Memory Consuming Features | Teradata Vantage - 17.00 - 17.05 - About Memory-Consuming Features - Advanced SQL Engine - Teradata Database

Teradata Vantage™ - Database Administration

Product
Advanced SQL Engine
Teradata Database
Release Number
17.00
17.05
Published
June 2020
Language
English (United States)
Last Update
2021-01-22
dita:mapPath
rgu1556127906220.ditamap
dita:ditavalPath
lze1555437562152.ditaval

Certain features may require more memory in order to show their optimal performance benefit. Of particular mention are:

  • Array INSERT
  • Compression
  • Cylinder read
  • Geospatial data type
  • Global and Persistent Data (GLOP)
  • Hash joins and product joins
  • Join index
  • Large objects (LOBs)
  • Row and column partitioning
  • Table functions
  • Stored procedures
  • User-defined functions (UDFs)
  • XML data type
  • 1 MB response buffer
  • 1 MB data blocks
  • Larger than 1 MB plan cache

Performance gains from implementing these features may be countered by their impact on memory usage. In turn, you may experience more segment swaps and incur additional swap physical disk I/O. To counter this, you can lower the FSG cache percent to ensure that 135 MB per AMP is allocated in OS memory.

Lowering the FSG cache percent may cause fewer cache hits on table data and instead cause a different type of additional physical disk I/O. In general, additional I/O on table data is a lesser performance issue than swapping I/O, but it can still have an impact on performance.