All-flash storage array optimized for Big Data

To better support big data applications, Fujitsu has launched an all-flash array and big data all-in-one machine optimized for big data. This ensures the high performance and high reliability of the entire system, this further improves the efficiency of data processing and analysis.

Big Data is another important change that will change traditional business models and IT application methods after cloud computing. From the storage point of view, Fujitsu is gradually turning its product focus to big data. Recently, it launched the latest all-flash storage array eternus dx200f and big data-oriented MHA.

The all-flash array eternus dx200f is an entry-level storage product for small and medium-sized enterprises. Although it is an entry-level product, the performance of eternus dx200f is no inferior to that of high-end storage products. SPC-1 benchmark test data shows that eternus dx200f iops value exceeds 0.2 million, in the case of full load, the delay is only 0.63 Ms.

With the increasing popularity of cloud computing applications, the increasing number of mobile devices, and the rapid growth of massive data, enterprises' data processing becomes increasingly complex. How to efficiently and reliably process massive data, analyze data in a timely manner, and use data effectively is an important topic for enterprises. When Oracle databases are widely used by many enterprises, one of the core values of Fujitsu eternus dx200f all-flash storage array is to improve the access performance of Oracle RDBMS.

Fujitsu eternus dx200f was designed to achieve rapid response and processing for big data. Therefore, eternus dx200f has outstanding performance in terms of performance. Based on Fujitsu's other classic entry-level storage system eternus dx200 S3, Fujitsu has designed eternus dx200f and comes standard with SSD, which can be expanded to up to 24 SSDS. It has multiple high-speed interfaces, including 16 Gb/s FC and 10 Gb/s iSCSI.

Because all SSD is used, eternus dx200f processes extremely fast and has a shorter response time, which can effectively shorten the data batch processing time. Compared with eternus dx200 S3, eternus dx200f reduces power consumption by 90% and saves up to 50% of the occupied space.

In addition to high performance, eternus dx200f can better meet the big data processing needs. For example, eternus dx200f can copy data and supports automatic configuration streamlining. Eternus dx200f adheres to the high-reliability technology widely acclaimed in the eternus DX series, and implements redundancy for controllers, power supplies, and other main components. Even if a fault occurs, you can also change components online without interrupting your business. Used in combination with Fujitsu Storage Management Software eternus SF 16, eternus dx200f can allocate resources based on business priorities to achieve efficient management of the overall storage environment.

Nowadays, big data has been applied in many industries, especially in risk prediction, cost control, and market prediction. With big data, Fujitsu has made breakthroughs in improving social infrastructure and achieving precise marketing for enterprises. To help enterprise users make full use of big data, Fujitsu launched the big data all-in-one MHA. MHA is an integrated big data management and analysis platform for software and hardware. It is ready to use out-of-the-box and has a built-in high security and high availability mechanism. It features high optimization, smooth resizing, and fast scaling.

Big data processing and analysis puts forward new requirements for IT infrastructure. First, better computing capability. The big data IT infrastructure must be capable of distributed horizontal scaling with higher computing density and performance. Second, better storage capabilities. The big data storage platform can better realize storage layering, process structured and unstructured data on a single platform, and achieve unified management. Third, the big data all-in-one machine can achieve simpler integration and provide an all-in-one solution for Infrastructure establishment, optimization, and integration. The Fujitsu big data all-in-one MHA is built based on the above principles. It integrates Fujitsu's innovative, high-performance hardware with distributed file systems, parallel computing systems, and visual analysis software, it provides users with a big data processing platform that is easier to use.

MHA uses hardware optimized for big data, including the master core node, Cluster Expansion node, data storage and archiving platform eternus DX S3, etc, the entire hardware platform has higher reliability and higher computing and storage density. In terms of software, MHA mainly includes two parts: basic data layer software and statistical Mining Layer software. Distributed Data Analysis and mining software belongs to the basic data layer software. The statistical Mining Layer software includes memory data analysis acceleration software, real-time parallel databases, and distributed data backup software.

Currently, the Fujitsu big data all-in-one MHA is mainly used in four fields: Distributed Data Analysis and mining, real-time memory data warehouse, memory data analysis acceleration, and data analysis visualization. In terms of big data, Fujitsu also has good support for hadoop, for example, the Fujitsu big data infrastructure platform based on Apache hadoop 2.0 includes a memory-based hadoop computing, storage appliances, and business-oriented hadoop storage and backup systems.

This article from the "Guo Tao's storage world" blog, please be sure to keep this source http://gtstorageworld.blog.51cto.com/908359/1535634