.

The future of big data in Enterprises

The ability to manage big data will become the core capabilities of enterprises that are increasingly using new forms of information, such as text and social media. This capability will help enterprises find the best model to support business decisions, the so-called model-based strategy. As a change engine, pattern-based strategies will make full use of patterns to find all dimensions in the process. Then, it provides the foundation for modeling new business solutions, so that enterprises can better adapt to the new environment. The ability to process and utilize big data will become a priority for many companies. Otherwise, they will be subject to the data and their competitors in the next few years. (Li Zhi/compilation)