Large Data Age: Inventory Major manufacturers large data layout

Emerging big data companies have sprung up. The rapid rise of major manufacturers, "large Data" is the program to strive in the future huge market demand relies on its own innovation for customers to create unique value.

IBM offers biginsights, Bigsheets and Bigcloud

Just a few years ago, IBM began experimenting with Hadoop in its labs, but it incorporated products and services into the commercial version last year, and Oracle and Microsoft announced that they would be actively accepting the platform. IBM launched the Infosphere biginsights software last May. The package includes the Apache Hadoop release, the Pig programming language for MapReduce programming, connectors for IBM's DB2 database, and IBM Bigsheets, a browser-based, Use the spreadsheet metaphor (Spreadsheet-metaphor) interface to explore and analyze data in Hadoop.

IBM then delivered biginsights and Bigsheets as a service in October through its Smart Cloud Enterprise (SmartCloud Enterprise) infrastructure. This service is based on the basic and Enterprise editions; The selling point is that customers do not have to buy supporting hardware or need it expertise to learn and experiment with large data processing and analysis capabilities. According to IBM, customers can build Hadoop clusters in less than 30 minutes and move data into clusters, costing each cluster 60 cents per hour.

Oracle: Database + Large Data machine

Oracle's Big Data strategy is clear and straightforward. The combination of NoSQL database and Big Data appliance provides customers with the ability to handle unstructured mass data directly. Oracle Big Data Appliance Oracle-sun distributed computing platform and Cloudera Apache Hadoop release, Cloudera Manager management Console, The open source distribution of R analytics software and Oracle NoSQL database are combined. Oracle also includes connectors that allow data to be transmitted back and forth between large data machines and Oracle Exadata or traditional Oracle database deployment environments. Oracle provides a gleam of support for this integrated software and hardware "Engineering integration System", but even with the tricky Hadoop puzzle, Oracle can take advantage of Cloudera's expertise, and it can also introduce clients using Cloudera's Hadoop training and consulting services.

Large data machines are configured with a full rack (full-rack) configuration with 864GB main memory, 216 processor cores, 648TB raw disk storage capacity, and 40,000 gigabit inifiniband internal connections between nodes. The total price of hardware and software will reach 450,000 U.S. dollars, each year to receive 12% of hardware and software support fee. The price is competitive, equivalent to less than 700 dollars per TB.

Microsoft: facing open Suspense

Microsoft launched a beta Hadoop service based on the Azure cloud platform last year, and this year it has pledged to launch a Windows-compliant large data solution based on Hadoop (Big Data Solution), part of Microsoft's SQL Server 2012 version. Microsoft announces two Community Technology Preview Connector components for large data processing based on Hadoop, one for SQL Server and the other for SQL Server Parallel Data Warehouse (PDW). This connector is a command-line tool that is deployed in a Linux environment.

The SQL Server Hadoop connector is the most important step on the way to Microsoft's big data. But because Hadoop, Linux and Sqoop are open source technologies, this means that Microsoft is open to the open source world on a large scale, which deserves the attention of users. In addition, Microsoft has announced the introduction of LINQ Packs, LINQ to HPC, Project "Daytona", and Excel Datascope, which will be designed for researchers and business analysts to make data analysis on Windows Azure.

EMC: A single data analysis platform

Greenplum has 430 million dollars in revenue for large data and is currently owned by EMC. The EMC Greenplum Unified Analysis Platform (UAP) is a single software platform where data teams and analysis teams can seamlessly share information, collaborate on analysis, do not need to work on different islands, or transfer data between different islands. Because of this, UAP includes the ECM Greenplum relational database, EMC Greenplum HD, and EMC Greenplum Chorus, which is a collaborative, social-like interface that can be handled by the data analysis Team, Whether a team member is a Ph. D. data scientist, data integration expert and business intelligence analyst, or a database administrator and business unit user and manager.

EMC's hardware for large data is a modular EMC Data computing Device (DCA) that can run and extend Greenplum relational databases and Greenplum HD nodes within a single device. DCA provides a shared command center interface that allows administrators to monitor, manage, and configure the performance and capacity of Greenplum databases and Hadoop systems. UAP software unifies data access, management, and workflow, and is associated with other data sources and data processing methods, and is expected to increase dramatically as the Hadoop platform matures.

(Responsible editor: The good of the Legacy)