FIX: Replication of Japanese DBCS Character Fails (160146)

SYMPTOMS

When you use SQL Server replication with the code page 932 installed, some replication commands may not be successfully run, resulting in some Japanese characters being mistakenly replicated in the subscribing table.

CAUSE

When SQL Server replicates, the T-SQL command performed is stored in the command column of the MSJob_commands table. If the performed T-SQL command is greater than 255 bytes, the T-SQL command is stored into two or more rows in the MSJob_commands. If the 255th byte of data is a Double-byte Character Set (DBCS) character, SQL Server truncates the lead byte of this double-byte character, and places the trailing byte at the next row. For example, if the double-byte character "8341" is at that critical location, it will be replicated as "2041". This results in " A" appearing in the subscribing table. In some cases, the distribution task fails as well.

STATUS

Microsoft has confirmed this to be a problem in Microsoft SQL Server version 6.5. This problem has been corrected in U.S. Service Pack 2 for Microsoft SQL Server version 6.5. For more information, contact your primary support provider.