Why HUAWEI CLOUD?

Heterogeneous data migration

Huawei Cloud provides services that can migrate nearly 20 types of data sources between on-premises data centers and the cloud. It also provides a wizard-based GUI that frees developers from complex programming and helps them quickly create migration tasks.

Easy spin-up and management of big data clusters

Huawei Cloud provides Hadoop-based high-performance big data components, such as Hudi, ClickHouse, Hive, Spark, Flink, Kafka, and HBase, for data lakes, data warehouses, and BI and AI applications.

Simplified big data operations platform

Huawei Cloud provides a one-stop data operations platform that manages data modeling, data development, job scheduling, data quality, data assets, and the full data lifecycle.

Challenges

As a company that pursues innovation, Dadi Cinema Group was determined to build a state-of-the-art financial settlement system in 2022 to secure payment transactions between their accounting department and around 300 marketing channels that can be reconciled and matched in T+1 manner. The heterogeneous data structures are much more complex than what conventional databases can handle, so data must be converged to a centralized data lake sustainably and analysed in data warehouses dynamically.

Solutions

  • DataArts Studio
  • DataArts Studio provides ER modeling and dimensional modeling methods for building hierarchical data warehouses, so that financial reconciliation, journal entry, order assembling, order splitting, and difference checking tables can be modeled, manipulated, and queried effectively.

  • MapReduce Service (MRS)
  • MRS provides a distributed analytics framework with high-profile cluster nodes, hardened metadata stores, HAed job history servers, and scenario-based task flows with sorting disabled. MRS enhances parallel computing power and accelerates the processing of large datasets.

  • CDM
  • CDM provides the pre-configured JDBC connector for relational databases or data warehouses and native big data connectors for the Hadoop ecosystem and data lake. With the connectors, either database tables or incremental files from various data sources can be integrated to the data lake timely and efficiently.

Customer Benefits

Benefits

By building a data platform from scratch, Dadi Cinema Group realized automatic reconciliation in their settlement system and increased the number of third-party channels by 265%, saving about 22,965 labor hours each year.

116 original data tables, 33 lookup tables, and 93 data standards were created as standardized and unified data assets. All the assets are organized into a four-layer architecture that complies with the enterprise business operations.

Tens of millions of orders can be processed by sophisticated data models within 3 hours every day. Processing results are exported as accurate and timely data services for downstream systems such as BI data analysis, data reports, and a wide range of applications.