Data storage and data retrieval method and apparatus

By adopting a hybrid storage method of MVCC row and column formats in the distributed data system, and combining it with node cluster optimization for query routing, the problem of high data replication and maintenance costs is solved, and the effect of efficient transaction processing and analysis query is achieved.

CN115398414BActive Publication Date: 2026-07-10HUAWEI CLOUD COMPUTING TECHNOLOGIES CO LTD
View PDF 1 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
HUAWEI CLOUD COMPUTING TECHNOLOGIES CO LTD
Filing Date
2021-02-07
Publication Date
2026-07-10

AI Technical Summary

Technical Problem

Existing technologies suffer from high data replication and maintenance costs, poor real-time performance, and resource waste when handling transactions and analytical queries in distributed data systems. This is especially true when using ETL processes and different storage formats, which leads to data inconsistency and low query efficiency.

Method used

It adopts a hybrid storage method of multi-version concurrency control (MVCC) row format and column format. The ingestion module stores the data in MVCC row format and streams it into MVCC column format and column format based on transformation parameters to maintain data consistency and compatibility. At the same time, node clusters are used to optimize query routing and handle transaction and analysis queries respectively.

Benefits of technology

It enables efficient processing of transactions and analytical queries in a single data storage system, reduces data duplication and management overhead, ensures service level agreements for high-priority tasks, and improves query response speed and resource utilization efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115398414B_ABST
    Figure CN115398414B_ABST
Patent Text Reader

Abstract

In one aspect, data is received to be written in a data store (106) in a first format, wherein the first format is a multi-version concurrency control row format (206). The data is stored in the data store (106) in the first format and converted from the first format to a second format based on first conversion parameters, wherein the second format is a multi-version concurrency control column format (208). The data is stored in the data store (106) in the second format and converted from the second format to a third format based on second conversion parameters, wherein the third format is a column format (210). The data store stores non-overlapping data in the first format, the second format, and the third format. In another aspect, data is retrieved from the data store (106) by a node cluster corresponding to a type of query.
Need to check novelty before this filing date? Find Prior Art