http://www.sqler.com/137190#0
INSERT SELECT :
non-FULL recovery model
AND not replicated
AND (
(Heap AND TABLOCK)
OR (B-tree AND empty AND TABLOCK)
OR (B-tree AND empty AND TF-610)
OR (B-tree AND nonempty AND TF-610 AND new key-range)
)
With this trace flag on, insertions into a B-tree
don’t have to use the TABLOCK hint to benefit from
minimal logging. Also, insertions into nonempty
B-trees can be done in minimally logged mode—at
least partially.
1) TF-610 옵션 설정 방법
DBCC TRACEON(610, -1);
2) non-FULL recovery Model
SIMPLE recovery model, Bulk_LOGGED
3) B-tree 의미
clustered or nonclustered index
INSERT SELECT :
non-FULL recovery model
AND not replicated
AND (
(Heap AND TABLOCK)
OR (B-tree AND empty AND TABLOCK)
OR (B-tree AND empty AND TF-610)
OR (B-tree AND nonempty AND TF-610 AND new key-range)
)
With this trace flag on, insertions into a B-tree
don’t have to use the TABLOCK hint to benefit from
minimal logging. Also, insertions into nonempty
B-trees can be done in minimally logged mode—at
least partially.
1) TF-610 옵션 설정 방법
DBCC TRACEON(610, -1);
2) non-FULL recovery Model
SIMPLE recovery model, Bulk_LOGGED
3) B-tree 의미
clustered or nonclustered index
'SQL Server 2008' 카테고리의 다른 글
MSSQL의 잘못된 로우수 해결 방법 (0) | 2010.04.22 |
---|---|
T-SQL Debugger is back in SQL Server 2008 ... debugging now made easy (0) | 2010.04.08 |
SQL2008의 새로운 기능 - 필터링된 인덱스 (0) | 2010.01.13 |
How to Read the Results of SELECT @@VERSION in SQL Server (0) | 2010.01.13 |
SQL2008의 새로운 기능 - 압축편 (1) | 2010.01.12 |