Re: Calling function from select statement is slower then using subselect in select

hpuxrac schreef:

> John wrote:
> > Hi,
> >
> > Can anyone tell me what
> > TABLE ACCESS (BY USER ROWID)
> > means? Thanks!
> >
> > 0 SELECT STATEMENT Optimizer=RULE
> > 1 0 SORT (AGGREGATE)
> > 2 1 NESTED LOOPS
> > 3 2 TABLE ACCESS (FULL) OF 'TBRACCD'
> > 4 2 TABLE ACCESS (BY INDEX ROWID) OF 'TBBDETC'
> > 5 4 INDEX (UNIQUE SCAN) OF 'PK_TBBDETC' (UNIQUE)
> > 6 0 NESTED LOOPS (OUTER)
> > 7 6 NESTED LOOPS
> > 8 7 NESTED LOOPS
> > 9 8 NESTED LOOPS
> > 10 9 TABLE ACCESS (BY INDEX ROWID) OF 'TBRACCD'
> > 11 10 INDEX (RANGE SCAN) OF 'TBRACCD_CREDITS_INDEX' (N
> > ON-UNIQUE)
> > 12 9 TABLE ACCESS (BY INDEX ROWID) OF 'TBBDETC'
> > 13 12 INDEX (UNIQUE SCAN) OF 'PK_TBBDETC' (UNIQUE)
> > 14 8 TABLE ACCESS (BY INDEX ROWID) OF 'SPRIDEN'
> > 15 14 INDEX (RANGE SCAN) OF 'SPRIDEN_KEY_INDEX' (UNIQUE)
> > 16 7 INDEX (RANGE SCAN) OF 'SPRIDEN_KEY_INDEX' (UNIQUE)
> > 17 6 TABLE ACCESS (BY USER ROWID) OF 'SPRADDR'
> >
> >
> > John
>
> Why don't you give us a complete SQL statement and include some details
> about your environment ( oracle release etc )?
>
> Even a better idea, maybe you should start a new thread with your
> question instead of adding it to an existing ( old and apparently
> unrelated ) thread.

Actually he copied the execution plan posted by OP in the very first message in this thread, which also includes the SQL involved.

To answer his question:
TABLE ACCESS (BY USER ROWID) indicates that table rows are retrieved via user-defined rowid's. I guess this is described in Oracle's Performance Tuning Guide.

-- 
Jeroen