Provincial tax big data platform localization replacement

Launch the Provincial Tax Board's CDH localization alternative program in 2022;

With the increase of user usage and business scenarios of the provincial tax big data platform, more diversified requirements have been put forward for the capability of the big data platform, with higher and higher requirements for data timeliness, more and more demands for high concurrency, and more and more scenarios for exploration and practice. The increasing demand for platform resources and functions has led to a strain on the computing power of big data platforms.

Through the construction of the big data center of the provincial Tax Bureau, the deployment of the big data platform and analysis tool set, the formation of a safe and reliable, efficient operation, complete data and stable operation of the big data support platform;

Fully compatible with CDH, it can realize the smooth migration of existing production and meet the requirements of localization;

USDP has a wealth of big data ecological technology, and the platform has a good storage and computing power expansion ability;

Good smooth upgrade ability, effectively support the evolution of tax bureau from data warehouse architecture to lake warehouse architecture;

Compatible with Atlas data collection and analysis, providing ecological and technical support for dolphin scheduling;

From closed ecological technology to ecological open source compatibility, explore and practice the flexibility of new technology to business;

Kerberos, Ranger, and LDAP security components are used to improve the security protection capability of big data services.

Improve the tax bureau multi-source heterogeneous data collection and unified storage capabilities;

The quasi-real-time data warehouse is constructed by means of pre-calculation and pre-processing.

Continuously optimize and improve Impala query efficiency to support the province's tax analysis query needs;

Unified management of blood relationship between Hadoop and non-Hadoop data;

Perfect the data directory structure, specify the data specification, and increase the data unified management;

Data interconnection between Iceberg and tax related systems is realized.

Take full advantage of Hudi time travel to assist data lifecycle management;