X hits on this document

100 views

0 shares

0 downloads

0 comments

22 / 48

Microsoft’s SAP ERP (R/3) System (cont)

Cached Query Scenario 1:

Table with columns c1 to c90

Table has 1 Mio rows

Clustered primary key index on columns c1 to c5

Nonclustered index on c1, c6 where c6 has two values only:

Value ‘X’ in 999,000 rows and ‘ ‘ in 1,000 rows

Column c1 has one value only: ‘300’

Query issued like: select * from t1 where c1=@P1 and c6 = @P2 where @p1=‘300’ and @P2=‘X’

SQL Server decides Clustered index access and caches Query with this plan

However most of the users would like to search for the rows with the value ‘ ‘  and now experience the slow plan over the clustered index seek instead of a faster plan over the non-clustered index

DBA-416M• ISV configurations & implementations – Lessons Learned

Document info
Document views100
Page views100
Page last viewedTue Dec 06 15:21:08 UTC 2016
Pages48
Paragraphs862
Words4327

Comments