Intel® 2020 Select Solution Based on GBase 8a MPP Cluster

Published on 2020-08-21

As digital transformation deepens, industries such as finance, telecommunications, energy, and government are experiencing rapid growth in stored data. Extracting value from this massive data has become a key direction for digital initiatives. However, massive data places heavy performance pressure on data processing, calling for continuous innovation in database technology.

GBase 8a MPP Cluster, developed by General Data Technology Co., Ltd. (hereinafter referred to as GBase), is designed to support structured big data processing. With its unique flat architecture, high availability, and dynamic scalability, GBase 8a MPP Cluster offers a cost-effective, large-scale distributed parallel database management solution for ultra-large data management.

The Intel® Select Solution based on GBase 8a MPP Cluster utilizes Intel® Xeon® Scalable processors, Intel® SSD D3-S4510 Series, and Intel® Ethernet Adapter X710-DA2 to significantly enhance the performance of GBase 8a MPP Cluster, fully meeting the big data analytics needs of data-intensive industries.

Zhang Shaoyong, Director of Data Analysis Product Department at GBase, attended

GBase 8a MPP Cluster Provides Critical Support for Industry Big Data

In the big data era, industries accumulate massive amounts of structured data throughout long-term business operations, and this data continues to grow rapidly. Typical structured data includes banking and telecom transaction data, transaction data from traditional data centers, management data, financial data, foundational information data, as well as emerging IoT and internet data. For these rapidly expanding databases, improving data processing performance is crucial for ensuring business continuity and unlocking data value.

For structured data processing, GBase 8a MPP Cluster offers significant performance advantages over traditional databases. It can support datasets ranging from terabytes to tens of petabytes, with concurrent users exceeding 300. IO wait time can be as low as 10% of traditional databases, and data analysis speed can be over 10 times faster than ordinary databases. In terms of scalability, GBase 8a MPP Cluster uses local resources on each server and, based on a node-to-node flat architecture, provides excellent horizontal scalability. Cluster computing performance and storage capacity increase nearly linearly with cluster expansion, and users do not need to stop services during the entire expansion process.

GBase 8a MPP Cluster also features a new columnar storage engine, extremely high data storage compression ratios, maintenance-free coarse-grained indexes, and many other big data processing technologies and features. Combined with MPP's efficient distributed computing model and a cost-based distributed intelligent optimizer, GBase 8a MPP Cluster can support petabyte-scale structured data analysis applications. Additionally, GBase 8a MPP Cluster utilizes intra-cluster replica synchronization technology, active-active cluster technology, cross-domain cluster data transfer synchronization, and virtual cluster technology to ensure high cluster availability and multi-cluster support under big data conditions, enabling multi-data center deployments.

GBase 8a MPP Cluster meets the structured data processing needs of data-intensive industries such as finance, telecommunications, energy, government, and IoT. It supports applications including data query, statistics, analysis, mining, and backup, while also serving as the underlying database for data warehouse systems, BI systems, and decision support systems.

Intel® Select Solution Based on GBase 8a MPP Cluster

The Intel® Select Solution based on GBase 8a MPP Cluster helps data-intensive industry users improve data analytics performance while significantly reducing application response time. This solution fully leverages the advantages of Intel® Xeon® Scalable processors, Intel® SSD D3-S4510 Series, and Intel® Ethernet Adapter X710-DA2, delivering a high-performance, highly scalable, and highly available database solution. Specifically, the solution offers the following features:

● By configuring selected Intel® Xeon® Scalable processors, the potential of GBase 8a MPP Cluster can be fully realized, with excellent total cost of ownership (TCO). Industry users can build GBase 8a MPP Cluster solutions using the verified and tested Intel® Xeon® Scalable processors.

● The solution covers both software and hardware, and has undergone thorough compatibility and stability verification, helping industry users significantly reduce upfront hardware selection and testing costs and achieve rapid system deployment.

