HomeMogDBMogDB StackUqbar
v5.0

Documentation:v5.0

Supported Versions:

Other Versions:

Data Compression in OLTP Scenarios

Availability

This feature is available since MogDB 3.0.0.

Introduction

The feature supports row-store data compression in OLTP scenarios, provides a general compression algorithm, and implements transparent compression of data pages and maintenance of page storage locations to achieve high compression and high performance. Disk persistence is implemented using two types of files: compressed address file (with the file name extension .pca) and compressed data file (with the file name extension .pcd).

Benefits

Typically, it is applicable where the database disk space needs to be reduced.

Description

Data compression in OLTP scenarios can reduce the disk storage space of row tables and index data and improve performance in I/O-intensive database systems.

Constraints

  • Only heap-organized data table compression is supported, i.e., normal row-storage table, Btree index compression.
  • The operating system must support punch hole operations.
  • The data backup media must support punch hole operation.
  • Do not support the modification of compression-related parameters. Do not support the conversion of uncompressed tables into compressed tables.
  • Compression and decompression operations will have a certain impact on CPU and performance. The advantage is that it increases the storage capacity of the disk, improves disk utilization, and saves disk IO and reduces disk IO pressure.

Dependencies

  • Requires the database to support double write operations.
  • Compression using open source compression algorithms PGLZ, ZSTD.

Parameters Related to Efficient Data Compression Algorithms

Copyright © 2011-2024 www.enmotech.com All rights reserved.