Rapids Hadoop

Rapids Data one-stop big data real-time analysis platform comprises five types of application products

Provide a full range of real-time big data support for your business: massive data analysis, second-level execution response, real-time analysis and processing, security and autonomous control

Rapids DB Cloud Image

Rapids Hadoop Image

Rapids Stream DB Image

Rapids Parallel R

Rapids DBaas Cloud Image

Rapids x86 Image

RapidsDB
Rapids Hadoop
Rapids StreamDB
Rapids ParallelR
Rapids DBaas Cloud
x86 All-In-One Machine

RapidsDB - standard-based, all-memory big data analysis processing platform

Big data processing requests can be distributed intelligently to high-speed memory processing clusters for computing and real-time analysis and processing. Data that requires high-speed processing can be quickly loaded asynchronously into the processing cluster memory, which can be easily extended to hundreds of physical nodes. A federated analytic platform unique in the industry for data analysis and processing of any data source with industry-standard SQL and JDBC interfaces to centrally provide platform-level assurance for big data while dispensing with unnecessary data migration.

  • Support cross-partition multi-table correlation analysis
  • Support cross-window correlation analysis
  • Support TPC-H and TPC-DS levels of complicated analysis and inquiries
  • Support SQL2013 and JDBC2.0
  • Seamless realization of high-performance SQL-on-Hadoop

Rapids Hadoop

Seamless realization of high-performance SQL-On-Hadoop.

Through the federated analysis engine HDFS Connector, support extraction of CSV format data from Hadoop in the form of data streaming for real-time correlation analysis in Rapids DB.

  • Support complex data analysis and processing of Hadoop data sources with industry-standard SQL and JDBC interfaces
  • Preset Hadoop configuration optimization parameters that promote high-speed deployment of super massive clusters
  • Support consolidation, monitoring and management of unified Rapids DB and Hadoop data flow

Rapids StreamDB - standard-based, all-memory big-data analysis processing platform

Very high-frequency data streaming (real-time big data) can be distributed to high-speed memory processing clusters to achieve real-time analysis and processing, breaking through the performance bottleneck of the traditional framework of storage first before analysis.

An industry leader in using industry-standard SQL and JDBC interfaces, unified JDBC interfaces to deal with streaming data, greatly reducing the development costs and expanding the possibility of streaming data-based industry cooperation.

  • Support streaming data and quasi - static data for SQL standard-based correlation analysis and processing
  • Supports cross-flow data window correlation analysis
  • Support cross-streaming data window correlation analysis

Rapids ParallelR- big data real -time machine learning algorithm library

R is the industry’s most popular machine learning statistical analysis environment (https://www.r-project.org/). By transplanting and embedding R into Rapids DB, RAPIDS PRALLELR provides smooth integration of the R language and SQL-based development model and the R operating environment in the Rapids cluster within the distributed R computing framework.

Support for over 20 specific algorithms of the following types.

  • Statistical analysis
  • Integration algorithm
  • In-depth neural network algorithm
  • Reducing latitude
  • Data feature analysis, precision marketing, production quality control

Rapids DBaaS Cloud System Architecture

Rapids DBaas Cloud is composed of a series of physical hosts in a cloud-based architecture to form an operational resource pool that is centrally managed and regulated by the Rapids Management Server which processes and verifies it. Once the request is approved, it is sent to a member server in the pool of computing resource operations in order to create a cluster for the user to use.

  • Fast and standardized deployment
  • Improved installation, configuration and deployment efficiency of large-scale cluster
  • Easy to expand
  • Improved equipment resource utilization
  • Polling of hardware resources enables applications to maximize the use of available hardware resources;
  • The introduction of life cycle management enables timely recovery of idled resources.

x86-based all-in-one machine – low development and operating costs

Support industry-standard x86 servers in the industry, which greatly reduces development costs and subsequent operating costs of high-performance operations.

Teaming up with the industry-leading server manufacturers, Sugon, Inspur and Taiji, to deliver systems that have the Rapids Data Platform and Rapids Hadoop pre-installed to provide all-in-one memory and storage-based machines.

  • Promised Performance Metrics