Big Data Servers
Huawei FusionCube BigData Machine is a hardware platform to accelerate big data business, which can be seamlessly connected with mainstream big data platforms. Based on Huawei’s innovated acceleration technologies, the BigData Machine provides the high-density data storage solution, Spark real-time analysis acceleration solution. These solutions are widely applied in public security, precision marketing, risk control, and real-time O&M, helping customers reduce TCO.
Better compatibility, excellent performance, easy deployment, and reduced TCO
- Better compatibility: Supports HortonWorks – Cloudera – FusionInsight – OpenSource Big Data platforms.
- Full solution: Provides Hadoop – Spark acceleration solutions.
- Excellent performance: Saves 30% on data node servers, Spark real-time analysis speed increased 40%.
Pain points of current High Density Storage solutions
- PB-level massive cold data archiving and storage occupies huge disk space at a high cost.
- Compressing data by the built-in soft compression algorithm and storing data consumes plenty of CPU resources with a low compression ratio.
Huawei solution to the problem
- Dedicated hardware compression releases 30% CPU resources.
- The algorithm with a high compression ratio improves data node storage capacity by 30%.
Pain points of current Spark Real-Time Analysis Acceleration solutions
- During the Shuffle process, temporary data and memory overflow data must be exchanged in the HDFS file system.
- The performance of HDFS, which is made up by HDDs, affects real-time analysis speed. Huawei solution benefits
Huawei solution to the problem
- Real-time analysis is accelerated by 40%.