FusionInsight is a unified platform for enterprise-level big data storage, query, and analysis with enhanced functions based on Apache open source community software, launched by Huawei for customers in many industries. It takes massive data processing engines and real-time data processing engines as the core, and creates agile, intelligent and trusted platform software and modeling middleware for the operation, maintenance, application development and other needs of data-intensive industries such as finance and operators. And OM system, allowing enterprises to discover new value points and business opportunities from various kinds of complex and disorderly massive data faster, more accurately and more stably.
- Agility: A comprehensively optimized big data platform that realizes rapid processing of massive data.
With an open architecture, it supports business needs such as batch processing, micro batch processing, and real-time processing. At the same time, comprehensive optimization of data processing, business development, business transplantation, etc., can easily realize rapid business transplantation and rapid data processing. - Wisdom: Modeling framework with over one million dimensions, accurate insight into user behavior and characteristics,
support for data modeling of user data with over one million dimensions, deep insight into user behavior and characteristics, business decisions for enterprises, and the development of new business opportunities. Mining to provide more scientific and comprehensive business support. - Trusted: Financial-level trusted solutions, which effectively support enterprise-level core business data processing
. Designed and developed according to the requirements of high reliability, high stability and high security of products in data-intensive industries such as finance and operators, and can effectively Support the processing and analysis of enterprise-level core business data. -
Classification
Indicator item
Specification
illustrate
Parallel computing engine (MapRdeuce) performance indicators
WordCount: Average processing power per node (GB/min)
8GB/min
12-node cluster;
typical configuration of nodes:
CPU: 2*E5-2650
Memory: 128G
Hard disk: SATA diskTerasort: Average processing power per node (GB/min)
6GB/min
Parallel computing engine (Ss
park) performance indicatorsWordCount: Average processing power per node (GB/min)
27GB/min/Node
Terasort: Average processing power per node (GB/min)
6GB/min/Node
Hive Performance Metrics
Processing Power - HiveAggregation: Average processing power per node (GB/min)
8GB/min
Processing Power - HiveJoin: Average processing power per node (GB/min)
2GB/min
HBase performance metrics
100% random read: the average number of records read per node (1KB per record), and the response time is less than 50MS
30,000 Records/s
100% random write: the average number of records written per node (1KB per record), and the response time is less than 50MS
37,000 Records/s
Sequential scan: the average number of scan records per node (1KB per record), and the response time is less than 50MS
10,000 Records/s