Supported collation options when installing Microsoft SQL Server Analysis Services

Products

• Information Centric Analytics

Issue/Introduction

The SQL collation specified in the Server Requirements section of the Symantec Information Centric Analytics Administrator Guide is Latin1_General_CP1_CI_AS; however, this collation is not available when installing Analysis Services. The closest match is Latin1_General. Is this collation supported?

Environment

Release : 6.x

Component : Microsoft SQL Server Analysis Services

Resolution

Broadcom supports Analysis Services' Latin1_General collation for use with Information Centric Analytics (ICA). The administrator guide will be updated to clarify this point.

NOTE: This applies only to Analysis Services; Microsoft SQL Server must use the Latin1_General_CP1_CI_AS collation.

Additional Information

Per Microsoft:

Analysis Services supports two collation types:

• Windows collations (versions _90 and _100)
  
  Windows collation versions are _90 (an unmarked older version) and the newer _100 version. Only the _100 version shows the version number in the collation name:
  • latin1_general
  • latin1_general_100

A Windows collation sorts characters based on the linguistic and cultural characteristics of the language. In Windows, collations outnumber the locales (or languages) used with them, due to many languages sharing common alphabets and rules for sorting and comparing characters. For example, 33 Windows locales, including all the Portuguese and English Windows locales, use the Latin1 code page (1252) and follow a common set of rules for sorting and comparing characters.

• Binary collations (either BIN or BIN2)
  
  Binary collations sort on Unicode code points, not on linguistic values. For example, Latin1_General_BIN and Japanese_BIN yield identical sorting results when used on Unicode data. Whereas a linguistic sort might yield results like aAbBcCdD, a binary sort would be ABCDabcd because the code point of all uppercase characters is collectively higher than the code points of the lowercase characters.