Teradata 16 New Features

Teradata 16 New Features

Brief overview of some of the new features added in Teradata 16 version.

➠ Two new string functions were added

• PIVOT ~ Transpose Rows to Columns
• UNPIVOT ~ Transpose Columns to Rows

  Visit PIVOT page and UNPIVOT page to check syntax and examples for above functions.



➠ New DATASET datatype: DATASET datatype is a new Complex Datatype (CDT) representing self-describing files interpreted based on a schema.

• The DATASET datatype allows variable types of storage formats but as of now only Avro is supported for Release 16.00.
  
• DATASET datatype can hold documents of size up to 16 MB.
  
• Functions are available to process, shredd, and publish data from DATASET datatype.

➠ COUNT DISTINCT Performance Improvement: Queries with COUNT DISTINCT are now processed more efficiently by Teradata.

• Benefits:
  
  • The primary benefit of this feature is an improvement in the performance of queries that use COUNT DISTINCT.

➠ Global Space Accounting: The space accounting in Teradata has been enhanced to manage space limits at a global level, rather than AMP-level limit. This enhancement is referred as global space accounting.

• System limits are now soft limits rather than hard limits(equal space allocated to each AMP) in earlier version.
  
• The soft limit is specified as a percentage and can be set at the system level.
  
• The soft limit setting applies to all types of space namely permanent, spool, and temporary space.
  
• Attribute 'Skew limit percentage' can be added to database or user that allows maximum AMP space usage to exceed the per-AMP quota.

➠ Unicode Pass Through: This is a Unicode error handling enhancement. The feature help users to allow Pass Through Characters (PTCs) to be imported into and exported from Teradata.

• PTCs are the set of all characters which are not supported by Teradata.
  
• The feature can be enabled or disabled at the session level and is valid only for Unicode sessions. It can be enabled/disabled using below command.

• SET SESSION CHARACTER SET UNICODE PASS THROUGH [ON|OFF]
  

  
  
• Benefits:
  
  • Increases the range of characters available to Unicode users from 61K to 1,112K including Emoji and other ideographs that could not be stored in Teradata previously.
    
  • Reduces or eliminates cases where translation errors require users to cleanse their Unicode data before submitting it to Teradata again.
    
  • Client load filtering is simplified which improves load performance.
    
  • Data is not lost as would be the case when replacement characters are used to replace unsupported characters.
    
    
  • Restrictions:
    
  • Legacy application such as Fastload and Multiload cannot be used load or unload data containing PTCs. Instead users should use Teradata Parallel Transporter(TPT) for this purpose.
    
  • This feature allows Teradata to store characters that were not supported previously. Therefore, data on a Teradata system with this feature enabled cannot be transported to a Teradata system with earlier versions.

➠ 1 MB PERM and Response Rows: The maximum row size of permanent and response spool rows is increased to 1 MB on large-cylinder systems.

• Benefits:
  • Allows application to handle larger data sets, including larger SELECT results.
  
  

➠ Compression Table-Level Option (Compression TLA): More choices has been added to define data block compression for tables, join indexes & hash indexes.

• Since these are table level attributes therefore these option take precedence over system level settings.

  New choice includes following

• ALWAYS value for the BLOCKCOMPRESSION option
• BLOCKCOMPRESSIONALGORITHM option
• BLOCKCOMPRESSIONLEVEL option

Source Teradata Site: Check Teradata 16 Release Summary for more detail.