Article ID: 251094 - View products that this article applies to.
This article was previously published under Q251094
This article has been archived. It is offered "as is" and will no longer be updated.
BUG #: 57007 (SQLBUG_70)
When updating existing statistics using sampling, two things may happen:
When updating existing statistics for non-unique clustered indexes on variable-length columns using sampling, either default or explicit, with a sample size of less than 100%, statistics may become inconsistent. This problem may also occur if SQL Server auto-updates the statistics on this index because the Auto Update Statistics database option is set for that database.
The workaround is to implement one or more of the following:
Microsoft has confirmed this to be a problem in SQL Server 7.0. This problem has been corrected in U.S. Service Pack 2 for Microsoft SQL Server 7.0. For more information, click the following article number to view the article in the Microsoft Knowledge Base:
254561For more information, contact your primary support provider.
(http://support.microsoft.com/kb/254561/ )INF: How to Obtain Service Pack 2 for Microsoft SQL Server 7.0 and Microsoft Data Engine (MSDE) 1.0
For example, consider the following table with a clustered index:
The statement that causes inconsistent data to be generated is:
REFERENCESFor information on how to disable automatic updating of statistics for a specific index, see the Books Online topic sp_autostats (T-SQL).