● For enterprises needing to improve the flexibility of large-scale distributed database systems, this solution provides a detailed reference architecture. Even if the cluster involves hardware from multiple brands or requires hardware vendor changes during future expansion, customers can still freely scale and choose based on this configuration architecture.

Hardware Selection

The Intel® Select Solution based on GBase 8a MPP Cluster combines Intel® Xeon® Scalable processors, Intel® SSD D3-S4510 Series, and Intel® Ethernet Adapter X710-DA2. Industry users can use this foundation to rapidly deploy high-performance GBase 8a MPP Cluster distributed analytical database systems.

● Intel® Xeon® Scalable processors: Intel® Xeon® Scalable processors feature new capabilities including Intel® Advanced Vector Extensions 512 (Intel® AVX-512), delivering workload-optimized performance for advanced analytics, high-performance computing (HPC) applications, and data compression to improve throughput. These breakthrough technologies bring cost-effective, large-capacity persistent memory capabilities to Intel's data-centric computing portfolio, significantly enhancing high-speed data processing for critical computing applications such as databases and data warehouses.

● Intel® SSD D3-S4510 Series: Based on 64-layer Intel® 3D NAND TLC media, the Intel® SSD D3-S4510 Series significantly outperforms traditional mechanical drives. In GBase 8a MPP Cluster deployments, it serves as a storage medium, controlling costs while satisfying performance demands. Moreover, its flexible workload capability allows it to cover a wider range of workloads, offering flexible capacity, durability, and energy efficiency.

● Intel® Ethernet Adapter X710-DA2: The Intel® Ethernet 710 Series accelerates the performance of the Intel® Select Solution based on GBase 8a MPP Cluster. The 710 Series provides professional 10/25GbE performance, supporting single or dual 1/10/25GbE port connections, delivering outstanding performance in a PCI Express v3.0 x8 slot, and supporting various advanced features such as Virtual Machine Device Queues (VMDq) and Single Root I/O Virtualization (SR-IOV). Additionally, compared to 40G network adapters, the 710 Series effectively reduces costs while meeting bandwidth requirements, demonstrating greater economic and practical value.

Performance Validated through Benchmark Testing

The Intel® Select Solution based on GBase 8a MPP Cluster offers two hardware configuration recommendations, "Base" and "Plus," catering to customers prioritizing optimal cost and optimal performance respectively, with a consistent software recommendation (configuration list in Appendix 1). We recommend users adopting the Intel® Select Solution to use these recommended configurations or higher ones.

The Intel® Select Solution based on GBase 8a MPP Cluster has passed benchmark performance validation testing and achieved workload-optimized performance metrics. The performance tests for this solution used the Decision Analysis* business intelligence computing test and the Decision Support System* benchmark. The Decision Analysis test is commonly used to evaluate decision support technology application performance, providing a comprehensive assessment of overall commercial computing capability. The Decision Support System test measures query response time under a single-user model, query throughput under a multi-user model, and data maintenance performance in more complex multi-user decision-making scenarios.

To test the performance difference between the "Base" and "Plus" configurations of the Intel® Select Solution based on GBase 8a MPP Cluster, both Decision Analysis* and Decision Support System* were used to test their query and load times (shorter is better). The test results are shown in Table 1.

Table 1: "Base" vs. "Plus" Configuration Test

Test results show that in the Decision Analysis test, the "Plus" configuration achieved over 200% performance improvement compared to the "Base" configuration. In the Decision Support System test, the improvement was greater than 150%. The "Base" configuration represents the baseline hardware required to meet the general performance requirements of the Intel® Select Solution based on GBase 8a MPP Cluster, while the "Plus" configuration enables system integrators, solution providers, and industry users to achieve higher performance in GBase 8a MPP Cluster deployments. Industry users can choose the GBase 8a MPP Cluster configuration that best suits their business needs based on actual workloads and requirements.

