GB/T 38673-2020 in English
VALIDInformation technology -- Big data -- Basic requirements for big data systems
- Issued on:2020-04-28
- Implemented on:2020-11-01
- File Format:PDF
- Delivery:Via email within 1~3 business days
$194.00
《GB/T 38673-2020信息技术 大数据 大数据系统基本要求》由TC28（全国信息技术标准化技术委员会）归口，主管部门为国家标准化管理委员会。
Introduction
Analysis of the core framework of the standard
GB/T38673-2020 has built a reference framework for a big data system that includes nine functional modules, including data collection module and preprocessing module. The technical requirements cover the entire life cycle of the system. For the first time, the standard clearly defines non-functional requirements as a separate chapter, reflecting the comprehensive control of system quality.
Comparative Analysis of Functional Modules
| Module | Core Requirements | Technical Implementation | Typical Solutions |
|---|---|---|---|
| Data Storage | Support Three Data Structures + Four Storage Modes | Distributed File/Columnar/Structured/Graph Storage | HDFS/HBase/Neo4j |
| Data Processing | Batch Processing + Stream Processing + Graph Computing Fusion | DAG Execution Engine + Sliding Window | Spark/Flink/GraphX |
| Data Analysis | 7 types of analytical capabilities + full process of machine learning | OLAP+ad hoc query+algorithm library | Presto/TensorFlow |
Evolution of key technologies
The standard reflects three major technology trends: batch-stream fusion architecture (6.4.e), multi-mode storage engine (6.3) and dynamic resource scheduling (6.8.g). Taking the financial risk control scenario as an example, after a bank adopted the Lambda architecture recommended by the standard, the delay of real-time transaction analysis was reduced from minutes to seconds.
Implementation suggestions
- Phase-based construction: Prioritize meeting the "should" level requirements of basic modules such as data collection, storage, and processing
- Security compliance: Pay special attention to the data encryption requirements in Article 7.3.4, and it is recommended to use the national encryption algorithm
- Performance optimization: Establish a hierarchical indexing strategy according to Article 6.5.a.3
Industry application cases
Smart city project: A big data platform built by a municipality based on this standard processes an average of 10PB of IoT data per day, realizes multi-department data integration through the data loading mechanism in Article 6.2.d, meets the high availability requirements in Article 7.1.1, and the annual system failure time is <15 minutes.
Loading PDF document...
Error loading PDF. Please make sure the file is valid and try again.
We also recommend
-
GB/Z 191-2026 in English
Intelligent computing—Technical framework of privacy-preserving computation
2026-07-02 -
GB/T 20538.2-2023 in English
Business data and process of electronic commerce eXtensible Markup Language(ebXML)—Part 2:Business information entity directory
2023-09-07 -
GB/T 44272-2024 in English
Information technology—Open source—Open source license framework
2024-08-23 -
GB/T 45181-2024 in English
Security-related misbehavior detection mechanism for connected vehicles
2024-12-31 -
GB/T 46883-2025 in English
Industrial internet platform—Specification for digital service of industrial park
2025-12-31 -
GB/Z 20539-2006 in English
Modeling guide for business process and information in e-Business
2006-09-18 -
GB/T 38633-2020 in English
Information technology—Big data—Functional requirements for system operation and management
2020-04-28
