October 14, 2008 at 11:18 am
Jesse,
I ran the SQL - I have one entry with a score of 119,567. It is displaying equality_columns of StrategyCode and HistoryDate on one table - no inequality columns - so I'd create an index on StrategyCode, HistoryDate?
I have another table where the sole entry is under inequality_columns - RRCode - so I'd make an index on RRCode?
I have a third table where I have TransactionAction, TransactionEntityType under equality_columns and ProcessStatusFlag under inequality_columns.
Is this telling me to make 2 indexes - or 1 index of TransactionAction, TransactionEntityType and ProcessStatusFlag?
TIA,
Doug
November 10, 2008 at 8:42 am
Yes for the first two:
StrategyCode, HistoryDate
RRCode
3rd:
one index on TransactionAction, TrasactionEntityType,PrcessStatusFlag
Don't neglect the includes
The missing index views don't account for clustered index inheritance. Clustered index seek columns are appended to the end as seeks to every non-unique nonclustered index, and as includes to every unique nonclustered index.
SQL Server will never seek any further keys on and index beyond the one it has to seek on an inequality operation. So put all the equalities first, and then the inequalities. Preferabbly find the one most-used (unless some other queries seek it as an equality) and relegate the rest to includes since they'll never get used.
November 10, 2008 at 8:53 am
Jesse,
I lost the link to this topic - do you have the URL of your script?
Thanx,
Doug
November 11, 2008 at 10:22 am
http://www.sqlservercentral.com/scripts/Index+Management/63937/
There seems to be one other topic that has the same problem also - What to do with Multiple Similar Index Reccommendations
Viewing 4 posts - 1 through 3 (of 3 total)
You must be logged in to reply to this topic. Login to reply