Technologies within the Intel® Select Solution Based on GBase 8a MPP Cluster

Beyond the Intel hardware foundation for the Intel® Select Solution based on GBase 8a MPP Cluster, integrated Intel technologies within Intel® Xeon® Scalable processors further enhance performance and reliability:

● Intel® Platform Trust Technology (Intel® PTT): This technology integrates directly into the chipset's Trusted Platform Module (TPM) to protect encryption keys. The new Intel® PTT feature is offered as an option rather than as a discrete chip, simplifying integration and activation.

● Intel® Hyper-Threading Technology (Intel® HT Technology): Enables more efficient use of processor resources by allowing each core to run multiple threads. As a feature, it also improves processor throughput and overall performance of threaded software, maintaining system responsiveness while running multiple demanding applications.

● Intel® Turbo Boost Technology: For peak demands, Intel® Turbo Boost Technology allows the processor to operate above its base frequency, accelerating processor and graphics performance to handle peak loads easily.

● Intel® Intelligent Power Technology: Allows the processor to quickly select its optimal operating frequency and voltage for best performance and power efficiency without operating system intervention.

The Intel® Select Solution Based on GBase 8a MPP Cluster Meets Massive Data Processing Needs of Industry Users

The Intel® Select Solution based on GBase 8a MPP Cluster has been verified and optimized, delivering workload-optimized configurations for Intel® Xeon® Scalable processors. When industry users need to deploy GBase 8a MPP Cluster solutions, they can avoid spending extensive time on debugging and performance optimization by using the Intel® Select Solution, gaining pre-tuned and tested configurations to meet massive data storage and processing requirements, ensure business continuity, and lay a solid foundation for digital transformation.

Appendix 1: Intel® Select Solution Based on GBase 8a MPP Cluster: Recommended Configuration List

Configuration 1 - Base: Testing conducted by Intel and GBase as of May 30, 2019. 4 nodes, Dual Intel® Xeon® Gold 5118 processors, 12 cores, HT enabled, Turbo Boost enabled, total memory 192 GB (12 slots/ 16GB/ 2400 MHz), 1x 480 GB Intel® SSD DC S4500, 12x 1TB SATA 10000RPM HDD, Microcode:0x200005e, CentOS 7.4, 3.10.0-957.12.2.el7.x86_64, GBase 8a MPP Cluster 8.6.2.

Configuration 2 - Plus: Testing conducted by Intel and GBase as of May 30, 2019. 4 nodes, Dual Intel® Xeon® Gold 6140 processors, 18 cores, HT enabled, Turbo Boost enabled, total memory 192 GB (12 slots/ 16GB/ 2400 MHz), 1x 480 GB Intel® SSD DC S4500, 12x 960 GB Intel® SSD DC S4510, Microcode:0x200005e, CentOS 7.4, 3.10.0-957.12.2.el7.x86_64, GBase 8a MPP Cluster 8.6.2.

Specific component performance was tested in specific systems. Any difference in hardware, software, or configuration may affect actual performance. Consider other sources of information to evaluate performance when considering a purchase. For more complete information about performance and benchmark results, please visit http://www.intel.cn/content/www/cn/zh/benchmarks/benchmark.html.

The software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests such as SYSmark* and MobileMark* are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more information, please visit http://www.intel.cn/content/www/cn/zh/benchmarks/benchmark.html.

The cost reduction scenarios described are intended as examples of how a specified Intel-architecture-based product, in the specified circumstances and configurations, may affect future costs and provide cost savings. Circumstances will vary. Intel does not guarantee any costs or cost savings.

Intel technologies' features and benefits depend on system configuration and may require enabled hardware, software or service activation. Performance varies depending on system configuration. No computer system can be absolutely secure. Check with your system manufacturer or retailer or learn more at intel.com.

Intel disclaims all express and implied warranties, including without limitation, the implied warranties of merchantability, fitness for a particular purpose, and non-infringement, as well as any warranty arising from course of performance, course of dealing, or usage of trade.

Intel, the Intel logo, Xeon, and Xeon are trademarks of Intel Corporation in the U.S. and/or other countries.

*Other names and brands may be claimed as the property of others.